How do cyanobacteria glide? David G. Adams The mechanisms by which bacteria move are not all fully understood. David Adams considers the motility of cyanobacteria. Bacterial gliding is a mystery. It constitutes Surface waves (a) Gone of the two means of locomotion in bacteria, From the early work of the other being swimming which is a well Jarosch, Castenholz and characterized system employing flagella driven by rotary Halfen arose a theory to motors in the cell wall. By contrast, the mechanism of explain the mechanism of gliding remains elusive, even though the process was gliding in Oscillatoria and first described well over a century ago. Although widely related cyanobacteria. differing groups of bacteria are capable of gliding, there They proposed that the are common features. Movement occurs in a direction driving force for gliding parallel to the cell’s long axis, is associated with the was the distortion of production of polysaccharide slime and requires proteinaceous fibrils in the attachment of the cell to a surface. No obvious structures cell wall. Such distortions are associated with gliding, yet the fact that all gliding might be propagated bacteria are Gram-negative implies that the outer rhythmically as waves membrane is important to the process. Nevertheless, it is moving from one end of the likely that there is more than one mechanism for gliding, filament to the other. (b) even amongst the cyanobacteria. These waves would be transmitted through the Cyanobacterial motility outer membrane and, Of all gliding bacteria the cyanobacteria are the most by interaction with the widespread and of immense environmental importance. substratum, cause the They show several forms of motility. Some unicellular filament to move in members of the genus Synechococcus can swim without the the opposite direction means of flagella, whereas some Synechocystis strains move (Fig. 2). Reversals of slowly on surfaces by a form of gliding, sometimes the direction of movement referred to as twitching, which requires type IV pili. could result from reversals Some filamentous non-motile forms, such as Nostoc spp., in the direction of wave produce specialized gliding filaments known as propagation. This theory hormogonia, which constitute a brief, dispersive stage of was supported by electron their life cycles. By contrast, many other filamentous micrographs showing a layer of fibrils in the cell wall, ABOVE: cyanobacteria, such as Oscillatoria spp. (Fig. 1a), possess between the outer membrane and the peptidoglycan, Fig. 1. Photomicrographs of two motile cyanobacteria, Oscillatoria permanent gliding motility. Members of the family arranged helically around the filament at an angle of sp. (a) and Spirulina sp. (b). Bars, Oscillatoriaceae can travel at up to 10 µm s–1 and the approximately 30° to the filament’s long axis. This 20 µm. filaments rotate as they glide, the direction of rotation helical arrangement of the fibrils might provide an PHOTOS PAULA DUGGAN being characteristic of the species. explanation for the rotation of filaments of cyanobacteria There have been many attempts to explain the such as Oscillatoria spp. as they glide. LEFT: mechanism of force generation in cyanobacterial gliding. Recent evidence has supported the idea of an array of Fig. 2. Schematic diagram to explain how fibrils in the cell wall The two most likely hypotheses, surface waves and slime helically arranged fibrils between the outer membrane might provide the driving force for extrusion, are considered here. and the peptidoglycan of these cyanobacteria (Fig. 3), gliding. Distortions of the fibrils although the fibril diameter is larger (25–30 nm) than between the outer membrane (OM) previously described (8–10 nm). The fibrils seen in situ and peptidoglycan (not shown) appear to be continuous (Figs 4 and 5), but they fragment might be propagated rhythmically, as waves moving from one end of when released from the cell wall and can be isolated and the filament to the other. These purified (Fig. 6). The diameter of the fibril fragments waves would be transmitted from a wide range of motile strains is very similar, but the through the outer membrane (small length varies. The composition of the fibrils has not been arrows) and, by interaction with the substratum, cause the filament to conclusively determined, but preliminary evidence move in the opposite direction suggests a glycoprotein, which may explain their (large arrow heads). extreme resistance to solubilization with all the commonly used reagents. The work of Hoiczyk and Baumeister has revealed a OM second array of fibrils in all rotating species. This layer consists of helically arranged fibrils, 8–12 nm in diameter, sitting on top of an S-layer anchored to the outer membrane (Fig. 7). This outermost fibrillar layer Fibril consists of a glycoprotein known as oscillin, which MICROBIOLOGY TODAY VOL 28/AUG 01 131 (a) (a) S NEAR RIGHT (TOP): (b) Fig. 3. Transmission electron OM micrographs of transverse thin sections of a motile Oscillatoria sp. PG The ‘corrugations’ in the cell wall (b) are caused by the presence of the fibrillar layer (F) between the peptidoglycan (PG) and the outer OM membrane (OM). The fibrils cover F CM the entire filament surface, part of which is shown in (a) and at a F higher magnification in (b). Bars, 50 nm. PHOTOS MATTHEW BEAN BELOW: Fig. 4. Transmission electron PG complex organelle-like structures (junctional pore micrograph of a negatively stained complexes; JPCs) that span the peptidoglycan and outer sample of a motile Oscillatoria sp. membrane (Fig. 7b). Propulsion of the filament would An actively motile sample was result from the adherence of the slime to both the crushed between glass slides and negatively stained. The presence of contains multiple repeats of a calcium-binding motif. filament surface and the substratum, combined with its the helically arranged fibrillar layer Oscillin shares homology with another glycoprotein, extrusion from a row of JPCs on one side of each septum. covering the whole cell results in a SwmA, which also possesses calcium-binding motifs and Switching slime extrusion to the JPCs on the other side criss-cross appearance (see Fig. 5 is required for swimming in Synechococcus. SwmA is not of the septum would result in a reversal in the direction of for an explanation). Bar, 500 nm. required for translocation but is thought to be involved gliding. Rotation of the filament would result from the PHOTO DENISE ASHWORTH. REPRINTED WITH PERMISSION FROM in the generation of thrust. Oscillin is likely to play a helically arranged oscillin fibrils and the direction of J BACTERIOL 181, 884–892 (1999) passive role in gliding, perhaps serving as a screw thread rotation would be determined by the orientation of the to guide rotation of the filament as it moves. fibrils. FAR RIGHT (TOP): There is a possible precedent for the powering of Fig. 5. Schematic diagrams Slime extrusion motility by slime extrusion. In the bacterium Acetobacter showing the fibrillar array in the cell wall of Oscillatoria spp. (a) Secretion of mucilage is commonly associated with xylinum cellulose extrusion from pores is thought to drive Part of filament showing cell septa gliding in filamentous cyanobacteria and provides the motility, but 200 times more slowly (0.05 µm s–1) than (S). Each fibril traces a helical path second possible explanation for the generation of thrust. Oscillatoria. However, gliding of cyanobacteria of the at 25–30° to the long axis of the This mucilage is extruded from a row of pores, known as genus Spirulina is difficult to explain by the slime filament. Only a small number of fibrils and only one turn of the helix junctional pores, on either side of the cell septum and can extrusion hypothesis. As the name implies, the filaments are shown. Fibrils at the front and be visualized using India ink (Fig. 7a). The recent work of Spirulina are spiral (Fig. 1b). However, the JPCs do not back of the filament run in different of Hoiczyk and Baumeister has revealed these pores to be cover the entire circumference of the septum, but occupy directions, resulting in the criss- the portion which is inside the spiral. In this case it is cross effect seen in Fig. 4 when difficult to envisage the extruded slime interacting with the filament has been flattened by crushing. (b) Cross section of the the substratum which would be in contact with the cell wall showing the arrangement external surface of the spiral. of the fibrils (F) in relation to the cytoplasmic membrane (CM), Hormogonia peptidoglycan (PG) and outer membrane (OM). Hormogonia motility differs from that in the REPRINTED WITH PERMISSION FROM Oscillatoriaceae by being only short-lived and not J BACTERIOL 181, 884–892 (1999) involving rotation of the filament as it glides. This raises the possibility that the mechanism differs from that in the permanently motile strains. Pili have been observed on the surface of hormogonia during their motile period and at least one gene, pilT, associated with pilus biogenesis, is highly expressed during the early stages of hormogonia formation. Hormogonia gliding may, therefore, like that of some Synechocystis strains, involve pili. Hormogonia are the infectious agents in the establishment of cyanobacteria–plant symbioses and they are attracted to the plant structures that house the symbiotic colonies by chemotaxis to chemicals released by the plant. Whatever the mechanism that drives hormogonia motility, it can be influenced by these plant 132 MICROBIOLOGYTODAY VOL 28/AUG 01 compounds that double the speed of gliding and may LOWER LEFT: (a) control the frequency of reversals. Fig. 6. Transmission electron micrograph of negatively stained fibril fragments isolated from Concluding comments Spirulina sp. Bar, 100 nm. So far, ultrastructural studies have not positively PHOTO MATTHEW BEAN identified structures associated with gliding. However, they have given us a far better picture of the surprising LEFT: complexity of the cell wall in gliding cyanobacteria.