Bacterial Hydrophobicity, an Overall Parameter for the

Transcription

1 APPLIED AND ENVIRONMENTAL MICROBIOLOGY, Jan. 1989, p /89/1142-6$2./ Copyright C) 1989, American Society for Microbiology Vol. 55, No. 1 Bacterial Hydrophobicity, an Overall Parameter for the Measurement of Adhesion Potential to Soil Particles THOR AXEL STENSTROM National Bacteriological Laboratory, S Stockholm, Sweden Received 28 June 1988/Accepted 17 October 1988 The adhesion of Salmonella typhimurium to the mineral particles quartz, albite, feldspar, and magnetite was shown to correlate with the hydrophobicity of the cell surface as measured by hydrophobic interaction chromatography. The same effects were also seen for seven other selected test strains, including Streptococcus faecalis, Streptococcus faecium, Escherichia coli, Citrobacter freundii, Shigella sonnei, and Shigella boydii. When the test strain of Salmonella typhimurium, was repeatedly cultivated in Luria broth, thus selecting for different degrees of fimbriation and roughness of the cell surface, varied cell hydrophobicity but constant negative and positive charge values were obtained. High hydrophobicity values always coincided with enhanced adhesion to the mineral particles. The negative charge of the bacterial surface as measured by electrostatic interaction chromatography appeared to play no role in the adhesion event. However, the positive charges on the cell surface contributed to the adhesion process. This was especially evident for cells exhibiting a high degree of hydrophobicity. Alteration of the ph between 4 and 9 did not significantly affect the adhesion process. 142 The complexity and heterogenicity of soil ecosystems have made it difficult to study events like microbial adhesion. Not only are the surface characteristics and activity of the microbial cells of importance, but the surface charge, wettability, surface texture, and diversity of different types of soil particles also may influence the adhesion process. The variability in ionic composition of the conditioning film, owing to the physicochemical composition of percolating water, and the decomposing activity of the diverse types of microorganisms that are present around the soil particles also greatly affect microbial adhesion. In environmental health and traditional risk assessment, the complexity of the retention of microbes by soil is largely ignored. This has limited the possibilities of comparing different investigations. In addition, there are several problems with respect to studies that include a mechanistic approach, in which as many factors as possible are held constant to elucidate the importance of one or a few factors in the adhesion process. This has led some investigators (15, 2) to emphasize the microbial surface as a static or constant factor. Variations in surface characteristics such as charge and hydrophobicity sometimes have been considered to be species- or strain-dependent factors. This variation will play a less important role, at least in relation to bacterial surface charge and in high-ionic-strength environments. This may be due to elevated concentrations of ions, for example, in marine environments (7, 17), or as a result of the presence of trivalent or tetravalent counterions (25). Both these factors reduce the electrical double layer and the repulsive forces, thus leading to less effects of variations of the cell surface charges. In low-ionic-strength environments, the microbial surface hydrophobicity and charges are of greater importance. In my studies of the binding of Salmonella typhimurium to inanimate surfaces, I have attempted to demonstrate how variations resulting from the alteration of the outer layer of the bacteria, spontaneous mutations, and the presence of fimbriae alter the overall surface characteristics, such as hydrophobicity and positive and negative charges. This renders an altered adhesion at the air-water interface (8) or to mineral particles (26). These studies indicated that variations in surface hydrophobicity resulting from the abovementioned surface alterations correlated with the degree of adhesion to the minerals. The aim of this study was to attempt to evaluate specifically the correlation between cell surface hydrophobicity and charge with regard to the adhesion of Salmonella typhimurium to mineral particles. In addition, studies were designed to investigate whether these factors could be used as a general assessment of bacterial adhesion to minerals in low-ionic-strength environments. MATERIALS AND METHODS Microorganisms and cultivation procedure. The nonflagellated strain of Salmonella typhimurium SH6749, originally obtained from P. H. Makela (Helsinki, Finland), was cultivated in Luria broth (12) at 35 C. Static aerobic cultivation in shallow liquid medium for 48 h produced type 1 fimbriae, while cultures shaken at 15 to 2 rpm for 24 h gave nonfimbriated cells. After harvesting and washing, the occurrence of fimbriae was checked in twofold dilutions by monitoring the capacity to agglutinate yeast cells (1). The presence of fimbriae was verified by electron microscopy in control experiments. Cultivation in Luria broth produced rough strains. With the addition of.1% (wt/vol) galactose and.1% (wt/vol) glucose to the medium, smooth bacteria were isolated (11). However, repeated recultivations of the smooth strain in this medium gave rise to bacteria that became successively more hydrophobic. Based on the media used and the number of recultivations, more than 5 culture preparations of Salmonella typhimurium with different cell surface hydrophobicities were obtained, representing the base for the subsequent test data for Salmonella typhimurium in this study. Other bacteria used were Streptococcus faecalis ATCC 29212; Streptococcus faecium ATCC 843; Escherichia coli C192 and E. coli NK (clinical isolate); Shigella sonnei 1 and Shigella boydii 15 (clinical isolate); and Citrobacterfreundii (well-water isolate). Unless otherwise stated, the isolates used were from the cultural collection at The National Bacteriological Laboratory, Stockholm, Sweden. These Downloaded from on October 21, 218 by guest

2 B~-A_A_ VOL. 55, 1989 HYDROPHOBIC ADHESION OF BACTERIA TO SOIL 143 strains were cultured in Luria broth without the addition of TABLE 1. Variation in binding of Salmonella typhimurium to the galactose and glucose. hydrophobic Sepharose (octyl), an anion-exchange resin (DEAE), The bacteria were metabolically labeled by adding and a cation-exchange resin (CM) [3H]leucine (.35,iCi/ml) to the Luria broth. The cells were % Adhered cells (mean + SD) harvested in the early stationary phase by centrifugation. Group n a They were washed Octyl DEAE CM once and suspended in phosphate-buffered saline solution (PBS) (NaCl, 8.5 g; Na2HPO4 2H2O, g; NaH2PO4 2H2,.39 g; distilled water, 1, ml [ph 7.4]) or in physiological saline. The appropriate solution 3 79± was further used as the eluant in subsequent experiments. " n, Number of parallel runs within each group. Bacterial surface characterization. The surface characterization was performed by hydrophobic interaction chromatography (HIC) and electrostatic interaction chromatography (ESIC). Tritium-labeled cells (2 x 18 to 5 x 18/ml) RESULTS were added to Pasteur pipettes packed with octyl-sepharose Variability and correlation between cell surface hydrophobicity and charge of SalmoneUla typhimurium. During cultiva- CL-4B (Pharmacia, Uppsala, Sweden) for HIC and with DEAE-Sepharose CL-6B (cation exchanger) and CM-Sepharose CL-6B (anion exchanger) for ESIC. Determinations of Salmonella typhimurium cells that expressed various detion, different culture conditions led to the production of radioactivity and control experiments were described previously (4). grees of fimbriation and lipopolysaccharide composition. This resulted in large differences in the adhesion of the Surface bacteria to characterization of the mineral particles. Quartz, hydrophobic resin. The potassium-feldspar, albite, and magnetite were obtained from cultivations. Low adhesion values to the hydrophobic resin degree of adhesion varied from 3 to 88% of added bacteria in the different the Swedish Geological Survey. The minerals were initially always coincided with a low binding to positively and ground from pieces of uncontaminated materials and then negatively charged surfaces (Table 1), while intermediate sieved. The size fraction between.5 and.63 mm was and high binding to the hydrophobic surface gave higher but used. varying degrees of binding to the charged surfaces (Table 1). The specific surface area and the zeta potential at ph 7. A direct correlation between the degree of binding of Salmonella typhimurium to the hydrophobic resin and to the of these minerals have previously been reported (26). The zeta potential measurements in this investigation were performed as described by Parreira (19) with Ag/AgCl elec- values to the hydrophobic resin above 1% always gave anion-exchange resin was not evident (Fig. 1). Adhesion trodes in distilled water. The ph was measured and adjusted greater than 4% binding to the anion-exchange resin (Fig. 1) with HCl and KOH before and after each experiment. The (Table 1). A covariation seemed to occur between the degree ph values given in Fig. 5 were obtained after each experimentchange resin and to the hydrophobic resin (Fig. 2). The of binding of Salmonella typhimurium to the cation-ex- The ph values of the fluid in which the mineral particles highly hydrophobic group corresponded to two separate were suspended were determined by shaking 1. g of the cation-exchange resin-binding groups: % and 6 + 4% particles in 3 ml of PBS for 3 min. The ph was measured in (Table 2). the supernatant. The deviation was never greater than.1 Surface properties of Salmonella typhimurium and binding ph unit. to mineral particles. Adhesion of Salmonella typhimurium to Adhesion assays. The minerals were suspended in the PBS, hydrophobic resin directly correlated with adhesion to and.75 g of each type was packed into clean Pasteur pipettes as for the HIC and ESIC determinations. Duplicate columns were always used for each mineral, and a maximum 1.- variation between the parallel columns of 5% was accepted. l a El a The contact time between bacteria and column material 8 ; always exceeded 1 min. Control experiments showed that n a - "I this time was necessary to ensure reproducible results. The - experiments reported here were always run parallel to the x 6 Eu U Sca HIC and ESIC determinations with an equivalent number of co bacterial cells. Elutions were performed with two 3-ml m 4 portions of PBS per column. Control experiments showed to-- F. * s that additional volumes of PBS removed less than an additional 3% of the total number of attached bacteria. 2 The eluates were collected in scintillation vials. The degree of adhesion was expressed as the percentage of A V *. radioactivity absorbed to the mineral particles, calculated as. W. the amount that was applied to the column minus the amount of radioactivity found in the eluate. This mode of expression Adhesion to octyl-sepharose (%) was necessary because the minerals gave different degrees of FIG. 1. Correlation between the quenching. degree of adhesion of Salinonella typhimurium to columns of octyl-sepharose and to anionexchange (DEAE) beads. The values are expressed as percent Control experiments that included the determination of CFU on agar plates were included and showed comparable adhesion of added bacteria. All data points given represent separate results. experiments and are means of duplicate investigations. Downloaded from on October 21, 218 by guest

3 144 STENSTROM APPL. ENVIRON. MICROBIOL. S 1 I- m *6 CO 6 o z. 4 O 2 'S E E ** 4. % ON 91 A ^ e ^ n n% 4 A Ho I U Adhesion to octyl-sepharose (%) FIG. 2. Correlation between the degree of adhesion of Salmonella typhimurium to columns of octyl-sepharose and to cationexchange (CM) beads. The values are expressed as percent adhesion of added bacteria. All data points given represent separate experiments and are means of duplicate investigations. Data points representing greater adhesion to both the cation exchanger and the octyl-sepharose are indicated in Fig. 3 (U). the mineral particles. This is shown in Fig. 3 for quartz, but similar correlations were also found for albite, feldspar, and magnetite. The differences in the degree of binding of Salmonella typhimurium to the cation-exchange resin were also reflected in the binding to the mineral particles (Table 2). The group of cells giving a higher binding to the cationexchange resin as well as a high binding to the hydrophobic resin are indicated in Fig. 3. A 4 to 5% increase in the adhesion of Salmonella typhimurium cells to quartz, feldspar, and magnetite was found when the binding to the hydrophobic resin changed from 26 to 82% with constant positive and negative cell surface binding values (Fig. 4). Influence of ph on adhesion of Salmonella typhimurium to mineral particles. The effect of different ph values on the surface properties and adhesion of Salmonella typhimurium was measured in the ph range of 4 to 9. Nonpiliated smooth and piliated rough Salmonella typhimurium represented the most hydrophilic and the most hydrophobic bacterial surfaces, respectively, that were tested. The ph range had a TABLE 2. Effect of variation in the positive charge on the adhesion of Salmonella typhimurium to mineral particles' % Total cells adsorbed (mean + SD) at: Surface High positive Medium positive charge charge CM resin Octyl resin 78 ± DEAE resin Quartz Magnetite 77 ± 8 64 ± 5 Feldspar Biotite 94 ± 1 86 ± 3 " The values given are based on 16 quantifications within each group, represented by the two groups of data in Fig. 2, with the greater adhesion to the cation exchanger and octyl-sepharose (U, high positive charge) and medium adhesion to the cation exchanger but high adhesion to octyl- Sepharose (El, medium positive charge). C 4, 2 4 Adhesion to octyl-sepharose (%) FIG. 3. Correlation between the degree of adhesion of Salmonella typhimurium to octyl-sepharose and to quartz particles. The values are expressed as percent adhesion of added bacteria. All data points given represent separate experiments and are means of duplicate investigations. The regression line for all data points is given. Values corresponding to greater adhesion to the cation exchanger in Fig. 2 are indicated (U). marginal effect on the adhesion of the bacteria (Fig. 5). The surface characteristics of the bacteria were of significantly greater importance, as can be seen by comparing Fig. 5a and b with Fig. 5c and d. The variation in ph did not have any profound effect on the zeta potential of the mineral particles except for magnetite, as judged from the streaming potential measurements (Fig. 6). When approaching the isoelectric point in the lower ph range, a lower zeta potential was recorded. Magnatite Felspar Quartz Octyl DEAE Cm Magnetite Feispar Quartz Octyl DEAE Cm MMMMMM MIM's'W"MOMM"t- 2 mw=isl- ~!I' I MO/I _ _ Adhesion (%) FIG. 4. Effect of cell surface hydrophobicity on adhesion of Salmonella typhimurium to mineral particles when positive and negative charges are unchanged. The values are means of eight separate experiments. The bars indicate the standard deviation. Downloaded from on October 21, 218 by guest

4 VOL. 55, 1989 HYDROPHOBIC ADHESION OF BACTERIA TO SOIL 145 A )O - Adhesion of smooth nonpillated a S. typhimurium to resins to - 4c 1C2 o - Adhesion of smooth nonpillated S.typhimurium to minerals a. b B ō Adhesion of rough pillated S.typhimurium to minerals. - I II. I ph ph FIG. 5. Effect of ph on adhesion of smooth nonpiliated (a and b) and rough piliated (c and d) Salmonella typhimurium to octyl-sepharose (*), to cation (x)- and anion (A)-exchange beads (a and c), and to the mineral particles quartz (O), feldspar (-), and magnetite (A) (b and d). The values are expressed as percent adhering cells and are means of parallel investigations. Correlation between adhesion to mineral particles and surface properties of other bacteria. Seven different strains of bacteria representing four species of the family Enterobacteriaceae and two species of Streptococcus were tested with respect to the correlation between surface properties and adhesion to minerals. The results showed a clear positive correlation between cell surface hydrophobicity and adhesion to mineral particles (Table 3). A covariation with charge existed for Shigella sonnei. E - C Is N o l. le Adhesion of rough pillated S. typhimurlum to resins ph FIG. 6. Zeta potential versus ph of the mineral particles albite. biotite, feldspar, magnetite, and quartz in streaming potential measurements. Values are means of duplicate investigations. 1 c DISCUSSION The present study showed that a large variability in cell surface characteristics such as cell surface hydrophobicity and charge may exist within a given bacterial strain. This variability may express various amounts of lipopolysaccharides (8), different types and number of cell surface appendages (6, 9, 26, 27, 33), or capsular material (22). Since the expression of fimbriae of E. coli (5, 18) and flagellae of Salmonella typhimurium (24, 34) may be under phase variation, this may add to the variability. In addition, the growth conditions (19, 28) may induce changes in the bacterial cell surface layers which will influence the overall degree of hydrophobicity and charge (14). TABLE 3. Adhesion of different bacterial species to octyl-sepharose, an anion-exchange resin (DEAE), and a cation-exchange resin (CM) as well as to quartz and magnetite mineral particles Bacterial strain d % Adhesion" to: Octyl DEAE CM Quartz Magnetite Shigella bovdii Citrobacter freundii Shigella sonnei Escherichia coli C Escheerichia coli NK Streptococcus facecium Streptococc(us faecalis " Values are expressed as the percent adhesion of 18 bacteria added to the inanimate surfaces and are mean values of duplicate investigations. Downloaded from on October 21, 218 by guest

5 146 STENSTROM The importance of hydrophobicity and charge for the nonspecific adhesion of microorganisms to solid surfaces has recently attracted considerable interest (for reviews, see references 1, 23, and 25). Adhesion of Salmonella typhimurium to octyl-sepharose as a method to estimate cell surface hydrophobicity, both in the presence of fimbriae and in association with the roughness of the cell surface, was directly related to adhesion to the mineral particles in this study. Furthermore, this effect was illustrated for the other bacterial species that were included in this study. Hydrophobic interaction may therefore be used as a generalized assay to measure the adhesion potential of bacteria to low-energy solid surfaces. The hydrophobic interaction, which traditionally has been considered to operate over short distances, has gained a renewed interest, partly owing to the discovery of long-range "hydrophobic interfacial forces" (2, 3, 3). These long-range hydrophobic interactions may be due to a local dehydration or rearrangement of the water molecules (1, 21, 3, 31), thereby considerably facilitating adhesion of bacterial cells to solid surfaces. This effect, as well as the perturbating effect of thin fimbriae, have introduced a better explanation for many adhesion events than that of the traditional DLVO theory (9). The role of negatively charged groups on the bacterial surface, as measured by ESIC, in the adhesion process is less clear. Apparently, these groups did not counteract the adhesion, as could be expected if the electrostatic repulsion forces constituted an important factor. No general correlation was evident between the extent of negative charge and the cell surface hydrophobicity (Fig. 1), nor did the greater negative zeta potential on the mineral surface affect the degree of repulsion. The evidence for the latter statement is partly based on the fact that magnetite expresses a lower zeta potential (Fig. 6) without a large change in bacterial adhesion as compared with adhesion to quartz, feldspar, or albite. Biotite also exhibited lower zeta potential, as measured by the streaming potential method. An explanation for the larger adhesion to this mineral may be that fixed positively charged groups probably are exposed at the edges in the crystal lattice and may play a more important role in the adhesion process than the overall zeta potential or negative charge. This is supported by the observation that when magnetite and the mica minerals were ground further and the zeta potential measured by particle microelectrophoresis in the size fraction below 1,um, a higher zeta potential was noted for these materials than for quartz, albite, and feldspar. This may have resulted in a higher number of exposed positive charges per surface area. This is furthermore supported by the early work of Marshall and co-workers (for a review, see reference 13) showing an adsorption of clay particles to the surface of Rhizobium cells, resulting from positive charges on platelet clay edges and the anion-exchange capacity of clays, resulting from breakage at the edges to expose trivalent or tetravalent positive ions or from the exchange of OH groups (25). The clay fraction of the tested minerals was not included in the present study. Other investigators have also concluded that the negative surface charge may play a minor role (27, 29) in the adhesion-repulsion processes of bacteria at a solid surface. However, Van Loosdrecht et al. (29) pointed out that the surface electrokinetic potential becomes more influential for more hydrophilic cells and concluded that this allows the bacteria to adhere in the so-called secondary minimum. The same effects may occur for the cells that displayed an intermediate degree of hydrophobicity in this study (Table 1) (Fig. 1), APPL. ENVIRON. MICROBIOL. since a large variation was shown in the degree of negative charge, while a relatively constant cell surface hydrophobicity was observed. The relatively minor importance of the charge-dependent interaction was also reflected by the ph experiments, which revealed that changes in ph in the range between 4 and 9 did not markedly affect the ionization- and charge-dependent adhesion, respectively, of either bacterial or mineral surfaces. This is in agreement with ph and electrophoretic mobility experiments performed with Rhizobium species (13). At ph values below 4 and at higher ionic strength, the charge effects will be more pronounced since both the isoelectric point (pl) and zero charge values are in this range. Even though the cells express a net negative charge, localized positive charges have to be considered. These have been shown to be expressed on the cell surface (16, 32), and suggestions concerning their synergistic role in fimbriaemediated adhesion of Salmonella typhimurium to mineral particles have been presented (26). The role of positive charges in adhesion has so far not been experimentally verified. In this study, the positive charges were found to play a less important role than hydrophobicity for adhesion of Salmonella typhimurium to mineral particles. In the upper hydrophobicity range, positive charges appeared to be more dominant (Fig. 2), while lower effects were found for the more hydrophilic cells. Since, in this study, more hydrophobic cells reflected partly fimbriation and partly the roughness of the surface, two explanations may be plausible. The positive charges may reflect either fixed charges at the top of the hydrophobic fimbriae or other cell surface proteins, exposed by the reduction in the length of the lipopolysaccharide molecules. Furthermore, it is evident that localized positively charged groups are more essential in the adhesion process than an overall negatively charged cell surface. ACKNOWLEDGMENTS This work was supported by the National Swedish Environmental Protection Board. I thank Ingela Blomen for excellent technical assistance and S. Kjelleberg for valuable discussions. LITERATURE CITED 1. Busscher, H. J., and A. H. Weerkamp Specific and non-specific interactions in bacterial adhesion to solid substrata. FEMS Microbiol. Rev. 46: Claesson, P. M., C. E. Blom, P. C. Herder, and B. W. Ninham Interaction between water-stable hydrophobic Langmuir- Blodgett monolayers on mica. J. Colloid Interface Sci. 114: Claesson, P. M., P. C. Herder, C. E. Blom, and B. W. Ninham Interaction between a positively charged hydrophobic surface and a negatively charged bare mica surface. J. Colloid Interface Sci. 118: Dahlback, B., M. Hermansson, S. Kjelleberg, and B. Norkrans The hydrophobicity of bacteria-an important factor in their initial adhesion at the air-water interface. Arch. Microbiol. 128: Eisenstein, B. I., and D. C. Dodd Pseudocatabolite repression of type 1 fimbriae of Escherichia coli. J. Bacteriol. 153: Faris, A., T. Wadstrom, and J. H. Freer Hydrophobic adsorptive and hemagglutinating properties of Escherichia coli possessing colonization factor antigens (CFA/I or CFA/II, type 1 pili, or other pili.) Curr. Microbiol. 5: Gordon, A. S., and F. J. Millero Electrolyte effects on attachment of an estuarine bacterium. Appl. Environ. Microbiol. 47: Hermansson, M., S. Kjelleberg, T. Korhonen, and T. A. Sten- Downloaded from on October 21, 218 by guest

6 VOL. 55, 1989 strom Hydrophobic and electrostatic characterization of surface structures of bacteria and its relationship to adhesion to an air-water interface. Arch. Microbiol. 131: Jones, G. W., and R. E. Isaacson Proteinaceous bacterial adhesions and their receptors. Crit. Rev. Microbiol. 1: Korhonen, T. K Yeast cell agglutination by purified enterobacterial pili. FEMS Microbiol. Lett. 6: Korhonen, T. K., K. Lounatmaa, H. Ranta, and N. Kuusi Characterization of type 1 pili of Salmonella tvphimuniinm LT2. J. Bacteriol. 144: Lounatmaa, K., P. H. Makela, and M. Sarvar Effect of polymyxin on the ultrastructure of the other membrane of wild-type and polymyxin-resistant strains of Sailmonella. J. Bacteriol. 127: Marshall, K. C Interfaces in microbial ecology. Harvard University Press, Cambridge, Mass. 14. McEldowney, S., and M. Fletcher Effect of growth conditions and surface characteristics of aquatic bacteria on their attachment to solid surfaces. J. Gen. Microbiol. 132: Minagi, S., Y. Miyake, Y. Fujioka, H. Tsuru, and H. Suginaka Cell-surface hydrophobicity of Catndida species as determined by the contact-angle and hydrocarbon adherence methods. J. Gen. Microbiol. 132: Neihof, R. A., and W. H. Echols Chemical modification of the surfaces of bacterial cell walls. Physiol. Chem. Phys. 1: Neihof, R. A., and G. I. Loeb The surface charge of particulate matter in seawater. Limnol. Oceanogr. 17: Normark, S., M. Baga, M. Goransson, F. P. Lindberg, B. Lund, M. Norgren, and B. E. Uhlin Genetics and biogenesis of Escherichia coli adhesion. p In D. Mirelman (ed.), Microbial lectins and agglutinins. John Wiley & Sons. Inc.. New York. 19. Parreira, H. C Automatic recording apparatus for measurements of streaming potentials. J. Colloid Sci. 2: Pedersen, K Electrostatic interaction chromatography. a method for assaying the relative surface charge of bacteria. FEMS Microbiol. Lett. 12: Pringle, J. H., and M. Fletcher Influence of substratum hydration and adsorbed macromolecules on bacterial attachment to surfaces. Appl. Environ. Microbiol. 51: Rosenberg, E., N. Kaplan,. Pines, M. Rosenberg, and D. Gutnick Capsular polysaccharides interfere with adherence of Acinetobacteer calcoaceticus to hydrocarbon. FEMS HYDROPHOBIC ADHESION OF BACTERIA TO SOIL 147 Microbiol. Lett. 17: Rosenberg, M., and S. Kjelleberg Hydrophobic interaction: role in bacterial adhesion. Microb. Ecol. 9: Saier, M. H., Jr., M. R. Schmidt, and M. Leibowitz Cyclic AMP-dependent synthesis of fimbriae in Salmnonella typliimirioum: effect of cya and pts mutations. J. Bacteriol. 134: Stotzky, G Influence of soil mineral colloids on metabolic processes, growth, adhesion, and ecology of microbes and viruses. Mechanisms of adhesion to clays, with reference to soil systems, p In D. C. Savage and M. H. Fletcher (ed.), Bacterial adhesion: mechanisms and physiological significance. Plenum Publishing Corp., New York. 26. Stenstrom, T. A., and S. Kjelleberg Fimbriae mediated nonspecific adhesion of Salmonella tyvphimurium to mineral particles. Arch. Microbiol. 143: Van der Mei, H. C., A. H. Weerkamp, and J. H. Busscher Physicochemical surface characteristics and adhesive properties of Streptococcus salivwrius strains with defined cell surface structures. FEMS Microbiol. Lett. 4: Van Loosdrecht, M. C. M., J. Lyklema, W. Norde, G. Schraa, and A. J. B. Zehnder The role of bacterial cell wall hydrophobicity in adhesion. Appl. Environ. Microbiol. 53: Van Loosdrecht, M. C. M., J. Lyklema, W. Norde, G. Schraa, and A. J. B. Zehnder Electrophoretic mobility and hydrophobicity as a measure to predict the initial step of bacterial adhesion. Appl. Environ. Microbiol. 53: Van Oss, C. J., R. J. Good, and M. K. Chaudhury The role of van der Waals forces and hydrogen bonds in "hydrophobic interactions" between biopolymers and low energy surfaces. J. Colloid Interface Sci. 111: Van Oss, C. J., R. J. Good, and M. K. Chaudhury Determination of the hydrophobic interaction energy. Application to separation processes. Sep. Sci. Technol. 22: Weerkamp, A. H., H. C. Van der Mei, D. P. E. Engelen, and C. E. A. De Windt Adhesion receptors (adhesins) of oral streptococci. p In J. M. Ten Cate, S. A. Leach, and J. Arends (ed.), Bacterial adhesion and preventive dentistry. IRL Press. Ltd., Oxford. England. 33. Wicken, A. J Bacterial cell walls and surfaces, p In D. C. Savage and M. H. Fletcher (ed.). Bacterial adhesion: mechanisms and physiological significance. Plenum Publishing Corp.. New York. 34. Zieg, J., M. Hilmen, and M. Simon Regulation of gene expression by site-specific inversion. Cell 15: Downloaded from on October 21, 218 by guest