Marinomonas mediterranea is a lysogenic ... - Semantic Scholar
Microbiology (2003), 149, 2679–2686
DOI 10.1099/mic.0.26524-0
Marinomonas mediterranea is a lysogenic bacterium that synthesizes R-bodies Diana Herna´ndez-Romero,1 Patricia Lucas-Elı´o,1 Daniel Lo´pez-Serrano,2 Francisco Solano2 and Antonio Sanchez-Amat1 Correspondence Antonio Sanchez-Amat
Department of Genetics and Microbiology1 and Department of Biochemistry and Molecular Biology B2, University of Murcia, 30100 Murcia, Spain
[email protected]
Received 28 May 2003 Accepted 16 June 2003
The melanogenic marine bacterium Marinomonas mediterranea synthesizes R-bodies as revealed by transmission electron microscopy. These structures were previously described in some obligate symbionts of paramecia and some free-living bacteria, none of which was isolated from sea water. In other micro-organisms, the synthesis of R-bodies has been related to extrachromosomal elements. Accordingly, M. mediterranea induction by mitomycin C or UV radiation resulted in the production of defective phages resembling bacteriocins, indicating that it is a lysogenic bacterium. Two mitomycin-C-resistant strains defective in prophage replication have been isolated. These mutants, and the previously obtained strains ngC1, T102 and T103, the latter mutated in the ppoS gene encoding a sensor histidine kinase, are affected not only in phage replication but also in polyphenol oxidase activities and melanin synthesis, suggesting a relationship between the control of all these processes.
INTRODUCTION Marinomonas mediterranea is a melanogenic marine bacterium isolated and classified by our group (Solano et al., 1997; Solano & Sanchez-Amat, 1999). Melanins are protective polyphenolic pigments synthesized by different organisms through all the phylogenetic scale, from bacteria to mammals. Although melanins can be synthesized through different pathways, in higher organisms and a number of micro-organisms, melanin is made using L-tyrosine as the main precursor and the key enzyme is tyrosinase. This enzyme is a polyphenol oxidase (PPO) which catalyses two consecutive reactions of the melanogenic pathway: o-hydroxylation of L-tyrosine into 3,4-dihydroxyphenylalanine (L-DOPA) and its subsequent oxidation to yield L-dopaquinone (Hearing & Tsukamoto, 1991). After formation of this o-quinone, the pathway can proceed spontaneously since the latter compound is very reactive and undergoes a series of reactions involving oxidation, isomerization and polymerization that lead to an ill-defined final melanin pigment. Although the final pigment is protective, the high reactivity of quinones and other melanin intermediates makes such chemicals potentially harmful for cells during the process of pigment synthesis. One of the most efficient mechanisms in nature for cellular protection against the toxicity of those metabolites is the Abbreviations: DMPO, dimethoxyphenol oxidase; DOPA, 3,4-dihydroxyphenylalanine; DO, DOPA oxidase; DOSDS, DO activated by SDS; PPO, polyphenol oxidase; SST, Tris-buffered saline solution; TEM, transmission electron microscopy; TH, tyrosine hydroxylase; THSDS, TH activated by SDS.
0002-6524 G 2003 SGM
Printed in Great Britain
confinement of the process. In animal skin, melanin synthesis takes place in melanocytes and inside those cells is restricted to a specialized organelle, the melanosome (Hu, 1981; Riley, 1992). M. mediterranea was the first prokaryote described expressing two different PPOs. One of them is a tyrosinase, strongly activated in vitro by SDS, that is involved in melanin synthesis (Solano et al., 1997; Lo´pez-Serrano et al., 2002). The other PPO is a blue-copper membrane-bound laccase that is able to oxidize a wide range of aromatic phenols, including not only those characteristic for laccases, such as 2,6-dimethoxyphenol, but also those characteristic for tyrosinases, L-tyrosine and L-DOPA (Sanchez-Amat & Solano, 1997; Sanchez-Amat et al., 2001). Both PPO activities, as well as melanin synthesis are regulated by PpoS, a sensor histidine kinase, as revealed by the fact that strain T103, mutated in the gene encoding PpoS, shows lower levels of tyrosinase and laccase activities (Lucas-Elı´o et al., 2002). R-bodies are cytoplasmic inclusion bodies synthesized by only a few species of bacteria (Pond et al., 1989). They are highly insoluble protein ribbons that are coiled into cylindrical structures in the cell. They were first described in obligate symbionts of paramecia to which they confer the killer phenotype (Preer et al., 1974). R-bodies have also been described in some free-living Gram-negative bacteria belonging to the genus Pseudomonas (Lalucat et al., 1979; Wells & Horne, 1983; Espuny et al., 1991), in an unidentified soil bacterium, strain EPS-5028, (Fuste´ et al., 1986), and in the anoxigenic photosynthetic Rhodospirillum centenum (Favinger et al., 1989). None of those micro-organisms was 2679
D. Herna´ndez-Romero and others
isolated from sea water. It has been shown that the synthesis of R-bodies is determined by extrachromosomal elements such as plasmids (Kanabrocki et al., 1986; Heruth et al., 1994), or prophages, since viral particles have been isolated from bacteria producing R-bodies (Pond et al., 1989). Additionally, in two Pseudomonas strains producing R-bodies, mitomycin C induction resulted in the release of phage tail-like particles, indicative of defective prophages (Lalucat et al., 1979; Espuny et al., 1991). Morphologically similar defective viral particles have also been described in other micro-organisms as R-type bacteriocins. This is one of the subgroups into which Pseudomonas bacteriocins have been classified according to morphology and mode of action (Kageyama, 1975). However, they have also been described in other bacterial species (Strauch et al., 2001; Thaler et al., 1995). R-type bacteriocins are inducible by UV radiation or mitomycin C, and they resemble the tail of the T-even bacteriophages, showing a contractile sheath. It has been shown that they also contain single-stranded DNA (Lee et al., 1999). In this study we observed the synthesis of a novel highly organized cytoplasmic structure in M. mediterranea, and we have addressed the question of whether this structure is related to the process of melanin synthesis, or whether it constitutes the first example of R-body synthesis in a marine bacterium. In relation to this, mitomycin C and UV radiation induced the release of phage tail-like particles, indicating that M. mediterranea is a lysogenic bacterium. It has been observed that regulatory mutants in PPO activities and melanin synthesis, such as strain T103, and others isolated in this work, are also affected in R-body synthesis and prophage induction, suggesting a relationship between all these processes in M. mediterranea.
METHODS Bacterial culture conditions. M. mediterranea was usually grown
in the complex media Marine Broth or Agar 2216 (Difco), or MMC (Solano et al., 1997). When indicated, it was grown in MMM, a chemically defined medium described by Solano et al. (2000) in the absence or presence of 5 mM L-tyrosine to enhance melanin synthesis. In some experiments medium MN was used; this is identical to MMM except that copper sulfate is not added. All cultures were grown at 25 uC with aeration. Prophage induction. Prophage induction by mitomycin C was
performed by adapting protocols described by Jiang & Paul (1998). M. mediterranea was grown to exponential phase in MMC (OD600=0?2) at 25 uC with agitation. Then mitomycin C was added to a final concentration of 0?05 or 0?1 mg ml21. These cultures and the control were incubated again at 25 uC while monitoring the optical density. For UV radiation induction, M. mediterranea was grown in MN to an OD600 of 0?2. A sample (5 ml) of the culture was transferred to a glass sterile Petri dish and subjected to UV radiation doses of 1 mJ in a UV cross-linker Hoefer UVC 500. The irradiated cultures, and the non-irradiated controls, were further incubated in an Erlenmeyer flask, monitoring the OD600. Assay for bacteriocin activity. Test bacteria were inoculated in
double-layer Marine Agar medium, in which the upper layer, with 2680
0?7 % low-melting-point agarose, was poured at 30 uC to avoid heat shock to M. mediterranea and other marine bacteria. Samples of mitomycin-C- or UV-induced cultures were centrifuged (4000 g, 6 min) to remove cell debris and serially diluted in Tris-buffered saline solution (SST). Then 5 ml aliquots of 100–1023 dilutions were spotted onto the plates. A clear zone on the bacterial lawn would be indicative of inhibition of the test strain. Attempted prophage curing. To try to obtain M. mediterranea
strains cured of the prophage, strains resistant to the different prophage-inducing conditions were isolated. UV-resistant strains were first isolated by plating different dilutions of exponential cultures onto MN plates. These plates were subjected to UV irradiation in doses ranging from 0?5 to 2 mJ and the surviving strains were selected. A second approach was to select strains resistant to mitomycin C. Exponential-phase cultures in liquid medium were induced with mitomycin C as described above and surviving strains were selected by plating in medium MN at 2, 4, 6 and 24 h following mitomycin C exposure. A last approach was to plate M. mediterranea in MN or MMC with mitomycin C (0?1 mg ml21). Under these conditions resistant strains were detected with a frequency of approximately 161026 and selected for further studies. The defective nature of the prophage detected meant that it was not possible to use the standard criterion of reinfection by the viral particles to determine if a strain was cured of the prophage. In this work, the criterion used was the loss of the response to mitomycin C induction in MMC liquid culture. However, this character could be originated, not only by prophage curing but also by a defect in its replication. PPO enzymic determinations. Cell extract preparations and
quantitative spectrophotometric determination of PPO enzymic activities for different substrates was performed as described by Solano et al. (1997). Tyrosine hydroxylase (TH) and DOPA oxidase (DO) activities were determined by monitoring, respectively, the oxidation of 2 mM L-tyrosine and L-DOPA at 475 nm in 0?1 M phosphate buffer (pH 5?0). For TH activity, 25 mM L-DOPA was added to the assay mixture to eliminate the lag period (Solano et al., 1997). When required, the activities were also assayed in the presence of 0?02 % SDS (THSDS and DOSDS, respectively). Dimethoxyphenol oxidase (DMPO) and syringaldazine oxidase activities were determined by monitoring the oxidation of 2 mM dimethoxyphenol at 468 nm in 0?1 M sodium phosphate buffer (pH 5?0) or the oxidation of 50 mM syringaldazine at 525 nm at pH 6?5 (Solano et al., 1997), respectively. Reference cuvettes always had the same composition, except for the enzymic extract. In all cases, one unit was defined as the amount of enzyme that catalyses the appearance of 1 mmol product min21 at 37 uC. Specific activities were normalized (mg protein)21, measured using the bicinchoninic acid kit (Pierce Europe). Electron microscopy Ultrathin sections. Cells harvested in the stationary phase of growth
were prefixed in a solution of 2?5 % glutaraldehyde in SST (Solano & Sanchez-Amat, 1999). Next, the samples were fixed in 1 % osmium tetraoxide, followed by ethanol dehydration and embedding in Epon. After obtaining thin sections, the samples were stained with uranyl acetate and lead citrate. Small variations in the protocol, such as the addition of ruthenium red or avoiding the glutaraldehyde prefixation did not significantly affect the results obtained. Negative staining. M. mediterranea mitomycin-C-induced cultures were fixed with 3?6 % formaldehyde for 24 h and negatively stained for 1 min with 2 % uranyl acetate in Formvar-coated grids.
Microbiology 149
R-bodies and lysogeny in M. mediterranea
RESULTS Ultrastructural characterization of M. mediterranea The ultrastructure of M. mediterranea under different culture conditions was studied by transmission electron microscopy (TEM). A spherical to ellipsoidal cytoplasmic structure with a high degree of organization and generally attached to the cytoplasmic membrane was observed in the wild-type strain (Fig. 1a). The overall diameter of the structure was quite constant, ranging between 180 and 300 nm. The organization of the cytoplasmic structure was more easily appreciated in structures released after cell lysis (Fig. 1b). The two micrographs shown in Fig. 1(b) and 1(b9) seem to be perpendicular cuts of the same kind of structure. Three different zones could be observed from the outside to the inside. First, an outer envelope consisting of several concentric layers with a thickness of about 4 nm each (Fig. 1b). More internally, a less heavily stained inner envelope is clearly observed. The periodicity of these layers is 8 nm, and depending on the plane of the cut they can appear as parallel layers (Fig. 1b9). This inner area appears to be a wrapping of layers around the central core. Finally, the most internal area seems to contain granular material. Usually the structure appeared as spherical, but on some occasions it looked more like a spindle. This aspect may also be dependent on the plane of the cut. Ultrastructure synthesis under different growth conditions The structure detected in M. mediterranea bears some organizational similarity to the premelanosomes of higher organisms, a cellular organelle in which melanin synthesis takes place to protect the cell from the quinones generated in that process (Hu, 1981). To explore if this structure was related to melanogenesis, M. mediterranea wild-type was grown in several media in which a different degree of melanin synthesis takes place: complex media MMC and 2216, and mineral medium MMM plus 5 mM L-tyrosine (Solano et al., 2000). In addition, it was also cultivated in MMM without L-tyrosine, a condition under which melanins are not visually detected (Solano et al., 2000). In spite of the different growth conditions and levels of melanin synthesis the cytoplasmic structure was observed by TEM and there was no apparent morphological difference. Moreover, accumulation of electron-dense materials that could be considered to be melanins inside the structure was not detected. Ultrastructure of different M. mediterranea mutant strains In the second series of experiments, the appearance of the structure was studied in several PPO M. mediterranea mutants. Strain T101 (Tn101) is mutated, by transposon http://mic.sgmjournals.org
Fig. 1. Electron micrographs of M. mediterranea grown for 48 h at 25 6C in Marine Broth 2216. Typical thin sections of the wild-type strain are shown. (a) The arrow points to the novel cytoplasmic structure. (b, b9) Two thin sections of cytoplasmic structures released from lysed cells: oe, outer envelope; ie, inner envelope; ia, inner area. Bar, 0?2 mm. 2681
D. Herna´ndez-Romero and others
Table 1. Specific PPO activities (mU mg”1) in cellular extracts from M. mediterranea WT and different mutant strains Activity
WT
MIT1
TH THSDS DO DOSDS DMPO
14 106 27 242 75
1?2 18 5 30 82
MIT2
T103*
ngC1
T102
0?4 6?5 0?2 3 11?2
2?3 14?7 11?7 36?6 17?6
1 6 5 13 8
1?3 4?5 6?1 12?8 7?2
*Data from Lucas-Elı´o et al. (2002).
insertion, in the membrane-bound blue-copper PPO and retains the capacity to synthesize melanins (Solano et al., 2000). Strains ng56 and T105 (L19) are affected in SDSactivated tyrosinase activity and are unable to synthesize melanins in any of the media assayed (Solano et al., 1997; Lo´pez-Serrano et al., 2002). By electron microscopy it was observed that all these strains retained the capacity to synthesize the cytoplasmic structure (data not shown). On the contrary, we were unable to observe the highly organized structure in strain T103, a regulatory mutant affected in PpoS, a sensor histidine kinase that regulates the expression of both PPOs as well as melanin synthesis (Lucas-Elı´o et al., 2002).
transposon mutagenesis. Although strain T102 was obtained using the same protocol used to generate strain T103, the transposon was inserted into a different locus (P. Lucas-Elı´o and others, unpublished results). In these new regulatory mutants we were also unable to detect the synthesis of the cytoplasmic structure. Prophage induction The synthesis of R-bodies has been related in some strains to extrachromosomal elements (Pond et al., 1989). To explore the lysogenic state of M. mediterranea, cultures in MMC were induced by mitomycin C. After the addition of this compound the cultures lysed, suggesting phage induction (Fig. 2). Electron microscopy observation of these lysates revealed the presence of phage tail-like viral particles (Fig. 3). Different kinds of particles were detected: extended rods (Fig. 3a), contracted particles and empty sheaths (Fig. 3b). As shown in Fig. 3(b), the contracted particles tend to aggregate. The morphology of these particles is identical to the morphology of R-type bacteriocins which
In our collection of PPO mutants we had isolated two other strains, ngC1 and T102, affected in PPO regulation and melanin synthesis (Table 1). Strain ngC1 was isolated by nitrosoguanidine mutagenesis, and strain T102 by
Fig. 2. Evolution of the OD600 of M. mediterranea cultures in MMC. $, Control; m, treated with 0?05 mg mitomycin C ml”1; &, treated with 0?1 mg mitomycin C ml”1. The arrow indicates the time of mitomycin C addition. 2682
Fig. 3. Electron microscopy of phage tail-like particles detected in the supernatants of M. mediterranea cultures in MMC lysed after induction with 0?1 mg mitomycin C ml”1. (a) Extended rods; (b) contracted particles (cp) and empty sheaths (es). Bar, 0?1 mm. Microbiology 149
R-bodies and lysogeny in M. mediterranea
are considered defective prophages in which the head proteins are not expressed (Shinomiya & Ina, 1989). M. mediterranea was also induced by UV radiation. Preliminary experiments indicated that it was very sensitive to this treatment and 1 mJ doses produced a survival rate of 0?5 %. Liquid cultures treated with this dose lysed, and in the supernatants of those lysates viral particles morphologically identical to those described previously in mitomycinC-induced cultures were also observed (see Fig. 3). The possible bacteriocin activity of cultures induced by mitomycin C or UV was assayed against different bacterial strains, including the M. mediterranea wild-type strain and all its mutant strains isolated in our lab. Unfortunately, there are no other M. mediterranea strains described so far, so the closest relatives that we could assay were Marinomonas vaga ATCC 27119 and Marinomonas communis ATCC 27118, the type strains of two other species in this genus. Finally, the same assay was performed using other unrelated marine bacteria such as Vibrio parahaemolyticus BB22 and Vibrio harveyi 392, as well as Escherichia coli DH5a. No inhibitory activity against any bacterial strain was observed with UVinduced cultures, or cultures in MMC medium induced with mitomycin C, in spite of the presence of viral particles. However, cultures in MN induced by mitomycin C showed inhibitory activity at the lowest dilution against all the strains assayed, including E. coli. This result indicates that it is not a bacteriocin activity, but rather a consequence of an unrelated wide-spectrum antimicrobial compound that is under study in our lab. Characterization of mutants affected in prophage replication To look for M. mediterranea mutants that could be cured of the prophage, or affected in its replication, several protocols were applied to isolate strains surviving after prophage induction (see Methods). A total of 140 strains were characterized: 35 of them were isolated as resistant to UV radiation; 60 strains were selected after mitomycin C treatment in liquid medium, and finally 45 strains that grew in plates containing mitomycin C were also selected. It was observed that all these strains, except strains MIT1 and MIT2 which were selected on MN plates containing mitomycin, were inducible by mitomycin C in liquid medium. The phenotypic characterization of MIT1 and MIT2 revealed some relationship between PPO activity and prophage replication, since they were amelanogenic, as well as affected in PPO activities (Table 1). In this sense, strains MIT1 and MIT2 were very similar to the regulatory mutants ngC1, T102 and T103. These results prompted us to study the response to mitomycin C in the different PPO mutant strains in our collection. It was observed that mutants in PPO structural genes, for example strain T101, were lysed (Fig. 4b). On the contrary, similar to MIT1 and MIT2, the regulatory mutants such as ngC1, T102 and T103 were not inducible by mitomycin C (Fig. 4a). All these five mutant strains, MIT1, MIT2, ngC1, T102 and T103, http://mic.sgmjournals.org
Fig. 4. Evolution of the OD600 of cultures in MMC of M. mediterranea mutant strains (a) T103, PPO regulatory mutant, and (b) T101, mutated in the multipotent laccase structural gene. $, Control; &, treated with 0?1 mg mitomycin C ml”1. The arrow indicates the time of mitomycin C addition.
were unable to synthesize the cellular structure, indicating the relationship between its synthesis, prophage replication and PPO activity regulation. To explore if strains MIT1, MIT2 and T103 were cured of the prophage, they were used as hosts in the double-layer agar technique against M. mediterranea wild-type lysates obtained by UV or mitomycin C induction. Although different media and incubation conditions were assayed it was not possible to detect plaques indicative of phage replication (data not shown).
DISCUSSION In this study it has been shown that M. mediterranea synthesizes a highly organized cytoplasmic structure under different culture conditions. This structure resembles morphologically the R-bodies described in some prokaryotic cells. However, it also shows some similarities to the lamellar structures described in some animal melanocytes, specialized cells inside of which melanin synthesis takes 2683
D. Herna´ndez-Romero and others
place (Hu, 1981). Since M. mediterranea is a melanogenic micro-organism that expresses several PPO activities, several experiments were undertaken to explore the possible relationship between both processes. It was observed that melanin synthesis did not take place inside this structure, and also that this structure was made under conditions where no melanin synthesis takes place, such as growth in MMM. In addition, the synthesis of the structure was not affected in some amelanogenic mutants such as ng56 and T105. All these data suggest that the synthesis of this structure is not directly related to melanogenesis. However, it was intriguing to observe that all the mutants affected in the regulation of PPO activities and melanin synthesis, ngC1, T102 and T103, were also affected in the structure synthesis. Although the mutations in ngC1 and T102 have not been fully characterized at the molecular level yet, the mutant T103 is affected in the gene encoding PpoS, a membrane histidine kinase, characteristic of a two-component regulatory system (Lucas-Elı´o et al., 2002). This observation of a connection at the regulatory level between structure synthesis and PPO activity is related to the modification of prophage replication in these mutant strains as discussed below. The structure produced by M. mediterranea is morphologically very similar to prokaryotic R-bodies and particularly to those produced by Hydrogenophaga (Pseudomonas) taeniospiralis (Lalucat et al., 1979). The main difference detected under electron microscopy is that the Marinomonas R-body shows an outer envelope that, as far as we know, has not been described in other bacteria. This structural difference may explain why it is not observed as a refractile body under light microscopy. These structures are well preserved after being released from the cells as observed in Fig. 1(b), indicative of a proteinaceous nature as for other R-bodies (Pond et al., 1989). The R-bodies described are a heterogeneous group of structures not closely related with regard to the proteins associated with them or the genetic determinants for their synthesis (Kanabrocki et al., 1986). The killer phenotype of paramecia that contains bacterial endosymbionts synthesizing R-bodies has been related to the unrolling of this structure and the breakage of the phagosomal membrane of the sensitive paramecia that ingest them (Preer et al., 1974). All these endosymbionts were placed in the genus Caedibacter. However, recent studies have shown that in spite of sharing the capacity to synthesize R-bodies, members of this single genus are very distant phylogenetically (Beier et al., 2002). Other similar structures participating in defensive roles are the extrusomes, synthesized by the symbiotic bacteria of ciliates (Petroni et al., 2000). The diversity of bacteria producing R-bodies is also revealed by the fact that they are synthesized by non-symbiotic bacteria. In this case, no physiological role has been proposed for them. In fact they have been described in bacteria that show a variety of energy metabolisms, such as hydrogen-oxidizing (Lalucat et al., 1979), heterotrophic (Espuny et al., 1991) or photosynthetic (Favinger et al., 1989) bacteria, and have been isolated from different 2684
habitats. However, the synthesis of the R-body by M. mediterranea gives the first example of the production of this kind of structure by a marine bacterium. The origin of the R-bodies in the strains studied is considered to be extrachromosomal, determined either by plasmids as in Caedibacter taeniospiralis (Kanabrocki et al., 1986; Heruth et al., 1994) or encoded by bacteriophages (Pond et al., 1989). So far there is no evidence of the existence of indigenous plasmids in M. mediterranea. However, the results presented in this study clearly indicate that M. mediterranea is a lysogenic bacterium containing at least one prophage which is inducible by either UV or mitomycin C. Although this is the first communication of lysogeny in bacteria of the genus Marinomonas, this result is in agreement with recent studies indicating that lysogeny is a common phenomenon in the marine environment. Using induction with mitomycin C or UV radiation it has been shown that up to 40 % of marine bacteria contain inducible prophages (Jiang & Paul, 1998). In another study it was shown that prophages were induced from 52 % of water samples (Cochran & Paul, 1998). Microscopic evidence suggests that the Marinomonas prophage is defective. Different viral particles were observed under all the inducing conditions (Fig. 3) and by analogy with R-type bacteriocins to which they are identical, they may correspond to the contractile phage tail of a defective prophage in which the head proteins are not expressed (Lee et al., 1999; Shinomiya & Ina, 1989). To assay the bacteriocin activity of the viral particles detected in the M. mediterranea induced cultures, different bacterial strains were used, including two strains belonging to other Marinomonas species. Although, the bacteriocin activity was not detected this result should be cautiously interpreted since bacteriocins generally possess a very narrow spectrum of activity and, perhaps, could be detected with other bacterial strains not available at this time. The R-type bacteriocins of Pseudomonas aeruginosa are more thoroughly characterized, but they are also produced by different bacterial species (Strauch et al., 2001; Thaler et al., 1995). Moreover, the existence of defective prophages in R-body-producing bacteria has also been described in other micro-organisms (Lalucat et al., 1979; Espuny et al., 1991), indicates a relationship between these two processes. The results obtained in this work, showing that the strains defective in phage replication are also affected in R-body synthesis supports this relationship. In this regard, it is tempting to speculate that in those cases R-bodies could be the result of the polymerization of viral proteins in the cytoplasm of the host. On the other hand, the defective nature of the prophage may be related to difficulties in the isolation of bacterial strains that have been cured of it. In spite of numerous efforts, only two strains, MIT1 and MIT2, not inducible by mitomycin C were isolated. The mutations in those strains are not characterized. However, both strains share Microbiology 149
R-bodies and lysogeny in M. mediterranea
some characteristics with strain T103 that was previously isolated and characterized by our group (Lucas-Elı´o et al., 2002), and with two other strains obtained by different protocols such as nitrosoguanidine mutagenesis (strain ngC1) and transposon mutagenesis (strain T102). All five of these strains are amelanogenic, affected in the levels of PPO activity and not inducible by mitomycin C. These results clearly indicate that there are some common mechanisms co-regulating PPO activities and prophage induction that are affected in those strains. Strain T103 is mutated by transposon insertion in ppoS, a gene encoding a membrane histidine kinase, revealing that PpoS is regulating prophage replication in addition to the previously described PPO activities and melanin synthesis (Lucas-Elı´o et al., 2002). PpoS is a tripartite sensor kinase, a kind of sensor histidine kinase characterized by the regulation of complex traits involving different cellular processes (Appleby et al., 1996). The coincidence of the regulation of prophage induction and PPO activities by PpoS as observed in this study is intriguing and there are some possibilities that would deserve further study. For example, it cannot be ruled out that the induction of prophage replication could be a stress factor for M. mediterranea that in turn may induce PPO activities. However, the sequencing of the genomes of different micro-organisms has revealed many genes of viral origin, and in some cases those genes play a physiologically relevant role for the host. In this context prophage replication would induce the death of some Marinomonas cells, facilitating the release of PPO enzymes and the synthesis of extracellular melanins. These melanins may protect the surviving cells from UV radiation or act as a sink for extracellular oxidative radicals. Alternatively, it could be a mechanism facilitating the polymerization of extracellular compounds, making them unavailable for competing micro-organisms. In this sense melanin synthesis and PPO activities could have some physiological relevance at the population level.
ACKNOWLEDGEMENTS This work was supported by the grant PB2001-0140 from the CICYT, Spain. P. Lucas-Elı´o and D. Lo´pez-Serrano were recipients of pre-doctoral fellowships from, respectively, the Se´neca Foundation (Comunidad Auto´noma de la Regio´n de Murcia) and the University of Murcia. We are grateful to Professor J. Lalucat (IMEDEA, CSIC-UIB) for helpful suggestions. We also thank the Electronic Microscopy Service of the University of Murcia.
Cochran, P. K. & Paul, J. H. (1998). Seasonal abundance of lysogenic
bacteria in a subtropical estuary. Appl Environ Microbiol 64, 2308–2312. Espuny, M. J., Andre´s, C., Mercade, M. E., Robert, M., Manresa, M. A. & Guinea, J. (1991). R-bodies in Pseudomonas aeruginosa strain
44T1. Antonie van Leeuwenhoek 60, 83–86. Favinger, J., Stadtwald, R. & Gest, H. (1989). Rhodospirillum
centenum, sp. nov. a thermotolerant cyst-forming anoxygenic photosynthetic bacterium. Antonie van Leeuwenhoek 55, 291–296. Fuste´, M. C., Simon-Pujol, M. D., Marques, A. M., Guinea, J. & Congregrado, F. (1986). Isolation of a new free-living bacterium
containing R-bodies. J Gen Microbiol 132, 2801–2805. Hearing, V. & Tsukamoto, K. (1991). Enzymatic control of
pigmentation in mammals. FASEB J 5, 2902–2909. Heruth, D. P., Pond, F. R., Dilts, J. A. & Quackenbush, R. L. (1994).
Characterization of genetic determinants for R body synthesis and assembly in Caedibacter taeniospiralis 47 and 116. J Bacteriol 176, 3559–3567. Hu, F. (1981). Melanocyte cytology in normal skin, melanocyte nevi and malignant melanomas. In Pathology of Malignant Melanoma, pp. 1–21. Edited by A. B. Ackerman. New York: Masson Publishing. Jiang, S. C. & Paul, J. H. (1998). Significance of lysogeny in the
marine environment: studies with isolates and a model of lysogenic phage production. Microb Ecol 35, 235–243. M. (1975). Bacteriocins and bacteriophages in Pseudomonas aeruginosa. In Microbial Drug Resistance, pp. 291– 305. Edited by S. Mitsuhashi & H. Hashimoto. Tokyo: University of Tokyo Press.
Kageyama,
Kanabrocki, J. A., Quackenbush, R. L. & Pond, F. R. (1986).
Organization and expression of genetic determinants for synthesis and assembly of type 51 R bodies. J Bacteriol 168, 40–48. Lalucat, J., Meyer, O., Mayer, F., Pares, R. & Schlegel, H. G. (1979).
R-bodies in newly isolated free-living hydrogen-oxidizing bacteria. Arch Microbiol 121, 9–15. Lee, F. K. N., Dudas, K. C., Hanson, J. A., Nelson, M. B., LoVerde, P. T. & Apicella, M. A. (1999). The R-type pyocin of Pseudomonas
aeruginosa C is a bacteriophage tail-like particle that contains singlestranded DNA. Infect Immun 67, 717–725. Lo´pez-Serrano, D., Sanchez-Amat, A. & Solano, F. (2002). Cloning
and molecular characterization of a SDS-activated tyrosinase from Marinomonas mediterranea. Pigment Cell Res 15, 104–111. Lucas-Elı´o, P., Solano, F. & Sanchez-Amat, A. (2002). Regulation of
polyphenol oxidase activities and melanin synthesis in Marinomonas mediterranea: identification of ppoS, a gene encoding a sensor histidine kinase. Microbiology 148, 2457–2466. Petroni, G., Spring, S., Schleifer, K. H., Verni, F. & Rosati, G. (2000).
Defensive extrusive ectosymbionts of Euplotidium (Ciliophora) that contain microtubule-like structures are bacteria related to Verrucomicrobia. Proc Natl Acad Sci U S A 97, 1813–1817. Pond, F. R., Gibson, I., Lalucat, J. & Quackenbush, R. L. (1989).
R-body-producing bacteria. Microb Rev 53, 25–67.
REFERENCES
Preer, J. R., Preer, L. B. & Jurand, A. (1974). Kappa and other
endosymbionts in Paramecium aurelia. Bacteriol Rev 38, 113–163.
Appleby, J. L., Parkinson, J. S. & Bourret, R. B. (1996). Signal
Riley, P. A. (1992). Materia melanica: further dark thoughts. Pigment
transduction via the multi-step phosphorelay: not necessarily a road less traveled. Cell 86, 845–848.
Sanchez-Amat, A. & Solano, F. (1997). A pluripotent polyphenol
Beier, C. L., Horn, M., Michel, R., Schweikert, M., Go¨rtz, H.-D. & Wagner, M. (2002). The genus Caedibacter comprises endosymbionts
of Paramecium spp. related to the Rickettsiales (Alphaproteobacteria) and to Francisella tularensis (Gammaproteobacteria). Appl Environ Microbiol 68, 6043–6050. http://mic.sgmjournals.org
Cell Res 5, 101–106. oxidase from the melanogenic marine Alteromonas sp. shares catalytic capabilities of tyrosinases and laccases. Biochem Biophys Res Commun 240, 787–792. Sanchez-Amat, A., Lucas-Elio, P., Ferna´ndez, E., Garcia-Borro´n, J. C. & Solano, F. (2001). Molecular cloning and functional
2685
D. Herna´ndez-Romero and others characterization of a unique multipotent polyphenol oxidase from Marinomonas mediterranea. Biochim Biophys Acta 1547, 104–116.
isolation of a multipotent polyphenol oxidase mutant. J Bacteriol 182, 3754–3760.
Shinomiya, T. & Ina, S. (1989). Genetic comparison of bacteriophage
Strauch, E., Kaspar, H., Schaudinn, C., Dersch, P., Madela, K., Gewinner, C., Hertwig, S., Wecke, J. & Appel, B. (2001).
PS17 and Pseudomonas aeruginosa R-type pyocin. J Bacteriol 171, 2287–2292. Solano, F. & Sanchez-Amat, A. (1999). Studies on the phylogenetic
relationships of melanogenic marine bacteria: proposal of Marinomonas mediterranea sp. nov. Int J Syst Bacteriol 49, 1241–1246. Solano, F., Garcı´a, E., Pe´rez de Egea, E. & Sanchez-Amat, A. (1997). Isolation and characterization of strain MMB-1 (CECT
Characterization of enterocoliticin, a phage tail-like bacteriocin, and its effect on pathogenic Yersinia enterocolitica strains. Appl Environ Microbiol 67, 5634–5642. Thaler, J. O., Baghdiguian, S. & Boemare, N. (1995). Purification
and characterization of xenorhabdicin, a phage tail-like bacteriocin, from the lisogenic strain F1 of Xenorhabdus nematophilus. Appl Environ Microbiol 61, 2049–2052.
4803), a novel melanogenic marine bacterium. Appl Environ Microbiol 63, 3499–3506.
Wells, B. & Horne, R. W. (1983). The ultrastructure of Pseudomonas
Solano, F., Lucas-Elı´o, P., Ferna´ndez, E. & Sanchez-Amat, A. (2000). Marinomonas mediterranea MMB-1 transposon mutagenesis:
avenae. II. Intracellular refractile (R-body) structure. Micron Microsc Acta 14, 329–344.
2686
Microbiology 149
DOI 10.1099/mic.0.26524-0
Marinomonas mediterranea is a lysogenic bacterium that synthesizes R-bodies Diana Herna´ndez-Romero,1 Patricia Lucas-Elı´o,1 Daniel Lo´pez-Serrano,2 Francisco Solano2 and Antonio Sanchez-Amat1 Correspondence Antonio Sanchez-Amat
Department of Genetics and Microbiology1 and Department of Biochemistry and Molecular Biology B2, University of Murcia, 30100 Murcia, Spain
[email protected]
Received 28 May 2003 Accepted 16 June 2003
The melanogenic marine bacterium Marinomonas mediterranea synthesizes R-bodies as revealed by transmission electron microscopy. These structures were previously described in some obligate symbionts of paramecia and some free-living bacteria, none of which was isolated from sea water. In other micro-organisms, the synthesis of R-bodies has been related to extrachromosomal elements. Accordingly, M. mediterranea induction by mitomycin C or UV radiation resulted in the production of defective phages resembling bacteriocins, indicating that it is a lysogenic bacterium. Two mitomycin-C-resistant strains defective in prophage replication have been isolated. These mutants, and the previously obtained strains ngC1, T102 and T103, the latter mutated in the ppoS gene encoding a sensor histidine kinase, are affected not only in phage replication but also in polyphenol oxidase activities and melanin synthesis, suggesting a relationship between the control of all these processes.
INTRODUCTION Marinomonas mediterranea is a melanogenic marine bacterium isolated and classified by our group (Solano et al., 1997; Solano & Sanchez-Amat, 1999). Melanins are protective polyphenolic pigments synthesized by different organisms through all the phylogenetic scale, from bacteria to mammals. Although melanins can be synthesized through different pathways, in higher organisms and a number of micro-organisms, melanin is made using L-tyrosine as the main precursor and the key enzyme is tyrosinase. This enzyme is a polyphenol oxidase (PPO) which catalyses two consecutive reactions of the melanogenic pathway: o-hydroxylation of L-tyrosine into 3,4-dihydroxyphenylalanine (L-DOPA) and its subsequent oxidation to yield L-dopaquinone (Hearing & Tsukamoto, 1991). After formation of this o-quinone, the pathway can proceed spontaneously since the latter compound is very reactive and undergoes a series of reactions involving oxidation, isomerization and polymerization that lead to an ill-defined final melanin pigment. Although the final pigment is protective, the high reactivity of quinones and other melanin intermediates makes such chemicals potentially harmful for cells during the process of pigment synthesis. One of the most efficient mechanisms in nature for cellular protection against the toxicity of those metabolites is the Abbreviations: DMPO, dimethoxyphenol oxidase; DOPA, 3,4-dihydroxyphenylalanine; DO, DOPA oxidase; DOSDS, DO activated by SDS; PPO, polyphenol oxidase; SST, Tris-buffered saline solution; TEM, transmission electron microscopy; TH, tyrosine hydroxylase; THSDS, TH activated by SDS.
0002-6524 G 2003 SGM
Printed in Great Britain
confinement of the process. In animal skin, melanin synthesis takes place in melanocytes and inside those cells is restricted to a specialized organelle, the melanosome (Hu, 1981; Riley, 1992). M. mediterranea was the first prokaryote described expressing two different PPOs. One of them is a tyrosinase, strongly activated in vitro by SDS, that is involved in melanin synthesis (Solano et al., 1997; Lo´pez-Serrano et al., 2002). The other PPO is a blue-copper membrane-bound laccase that is able to oxidize a wide range of aromatic phenols, including not only those characteristic for laccases, such as 2,6-dimethoxyphenol, but also those characteristic for tyrosinases, L-tyrosine and L-DOPA (Sanchez-Amat & Solano, 1997; Sanchez-Amat et al., 2001). Both PPO activities, as well as melanin synthesis are regulated by PpoS, a sensor histidine kinase, as revealed by the fact that strain T103, mutated in the gene encoding PpoS, shows lower levels of tyrosinase and laccase activities (Lucas-Elı´o et al., 2002). R-bodies are cytoplasmic inclusion bodies synthesized by only a few species of bacteria (Pond et al., 1989). They are highly insoluble protein ribbons that are coiled into cylindrical structures in the cell. They were first described in obligate symbionts of paramecia to which they confer the killer phenotype (Preer et al., 1974). R-bodies have also been described in some free-living Gram-negative bacteria belonging to the genus Pseudomonas (Lalucat et al., 1979; Wells & Horne, 1983; Espuny et al., 1991), in an unidentified soil bacterium, strain EPS-5028, (Fuste´ et al., 1986), and in the anoxigenic photosynthetic Rhodospirillum centenum (Favinger et al., 1989). None of those micro-organisms was 2679
D. Herna´ndez-Romero and others
isolated from sea water. It has been shown that the synthesis of R-bodies is determined by extrachromosomal elements such as plasmids (Kanabrocki et al., 1986; Heruth et al., 1994), or prophages, since viral particles have been isolated from bacteria producing R-bodies (Pond et al., 1989). Additionally, in two Pseudomonas strains producing R-bodies, mitomycin C induction resulted in the release of phage tail-like particles, indicative of defective prophages (Lalucat et al., 1979; Espuny et al., 1991). Morphologically similar defective viral particles have also been described in other micro-organisms as R-type bacteriocins. This is one of the subgroups into which Pseudomonas bacteriocins have been classified according to morphology and mode of action (Kageyama, 1975). However, they have also been described in other bacterial species (Strauch et al., 2001; Thaler et al., 1995). R-type bacteriocins are inducible by UV radiation or mitomycin C, and they resemble the tail of the T-even bacteriophages, showing a contractile sheath. It has been shown that they also contain single-stranded DNA (Lee et al., 1999). In this study we observed the synthesis of a novel highly organized cytoplasmic structure in M. mediterranea, and we have addressed the question of whether this structure is related to the process of melanin synthesis, or whether it constitutes the first example of R-body synthesis in a marine bacterium. In relation to this, mitomycin C and UV radiation induced the release of phage tail-like particles, indicating that M. mediterranea is a lysogenic bacterium. It has been observed that regulatory mutants in PPO activities and melanin synthesis, such as strain T103, and others isolated in this work, are also affected in R-body synthesis and prophage induction, suggesting a relationship between all these processes in M. mediterranea.
METHODS Bacterial culture conditions. M. mediterranea was usually grown
in the complex media Marine Broth or Agar 2216 (Difco), or MMC (Solano et al., 1997). When indicated, it was grown in MMM, a chemically defined medium described by Solano et al. (2000) in the absence or presence of 5 mM L-tyrosine to enhance melanin synthesis. In some experiments medium MN was used; this is identical to MMM except that copper sulfate is not added. All cultures were grown at 25 uC with aeration. Prophage induction. Prophage induction by mitomycin C was
performed by adapting protocols described by Jiang & Paul (1998). M. mediterranea was grown to exponential phase in MMC (OD600=0?2) at 25 uC with agitation. Then mitomycin C was added to a final concentration of 0?05 or 0?1 mg ml21. These cultures and the control were incubated again at 25 uC while monitoring the optical density. For UV radiation induction, M. mediterranea was grown in MN to an OD600 of 0?2. A sample (5 ml) of the culture was transferred to a glass sterile Petri dish and subjected to UV radiation doses of 1 mJ in a UV cross-linker Hoefer UVC 500. The irradiated cultures, and the non-irradiated controls, were further incubated in an Erlenmeyer flask, monitoring the OD600. Assay for bacteriocin activity. Test bacteria were inoculated in
double-layer Marine Agar medium, in which the upper layer, with 2680
0?7 % low-melting-point agarose, was poured at 30 uC to avoid heat shock to M. mediterranea and other marine bacteria. Samples of mitomycin-C- or UV-induced cultures were centrifuged (4000 g, 6 min) to remove cell debris and serially diluted in Tris-buffered saline solution (SST). Then 5 ml aliquots of 100–1023 dilutions were spotted onto the plates. A clear zone on the bacterial lawn would be indicative of inhibition of the test strain. Attempted prophage curing. To try to obtain M. mediterranea
strains cured of the prophage, strains resistant to the different prophage-inducing conditions were isolated. UV-resistant strains were first isolated by plating different dilutions of exponential cultures onto MN plates. These plates were subjected to UV irradiation in doses ranging from 0?5 to 2 mJ and the surviving strains were selected. A second approach was to select strains resistant to mitomycin C. Exponential-phase cultures in liquid medium were induced with mitomycin C as described above and surviving strains were selected by plating in medium MN at 2, 4, 6 and 24 h following mitomycin C exposure. A last approach was to plate M. mediterranea in MN or MMC with mitomycin C (0?1 mg ml21). Under these conditions resistant strains were detected with a frequency of approximately 161026 and selected for further studies. The defective nature of the prophage detected meant that it was not possible to use the standard criterion of reinfection by the viral particles to determine if a strain was cured of the prophage. In this work, the criterion used was the loss of the response to mitomycin C induction in MMC liquid culture. However, this character could be originated, not only by prophage curing but also by a defect in its replication. PPO enzymic determinations. Cell extract preparations and
quantitative spectrophotometric determination of PPO enzymic activities for different substrates was performed as described by Solano et al. (1997). Tyrosine hydroxylase (TH) and DOPA oxidase (DO) activities were determined by monitoring, respectively, the oxidation of 2 mM L-tyrosine and L-DOPA at 475 nm in 0?1 M phosphate buffer (pH 5?0). For TH activity, 25 mM L-DOPA was added to the assay mixture to eliminate the lag period (Solano et al., 1997). When required, the activities were also assayed in the presence of 0?02 % SDS (THSDS and DOSDS, respectively). Dimethoxyphenol oxidase (DMPO) and syringaldazine oxidase activities were determined by monitoring the oxidation of 2 mM dimethoxyphenol at 468 nm in 0?1 M sodium phosphate buffer (pH 5?0) or the oxidation of 50 mM syringaldazine at 525 nm at pH 6?5 (Solano et al., 1997), respectively. Reference cuvettes always had the same composition, except for the enzymic extract. In all cases, one unit was defined as the amount of enzyme that catalyses the appearance of 1 mmol product min21 at 37 uC. Specific activities were normalized (mg protein)21, measured using the bicinchoninic acid kit (Pierce Europe). Electron microscopy Ultrathin sections. Cells harvested in the stationary phase of growth
were prefixed in a solution of 2?5 % glutaraldehyde in SST (Solano & Sanchez-Amat, 1999). Next, the samples were fixed in 1 % osmium tetraoxide, followed by ethanol dehydration and embedding in Epon. After obtaining thin sections, the samples were stained with uranyl acetate and lead citrate. Small variations in the protocol, such as the addition of ruthenium red or avoiding the glutaraldehyde prefixation did not significantly affect the results obtained. Negative staining. M. mediterranea mitomycin-C-induced cultures were fixed with 3?6 % formaldehyde for 24 h and negatively stained for 1 min with 2 % uranyl acetate in Formvar-coated grids.
Microbiology 149
R-bodies and lysogeny in M. mediterranea
RESULTS Ultrastructural characterization of M. mediterranea The ultrastructure of M. mediterranea under different culture conditions was studied by transmission electron microscopy (TEM). A spherical to ellipsoidal cytoplasmic structure with a high degree of organization and generally attached to the cytoplasmic membrane was observed in the wild-type strain (Fig. 1a). The overall diameter of the structure was quite constant, ranging between 180 and 300 nm. The organization of the cytoplasmic structure was more easily appreciated in structures released after cell lysis (Fig. 1b). The two micrographs shown in Fig. 1(b) and 1(b9) seem to be perpendicular cuts of the same kind of structure. Three different zones could be observed from the outside to the inside. First, an outer envelope consisting of several concentric layers with a thickness of about 4 nm each (Fig. 1b). More internally, a less heavily stained inner envelope is clearly observed. The periodicity of these layers is 8 nm, and depending on the plane of the cut they can appear as parallel layers (Fig. 1b9). This inner area appears to be a wrapping of layers around the central core. Finally, the most internal area seems to contain granular material. Usually the structure appeared as spherical, but on some occasions it looked more like a spindle. This aspect may also be dependent on the plane of the cut. Ultrastructure synthesis under different growth conditions The structure detected in M. mediterranea bears some organizational similarity to the premelanosomes of higher organisms, a cellular organelle in which melanin synthesis takes place to protect the cell from the quinones generated in that process (Hu, 1981). To explore if this structure was related to melanogenesis, M. mediterranea wild-type was grown in several media in which a different degree of melanin synthesis takes place: complex media MMC and 2216, and mineral medium MMM plus 5 mM L-tyrosine (Solano et al., 2000). In addition, it was also cultivated in MMM without L-tyrosine, a condition under which melanins are not visually detected (Solano et al., 2000). In spite of the different growth conditions and levels of melanin synthesis the cytoplasmic structure was observed by TEM and there was no apparent morphological difference. Moreover, accumulation of electron-dense materials that could be considered to be melanins inside the structure was not detected. Ultrastructure of different M. mediterranea mutant strains In the second series of experiments, the appearance of the structure was studied in several PPO M. mediterranea mutants. Strain T101 (Tn101) is mutated, by transposon http://mic.sgmjournals.org
Fig. 1. Electron micrographs of M. mediterranea grown for 48 h at 25 6C in Marine Broth 2216. Typical thin sections of the wild-type strain are shown. (a) The arrow points to the novel cytoplasmic structure. (b, b9) Two thin sections of cytoplasmic structures released from lysed cells: oe, outer envelope; ie, inner envelope; ia, inner area. Bar, 0?2 mm. 2681
D. Herna´ndez-Romero and others
Table 1. Specific PPO activities (mU mg”1) in cellular extracts from M. mediterranea WT and different mutant strains Activity
WT
MIT1
TH THSDS DO DOSDS DMPO
14 106 27 242 75
1?2 18 5 30 82
MIT2
T103*
ngC1
T102
0?4 6?5 0?2 3 11?2
2?3 14?7 11?7 36?6 17?6
1 6 5 13 8
1?3 4?5 6?1 12?8 7?2
*Data from Lucas-Elı´o et al. (2002).
insertion, in the membrane-bound blue-copper PPO and retains the capacity to synthesize melanins (Solano et al., 2000). Strains ng56 and T105 (L19) are affected in SDSactivated tyrosinase activity and are unable to synthesize melanins in any of the media assayed (Solano et al., 1997; Lo´pez-Serrano et al., 2002). By electron microscopy it was observed that all these strains retained the capacity to synthesize the cytoplasmic structure (data not shown). On the contrary, we were unable to observe the highly organized structure in strain T103, a regulatory mutant affected in PpoS, a sensor histidine kinase that regulates the expression of both PPOs as well as melanin synthesis (Lucas-Elı´o et al., 2002).
transposon mutagenesis. Although strain T102 was obtained using the same protocol used to generate strain T103, the transposon was inserted into a different locus (P. Lucas-Elı´o and others, unpublished results). In these new regulatory mutants we were also unable to detect the synthesis of the cytoplasmic structure. Prophage induction The synthesis of R-bodies has been related in some strains to extrachromosomal elements (Pond et al., 1989). To explore the lysogenic state of M. mediterranea, cultures in MMC were induced by mitomycin C. After the addition of this compound the cultures lysed, suggesting phage induction (Fig. 2). Electron microscopy observation of these lysates revealed the presence of phage tail-like viral particles (Fig. 3). Different kinds of particles were detected: extended rods (Fig. 3a), contracted particles and empty sheaths (Fig. 3b). As shown in Fig. 3(b), the contracted particles tend to aggregate. The morphology of these particles is identical to the morphology of R-type bacteriocins which
In our collection of PPO mutants we had isolated two other strains, ngC1 and T102, affected in PPO regulation and melanin synthesis (Table 1). Strain ngC1 was isolated by nitrosoguanidine mutagenesis, and strain T102 by
Fig. 2. Evolution of the OD600 of M. mediterranea cultures in MMC. $, Control; m, treated with 0?05 mg mitomycin C ml”1; &, treated with 0?1 mg mitomycin C ml”1. The arrow indicates the time of mitomycin C addition. 2682
Fig. 3. Electron microscopy of phage tail-like particles detected in the supernatants of M. mediterranea cultures in MMC lysed after induction with 0?1 mg mitomycin C ml”1. (a) Extended rods; (b) contracted particles (cp) and empty sheaths (es). Bar, 0?1 mm. Microbiology 149
R-bodies and lysogeny in M. mediterranea
are considered defective prophages in which the head proteins are not expressed (Shinomiya & Ina, 1989). M. mediterranea was also induced by UV radiation. Preliminary experiments indicated that it was very sensitive to this treatment and 1 mJ doses produced a survival rate of 0?5 %. Liquid cultures treated with this dose lysed, and in the supernatants of those lysates viral particles morphologically identical to those described previously in mitomycinC-induced cultures were also observed (see Fig. 3). The possible bacteriocin activity of cultures induced by mitomycin C or UV was assayed against different bacterial strains, including the M. mediterranea wild-type strain and all its mutant strains isolated in our lab. Unfortunately, there are no other M. mediterranea strains described so far, so the closest relatives that we could assay were Marinomonas vaga ATCC 27119 and Marinomonas communis ATCC 27118, the type strains of two other species in this genus. Finally, the same assay was performed using other unrelated marine bacteria such as Vibrio parahaemolyticus BB22 and Vibrio harveyi 392, as well as Escherichia coli DH5a. No inhibitory activity against any bacterial strain was observed with UVinduced cultures, or cultures in MMC medium induced with mitomycin C, in spite of the presence of viral particles. However, cultures in MN induced by mitomycin C showed inhibitory activity at the lowest dilution against all the strains assayed, including E. coli. This result indicates that it is not a bacteriocin activity, but rather a consequence of an unrelated wide-spectrum antimicrobial compound that is under study in our lab. Characterization of mutants affected in prophage replication To look for M. mediterranea mutants that could be cured of the prophage, or affected in its replication, several protocols were applied to isolate strains surviving after prophage induction (see Methods). A total of 140 strains were characterized: 35 of them were isolated as resistant to UV radiation; 60 strains were selected after mitomycin C treatment in liquid medium, and finally 45 strains that grew in plates containing mitomycin C were also selected. It was observed that all these strains, except strains MIT1 and MIT2 which were selected on MN plates containing mitomycin, were inducible by mitomycin C in liquid medium. The phenotypic characterization of MIT1 and MIT2 revealed some relationship between PPO activity and prophage replication, since they were amelanogenic, as well as affected in PPO activities (Table 1). In this sense, strains MIT1 and MIT2 were very similar to the regulatory mutants ngC1, T102 and T103. These results prompted us to study the response to mitomycin C in the different PPO mutant strains in our collection. It was observed that mutants in PPO structural genes, for example strain T101, were lysed (Fig. 4b). On the contrary, similar to MIT1 and MIT2, the regulatory mutants such as ngC1, T102 and T103 were not inducible by mitomycin C (Fig. 4a). All these five mutant strains, MIT1, MIT2, ngC1, T102 and T103, http://mic.sgmjournals.org
Fig. 4. Evolution of the OD600 of cultures in MMC of M. mediterranea mutant strains (a) T103, PPO regulatory mutant, and (b) T101, mutated in the multipotent laccase structural gene. $, Control; &, treated with 0?1 mg mitomycin C ml”1. The arrow indicates the time of mitomycin C addition.
were unable to synthesize the cellular structure, indicating the relationship between its synthesis, prophage replication and PPO activity regulation. To explore if strains MIT1, MIT2 and T103 were cured of the prophage, they were used as hosts in the double-layer agar technique against M. mediterranea wild-type lysates obtained by UV or mitomycin C induction. Although different media and incubation conditions were assayed it was not possible to detect plaques indicative of phage replication (data not shown).
DISCUSSION In this study it has been shown that M. mediterranea synthesizes a highly organized cytoplasmic structure under different culture conditions. This structure resembles morphologically the R-bodies described in some prokaryotic cells. However, it also shows some similarities to the lamellar structures described in some animal melanocytes, specialized cells inside of which melanin synthesis takes 2683
D. Herna´ndez-Romero and others
place (Hu, 1981). Since M. mediterranea is a melanogenic micro-organism that expresses several PPO activities, several experiments were undertaken to explore the possible relationship between both processes. It was observed that melanin synthesis did not take place inside this structure, and also that this structure was made under conditions where no melanin synthesis takes place, such as growth in MMM. In addition, the synthesis of the structure was not affected in some amelanogenic mutants such as ng56 and T105. All these data suggest that the synthesis of this structure is not directly related to melanogenesis. However, it was intriguing to observe that all the mutants affected in the regulation of PPO activities and melanin synthesis, ngC1, T102 and T103, were also affected in the structure synthesis. Although the mutations in ngC1 and T102 have not been fully characterized at the molecular level yet, the mutant T103 is affected in the gene encoding PpoS, a membrane histidine kinase, characteristic of a two-component regulatory system (Lucas-Elı´o et al., 2002). This observation of a connection at the regulatory level between structure synthesis and PPO activity is related to the modification of prophage replication in these mutant strains as discussed below. The structure produced by M. mediterranea is morphologically very similar to prokaryotic R-bodies and particularly to those produced by Hydrogenophaga (Pseudomonas) taeniospiralis (Lalucat et al., 1979). The main difference detected under electron microscopy is that the Marinomonas R-body shows an outer envelope that, as far as we know, has not been described in other bacteria. This structural difference may explain why it is not observed as a refractile body under light microscopy. These structures are well preserved after being released from the cells as observed in Fig. 1(b), indicative of a proteinaceous nature as for other R-bodies (Pond et al., 1989). The R-bodies described are a heterogeneous group of structures not closely related with regard to the proteins associated with them or the genetic determinants for their synthesis (Kanabrocki et al., 1986). The killer phenotype of paramecia that contains bacterial endosymbionts synthesizing R-bodies has been related to the unrolling of this structure and the breakage of the phagosomal membrane of the sensitive paramecia that ingest them (Preer et al., 1974). All these endosymbionts were placed in the genus Caedibacter. However, recent studies have shown that in spite of sharing the capacity to synthesize R-bodies, members of this single genus are very distant phylogenetically (Beier et al., 2002). Other similar structures participating in defensive roles are the extrusomes, synthesized by the symbiotic bacteria of ciliates (Petroni et al., 2000). The diversity of bacteria producing R-bodies is also revealed by the fact that they are synthesized by non-symbiotic bacteria. In this case, no physiological role has been proposed for them. In fact they have been described in bacteria that show a variety of energy metabolisms, such as hydrogen-oxidizing (Lalucat et al., 1979), heterotrophic (Espuny et al., 1991) or photosynthetic (Favinger et al., 1989) bacteria, and have been isolated from different 2684
habitats. However, the synthesis of the R-body by M. mediterranea gives the first example of the production of this kind of structure by a marine bacterium. The origin of the R-bodies in the strains studied is considered to be extrachromosomal, determined either by plasmids as in Caedibacter taeniospiralis (Kanabrocki et al., 1986; Heruth et al., 1994) or encoded by bacteriophages (Pond et al., 1989). So far there is no evidence of the existence of indigenous plasmids in M. mediterranea. However, the results presented in this study clearly indicate that M. mediterranea is a lysogenic bacterium containing at least one prophage which is inducible by either UV or mitomycin C. Although this is the first communication of lysogeny in bacteria of the genus Marinomonas, this result is in agreement with recent studies indicating that lysogeny is a common phenomenon in the marine environment. Using induction with mitomycin C or UV radiation it has been shown that up to 40 % of marine bacteria contain inducible prophages (Jiang & Paul, 1998). In another study it was shown that prophages were induced from 52 % of water samples (Cochran & Paul, 1998). Microscopic evidence suggests that the Marinomonas prophage is defective. Different viral particles were observed under all the inducing conditions (Fig. 3) and by analogy with R-type bacteriocins to which they are identical, they may correspond to the contractile phage tail of a defective prophage in which the head proteins are not expressed (Lee et al., 1999; Shinomiya & Ina, 1989). To assay the bacteriocin activity of the viral particles detected in the M. mediterranea induced cultures, different bacterial strains were used, including two strains belonging to other Marinomonas species. Although, the bacteriocin activity was not detected this result should be cautiously interpreted since bacteriocins generally possess a very narrow spectrum of activity and, perhaps, could be detected with other bacterial strains not available at this time. The R-type bacteriocins of Pseudomonas aeruginosa are more thoroughly characterized, but they are also produced by different bacterial species (Strauch et al., 2001; Thaler et al., 1995). Moreover, the existence of defective prophages in R-body-producing bacteria has also been described in other micro-organisms (Lalucat et al., 1979; Espuny et al., 1991), indicates a relationship between these two processes. The results obtained in this work, showing that the strains defective in phage replication are also affected in R-body synthesis supports this relationship. In this regard, it is tempting to speculate that in those cases R-bodies could be the result of the polymerization of viral proteins in the cytoplasm of the host. On the other hand, the defective nature of the prophage may be related to difficulties in the isolation of bacterial strains that have been cured of it. In spite of numerous efforts, only two strains, MIT1 and MIT2, not inducible by mitomycin C were isolated. The mutations in those strains are not characterized. However, both strains share Microbiology 149
R-bodies and lysogeny in M. mediterranea
some characteristics with strain T103 that was previously isolated and characterized by our group (Lucas-Elı´o et al., 2002), and with two other strains obtained by different protocols such as nitrosoguanidine mutagenesis (strain ngC1) and transposon mutagenesis (strain T102). All five of these strains are amelanogenic, affected in the levels of PPO activity and not inducible by mitomycin C. These results clearly indicate that there are some common mechanisms co-regulating PPO activities and prophage induction that are affected in those strains. Strain T103 is mutated by transposon insertion in ppoS, a gene encoding a membrane histidine kinase, revealing that PpoS is regulating prophage replication in addition to the previously described PPO activities and melanin synthesis (Lucas-Elı´o et al., 2002). PpoS is a tripartite sensor kinase, a kind of sensor histidine kinase characterized by the regulation of complex traits involving different cellular processes (Appleby et al., 1996). The coincidence of the regulation of prophage induction and PPO activities by PpoS as observed in this study is intriguing and there are some possibilities that would deserve further study. For example, it cannot be ruled out that the induction of prophage replication could be a stress factor for M. mediterranea that in turn may induce PPO activities. However, the sequencing of the genomes of different micro-organisms has revealed many genes of viral origin, and in some cases those genes play a physiologically relevant role for the host. In this context prophage replication would induce the death of some Marinomonas cells, facilitating the release of PPO enzymes and the synthesis of extracellular melanins. These melanins may protect the surviving cells from UV radiation or act as a sink for extracellular oxidative radicals. Alternatively, it could be a mechanism facilitating the polymerization of extracellular compounds, making them unavailable for competing micro-organisms. In this sense melanin synthesis and PPO activities could have some physiological relevance at the population level.
ACKNOWLEDGEMENTS This work was supported by the grant PB2001-0140 from the CICYT, Spain. P. Lucas-Elı´o and D. Lo´pez-Serrano were recipients of pre-doctoral fellowships from, respectively, the Se´neca Foundation (Comunidad Auto´noma de la Regio´n de Murcia) and the University of Murcia. We are grateful to Professor J. Lalucat (IMEDEA, CSIC-UIB) for helpful suggestions. We also thank the Electronic Microscopy Service of the University of Murcia.
Cochran, P. K. & Paul, J. H. (1998). Seasonal abundance of lysogenic
bacteria in a subtropical estuary. Appl Environ Microbiol 64, 2308–2312. Espuny, M. J., Andre´s, C., Mercade, M. E., Robert, M., Manresa, M. A. & Guinea, J. (1991). R-bodies in Pseudomonas aeruginosa strain
44T1. Antonie van Leeuwenhoek 60, 83–86. Favinger, J., Stadtwald, R. & Gest, H. (1989). Rhodospirillum
centenum, sp. nov. a thermotolerant cyst-forming anoxygenic photosynthetic bacterium. Antonie van Leeuwenhoek 55, 291–296. Fuste´, M. C., Simon-Pujol, M. D., Marques, A. M., Guinea, J. & Congregrado, F. (1986). Isolation of a new free-living bacterium
containing R-bodies. J Gen Microbiol 132, 2801–2805. Hearing, V. & Tsukamoto, K. (1991). Enzymatic control of
pigmentation in mammals. FASEB J 5, 2902–2909. Heruth, D. P., Pond, F. R., Dilts, J. A. & Quackenbush, R. L. (1994).
Characterization of genetic determinants for R body synthesis and assembly in Caedibacter taeniospiralis 47 and 116. J Bacteriol 176, 3559–3567. Hu, F. (1981). Melanocyte cytology in normal skin, melanocyte nevi and malignant melanomas. In Pathology of Malignant Melanoma, pp. 1–21. Edited by A. B. Ackerman. New York: Masson Publishing. Jiang, S. C. & Paul, J. H. (1998). Significance of lysogeny in the
marine environment: studies with isolates and a model of lysogenic phage production. Microb Ecol 35, 235–243. M. (1975). Bacteriocins and bacteriophages in Pseudomonas aeruginosa. In Microbial Drug Resistance, pp. 291– 305. Edited by S. Mitsuhashi & H. Hashimoto. Tokyo: University of Tokyo Press.
Kageyama,
Kanabrocki, J. A., Quackenbush, R. L. & Pond, F. R. (1986).
Organization and expression of genetic determinants for synthesis and assembly of type 51 R bodies. J Bacteriol 168, 40–48. Lalucat, J., Meyer, O., Mayer, F., Pares, R. & Schlegel, H. G. (1979).
R-bodies in newly isolated free-living hydrogen-oxidizing bacteria. Arch Microbiol 121, 9–15. Lee, F. K. N., Dudas, K. C., Hanson, J. A., Nelson, M. B., LoVerde, P. T. & Apicella, M. A. (1999). The R-type pyocin of Pseudomonas
aeruginosa C is a bacteriophage tail-like particle that contains singlestranded DNA. Infect Immun 67, 717–725. Lo´pez-Serrano, D., Sanchez-Amat, A. & Solano, F. (2002). Cloning
and molecular characterization of a SDS-activated tyrosinase from Marinomonas mediterranea. Pigment Cell Res 15, 104–111. Lucas-Elı´o, P., Solano, F. & Sanchez-Amat, A. (2002). Regulation of
polyphenol oxidase activities and melanin synthesis in Marinomonas mediterranea: identification of ppoS, a gene encoding a sensor histidine kinase. Microbiology 148, 2457–2466. Petroni, G., Spring, S., Schleifer, K. H., Verni, F. & Rosati, G. (2000).
Defensive extrusive ectosymbionts of Euplotidium (Ciliophora) that contain microtubule-like structures are bacteria related to Verrucomicrobia. Proc Natl Acad Sci U S A 97, 1813–1817. Pond, F. R., Gibson, I., Lalucat, J. & Quackenbush, R. L. (1989).
R-body-producing bacteria. Microb Rev 53, 25–67.
REFERENCES
Preer, J. R., Preer, L. B. & Jurand, A. (1974). Kappa and other
endosymbionts in Paramecium aurelia. Bacteriol Rev 38, 113–163.
Appleby, J. L., Parkinson, J. S. & Bourret, R. B. (1996). Signal
Riley, P. A. (1992). Materia melanica: further dark thoughts. Pigment
transduction via the multi-step phosphorelay: not necessarily a road less traveled. Cell 86, 845–848.
Sanchez-Amat, A. & Solano, F. (1997). A pluripotent polyphenol
Beier, C. L., Horn, M., Michel, R., Schweikert, M., Go¨rtz, H.-D. & Wagner, M. (2002). The genus Caedibacter comprises endosymbionts
of Paramecium spp. related to the Rickettsiales (Alphaproteobacteria) and to Francisella tularensis (Gammaproteobacteria). Appl Environ Microbiol 68, 6043–6050. http://mic.sgmjournals.org
Cell Res 5, 101–106. oxidase from the melanogenic marine Alteromonas sp. shares catalytic capabilities of tyrosinases and laccases. Biochem Biophys Res Commun 240, 787–792. Sanchez-Amat, A., Lucas-Elio, P., Ferna´ndez, E., Garcia-Borro´n, J. C. & Solano, F. (2001). Molecular cloning and functional
2685
D. Herna´ndez-Romero and others characterization of a unique multipotent polyphenol oxidase from Marinomonas mediterranea. Biochim Biophys Acta 1547, 104–116.
isolation of a multipotent polyphenol oxidase mutant. J Bacteriol 182, 3754–3760.
Shinomiya, T. & Ina, S. (1989). Genetic comparison of bacteriophage
Strauch, E., Kaspar, H., Schaudinn, C., Dersch, P., Madela, K., Gewinner, C., Hertwig, S., Wecke, J. & Appel, B. (2001).
PS17 and Pseudomonas aeruginosa R-type pyocin. J Bacteriol 171, 2287–2292. Solano, F. & Sanchez-Amat, A. (1999). Studies on the phylogenetic
relationships of melanogenic marine bacteria: proposal of Marinomonas mediterranea sp. nov. Int J Syst Bacteriol 49, 1241–1246. Solano, F., Garcı´a, E., Pe´rez de Egea, E. & Sanchez-Amat, A. (1997). Isolation and characterization of strain MMB-1 (CECT
Characterization of enterocoliticin, a phage tail-like bacteriocin, and its effect on pathogenic Yersinia enterocolitica strains. Appl Environ Microbiol 67, 5634–5642. Thaler, J. O., Baghdiguian, S. & Boemare, N. (1995). Purification
and characterization of xenorhabdicin, a phage tail-like bacteriocin, from the lisogenic strain F1 of Xenorhabdus nematophilus. Appl Environ Microbiol 61, 2049–2052.
4803), a novel melanogenic marine bacterium. Appl Environ Microbiol 63, 3499–3506.
Wells, B. & Horne, R. W. (1983). The ultrastructure of Pseudomonas
Solano, F., Lucas-Elı´o, P., Ferna´ndez, E. & Sanchez-Amat, A. (2000). Marinomonas mediterranea MMB-1 transposon mutagenesis:
avenae. II. Intracellular refractile (R-body) structure. Micron Microsc Acta 14, 329–344.
2686
Microbiology 149
