The ecology of the myxobacteria

What are myxobacteria?

Myxobacteria are Gram-negative eubacteria with relatively large (0.6–1.2 by 3–15 μm) rod-shaped cells. They live in soil and related habitats and are famous for three capabilities: (i) they move by gliding and their colonies are therefore thin, film-like swarms that spread gradually over the culture plate; (ii) they have sophisticated intercellular communication systems and a highly developed social life; and (iii) they show a remarkable morphogenetic potential, which is expressed on two levels. In a co-operative morphogenesis involving 10⁵–10⁶ cells, induced by starvation conditions, they may produce a fruiting body. Within the maturing fruiting body, a cellular morphogenesis takes place, during which the vegetative cells convert into desiccation-resistant myxospores. It might also be mentioned that myxobacteria have the largest genomes of all bacteria (9500–10 000 kbp) and a DNA with a G + C content of 66–72 mol%. Details can be found in several reviews (Reichenbach and Dworkin, 1992; Dworkin and Kaiser, 1993; Dworkin, 1996).

Ecological facts seem to explain the special traits of the myxobacteria

These peculiarities of the myxobacteria clearly must have a biological, and most likely an ecological, meaning. It appears that most of the phenomena mentioned above are best explained by the nutritional requirements and the habitats preferentially colonized by myxobacteria. All myxobacteria specialize in the degradation of biomacromolecules. Most species are proteolytic and bacteriolytic, attacking dead organic matter and dead and living bacteria, yeasts and other microorganisms with a host of excreted enzymes. One group of myxobacteria decomposes cellulose. Gliding motility may be useful to find and approach these non-diffusing substrates, particularly in a relatively dry and complex environment such as soil. The enzymatic attack, which involves the investment of substantial energy into the production of exoenyzmes, is more efficient if a community is at work, because then higher enzyme levels can be reached, and losses by diffusion can be minimized. A cell may profit from the efforts made by another cell and vice versa. The following experiment appears to support this assumption. When a medium containing 0.2% casein as the only nutrient is inoculated with Myxococcus xanthus cells, growth only occurs above a certain cell density (around 10³ cells ml⁻¹). Below that level, the cells cannot produce sufficient enzyme activity for casein hydrolysis (Rosenberg et al., 1977). Revealingly, the culture would grow at the same low cell density on a medium with hydrolysed casein (Casitone). Fruiting body formation, too, may be explained by the need to live in a community. While myxospores are already resistant to harmful environmental factors, the fruiting body guarantees that a new life cycle is started with a population rather than a single cell. Further, in the swarm colony, chemical and physical mechanisms (pheromones, pili, slime tracks) keep the cells together, although they can move much faster than the swarm expands. On the other hand, it is not trivial to produce single-cell colonies with myxobacteria. Individual cells normally do not grow readily into colonies even on agar media that allow luxuriant growth when inoculated massively. In some cases, results improve if spent medium of a liquid culture is added to the plating agar. This could mean that some quorum-sensing mechanism (Gray, 1997) is involved that prevents the futile growth of individual cells. Myxobacteria produce many different growth inhibitors. In fact, they have been shown to be one of the few bacterial groups that are a rich source of secondary metabolites (Reichenbach and Höfle, 1998). Conceivably, the organisms use that strategy to kill prey microorganisms and to defend their niche in the habitat, because their coherent colonies are essentially stationary, in a similar manner to the mycelia of the Actinomycetales.

Myxobacteria are very common and ubiquitous organisms

With their specific lifestyle, myxobacteria are obviously very successful. They are found nearly everywhere, from Antarctica to the tropics, and from sea level to high altitudes, in all vegetation belts from tropical rainforests to deserts. They may not live in all these places very conveniently and may simply have been blown to some of the more austere environments and survive there for some time, judging from experimental data for at least 10–20 years, owing to their desiccation and temperature-resistant myxospores. The richest myxobacterial floras are found in warm, semiarid areas, e.g. in the south-west of the USA, in northern India or in Egypt. A single spoonful of soil from such areas may yield 5–10 different species. Their myxospores and undegraded biomacromolecules in the habitat may give the myxobacteria a selection advantage in these areas. But they are also very common in central Europe, the mid-west of the USA and similar places. In more restrictive environments, such as the Sahara, in northern forests and at high altitudes, only few species are normally found, mostly Myxococcus, Corallococcus, Archangium and Nannocystis strains.

Myxobacteria are essentially soil bacteria

The typical habitat of myxobacteria is soil, from dune and desert sands to rich black chernozem. All known myxobacteria are strictly aerobic and live in the topmost layers of the soil. In addition to soil, myxobacteria colonize decaying plant material including rotting wood and bark from dead and living trees, decomposing lichens and insects, and dung, especially of certain herbivorous mammals, such as wild rabbits, hares, deer, sheep and goats. Rabbit dung is a classical source for the isolation of myxobacteria. When incubated in a moist chamber at room temperature or at 30°C for a couple of days, fruiting bodies will almost always appear on the surface of some dung pellets and can be used for isolation (technical details will be found in Reichenbach and Dworkin, 1992). Where are the myxobacteria on dung coming from? As aerobic organisms, myxobacteria are not likely to live within animals. But myxobacteria on dung collected on thick layers of snow suggest that they may pass through the digestive track of the animal (Rückert, 1975a; personal observation). This has indeed been shown by pipette feeding fruiting bodies of Chondromyces to white mice in the laboratory and demonstrating the organism, certainly absent in that environment, on the faeces of the animals (Kühlwein, 1950). Myxobacteria have also been isolated regularly from the gut content of carp, and it has been concluded that they pass through rather than inhabit the fish (Till, 1982).

Myxobacteria are also found in water bodies

Myxobacteria can also be isolated from fresh water, which is perhaps not so astonishing, because soil organisms notoriously exchange into water bodies, being regularly washed or blown in and often surviving there periodically or permanently (Brauss et al., 1967; Gräf, 1975; Trzilováet al., 1981). They may be found on drippling bodies in sewage plants. One species, Polyangium parasiticum, has been described as a parasite of the green alga, Cladophora, in the environments of Vienna (Geitler, 1925). It made holes in the cell wall of the alga, entered the cells and lysed its contents. Fruiting bodies were produced inside and on the surface of the algal filaments. Pl. parasiticum may have been a facultative parasite, but was never reported again.

In this connection, the reader must be warned that, for some time, Cytophaga and Cytophaga-like bacteria were classified as myxobacteria, among them many water bacteria including fish pathogens. We know today that those organisms are not related to the myxobacteria at all, but belong to the Cytophaga–Flavobacterium–Bacteroides phylum. The myxobacteria are found in the delta branch of the Proteobacteria.

Are there marine myxobacteria?

Considering the life cycle of myxobacteria with their desiccation-resistant myxospores, aquatic habitats appear not to be very typical for them. Yet myxobacteria have been isolated from marine sources. They have been found repeatedly in sand and debris taken from the intertidal range of the Atlantic Ocean (Brockman, 1967; Rückert, 1975b), and once even from samples collected at 50 m depths around the island of Helgoland (Rückert, 1984). But all those strains were isolated on low-salt media, and none of them tolerated the salt concentrations of seawater. My own experience is equally negative. Experiments to isolate myxobacteria on seawater agar from samples collected in the salt marshes on the island of Sylt in the North Sea were a complete failure. When isolating marine Cytophaga-like bacteria at the marine station of Roscoff in Brittany, we started hundreds of cultures on seawater agar with all kinds of samples from near the station, using techniques that should have yielded myxobacteria as well. But there were none. I obtained many myxobacteria from marine debris collected on the shores of the island of Fynen in Denmark, when low-salt media were used for isolation, just as described in the articles cited above. On the other hand, Japanese colleagues reported recently at a symposium on myxobacteria in Greece that they had isolated truly marine, halotolerant myxobacteria (R. Fudou et al., personal communication). Using molecular taxonomy, they found that those organisms belonged to the Nannocystis complex. Does this mean that there are myxobacteria in the Pacific and not in the Atlantic? This may sound weird, but many algae and seaweeds also show a restricted distribution.

Myxobacteria sometimes settle in special habitats

Dust and soil are transported by air currents and so are desiccation-resistant soil organisms. Air contamination by myxobacteria has been reported occasionally, if rarely (Wu et al., 1968). About half the leaves taken from various shrubs and trees in Germany yielded myxobacteria (Rückert, 1981). However, not all tree species seem to harbour myxobacteria.

Myxobacteria have also been found in samples taken deep within caves in areas apparently not touched by man before (Menne and Rückert, 1988). Only a few species have been discovered, as is typical of unfavourable habitats.

Somewhat unexpectedly, myxobacteria have also been isolated from bogs. Many myxobacteria were obtained from various parts of an alkaline bog (pH 6–8.7) near Vestaburg, MI, USA (Hook, 1977). Of particular interest was the abundant occurrence of Angiococcus (Cystobacter) disciformis specifically in the fossa, i.e. the ditch surrounding the bog (Hook et al., 1980). While the organism is also found in other places, it may be particularly well adapted to that habitat. Really astonishing was the discovery of myxobacteria (Myxococcus, Corallococcus and Polyangium sp.) in peat bogs in the Rhön mountains in Germany (pH 3.2–4.8) and of the Hautes Fagnes/Hohe Venn at the Belgian–German border (pH 3.7, water content 91%, oxygen 2.58 mg l⁻¹; Dawid, 1984). The samples were taken from the bog at a depth of 5–10 cm below the plant cover of about 10 cm. The myxobacteria were isolated at a neutral pH. It is not clear whether the organisms really grow, perhaps very slowly, in the bog or are washed down from the plant cover or from soil blown into the bog. The rather narrow spectrum of species may speak for the former assumption.

Very little is known about specific habitats of the various myxobacterial species. A reliable and convenient source of Myxococcus species is the dung of herbivorous mammals. Mx. fulvus is often found on rotting wood, and all species can also be isolated from soil. While it is plausible that harsh environments, such as high altitudes, deserts or peat bogs, select for tough species, mainly those with spherical myxospores (e.g. Norén, 1952), the distribution patterns of species are still enigmatic.

Most myxobacteria are strictly aerobic mesophiles with a narrow pH range, a low salt tolerance and, in the vegetative state, are highly sensitive to desiccation

Almost all myxobacteria grow well at 30°C, although their temperature range is much wider. Many strains are psychrotrophic and grow, albeit slowly, at low temperatures, e.g. in the refrigerator at 4–8°C (Zhukova, 1963; Menne and Rückert, 1988; personal observation). Also, isolation at 10°C brought forth many different myxobacteria (Krzemieniewska and Krzemieniewski, 1927). Myxobacteria have been isolated from soil collected along the Arctic coast of North America (Brockman and Boyd, 1963). Only enrichment cultures incubated for 6 weeks at 25–26°C, but not those incubated at 6–8°C, yielded myxobacteria. But the incubation time for isolating psychrophiles may simply have been too short. Myxobacteria were also obtained from Antarctic soils. In one study, the crude cultures were kept at 30°C and at room temperature and, understandably, only mesophilic myxobacteria were obtained (Rückert, 1985). In a second study, psychrophilic myxobacteria could be isolated after incubation times of between 8 weeks and 9 months at 4°C (Dawid et al., 1988). These organisms would not grow at 18°C. Survival of fruiting bodies for a long time in dry environments at low temperatures can be expected from the behaviour of myxobacteria in the laboratory. Vegetative cells also survive freezing but remain living for a longer time only at temperatures below −50 to −60°C.

In the laboratory, the useful temperature range for cultivating myxobacteria is 28–34°C, with generation times between 4 and 14 h. In nature, the organisms certainly grow normally at lower temperatures and probably much more slowly. Many strains still grow at 38–40°C (e.g. McCurdy, 1969; Gerth et al., 1994; unpublished data), but cultures get increasingly unstable at these high temperatures and quickly break down and lyse at the end of the growth cycle. Vegetative cells are killed at temperatures above 45°C, but myxospores suspended in water tolerate 58–60°C, which can be used to purify strains during isolation (for technical details, see Reichenbach and Dworkin, 1992). Dry myxospores may survive much higher temperatures (G. Rückert, personal communication). We obtained growing cultures from fruiting bodies of various genera that had been heated on filter paper in the oven at 140°C for 30 min, in some cases even for 45 min (unpublished data).

In culture, the pH range of myxobacteria is normally 6.8–7.8. The cellulose degraders may still grow at a somewhat lower pH but, even for them, the limit is 6.0–6.4. This may reflect in part the practice of using media with a pH of around 7 for isolation. Alkalophilic myxobacteria that do not grow below a pH of 9 have indeed been obtained from alkaline lakes of east Africa (sample pH 9.5), when alkaline media were used for isolation (W. Dawid, personal communication). So far, no acidophilic myxobacteria have been reported, although myxobacteria can be isolated from soils with an acid pH as low as 2.5 (e.g. Krzemieniewska and Krzemieniewski, 1927; Rückert, 1972; 1979; Dawid, 1979).

The salt tolerance of myxobacteria appears to be rather low. As already mentioned, myxobacteria may be isolated from marine habitats but, until recently, all strains thus obtained were not halotolerant. Soils from salt marshes or certain desert habitats with a high salt content have to be desalted before myxobacteria can be isolated from them (Rückert, 1983). Apparently some myxobacteria tolerate salt better than others. The addition of 0.5% NaCl to the isolation agar stimulates the development of Mx. virescens and Ar. gephyra (Rückert, 1978). On water agar with streaks of living Escherichia coli, a standard method for the isolation of myxobacteria, I obtained only Mx. virescens and Na. exedens when 1% NaCl was added to the medium. A reduction in the NaCl content to 0.8% also allowed Mx. fulvus and Corallococcus spp. to develop (unpublished results).

For an active life, myxobacteria of course need water. Vegetative cells are rather sensitive to desiccation, so that lyophilization is not practical for their preservation. In contrast, myxospores are completely desiccation resistant and may be kept dry at room temperature, e.g. in fruiting bodies on filter paper, for 10–25 years, the longest period we have been able to test so far. It appears that the deposition of substantial amounts of trehalose in myxospores of Mx. xanthus is responsible for desiccation resistance (McBride and Zusman, 1989). In contrast to endospores, myxospores only retain their resistance to harmful environmental factors in the dry state.

Myxobacteria may severely restrict other microorganisms in their common habitat

It is not easy to get a clear idea about the numbers of myxobacteria in soil. First, the organisms are attached to particles by excreted slime and, in addition, they are probably connected among themselves by pili, so that a suspension of individual cells for plating cannot be obtained. And, secondly, most myxobacteria would not readily produce colonies from single cells. Yet some estimates have been made. In southern England, between 2000 and 75 000 bacteriolytic myxobacteria have been counted per g of soil. The numbers increased dramatically in compost, where 500 000 cells of Mx. fulvus g⁻¹, as well as small numbers of other myxobacteria, have been found (Singh, 1947). A study of various types of soil in Yugoslavia showed 1500–80 000 bacteriolytic myxobacteria per g. The densest populations were seen in chernozem and alluvial soils with a high calcium carbonate content; the lowest in podzol and solonetz, which may be in response to pH and organic load (Sabados-Saric, 1957). In south-eastern Canada, the number of Myxococcus–Corallococcus was determined to be between 1000 and 450 000 g⁻¹ soil (McCurdy, 1969). In cultivated marsh soils on the Kola Peninsula, the cellulose-degrading So. cellulosum was present with up to 10 × 10⁶ cells g⁻¹ (Zhukova, 1959). All these estimates are probably too low and, in any case, refer only to a selection of species. We may conclude, nevertheless, that myxobacteria are present in soil in substantial numbers. Normally, several different species can be isolated from a soil sample the size of a pea, which suggests that many species are fairly common and contribute to the total myxobacterial population.

It has been known for many years that most myxobacteria are ‘micropredators’ (Singh, 1947) and attack and lyse other bacteria, whether living or dead (e.g. Pinoy, 1921; Beebe, 1941; Oetker, 1953; Kühlwein, 1955). From this, it has been concluded that myxobacteria should have a marked effect on the soil microflora and could perhaps affect soil fertility negatively. In fact, when Azotobacter and Mx. xanthus were inoculated concomitantly into soil, the number of Azotobacter decreased very substantially during incubation (Callao et al., 1966). After adding Micrococcus luteus to natural soil, a sequence of three predatory bacteria was observed. The last to appear was a myxobacterium with spherical myxospores, perhaps a Myxococcus species (Liu and Casida, 1983). Interestingly, the myxospores survived in acidic soil for a considerable time but did not germinate until the pH was adjusted to neutral and Micrococcus was added.

The lysis of cyanobacteria has been studied repeatedly in the hope that myxobacteria could be used to control water blooms (Burnham et al., 1981; 1984; Daft et al., 1985). Lysis occurred on agar plates as well as in liquid cultures. In liquid media, the Myxococcus strains used formed floating spherules and entrapped the cyanobacteria in layers of excreted slime. Predator–prey cycles with a periodicity of 9 days at 25°C were observed. As few as 50 Myxococcus cells 100 ml⁻¹ with 10⁷Phormidium luridum cells ml⁻¹ were sufficient to start a lytic myxobacterial population. In chemostat systems, the myxobacteria grew either on glass beads or on the wall of the container, and a continuous lysis of the cyanobacteria was demonstrated. Lysozyme activity was seen in the culture liquid.

Also, lysis of various yeasts has been observed many years ago (e.g. Oetker, 1953), and an excellent medium for cultivating practically all myxobacteria is yeast agar (VY/2 agar) prepared with 0.5% bakers' yeast. Lysis of Cladophora green algae in a natural habitat by Polyangium parasiticum has already been mentioned. An incident very similar to the Cladophora case has been reported for fungal hyphae (Rhizoctonia solani) and conidia (Cochliobolus miyabeanus). After the objects had been buried in soil for 4 weeks, they showed many small perforations in their walls, and a myxobacterium, apparently a Polyangium, could be isolated from them. This organism produced the same holes in the fungal cell walls in culture. Through them, groups of myxobacteria would enter the cells, lyse their contents and either leave again or produce fruiting bodies inside (Homma, 1984).

It should be mentioned in this connection that, thus far, no pathogenic myxobacteria have been observed. There are fish pathogens, doing considerable damage especially in aquaculture, that were classified for some time in the past as myxobacteria under the name of Chondrococcus columnaris. In fact, these organisms are Cytophaga-like bacteria.

A true symbiosis has been demonstrated recently for certain myxobacteria

It has been suggested repeatedly in the past that certain myxobacteria, particularly those with more complex fruiting bodies, may live in symbiosis with other bacteria. Definitive proof of this has been lacking until recently, because cultivation of myxobacteria was not well mastered until the 1960s.

When trying to purify Cm. crocatus isolates, it turned out that, with one single exception, all strains (so far nine) would only grow as long as a small, delicate, rod-shaped bacterium was present in the culture. Subcultures of Cm. crocatus without the companion soon withered away. All other Chondromyces species, at least the several dozen strains in our collection, can be cultivated without problems as pure strains. The Cm. crocatus companion, too, does not grow well without its Chondromyces partner. Very small colonies develop on various media after weeks of incubation. Good growth was only obtained when the companion was cultivated in a membrane reactor with the mixed culture in the other compartment. The Cm. crocatus strains can readily be grown together with the companion in large bioreactors, as the two partners were regulate their development mutually (Kunze et al., 1994; 1995). When the 16S rRNA genes of the two partners were sequenced, it was discovered that the companion was always the same species, regardless of whether the Cm. crocatus strains came from samples from Madeira, Malaysia or Brazil (Jacobi et al., 1996; 1997). The symbiont turned out to be a novel species, probably belonging to the genus Sphingobacterium. Interestingly, one Cm. crocatus strain showed a slight deviation in its 16S rRNA base sequence. Its companion also showed a slight deviation from the base sequences of the other companion strains. This suggests that the Chondromyces–Sphingobacterium pairs have lived together for a long time. How does Cm. crocatus find its partner when germinating in a new, distant place? With the aid of fluorescent probes, it was shown that the companion is already present within the sporangioles of the Chondromyces fruiting body (Jacobi et al., 1997). This is very unusual, as myxobacterial fruiting bodies are not normally contaminated when freshly produced.

A symbiosis between Cm. crocatus and a specific bacterium was described 80 years ago (Pinoy, 1913; 1921). From the description given of that ‘companion’, it must be concluded, however, that the author has not seen the real symbiont. He mentions that the companion is Gram positive, probably related to Micrococcus luteus, grows well on various media and is lysed by the developing Chondromyces, all of which does not fit the symbiont found in our cultures.

Myxobacteria excrete a host of chemicals for various purposes

Myxobacteria communicate with one another. The mechanisms are mechanical (e.g. by pili) as well as chemical. In Sg. aurantiaca, a pheromone has been demonstrated during fruiting body formation (Stephens et al., 1982), and its structure has been elucidated recently (Hull et al., 1998; Plaga et al., 1998; Morikawa et al., 1998). Myxobacterial cultures often produce an earthy smell, which, in Na. exedens, is caused by geosmin (Trowitzsch et al., 1981). As myxobacteria are so common in soil, they may contribute to its odour, which sometimes becomes a problem in drinking water supplies. Cultures of Cm. crocatus have an odour reminiscent of pyridine, which is so unique and typical that the organism can be recognized by it at once. As already mentioned, myxobacteria also produce and usually excrete many bioactive secondary metabolites (Reichenbach and Höfle, 1998). These compounds often show otherwise very rare mechanisms of action. Thus, 18 different new electron transport inhibitors acting on complex I or complex III have been discovered, eight compounds interfering with the actin or with the tubulin cytoskeleton, usually with extremely high efficiencies (disorazol at 1 pg ml⁻¹), and four compounds acting on nucleic acid polymerases. Most of these compounds appear to be developed to inhibit eukaryotic competitors in the habitat: probably fungi and protozoa.

Gliding motility is found in many, unrelated bacterial groups

In the past, gliding motility has often been taken as a proof of a phylogenetic relationship between organisms. But this is not the case. Gliding representatives are known from most bacterial phyla (e.g. Stackebrandt, 1986). In many cases, a biological meaning for the organism can be seen in gliding movements: organisms living in biofilms and shifting gradients, such as Beggiatoa (H₂S/O₂), Chloroflexus and cyanobacteria (light); organisms specializing in the degradation of non-diffusing substrates, such as Cytophaga and Cytophaga-like bacteria, Lysobacter and the myxobacteria. In spite of many efforts over 100 years of research, the mechanism of gliding is still not yet understood. Recently, a peculiar apparatus consisting of tiny rings connected by short longitudinal elements has been demonstrated in the surface of myxobacteria (Lünsdorf and Reichenbach, 1989; Freese et al., 1997). This apparatus appears to perform conformational changes that could create a contraction wave in the cell surface, and could thus represent the machinery of gliding motility.