Extracellular phospholipids of isolated bacterial

0 downloads 0 Views 192KB Size Report
vesicles and a bilipid layer covering the entire surface of that community. These supracommunity ... In. Extracellular phospholipids of bacterial communities. 149 ...
C 2004 Cambridge University Press Biofilms (2004) 1, 149–155  DOI: 10.1017/S147905050400136X Printed in the United Kingdom

Extracellular phospholipids of isolated bacterial communities * Corresponding author: Dr V. V. Tetz Department of Microbiology, Virology and Immunology St Petersburg State Pavlov Medical University 6/8 Tolstoy Street St Petersburg 197089 Russia T 7 812 347 6049 F 7 812 234 3146 E [email protected] [email protected] 1 Department of Microbiology, Virology and Immunology St Petersburg State Pavlov Medical University, 6/8 Tolstoy Street, St Petersburg 197089, Russia 2 Institute of Ecology and Genetics of Microorganisms UB RAS, Perm, Russia

V. V. Tetz1∗ , V. P. Korobov2 , N. K. Artemenko1 , L. M. Lemkina2 , N. V. Panjkova2 and G. V. Tetz1

ABSTRACT We have made a comparative analysis of the extracellular phospholipid composition of biofilms of Gram-negative and Gram-positive bacteria. The surface film of a growing bacterial community contains small membrane vesicles and a bilipid layer covering the entire surface of that community. These supracommunity films containing the bilipid layer can cover the entire surface of a Petri dish and form a boundary between bacterial communities and the environment. A mixed bacterial lawn, formed by unrelated bacteria, also becomes covered with a single film containing a lipid bilayer. The phospholipid compositions of the bacterial cell and surface film bilipid layer reflect the nature of the bacterial strains forming the community, but have some specific differences.

I NTR OD U CTI ON It is generally assumed that all types of biological membranes are well understood. Eukaryotic cells contain a plasma membrane and distinct membrane-bound organelles. In animal and plant cells membranes are found in the nuclear envelope, mitochondria, endoplasmic reticulum, peroxisomes, Golgi complex, lysosomes and chloroplasts. All prokaryotes have cytoplasmic membranes, and Gramnegative bacteria (including spirochaetes, rickettsia and chlamydia) also possess a cell wall outer membrane. Only a small number of known bacterial species have intracellular membranes. Little is known about structures containing bilayer lipids existing outside of eukaryotic and prokaryotic cells. The membrane matrix vesicles in tissues of some animals and humans and also vesicles that are formed in the nutrient broth during growth of some Gram-negative and Gram-positive bacteria were, until recently, the only examples of extracellular membranes (Anderson, 1976; Felix & Fleisch, 1976; Mayrand & Grenier, 1989). We have studied the morphology, physiology and ultrastructure of various types of isolated communities of Gram-negative and Gram-positive bacteria cultivated on solid media and in liquids. We have investigated “classical” colonies, colony-like communities and mixed bacterial communities (Tetz et al., 1990,1993a,b; Tetz, 1996, 1999; Tetz & Rybalchenko, 1997). All types of bacterial communities studied have extracellular bilipid components but until now we had no information on their chemical structure. A biofilm is another type of bacterial community, the object of otherwise extensive investigation (Costerton et al., 1999; Davey & O’Toole, 2000; O’Toole et al., 2000; O’Gara & Humphreys, 2001), where the

extracellular bilipid components have not yet been studied. The aim of this investigation, therefore, was the comparative analysis of the ultrastructures and phospholipid compositions of the extracellular bilipid components of lawns of Gram-negative and Gram-positive bacteria used as a simple model for biofilm formation.

MATE R IALS AN D M ETH OD S Organisms Bacterial strains Escherichia coli JC10240, B, K-12, ATCC (F), Shigella flexneri VT100 and Staphylococcus aureus ATCC 209P (Tetz et al., 1990, 1993a,b) were used. Bacteria were grown on LB liquid medium at 37 ◦ C for 15– 18 h to an optical density at 560 nm of 0.5–0.6 and were then transferred to a rotating shaker and cultivated at 37 ◦ C and 150 r.p.m. (revolutions per minute) to a density of 1.0– 1.1. Lawns were grown in 90 mm diameter Petri dishes containing 12.5 ml of LB agar. To ensure high-quality lawns, the surface of the solid medium was layered with 1 ml of the inoculation culture, exposed to a sterile airflow to dry out, and then transferred to an incubator. Bacteria were cultivated at 37 ◦ C for 24–48 h. Mixed bacterial lawns were obtained by the same method, but in this case a mixture of the two bacteria was plated onto agar. Phospholipid extraction and investigation Two different methods were used for separating the bacterial membrane from the extracellular lipids. In

Extracellular phospholipids of bacterial communities

149

150

V. V. Tetz et al.

method 1, bacterial cells were separated from the fraction containing extracellular lipids, which included membrane vesicles and the bilipid layers of the surface film. In method 2, the bacterial membranes and the bilipid layer of the surface film were isolated from the membrane vesicles of the matrix.

Method 1

To investigate the composition of the phosphoruscontaining components of the lipid extracts from cells and films, labelled phosphate – Na2 H33 PO4 (Isotope, Russia) – with a total activity of 1.1–1.2 MBq was added to the inoculum. After bacterial growth and lawn formation, the culture was layered with 10 ml of an isotonic solution of sodium chloride. Cells were then collected together with the surface film, and the whole suspension was transferred quantitatively into tubes for low-speed centrifugation. Bacteria were precipitated by centrifugation at 3000 g for 30 min, and washed twice with the same volume of isotonic sodium chloride solution. The supernatant obtained was used as the source for surface film components and the intercellular matrix. The precipitated bacteria were used to obtain membrane phospholipids. Lipids were isolated by the method of Bligh & Dyer (1959). Extracts obtained were evaporated, redissolved in a mixture of chloroform–methanol (50:50, v/v), and the lipid content analysed using high-performance thinlayer chromatography on Sorbfil plates of Sorbpolymer (Russia) with a silica gel particle size of 8–12 µm. Two-dimensional separation was performed using the following solvent systems: first direction, chloroform– methanol–ammonium (70:30:2, by vol.); second direction, chloroform–methanol–water (65:25:4, by vol.). To visualize separated phospholipids after the chromatography, the plates were sprayed with 10% CuSO4 in 8% phosphoric acid and heated at 180 ◦ C for 30 min. Identification of areas with glycolipids was made by treating the plates with Schiff’s reagent. Amino-containing phospholipids were detected on identical plates by treating with a mixture of 1% ninhydrine solution in acetone and 0.1% cadmium acetate in 30% acetic acid (5:1, v/v) (Smalley & Birss, 1987). The mobility of the phospholipid markers was examined using the same two-dimensional separation. Preparations of phosphatidyl glycerol (Fluka, Switzerland), lysophosphatidyl glycerol, phosphatidyl ethanolamine, lysophosphatidyl ethanolamine, phosphatidic acid (all from Sigma, USA) and cardiolipin (Medpreparations, Ukraine) were used as standards. The incorporation of radioactive label into phosphorus-containing components of lipid extracts was determined by autoradiographic superposition of film on the Biomax MR (Kodak) plates followed by exposure for 3–7 days at ambient temperature. Areas of labelled compounds were visualized according to a procedure specially developed by Kodak for autoradiographic films. Processing of the radiochromatograms was made after scanning of ready-made films with a DeskScan II device (Hewlett-Packard, Switzerland). For quantification of radioactivity, spots were carefully

scraped off and quantitatively transferred into vials containing 10 ml of the scintillation liquid Ecolite (ICN, USA) for counting. The level of sample radioactivity was measured with a Beta-2 setup (Russia). Method 2

Isolation of the bilipid component of the surface film without the matrix vesicles was achieved as described for method 1. For this purpose the lawn was layered carefully with 15 ml of 0.25% glutaraldehyde in 0.1 M phosphate buffer, pH 7.2. After 60 min of incubation at 37 ◦ C the film emerged from the lawn surface and was collected for analysis. Thus we isolated the bilipid components of the surface film from the membrane vesicles. Residual bacteria in the lawn were collected separately after removal of the glutaraldehyde. Lipids were analysed as described for method 1.

Transmission electron microscopy Pieces of agar with confluently growing bacteria were cut out and fixed in situ for 24 h at 4 ◦ C in 2.5% (v/v) glutaraldehyde in an 0.005 M cacodylate buffer, pH 7.2. The fluid fixative was introduced into the bottom of the agar plates underneath the bacterial communities. These were fixed by diffusion of the fluid fixative through the agar as well as by its vaporization. This method preserved the structure of bacterial communities and prevented the appearance of artifacts. A block consisting of a fixed bacterial community on a piece of agar was dissected and washed in the same buffer, and then postfixed for 24 h in 1% (w/v) osmium tetroxide. Specimens were dehydrated in a graded ethanol series and embedded in Spurr medium (Spurr, 1969). For the isolation of surface film, the bacterial lawn was covered by 2.5% solution of glutaraldehyde without prefixation by diffusion and vaporization. After 40 min of incubation at room temperature, the supernatant surface film was removed and postfixed as described above. Ultrathin sections were prepared using an LKB-8800 (LKB, Sweden) ultramicrotome and were stained as described by Reynolds (1963). The specimens were examined through a JEM-100C (Jeol, Japan) electron microscope at 80 kV.

R E S U LTS Material was studied, at different stages of growth, from more than 100 bacterial lawns of different strains of Grampositive bacteria, Gram-negative bacteria and a mixture of these strains. The bacterial lawns tested had similar ultrastructures, consisting of 20–40 layers of cells after 24 h of growth (Fig. 1). At this time no disrupted cells were found on the lawns. Escherichia coli and S. aureus formed a lawn when plated together, Electron microscopy revealed that all the communities were covered by a surface film that contained bilipid layer components. The thicknesses of this layer in the surface films of different strains of Gram-positive and Gram-negative bacterial communities

Extracellular phospholipids of bacterial communities

Fig. 1: Ultrastructure of a 24 h bacterial lawn formed by S. aureus Arrow shows surface film. Magnification 6800×.

were practically identical (approximately 8.0 nm) and corresponded to the thickness of plasma membranes of the individual bacterial strains. The surface film was a stable structure occupying a large area. In the case of confluent growth it covered the entire surface of the Petri dishes and could be removed and studied separately. These surface films formed the external boundary of the bacterial community investigated, and in sections derived from different parts of the communities no bacterial cells were observed outside the surface film. An electron micrograph of part of the surface film that contains the bilipid layer is shown in Fig. 2. Bacterial lawns that are formed by mixed unrelated bacteria have a single film for the whole community. The second type of biofilm component that contained an extra-bacterial bilipid layer were the vesicles of different sizes found in the matrix of the lawns formed by Gram-positive bacteria, Gram-negative bacteria or a mixture of the strains.

Isolation of surface film by treating with glutaraldehyde enabled separation of the lipids in the surface film from those in the extracellular matrix of the lawn. Chromatograms of cell lipid extracts treated with glutaraldehyde indicated that the compositions of bacterial phospholipids from Gram-positive S. aureus and Gram-negative E. coli differed greatly (Fig. 3). In the case of the mixed lawn, nearly all the major phospholipid components typical for each bacterium were detected in the cell lipid extracts. The compositions of phospholipids in biofilms coating bacterial lawns formed by one or two strains of bacteria differed significantly from that typical for the individual strains forming the lawns (Fig. 3). Marked differences were found between the composition of lipids from individual strains of bacterial cell membranes and that of the surface films coating cell lawns that are formed by bacteria of one type or a mixture of different microorganisms. Surface films formed on mixed lawns had a number of

151

152

V. V. Tetz et al.

Fig. 2: Isolated surface film of a S. aureus lawn. Magnification 100 000×.

Table 1: Relationships between phosphorus-containing components in lipid extracts from bacterial cells and surface films (lawn extracellular matrix) grown as lawns on solid nutrient medium (%)

E. coli

E. coli + S. aureus

S. aureus

Phospholipid fractions

Cells

Film

Cells

Film

Cells

Film

1. Phosphatidyl glycerol (PG) 2. Lysophosphatidyl glycerol (LPG) 3. Phosphatidyl ethanolamine (PEA) 4. Lysophosphatidyl ethanolamine (LPEA) 5. Phosphatidic acid (PA) 6. Cardiolipin (CL) Low mobility unidentified lysophospholipid fractions: a b c d

10.0 1.4 80.3 1.8 0.1 4.6

13.4 — 70.4 1.2 0.9 12.2

71.6 5.6 2.7 — — 7.7

32.2 0.7 0.9 — — 65.5

14.0 0.8 65.8 0.5 < 0.1 11.9

3.2 0.5 62.6 0.8 0.5 31.0

— 1.1 — —

— — — —

3.3 7.1 0.8 0.6

0.6 — — —

0.9 0.6 0.7 —

1.3

3.5

1.3

0.2 4.4

0.8 0.7

Total number of lysophospholipids

4.3

1.2

17.4

Highly mobile phosphorus-containing unidentified lipids: e f

0.1 0.4

0.3 1.5

— 0.6

— —

— — — —

See also Fig. 4.

components in addition to those lipids typical of the surface films on the lawns formed by individual strains. Unfortunately, the interpretation of the results obtained with glutaraldehyde treatment is complicated because the treatment considerably changed the mobility of the isolated lipids; this nearly prevented the identification of phospholipid fractions based on a comparison of experimental chromatograms with those of separated standard preparations. However, we have managed to

study the phospholipid content of bacterial cell membranes and the lawn bilipid component of the extracellular matrix without glutaraldehyde treatment. The results indicate the differences between the phospholipid composition of the bacterial cell membranes and those of the extracellular vesicles and bilipids of the films coating the lawns (Table 1). A number of the isolated components were not identified, but significant enrichment of the films with glycolipid compounds

Extracellular phospholipids of bacterial communities

Fig. 3: Chromatography of phospholipids with glutaraldehyde treatment. (a) S. aureus; (b) surface film of S. aureus; (c) E. coli; (d) surface film of E. coli; (e) mixed lawn of E. coli and S. aureus; (f ) surface film of mixed lawn (E. coli and S. aureus). S, start.

was recorded. The level of glycophospholipids in the extracellular matrix and surface film was much lower than that in the bacterial cell membranes.

D I S CU S S I ON Our results indicate that a bacterial lawn grown on the surface of solid media is an example of an isolated bacterial community. All the bacterial lawns tested were covered by surface films that formed the boundary between the cells and the environment. These lawns were formed by one or two or more unrelated bacterial strains. A lawn as an isolated community of large area is similar to a biofilm (Costerton et al., 1999; Davey & O’Toole, 2000). Two different extracellular structures containing bilipid components – membrane vesicles and the layer of surface film – were found in all bacterial lawns tested. As no disrupted bacteria were found in the lawns under electron microscopy, we assumed that material from dead cells is not likely to play a part in the formation of the bilipid structures. All the communities studied (formed by Gram-negative bacteria, Gram-positive bacteria or a mixture of both) had vesicles of different sizes surrounded

by a bilipid layer in the intracellular matrix. The bilipid component was found in the surface film of all the communities of Gram-negative and Gram-positive bacteria tested, and electron micrographs of isolated surface film suggest that this component is present over a large area. We have found no differences among the structures of the surface films of Gram-negative bacteria, Gram-positive bacteria or communities formed by mixing the strains. The structures of the bilipid components of the bacterial cell membrane, vesicles and surface films were also practically identical. The isolation of bilipid layers after glutaraldehyde treatment indirectly indicates the presence of some proteins in these surface films, as glutaraldehyde targets proteins and does not form complexes with lipids. The phospholipid composition of the bacterial cell membranes and the extracellular bilipid layers of vesicles and surface films have some individual features. The predominant phospholipids in E. coli cells are phosphatidyl ethanolamine (80.3%) and phosphatidyl glycerol (10.0%), and in S. aureus cells phosphatidyl glycerol (71.6%) and cardiolipin (7.7%). Data obtained on the content of major phospholipids in these structures are characteristic for the bacteria of the genera Escherichia (Raetz, 1978) and Staphylococcus (Nahaie et al., 1984). The peculiarities in the phospholipid composition of the E. coli strain used are: relatively low content of phospholipid lysoderivatives (4.3% in total) and the occurrence of trace amounts (0.5%) of phosphorus-containing lipids with high mobility (fractions e and f, Table 1) whose origin has not been identified. Analysis of the extracellular lipid components of the E. coli lawn revealed some specific differences in their phospholipid components: there was a decline in the amounts of lysophosphatidyl ethanolamine and phosphatidyl ethanolamine, and a sharp rise in the relative content of cardiolipin and e + f fractions (Table 1) by 2.5- and 3.5-fold, respectively, as compared with the phospholipids in the cells forming the lawn. Results of the assay of cell phospholipids from the S. aureus 209P lawn showed a high content of lysophospholipids: nearly 17% of labelled phosphate was incorporated into total cell phospholipids from the lawn. The relative content of phospholipids in the lawn extracellular components differed greatly from those in the individual bacterial strains forming these lawns. Thus the percentage of phosphatidyl glycerol was reduced more than 2-fold, that of phosphatidyl ethanolamine by 3-fold, and that of lysophosphatidyl glycerol by 8-fold. Lysoform comprised 1.3% of the total lipids, i.e. 10 times less than in the bacterial cell membranes. Most importantly there is a nearly 9-fold increase in the relative content of cardiolipin in the extracellular components. These differences in phospholipid composition suggest a biological significance. An elevated cardiolipin content in lipids from the extracellular space in bacterial lawns is apparently a prerequisite for spontaneous formation of a bilipid structure with high integrity that could function as a protective barrier between the bacterial cell community and the surrounding space. It is known that an increase in cardiolipin content results in an increase in membrane rigidity. It is most evident in the bacterial cell

153

154

V. V. Tetz et al.

Fig. 4: Glycolipid and aminophospholipid components of cell membrane extracellular bilipid structures. For lipid abbreviations, see Table 1. S, start.

response to unfavourable environmental factors, as the cell membranes are enriched with cardiolipin under these conditions. A low lysophospholipid content is apparently another important characteristic of extracellular lawn components. It is known that the occurrence of free lysophospholipids that takes place under the activation of endogenous phospholipases or is a consequence of cell membrane attack by exogenous phospholipase A2 causes the release of a considerable number of free lysoforms, in particular lysolecithin. The result of this uncontrolled event is complete disorganization of the membrane structure, followed by cell lysis. A direct detergent effect of lysolecithin has been detected in bacteria (Mandersloof et al., 1975; Kondo & Kanai, 1985). Dynamics of the events under this detergent action of lysophospholipids was evident in the model membrane systems. Even with an insignificant increase in lysophospholipid content of the membrane, a decrease in the thickness of the lipid bilayer occurs (Mandersloof et al., 1975) that may not actually cause disruption. However, higher concentrations of lysoforms result in the formation of intramembrane micelles and the rupture of membrane integrity (Inoue & Kitagawa, 1974). An important outcome of the elevated number of phospholipid lysoforms in the membrane composition is the augmentation of membrane sensitivity to osmotic alterations (Inglefield et al., 1976). Therefore, the increase in cardiolipin content and decrease in the level of lysophospholipids in films coating the lawns should result in an enhanced strength of the barrier. This suggestion for the possible elevated strength of the surface film is supported indirectly by the comparative analysis of cell glycolipid components and extracellular lipids (Fig. 4). In this context, the identification of a considerable number of glycolipids in the biofilms is not unexpected. The increase in both their content and their diversity is important. These observations are an additional support for the elevated stability of the films coating the lawns as the rise in glycolipid concentration has a marked effect on the temperature of phase transitions in membrane formation, thereby providing higher stability for the membrane structure (Clowes et al., 1971; Bertoli et al., 1981). Our results indicate the existence

of different extracellular components containing bilipid layers in bacterial lawns. These lipids form vesicles that are a part of the surface film. The bilipid extracellular components have a structure that is nearly identical in the Gram-negative and Gram-positive bacterial communities investigated, both individually and in co-culture. Extracellular components have individual lipid compositions that reflect the origin of the bacteria forming the community. The composition of lipids from extracellular bilipid layers indicates their increased stability, as compared with bacterial cell membranes. It is possible to speculate on the role of these extracellular bilipid components of bacterial communities in the evolution of life. The fusion of membrane vesicles of ancient Gram-positive bacteria on the bacterial surface could have led to the formation of an outer membrane and to the appearance of the first cell wall of a Gramnegative bacterium. Another suggestion rests on the fact that mixed bacterial communities surrounded by such supracommunity bilipid layers are very similar to eukaryotic cells. It is possible that eukaryotic cell formation was initiated not in accordance with the “endosymbiotic” theory (where one cell serves as a “host”) but by a cooperative symbiosis that included the formation of a common surface membrane surrounding communities of unrelated bacteria. Thus it may be that bilipid layers in the surface film of ancient mixed bacterial communities were precursors for the eukaryotic plasma membrane and that ancient mixed bacterial communities were precursors for the first eukaryotic cells.

ACK N OWLE D G E M E NTS We thank Komissarchik Yan Yu. and M. S. Brudnaya for their excellent assistance in the electron microscopy investigations.

R E FE R E N CE S Anderson, H. C. (1976) Matrix vesicle calcification. Federation Proceedings 35, 105–108 Bertoli, E., Masserini, M., Sonnino, S., Ghidoni, R., Cestaro, B. & Tettamanti, G. (1981) Electron paramagnetic resonance studies on the fluidity and surface dynamics of egg phospatidylcholine vesicles containing gangliosides. Biochimica et Biophysica Acta 467, 196–202 Bligh, E. G. & Dyer, W. J. (1959) A rapid method of total lipid extraction and purification. Canadian Journal of Biochemistry and Physiology 37, 911–917 Clowes, A. W., Cherry, R. J. & Chapman, D. (1971) Physical properties of lecithin-cerebroside bilayers. Biochimica et Biophysica Acta 249, 301–317 Costerton, J. W., Stewart, P. S. & Greenberg, E. P. (1999) Bacterial biofilms: a common cause of persistent infections. Science 284, 1318–1322 Davey, M. E. & O’Toole, G. A. (2000) Microbial biofilms: from ecology to molecular genetics. Microbiology and Molecular Biology Reviews 64, 847–867 Felix, R. & Fleisch, H. (1976) Role of matrix vesicles in calcification. Federation Proceedings 35, 169–171

Extracellular phospholipids of bacterial communities

Inglefield, P. T., Lindblom, K. A. & Gottlieb, A. M. (1976) Water binding and mobility in the phosphatidilcholine/cholesterol/ water lamellar phase. Biochimica et Biophysica Acta 419, 196–205 Inoue, K. & Kitagawa, T. (1974) Effect of exogenous lysolecithin on liposomal membranes. Its relation to membrane fluidity. Biochimica et Biophysica Acta 363, 361–372 Kondo, E. & Kanai, K. (1985) Mechanism of bactericidal activity of lysolecithin and its biological implication. Japanese Journal of Medical Science and Biology 38, 181–194 Mandersloot, J. G., Reman, F. C., Van Deenen, L. L. & De Gier, J. (1975) Barrier properties of lecithin/lysolecithin mixtures. Biochimica et Biophysica Acta 382, 22–26 Mayrand, D. & Grenier, D. (1989) Biological activities of outer membrane vesicles. Canadian Journal of Microbiology 35, 607–613 Nahaie, M. R., Goodfellow, M., Minnikin, D. E. & Hajek, V. (1984) Polar lipid and isoprenoid composition in the classification of Staphylococcus. Journal of General Microbiology 130, 2427–2437 O’Gara, J. P. & Humphreys, H. (2001) Staphylococcus epidermidis biofilms: importance and implications. Journal of Medical Microbiology 50, 582–587 O’Toole, G. A., Kaplan, H. B. & Kolter, R. (2000) Biofilm formation as microbial development. Annual Review of Microbiology 54, 49–79 Raetz, C. R. (1978) Enzymology, genetics, and regulation of membrane phospholipids synthesis in Escherichia coli. Microbiological Reviews 42, 614–659

Reynolds, E. S. (1963) The use of lead citrate at high pH as an electron-opaque stain in electron microscopy. Journal of Cell Biology 17, 208–212 Smalley, J. W. & Birss, A. J. (1987) Trypsin-like enzyme activity of the extracellular membrane vesicles of Bacteroides gingivalis W50. Journal of General Microbiology 133, 2883–2894 Spurr, A. P. (1969) A low viscosity epoxy resin embedding medium for electron microscopy. Journal of Ultrastructure Research 26, 31–43 Tetz, V. V. (1994) Colony-like communities of bacteria. Microbios 80, 63–65 Tetz, V. V. (1996) The effect of antimicrobial agents and mutagen on bacterial cells in colonies. Medical Microbiology Letters 5, 426–436 Tetz, V. V. (1999) Formation and structure of mixed bacterial communities. Acta Pathologica, Microbiologica et Immunologica Scandinavica 107, 645–654 Tetz, V. V. & Rybalchenko, O. V. (1997) Ultrastructure of colony-like communities of bacteria. Acta Pathologica, Microbiologica et Immunologica Scandinavica 105, 99–107 Tetz, V. V., Rybalchenko, O. V. & Savkova, G. A. (1990) Ultrastructural features of microbial colony organization. Journal of Basic Microbiology 30, 597–607 Tetz, V. V., Rybalchenko, O. V. & Savkova, G. A. (1993a) Surface films of Escherichia coli colonies. FEMS Microbiology Letters 107, 231–239 Tetz, V. V., Rybalchenko, O. V. & Savkova, G. A. (1993b) Ultrastructure of surface film of bacterial colonies. Journal of General Microbiology 139, 855–858

155