University of Groningen Macroscopic and microscopic ...

University of Groningen

Macroscopic and microscopic approaches toward bacterial adhesionVadillo Rodríguez, Virginia

IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite fromit. Please check the document version below.

Document VersionPublisher's PDF, also known as Version of record

Publication date:2004

Link to publication in University of Groningen/UMCG research database

Citation for published version (APA):Vadillo Rodríguez, V. (2004). Macroscopic and microscopic approaches toward bacterial adhesion. s.n.

CopyrightOther than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of theauthor(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).

The publication may also be distributed here under the terms of Article 25fa of the Dutch Copyright Act, indicated by the “Taverne” license.More information can be found on the University of Groningen website: https://www.rug.nl/library/open-access/self-archiving-pure/taverne-amendment.

Take-down policyIf you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediatelyand investigate your claim.

Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons thenumber of authors shown on this cover page is limited to 10 maximum.

Download date: 25-04-2022

https://research.rug.nl/en/publications/7629cb61-0b4c-4253-b8f2-359b9dae1b64

Macroscopic and microscopic approaches toward bacterial adhesion

Cover: Satellite view of Groningen symbolizing our macroscopic view of bacteria combined with an AFM tip representing our approach to understand their (bacteria) microscopic details. Omslag: Satelliet opname van Groningen als het symbool van ons macroscopisch beeld van de bacterie, gecombineerd met een AFM tip wat staat voor onze manier van kijken naar de microscopische details van de bacterie.

Financial support for the printing of this thesis by the Institute of Biomedical Materials

Science and Applications is gratefully acknowledged Copyright © 2004 by Vadillo Rodriguez, V ISBN 90 367 2116 4 Printed by Stichting Drukkerij de Regenboog, Groningen, The Netherlands

RIJKSUNIVERSITEIT GRONINGEN


Proefschrift

ter verkrijging van het doctoraat in de

Medische Wetenschappen

aan de Rijksuniversiteit Groningen

op gezag van de

Rector Magnificus, dr. F. Zwarts,

in het openbaar te verdedigen op

woensdag 6 oktober 2004

om 16.15 uur

door

Virginia Vadillo Rodríguez

geboren 22 december 1976

te Cáceres

Promotores: Prof. Dr. Ir. H.J. Busscher Prof. Dr. H.C. van der Mei Prof. Dr. Ir. W. Norde

Beoordelingscommissie:

Prof. Dr. D. Hoekstra Prof. Dr. J.M. van Dijl Prof. Dr. P.G. Rouxhet

Paranimfen:

Joop de Vries Jerome D. Swinny

Dedicado a mis padres Manuel y Virginia,

y a mi hermano Manuel

“En verdad, igual que el sol amo yo la vida y todos lo mares profundos.

Y esto significa para mi conocimiento: todo lo profundo debe ser elevado – ¡hasta mi altura!”

Prof. Friedrich Nietzsche

(Así habló Zaratustra. Un libro para todos y para nadie)

“Verily, like the sun I love life and all deep seas.

And this is what knowledge means to me: all that is deep shall rise up – to my height! ”

Prof. Friedrich Nietzsche

(Thus spoke Zarathustra. A book for all and none)

Table of contents

1. Macroscopic and microscopic adhesive properties of microbial cell surfaces

1

2. Softness of the bacterial cell wall of Streptococcus mitis as probed by micro-electrophoresis

11

3. A comparison of AFM interaction forces between bacteria and silicon nitride substrata for three commonly used immobilization methods

19

4. On relations between microscopic and macroscopic physico-chemical properties of bacterial cell surfaces –an AFM study on Streptococcus mitis strains

29

5. Relations between macroscopic and microscopic adhesion of Streptococcus mitis strains to surfaces

43

6. Atomic force microscopic corroboration of bond aging for adhesion of Streptococcus thermophilus to solid substrata

59

7. Dynamic cell surface hydrophobicity of lactobacillus strains with and without surface layer proteins

67

8. Role of lactobacillus cell surface hydrophobicity as probed by AFM in adhesion to surfaces at low an high ionic strength

73

General Discussion 89 Summary 99Samenvatting 103Resumen 107 Reference list 113Acknowledgments 123

1 Macroscopic and microscopic adhesive properties of microbial cell surfaces Virginia Vadillo Rodríguez, Henk J. Busscher and Henny C. van der Mei*

1.1 INTRODUCTION The study of microbial adhesion encompasses a broad range of scientific disciplines, ranging from medicine, dentistry and microbiology to colloid and surface science. Initially, the involvement of colloid and surface scientists originated from the simple realization that microorganisms are, with respect to their dimensions, colloidal particles and that their adhesion should be predictable by surface thermodynamics (Absolom et al., 1983) or Derjaguin-Landau-Verwey-Overbeek (DLVO)-theory like approaches (Bos et al., 1999). Considering the ubiquitous nature of microbial adhesion, such generalized predictive models would be extremely valuable.

* Reproduced with permission of IWA Publishing from “Biofilms in Medicine, Industry and Environmental Biotechnology” (2003) edited by P. Lens, V. O’Flaherty, A.P. Moran, P. Stoodley and T. Mahony

Macroscopic and microscopic approaches toward bacterial adhesion 2

Microorganisms have, from a physico-chemical point of view, generally been considered to be similar to inert polystyrene particles. However, microorganisms are not smooth particles and, in contrast to polystyrene particles, carry long, usually very thin surface structures protruding from the cell surface and radiating outwards into the surrounding liquid. These structures are responsible for adhesion to a variety of surfaces. Bacteria in particular carry a wide range of surface structures that have been described on the basis of their ultrastructure and distribution on the cell surface. There are many morphologically distinct types of surfaces structure and almost every bacterial strain or species carries its own type of surface appendage. Yet, the function of these surface structures in specific adhesion processes to different substratum surfaces as well as their influence on overall physico-chemical cell surface characteristics, remains to be identified for most bacterial strains.

Application of physico-chemical models to explain microbial adhesion to solid substratum have been successful for a limited number of strains and species (Van Loosdrecht et al., 1989), despite the macroscopic nature of the input data (i.e. Hamaker constants, acid-base properties, zeta potentials and contact angles). Therefore, relevant physico-chemical measurements on microbial cell surfaces require a microscopic resolution that can not be accomplished with most currently employed methods.

The aim of this chapter is on the one hand to point out the merits of a macroscopic physico-chemical approach toward microbial adhesion, while on the other hand emphasizing the need of a microscopic physico-chemical surface characterization. 1.2 MACROSCOPIC PHYSICO-CHEMICAL CHARACTERISTICS OF MICROBIAL CELL SURFACES Bacterial cell surface hydrophobicity and charge are commonly accepted as influential on bacterial interactions with their environment. Therefore contact angles and zeta potentials of microbial cell surfaces are frequently measured for use as input data for predictive, physico-chemical models of their adhesion to surfaces. Furthermore, at a similar overall level as cell surface hydrophobicities by water contact angles and zeta potentials, chemical composition data of cell surfaces are being obtained by unexpected techniques, like X-ray photoelectron spectroscopy (XPS) (Amory et al., 1988) or secondary-ion mass spectroscopy (SIMS) (Tyler, 1997). 1.2.1 Cell surface hydrophobicity Hydrophobicity is a general term utilized to describe the relative affinity of a surface for water. Microbial cell surface hydrophobicity can only be measured by placing water droplets on carefully prepared and dried microbial lawns (Busscher et al., 1984). If water molecules have a greater preference to surround each other than to contact a microbial cell surface, the surface appears as hydrophobic and water droplets do not spread. In contrast, if water molecules favor a microbial cell surface rather than each other, a water droplet spreads and the surface appears hydrophilic.

Macroscopic and microscopic cell surfaces properties 3

Spreading is determined in part by the intermolecular forces, such as Lifshitz-Van der Waals and acid-base forces (Van Oss, 1995), as can be calculated amongst others, from the Hamaker constants and measured contact angles with liquids. The only non-trivial step in contact angle measurements on microbial lawns involves the degree of drying of the lawns. It must be realized that drying of the lawns determines the degree of collapse of microbial surface appendages and therewith the contact angle measured (Busscher et al., 1984). Usually contact angles measured on microbial lawns increase as a function of drying time until a so-called “plateau” is reached. As long as the lawns are dried to the plateau for water contact angles, the cell surface is assumed to be in a physiologically relevant state. However, theoretically, cell surface hydrophobicity can not be measured solely by water contact angles as this does not allow to account for acid-base interactions as occurring between water molecules and between water molecules and the microbial cell surface. Indeed, whereas water contact angles have been designated as the intrinsic surface hydrophobicity of a microbial cell surface, the thermodynamic cell surface hydrophobicity strictly speaking reflects its surface free energy, including Lifshitz-Van der Waals and acid-base components. Calculation of the thermodynamic cell surface hydrophobicity requires contact angle measurements with at least three different liquids with varying acid-base properties, as water, formamide, methyleneiodine and/or α-bromonaphthalene. (Van Oss et al., 1987).

The cell surface hydrophobicity describes a macroscopic surface property. The hydrophobicity of surface appendages, like fibrils or fimbriae may be substantially different from the overall cell surface hydrophobicity. However, despite being an overall cell surface property, microbial contact angles vary greatly between different strains and even the presence of tufts of fibrils as on Streptococcus sanguis strains is reflected in the measured values of contact angles (Busscher et al., 1991). Both Streptococcus salivarius HBC12 and S. sanguis CR311 VAR3 have bald cell surfaces

Figure 1.1. Electron micrograph of S. sanguis PSH 1b labeled with colloidal gold at pH 10.2. Gold is attached only to the ends of the long fibrils. Cells were stained with 1% methylamine tungstate. Bar marker is 200 nm (adapted from Van der Mei et al. (1998b) with permission).


and are hydrophilic (water contact angles 21 and 31 degrees, respectively) compared with their peritrichously fibrillated (S. salivarius HB) and tufted (S. sanguis CR311) parent strains. Hence, hydrophobicity is conveyed to the cell surface by fibrils and fimbriae and unlikely by the bald cell surface. Handley et al. (1991), for instance, demonstrated that hydrophobic, colloidal gold particles only adsorbed to the tip of fibrils on S. sanguis PSH 1b and concluded that the cell surface hydrophobicity was confined to the ends of the long fibrils (see Figure 1.1).

Therefore, contact angles measured with liquid droplets on a microbial lawn are essentially representative of a fuzzy coat of cellular surface material, collapsed into a lawn. Therewith results are useful to interpret the long-range interactions between an organism and a substratum surface, but not necessarily for the interpretation of short-range interactions, which may be dominated by structural and chemical cell surface heterogeneities (Van der Mei et al., 1998b). 1.2.2 Zeta potentials Particulate electrophoresis is the most common method to determine bacterial zeta potentials and surface charge densities (James, 1991). The measurement of microbial electrophoretic mobilities and the derivation of zeta potentials thereof by particulate microelectrophoresis proceeds according to the standard methodologies in physico-chemistry (James, 1991), although the calculation of zeta potentials from measured electrophoretic mobilities is not always straightforward (Van der Wal et al., 1997). Furthermore, it is important to realize that the plane of shear may be removed far away from the bacterial cell wall if long fibrils or fimbriae are present (e.g. as on Streptococcus mitis strains) and consequently, in these cases, the zeta potentials measured are not representative for the one of the true cell surface (Figure 1.2).

Moreover, long appendages may collapse onto the cell surface upon increasing the ionic strength, as demonstrated e.g. by dynamic light scattering (Van der Mei et al., 1994). Depending on ionic strength of the liquid, also electro-osmotic flow of fluid may occur within these polyelectrolyte layers.

Fibrils, fimbriae and even extracellular surface polymers around bacterial cells may be considered as a polyelectrolyte layer, possessing a number of fixed charges, i.e. ionic groups that are covalently linked to the polymer and thus have a strong impact on the electrostatic interactions of bacteria with surfaces. Hayashi et al. (2001) and Poortinga et al. (2001) have both described that this electrostatic repulsion is often overestimated due to the neglect of bacterial cell surface softness, i.e. the ease with which electro-osmotic fluid flow develops in the surface layer. Soft, ion-penetrable cell surfaces experience less electrostatic repulsion than similarly charged, hard, ion-impenetrable surfaces, since their diffuse layer charges are driven into the ion-penetrable cell walls causing an effective decrease in surface potential (see also Figure 1.2) and, hence, electrostatic repulsion. Recently, Morisaki et al. (1999) explained adhesion of a negatively charged marine bacterium Vibrio alginolyticus, onto a negatively charged substratum by considering the softness of the strains.

When analyzing microbial adhesion data in terms of electrostatic interactions, it should not be a priori assumed that the electrostatic interaction is repulsive, simply


hard layer model

Figure 1.2. Schematic representation of the surfaces of a negatively charged ion-impenetrable (a) and ion-penetrable bacterium (b). The electrokinetic potential decreases exponentially from the ion-impenetrable bacterial core surface and the slip plane, determining its zeta potential close to the surface. The ion-penetrable bacterium is covered by a polyelectrolyte layer with fixed negative charges, through which electrophoretic fluid flow is possible. The slip plane is assumed to remain at approximately the same position as for the ion-impenetrable bacterium inside the soft layer. Therefore, the zeta potential is more negative than the potential Ψ0 at the outside of the soft layer (taken from Kiers et al. (2001) with permission).

Distance

slip plane

Elec

tric p

oten

tial, ψ

soft layer model

Ψ 0

Elec

tric

pote

ntia

l, ψ

Distance

slip plane

B

acte

rial c

ore

zeta potential

zeta potential

Bac

teria

l cor

e


because the zeta potentials of the interacting surfaces are both negative. Indeed, nearly all biological surfaces carry a net negative charge. However, at a more microscopic level than the macroscopic level of particulate microelectrophoresis, microbial cell surfaces may have positively charged domains mediating adhesion through local electrostatic attraction despite overall repulsion, like described e.g. the interaction between Treponema denticola and human erythrocytes (Cowan et al., 1994). Obviously, a minor number of positively charged sites, while instrumental for adhesion, does hardly affect the macroscopic cell surface charge density.

1.2.3 Chemical composition of microbial cell surfaces X-ray photoelectron spectroscopy (XPS) provides a mean to obtain the chemical composition of the outermost microbial cell surfaces (Amory et al., 1988). XPS spectra of microbial cell surfaces are fairly similar, with carbon, nitrogen, oxygen and phosphorous being the main elements detected, albeit in different amounts on different isolates. Decomposition of C1s and O1s electron binding energies has furthermore indicated the presence of lipids, proteins and polysaccharides. As XPS is a high vacuum technique, an extensive sample preparation, including washing, centrifuging and freeze-drying, is involved before microbial cell surfaces can be studied by XPS (Rouxhet et al., 1994). These steps obviously bring the cell surface in a state that is far remote from its physiological one. Some authors believe that the integrity of the vulnerable cell surface of especially Gram-negative bacteria, as compared with Gram-positive bacteria, is disrupted by this extensive preparation (Marshall et al., 1994) with a potential impact on the results.

However, the relevance of XPS for the analysis of microbial cell surfaces is supported by the relationships with other physico-chemical cell surfaces properties, preferably measured on cells in a more physiologically relevant state than in their dehydrated, freeze-dried state as for XPS. For instance, combinations of contact angle data on a collection of widely different streptococcal strains and XPS have demonstrated that hydrophobicity is conveyed to the cell surface by nitrogen-rich groups, concurrent with the possession of a high isoelectric point (the pH at which the zeta potential is zero) (Van der Mei et al., 1988b). Furthermore, the presence of tufts of fibrils on S. sanguis CR311 increased the N/C from 0.066 of a bald variant to 0.085 for the parent strain, indicating a nitrogen (protein) rich composition of the tufts (Busscher et al., 1991). In the same way, the progressive removal of fibrils from the cell surface of S. salivarius HB was accompanied by a significant decrease in the N/C elemental surface composition ratio (Van der Mei et al., 1988a). However, like almost all physico-chemical methods for the study of microbial cell-surfaces properties, the spatial resolution of XPS is inadequate to deal with chemical and structural heterogeneities, like sparsely or unevenly distributed fibrils or fimbriae on cell surfaces.

Secondary-ion mass spectroscopy (SIMS) is a surface sensitive technique, that also probes the chemical composition of a surface, but through a different principle as XPS. XPS involves the bombardment of a surface with X-rays and the subsequent measurement of the photo-emitted electrons. Since these photo-emitted


electrons have discrete kinetic energies that are characteristic of the emitting atoms and their bonding states, they can be applied for chemical analysis. In SIMS, the surface is bombarded with a focused beam of primary ions, the impact of which produces secondary ions that are collected and focused in a mass spectrometer where they are separated according to their mass. Tyler (1997) has shown that SIMS can also be applied to probe microbial cell surface chemistry. SIMS spectra of four freeze-dried strains, S. salivarius HB and three mutants, indicated the presence of proteins, hydrocarbons and carbohydrates on the bacterial cell surfaces, as well as of proteins and teichoic acid on the cell wall. The correlation between SIMS spectra and previous XPS analysis on those strains was excellent. Taking into account that SMIS is not only capable of providing accurate analysis of surface chemistry, but is also sensitive to the composition and orientation of bio-molecules, the potential of this technique to characterize bacterial cell surfaces seems promising but needs to be further explored. 1.3 MICROSCOPIC PHYSICO-CHEMICAL PROPERTIES OF MICROBIAL CELL SURFACES BY ATOMIC FORCE MICROSCOPY Even though structural and chemical heterogeneities on microbial cell surfaces have an impact on the overall cell surfaces, methods to obtain detailed knowledge on cell surface heterogeneities are still lacking. Indeed, considering the importance of structural and chemical heterogeneities in microbial adhesion, the development of a generalized model for microbial adhesion to surfaces seems still far beyond reach. However, the introduction of the atomic force microscope (AFM) and its application to biological surfaces (Dufrêne, 2001) has offered new possibilities to obtain microscopic physico-chemical properties of bacterial cell surfaces.

AFM provides exciting possibilities for probing the structural and physical properties of living microbial cells. Using topographic imaging, cell surface nano-structures (e.g. appendages and flagella) can be directly visualized (see for instance Figure 1.3 and the changes of cell surface morphology occurring during physiological processes can be determined.

Force-distance curves, shown in Figure 1.4, can provide complementary information on surface forces and adhesion mechanisms at a square nano-meter scale, yielding new insight into the mechanisms of biological events such as microbial adhesion and aggregation.

Initial AFM studies on bacterial cell surfaces have focused on probing surface morphology and surface forces. For instance, Razatos et al. (1998) showed that the adhesion force between a silicon nitride AFM tip and Escherichiae coli was affected by the length of lipopolysaccharide molecules on the cell surface and by the production of a capsular polysaccharide. Furthermore, it was discovered using AFM that E. coli JM109 and K12J62 have different surface morphologies dependent on environmental conditions, while lysozyme treatment led to the loss of surface rigidity and eventually to dramatic changes of bacterial shape (Bolshakova et al., 2001). Camesano and Logan (2000) concluded that the interaction between


0 2 µm

300 nm 150 nm 0 nm

Figure 1.3. AFM contact mode topographic image of S. mitis T9 immersed in water. The image reveals characteristic topographic features on the right hand side of the cell surface, i.e. lines oriented in the scanning direction, attributable to fibrils.

Figure 1.4. Force-distance curve of S. mitis T9 in water. The solid line represents the approach curve, while the dashed line indicated the retraction curve. Upon approach, a long-range repulsion, starting at a separation of ~ 100 nm, was detected, while no jump-to-contact was observed. Upon retraction, multiple adhesion forces were found.

1200 10008006004002000

6

4 2 0

Forc

e (n

N)

Separation distance (nm)


negatively charged bacteria and the silicon nitride tip of an atomic force microscope was dominated by electrosteric repulsion. Only a limited number of studies have focused on the characterization of local properties of bacterial cell surfaces. Recently, for instance, the turgor pressure of a spherical bacterium, Enterococcus hirae, in deionized water was derived from the indentation depth caused by an AFM tip and found to be between 4 and 6 x 105 Pa (Yao et al., 2002). However, in order to develop a ubiquitously valid physico-chemical model for microbial adhesive interactions, microscopic characterization of properties as hydrophobicity, surface charge density and chemical composition on microbial cell surfaces is required. Due to the structural and chemical heterogeneities that the cell surfaces present, it would be of interest to collect an array of force curves over the entire cell surface. Such an array would produce information about the distribution of different surface properties. For example, using charged or chemically functionalized AFM tips to probe the surface, would allow localizing more specific interactions at a microscopic level. At present, the only charge maps for biological samples have been made for bacteriorhodopsin membrane patches (Butt, 1992) and phospholipid bilayer patches (Heinz and Hoh, 1999a) on hard substrates. From the known surface charge density of the substratum, it was possible to calculate a reasonable value for the surface charge density of the membrane (Butt, 1992). Hydrophobicity at a microscopic level has been probed on spore surfaces of the fungus Phanerochaete chrysosoporium (Dufrêne, 2000) by using chemically modified AFM probes, terminated with OH (hydrophilic) and CH3 (hydrophobic) groups. 1.4 TOWARD RELATIONS BETWEEN MICROSCOPIC AND MACROSCOPIC PROPERTIES Overall properties are a macroscopic expression of interactions taking place at a microscopic level. Extension of a microscopic property of microbial cell surface to the entire surface should theoretically lead to the corresponding macroscopic property. However, it will be a delicate task to amalgamate microscopic properties derived from AFM measurements into the macroscopic cell surface properties. The different conditions under which the macroscopic and microscopic properties are measured, should be taken into account. Properties as hydrophobicity and surface charge, from a macroscopic point of view, are determined in a two component system, i.e. bacterium and liquid medium. At the microscopic level, the AFM tip interacts with the bacterium in a medium, and consequently, bacterium and medium would respond to this third component, i.e. the tip. Therefore, hydrophobicity and charge mappings derived from AFM measurements depend on the properties of the AFM tip as well. Macroscopic and microscopic properties estimated in both systems are related, but to find out how, constitutes an enormous challenge. It involves an accurate knowledge of geometry and physico-chemical characteristic of the AFM tip as well as of a theory describing long and short-range interactions in such a system.


1.5 CONCLUSIONS AND AIM OF THIS THESIS Physico-chemical properties of microorganisms can vary widely and generalizations at the species or even strain level are virtually impossible. The degree of success of physico-chemical models to explain microbial adhesion frequently decreases as the complexity of cell surface appendages on the organisms under consideration increases. Understanding how microscopic properties can be amalgamated into the macroscopic properties previously determined by many different research groups all over the world for a large variety of different strains and species is an imperative next step in the characterization of microbial cell surfaces. Subsequently, a generalized physico-chemical theory to account for bacterial adhesion to substratum surfaces will become in reach. Recently, the introduction of the atomic force microscope and its application to biological surfaces has opened a new avenue to obtain microscopic physico-chemical properties of bacterial cell surfaces. It is a challenge for the future to develop models, based on these improved methodologies, that will allow predicting bacterial adhesion from the initial adhesion events.

This thesis is aimed to reach a microscopic characterization of physico-chemical properties of bacterial cell surfaces as well as to find out if and how microscopic properties could be amalgamated into macroscopic cell surface properties and related to macroscopic bacterial adhesion on solid substrata.

2 Softness of the bacterial cell wall of Streptococcus mitis as probed by micro-electrophoresis Virginia Vadillo Rodríguez, Henk J. Busscher, Willem Norde and Henny C. van der Mei*

2.1 INTRODUCTION Accurate quantification of the physico-chemical properties of bacterial cell surfaces is essential in order to reach a better understanding of bacterial adhesion to a substratum. Current methods to study bacterial physico-chemical characteristics of cell surfaces usually yield overall properties and they do not account for structural and chemical heterogeneities, such as the occurrence of fibrils, fimbriae or other surface appendages (Hermansson, 1999). These structural and chemical heterogeneities have a definite function in adhesion and it has been found, for instance, that fibrillated Streptococcus salivarius strains carrying the Lancefield * Reproduced with permission of John Wiley & Sons, Inc. from Electrophoresis (2002) 23: 2007-2011


Group K polysaccharide antigen are more adhesive to buccal epithelial cells than strains lacking the antigen (Weerkamp & McBride, 1980; Handley et al., 1987). Streptococcus parasanguis carries fimbriae that are held responsible for adhesion to saliva-coated hydroxyapatite (Fives-Taylor & Thompson, 1985). Yet, the influence of structural and chemical heterogeneities on overall physico-chemical cell surface properties, such as zeta potential and hydrophobicity, remains to be identified for most strains.

Zeta potential and hydrophobicity are macroscopic corollaries of the overall chemical composition of the bacterial cell surface and their role in adhesion is generally assumed to be confined to the interaction at larger separation distances (more than a few nanometers). Bacterial surface appendages and chemical heterogeneities often exert localized attraction and consequently function to bridge a gap between otherwise repelling substratum surfaces (Matthyse et al., 1981; Smit et al., 1986). Fibrils, fimbriae and even extracellular surface polymers around bacterial cells may be considered as a polyelectrolyte layer, possessing a number of fixed charges, i.e. ionic groups that are covalently linked to the polymer. Electro-osmotic flow of fluid may occur within these polyelectrolyte layers and this, in turn, has a strong impact on electrostatic interaction of bacteria with surfaces. The ease at which fluid can flow through the polyelectrolyte layer is referred to as the electrophoretic softness. Ohshima (1995) has forwarded a model that enables to derive the electrophoretic softness 1/λ and the fixed charge density ρfix of polyelectrolyte layers from the dependence of the electrophoretic mobility on ionic strength, according to ⎤⎡ 2 is in which µ is the electreciprocal Debye-lengthstrength. ρfix is related tocharge e and the valency

The theory assumes

However, the thickness omay increase with incrmolecular electrostatic isegments in the polyelpolymer chain. Then, aincrease of the polyelecAt first approximation,

⎥⎥⎦⎢

⎢⎣ +

+⎠⎞⎜

⎝⎛+=

κλκλ

κλ

ηλρfixµ

/12/112 (2.1)(2.1)

rophoretic mobility, η the viscosity of the solution and κ the of the polymer layer, which determined by the ionic the number density of charged groups N, the electrical unit z through

(2.2) ρfix= zeN

that ρfix and 1/λ are invariant with the ionic strength. f a polyelectrolyte layer may shrink and the charge density

easing ionic strength, both as a result of reduced intra-nteraction. On the other hand, dense packing of polymer ectrolyte layer suppresses the charge density along the dding low molecular weight electrolytes may cause an

trolyte fixed charge density that causes the layer to swell. these effects may be disregarded and ρfix and 1/λ, are

Electrophoretic softness of bacterial cell walls

13

estimated from a plot of the electrophoretic mobility against the ionic strength or, for that matter, the reciprocal Debye-length κ.

Accounting for the electrophoretic softness of bacterial cell surfaces in adhesion models leads to a reduction of electrostatic repulsions between bacteria and a substratum surface of the same charge sign (Morisaki et al., 1999; Poortinga et al., 2001). These decreased repulsions are more realistic. For instance, when a soft glass surface interacts with an ion-penetrable bacterium, their electric double layers overlap, whereby the diffuse double layer charges are driven into the ion-penetrable layers causing an effective decrease in surface potential and electrostatic repulsion (Poortinga et al., 2001).

Streptococcus mitis is one of the primary colonizers of hard surfaces in the oral cavity and colonizes both dental hard tissues as well as mucous membranes, most notable the cheek and tongue (Marsh & Martin, 1992). By comparison with other oral streptococci, S. mitis strains usually carry sparsely distributed but extremely long fibrils, as demonstrated by electron microscopy (Cowan et al., 1992), and their cell surfaces should be regarded as being soft.

The aim of this study is to determine the electrophoretic softness and fixed charge density in the polyelectrolyte layer of nine S. mitis strains according to a soft-particle analysis using measured electrophoretic mobilities (Ohshima, 1995). The results are compared with those reported for other bacterial strains having different surface morphologies. 2.2 MATERIALS AND METHODS

2.2.1 Bacterial strains, growth conditions and harvesting The nine S. mitis strains used in this study were all cultured in Todd Hewitt Broth (Oxoid, Basingstoke, UK). For each experiment, the strains were inoculated from blood agar in a batch culture. This culture was used to inoculate a second culture that was grown for 16 h prior to harvesting. Bacteria were harvested by centrifugation (5 min at 10,000 g), washed twice with demineralized water and resuspended in water. To break up bacterial chains, cells were sonicated for 30 s at 30 W (Vibra Cell model 375, Sonics and Materials Inc., Danbury, CT, USA). Sonication was done intermittently while cooling in an ice/water bath. These conditions were found not to cause cell lysis in any strain. 2.2.2 Particulate microelectrophoresis Electrophoretic mobilities were measured at 25 ˚C with a Lazer Zee Meter 501 (PenKem, Bedford Hills, NY, USA) equipped with an image analysis option for tracking and zeta sizing (Wit et al., 1997). Measurements were carried out in KCl solutions of various ionic strengths (pH 5) ranging from 1 mM to 150 mM. Prior to each measurement, an aliquot of the bacterial suspension was added to the appropriate KCl solution to obtain a density of approximately 1×108 cells ml-1. The pH of the solutions did not change upon addition of the bacterial cells. The Lazer Zee Meter was set at 150 V for determination of the electrophoretic mobilities of the bacteria.


2.3 RESULTS The electrophoretic mobilities of the nine S. mitis strains are summarized in Figure 2.1 as a function of ionic strength. Although the electrophoretic mobilities clearly become less negative with increasing ionic strength, they do not approach zero at elevated ionic strength, which is a characteristic feature of soft particles. The electrophoretic softness (1/λ) and fixed charge density (ρfix), characterizing the properties of a soft particle, as obtained from fitting the data in Figure 2.1 to Eq. 2.1, are summarized in Table 2.1. At low ionic strength, i.e. 1mM, the observed mobilities deviate considerably from the ones calculated by the soft-particle model, clearly because the model assumptions of constant fixed charge density and softness are not applicable in the low ionic strength region. Similarly, except for S. mitis 272 the mobilities at 150 mM appear to be slightly above the best-fit theoretical curves; nevertheless those values have been considered in our analysis. Note that for S. mitis T9 the electrophoretic mobility at 150 mM is zero, despite the fact that a stationary value appears to have been reached at lower ionic strengths.

Although the quality of the fit, as expressed by the coefficient of determination (r2), is generally high (see Table 2.1), it varies with the ionic strength range considered. For most strains the fit is the best for the ionic strength range 10 – 100 mM (r2 = 0.95), and reduces to 0.93 for the ionic strength ranges 15 – 150 mM and 10 – 150 mM. For some strains, i.e. for S. mitis 357, BMS and ATCC33399, the quality of the fit is the highest for the ionic strength range 15 – 150 mM. Fitting errors in electrophoretic softness and fixed charge density are around 0.3 nm and -0.3×106 C m-3, and slightly dependent on the ionic strength range applied.

Based on the analysis above, it is concluded that for each isolate an optimal ionic strength range should be chosen, based on the quality of the fit obtained, that yields the best estimate for the bacterial cell surface softness and fixed charge density. These values have been marked in bold in Table 2.1 and are the ones that are considered further in this study.

Figure 2.2 expresses the softness against the fixed charge density obtained. It is interesting to note that for the present collection of S. mitis strains the softest appear to possess the least fixed charge. 2.4 DISCUSSION The frequency of occurrence of extracellular surface structures on S. mitis strains is very high and various appendages with different lengths up to several microns have been detected. Generally, the density of appendages on cell surfaces may vary considerably between strains, species and even isolates, but for S. mitis isolates it has been described that the density of fibrils is sparse (Van der Mei et al., 1998b).

The present study clearly reveals that the cell surface of all nine S. mitis strains are electrophoretically soft with fixed negative charge densities. The degree of softness and charge density of the fluid-penetrable layer varies between the different strains. The data presented in Figure 2.2 clearly show that the softest strain has the lowest charge density. This correlation possibly reflects the density distribution of the fibrils on the different strains. For sake of comparison, results of


15

Ionic strength (M)0.00 0.02 0.04 0.06 0.08 0.10 0.12 0.14 0.16

Elec

troph

oret

ical

mob

ility

(x

10-8

m2 V

-1s-1

)

-4

-3

-2

-1

0

S. mitis T9 S. mitis 244

S. mitis 357

S. mitis 398

S. mitis ATCC33399

S. mitis BA

S. mitis BMS

S. mitis 272

S. mitis ATCC9811

Figure 2.1. Electrophoretical mobilities in KCl solution (pH 5) of nine S. mitis strains as a function of ionic strength. The solid lines represent the best fit to Eq. 2.1.

zeN (106C m-3)-5.0 -4.5 -4.0 -3.5 -3.0 -2.5 -2.0 -1.5 -1.0 -0.5

Elec

troph

oret

ic so

ftnes

s (nm

)

0.0

0.5

1.0

1.5

2.0

2.5

3.0

S. mitis BAS. mitis 272S. mitis 398S. mitis 244S. mitis T9S. mitis ATCC9811S. mitis 357S. mitis BMSS. mitis ATCC33399

Figure 2.2. The electrophoretic softness as a function of the fixed charge density for nine S. mitis strains (optimal values for these data were derived from Table 2.1).


Table 2.1. The cell surface softness 1/λ and charge density ρfix of nine S. mitis strains, as derived from a soft-particle analysis of electrophoretic mobilities measured in KCl over different ionic strength ranges, together with the quality of the fit, expressed as r2. Values associated with the best fit are marked in bold. All experiments were carried out in five-fold on separate bacterial cultures.

S. mitis strains

Ionic strength range (mM)

1/λ

(nm)

ρfix

(106 C m-3)a

r2

BA 10 – 100 2.5 -1.2 0.95 15 – 150 1.9 -1.8 0.92 10 – 150 2.2 -1.4 0.92 272 10 – 100 2.1 -1.6 0.99 15 – 150 1.9 -2.0 0.97 10 – 150 2.0 -1.8 0.98 398 10 – 100 2 .0 -2.7 0.95 15 – 150 1.2 -4.4 0.88 10 – 150 1.5 -3.3 0.88 244 10 – 100 1.8 -2.3 0.94 15 – 150 1.3 -3.2 0.82 10 – 150 1.5 -2.8 0.90 T9 10 – 100 1.7 -2.2 0.97 15 – 150 -b - - 10 – 150 - - -

ATCC9811 10 – 100 1.7 -2.2 0.95 15 – 150 1.2 -3.0 0.92 10 – 150 1.4 -2.5 0.93 357 10 – 100 1 .6 -2.7 0.90 15 – 150 1.0 -4.3 0.92 10 – 150 1.3 -3.2 0.90

BMS 10 – 100 1.6 -2.7 0.97 15 – 150 1.2 -3.9 0.98 10 – 150 1.4 -3.1 0.96

ATCC33399 10 – 100 1.5 -2.6 0.97 15 –150 1.1 -3.5 0.99 10 – 150 1.3 -2.8 0.97

a In the original Ohshima analysis, fixed charge densities are expressed as C m-3. The conversion between C m-3 and mM can be done dividing by the electric unit charge, Avogadro’s constant and a factor of 103

b No convergence of the fit


17

similar studies on other bacterial strains are given. The cell surface morphologies of these strains together with that of S. mitis are schematically depicted in Figure 2.3. The S. salivarius HB, has a dense pack of fibrils on the surface with lengths up to 178 nm (Van der Mei et al., 1987) and yields an electrophoretic softness of 1.4 nm, and a fixed charge density of -1.3×106 C m-3. On the other hand, a completely bald variant, devoid of any fibrillar surface appendages, had a softness of 0.7 nm and a fixed charge density of -1.5×106 C m-3 (Bos et al., 1998). Encapsulated staphylococci with a thick, contiguous polysaccharide-rich coat also presented themselves as electrophoretically soft when grown in liquid cultures (1.7 nm) and on solid agars (2.8 nm), while their fixed charge densities were relatively high, i.e. between -2.3×106 and -2.7×106 C m-3 (Kiers et al., 2001).

The bald variant of S. salivarius appears to be by far the least soft, as expected. The other strains show comparable softness in spite of different surface morphologies. It indicates that the Ohshima theory may be applied to indicate cell surface softness, but that it does not provide information on the surface morphology underlying the softness. Remarkably enough, the resistance to electro-osmotic flow, determining the electrophoretic softness, appears to be hardly affected by the density of the fibrillar layer and, moreover, is comparable to that in slime coatings. In this respect it should be realized that, in view of the values for the softness parameter, electro-osmotic flow occurs only through the very outer part (a few nm) of the extracellular layer that extends over a thickness of hundred(s) of nm. Hence, it seems that a tenuous distribution of extracellular structures, as is the case for the sparsely fibrillated S. mitis, suffices to immobilize the solvent flow.

The differences in fixed charge densities obtained for the different morphologies are even more difficult to understand. In particular, the finding that the fixed charge densities in the outer region of the sparsely fibrillated S. mitis is about twice as high as in the densely packed fibrillar coating of S. salivarius is highly unexpected. It suggests that the charge density in the fibrils of S. mitis is much higher. Assuming a

b c d a Figure 2.3. Schematic presentation of four different bacterial cell surface structures. (a) S. mitis, with long and sparsely distributed fibrils. (b) S. salivarius HB showing a dense pack of fibrils with different lengths. (c) S. salivarius HBC12, a completely bald, non-fibrillated organism. (d) Staphylococci surrounded by a thick slime layer.


value of 0.20 nm for the radius of ionic groups in the polyelectrolyte, a fixed charge density of -2.5×106 C m-3 (being the average derived for the S. mitis strains) corresponds to a volume fraction of only 5×10-2 % in the soft part of the layer taken by ionic groups. Considering that the electrophoretic softness analysis reveals the charge density at the outer periphery of the extracellular layer, it would be instructive to compare that charge density with the charge density in the cell wall. The latter may be obtained by proton titration of whole cell walls. The charge densities reported for whole cell walls is typically three to four times higher (Van der Wal et al., 1997; Carstensen & Marquis, 1968) than those derived from the softness analysis. The difference may be explained by the much higher density of the bacterial cell wall as compared to the extracellular layer.

Summarizing, we have demonstrated that S. mitis cell surfaces are electrophoretically soft. A comparison with surfaces of other bacterial strains that are reported to be soft as well reveals that the soft particle model as forwarded by Ohshima is not conclusive with respect to the surface structures causing the softness. The most probable reason is that for the various morphologies, in our case sparsely or densely packed fibrils or porous slimy coating, the electro-osmotic flow occurs only in the very outer region of a many-fold thicker extracellular surface layer. Nevertheless, irrespective of the surface morphology, establishing the surface softness is essential for the evaluation of the bacterial ζ-potential. Knowledge of that quantity is a necessity for quantitative assessment of electrostatic interaction in bacterial adhesion and aggregation.

3 A comparison of AFM interaction forces between bacteria and silicon nitride substrata for three commonly used immobilization methods Virginia Vadillo Rodríguez, Henk J. Busscher, Willem Norde, Joop de Vries, René J.B. Dijkstra, Ietse Stokroos and Henny C. van der Mei *

3.1 INTRODUCTION During the last years Atomic Force Microscopy (AFM) has been increasingly used in biosciences (Hamers, 1996; Bottomley et al., 1996). Theoretically, it combines the two most important aspects for studying structure-function relationships of biological specimens: high-resolution imaging with high signal-to-noise ratio in the molecular/sub-molecular scale and the ability to operate in aqueous environments,

* Accepted for publication in Applied and Environmental Microbiology


allowing the observation of dynamic molecular events in real-time and under physiological conditions. The AFM is surprisingly simple in its concept. A sharp tip located at the free end of a flexible cantilever scans over a surface. Interaction forces between the tip and the sample surface subsequently cause the cantilever to deflect. The deflection signal is acquired and digitized to provide a three-dimensional image of the surface.

Several biological specimens have been imaged, achieving a lateral and vertical resolution in the nanometer and sub-nanometer scale, respectively (Bustamante & Rivetti, 1996; Shao et al., 1996; Engel et al., 1997). However, when imaging living microbial cell surfaces, the softness of the cell surface together with the high pressure over the contact area between the tip and the cell can prevent high resolution imaging. The image contrast is indeed influenced by the probe geometry, the imaging parameters, the surface topography, the visco-elastic and physico-chemical properties of the cell surface. Additional problems arise from friction and lateral displacement of the organism under study, which makes immobilization strategies critical.

Beyond being an imaging device, AFM has evolved as an instrument for measuring molecular interaction forces (Rief et al., 1997a; Rief et al., 1997b). Biological interactions that have been investigated include antibody-antigen recognition, protein-ligand binding and complementary DNA base pairing (Browing-Kelly et al., 1997; Boland & Ratner, 1995; Chilkoti et al., 1995; Florin et al., 1994; Lee et al., 1994). It was further shown that AFM can be applied to measure interaction forces between bacteria and a substratum surface, including the contribution of bacterial polysaccharides to bacterium-surface interactions (Razatos et al., 1998; Camesano & Logan, 2000). AFM has also been used to characterize under aqueous conditions the supra-molecular organization of bacterial extracellular polymeric substances (EPS) adsorbed onto solid substrata (Van der Aa & Dufrêne, 2002). Moreover, AFM allowed calculation of the Young´s modulus of the dried sheath of the archaebacterium Methanospiriullum hungatei. Measurements of the rigidity of the bacteria by AFM technique enabled the determination of the bacterial turgor pressure (Arnoldi et al., 2000).

Immobilization of the organisms is not only critical when using the AFM as an imaging device, but also when it is used as a force probe. In order to probe the structure, function, physico-chemical and mechanical properties of bacterial cell surfaces under physiological conditions, it is required that immobilization does not affect the chemical and structural integrity of the cell surface. Yet, organisms under study need to be firmly anchored in order to withstand the lateral forces of the scanning tip. Different approaches have been used for bacterial immobilization in AFM. For instance, poly-L-lysine or poly(ethyleneimide) (PEI) can be used to create positively charged glass surfaces promoting irreversible adhesion of bacteria (Bolshakova et al., 2001; Velegol & Logan, 2002). Glass slides have also been treated with aminosilanes to immobilize bacteria through cross-linking carboxyl groups on their surfaces with amine groups coupled to the glass (Camesano et al., 2000). Razatos et al. (1998) developed a procedure for coating the silicon nitride

Bacterial immobilization for AFM 21

AFM tip with a confluent layer of bacteria in which a drop of glutaraldehyde-treated bacterial suspension is placed on a PEI-coated tip (Razatos et al., 1998). Occasionally, organisms have been immobilized by mechanical trapping on membrane filters with a pore size chosen slightly smaller than the dimensions of the bacterium (Boonaert et al., 2002; Dufrêne, 2000; Van der Mei et al., 2000a). Rarely, minute glass beads functionalized with amino groups were coated with bacteria and linked to the silicon nitride cantilever using a small amount of epoxy resin (Lower et al., 2000). Other methods for the immobilization of bacteria on surfaces for AFM may exist, but it is considered beyond the scope of this paper to give a comprehensive list of sample preparation techniques. Different immobilization strategies, however, are likely to yield different results in AFM as not all methods equally well preserve the integrity of the immobilized cells.

Therefore, this chapter compares the interaction forces obtained between Klebsiella terrigena and the silicon nitride tip of an AFM for three immobilization methods: a) mechanical trapping, b) adsorption to positively charged glass, c) fixation to the tip. 3.2 MATERIALS AND METHODS

3.2.1 Bacterial strain, growth conditions and harvesting The Gram-negative strain Klebsiella terrigena ATCC33527, occurring commonly in soil, water, grain, fruits and vegetable, was used in this study. K. terrigena was grown aerobically in Nutrient broth (OXOID, Basingstoke, UK) at 37 ºC. For each experiment, the strain was inoculated from Nutrient agar in a batch culture. This culture was used to inoculate a second culture that was grown for 16 h prior harvesting. Bacteria were harvested by centrifugation (5 min at 10,000 g), washed twice with demineralized water and resuspended in water or in 0.25 mM potassium phosphate buffer at pH 6.8. 3.2.2 Sample preparation Immobilization of K. terrigena was carried out by three different methods: (a) Bacterial cells were suspended in water to a concentration of 105 per ml after

which 10 ml of this suspension was filtered through an Isopore polycarbonate membrane (Millipore) with pore size of 0.8 µm, i.e. slightly smaller that the bacterial dimensions, to immobilize the bacteria through mechanical trapping (Kasas & Ikai, 1995). Filtration was carried out by placing a filter on a vacuum filtration flask, after which bacteria were added to the top of the filter and a vacuum was applied for approximately 10s. After filtering, the filter was carefully fixed with double-sticky tape on a glass slide and transferred to the AFM.

(b) Bacteria were also attached through electrostatic interactions (physical adsorption) to a glass slide, made positively charged through adsorption of poly-L-lysine hydrobromide. In order to coat a glass surface with poly-L-lysine hydrobromide, glass was first cleaned by sonicating for 2 min in 2 % RBS35 surfactant solution in water (Omnilabo International BV, The Netherlands), rinsed thoroughly with tap water, dipped in methanol, and again rinsed with


demineralized water, after which a drop of 0.01 % (w/v) poly-L-lysine hydrobromide solution was added. After air-drying, the slide was rinsed with demineralized water and dipped into the water bacterial suspension. After 15 min, the bacteria-coated slide was rinsed with demineralized water to remove loosely attached bacteria and transferred to the AFM (Camesano et al., 2000).

(c) Finally, bacteria were immobilized through glutaraldehyde fixation onto the silicon nitride tip of the AFM. This method requires pretreatment of both, the K. terrigena cells and the AFM cantilevers (Park Scientific Instruments, Mountain View, CA). Bacteria were first treated with 2.5 % v/v glutaraldehyde solution (pH adjusted to 6.8) for 2.5 h at 4 °C. After glutaraldehyde fixation, bacteria were washed in 0.25 mM potassium phosphate solution and pelleted by centrifugation at 10,000 g for 5 min. To prepare AFM cantilevers a drop of 1% v/v poly(ethyleneimide) solution was adsorbed onto the cantilevers for 2.5 h. The cantilevers were subsequently rinsed in demineralized water and stored at 4°C. A bacterial pellet was manually transferred onto the PEI-coated silicon nitride tip employing a micromanipulator, while viewing the procedure under an optical microscope. The bacteria-covered tip was further treated with a drop of glutaraldehyde (2.5 % v/v) at 4 °C to strengthen and anchor the pellet onto the tip. After incubation for 1 to 2 h, the cantilevers were rinsed in demineralized water and transferred to the AFM (Razatos et al., 1998).

3.2.3 Atomic force microscopy AFM measurements were made at room temperature under 0.25 mM potassium phosphate solution at pH 6.8, using an optical level microscope (Nanoscope III Digital Instrument). “V”-shaped silicon nitride cantilevers from Park Scientific Instruments (Mountain View, CA) with a spring constant of 0.06 N m-1 and a probe curvature of ∼ 50 nm were used. Individual force curves were collected over the top of trapped and physically adsorbed bacteria on randomly selected locations with z-displacements of 100 – 200 nm at z-scan rates ≤ 1 Hz. Similarly, force curves were collected between the bacteria-coated AFM tip and silicon nitride sheets (Onstream, The Netherlands). The slope of the retraction force curves in the region where probe and sample are in contact were used to convert the voltage into cantilever deflection. The conversion of deflection into force was carried out as has been previously described by others (Dufrêne, 2000).

Approach curves were fitted to an exponential function, where the interaction force F is described as F=F0 exp (-d/λ), where F0 is the force at zero separation distance between the interacting surfaces, d the separation distance and λ the decay length of the interaction force F. Retraction curves only showed a single adhesion peak in a number of cases. The percentage occurrence of an adhesion peak, its magnitude as well as the distance at which the adhesion peak appeared were recorded and averaged.

Results represent the average of at least 150 force-distance curves taken over 5 to 10 different organisms while measuring on 10 different locations per organism.

a)

b)

c)

Sepa

ratio

n di

stan

ce (n

m)

020

040

060

080

010

0012

00Se

para

tion

dist

ance

(nm

)0

200

400

600

800

1000

1200

Force (nN) 05101520

Se

para

tion

dista

nce

(nm

)0

200

400

600

800

1000

1200

Force (nN) -30

-20

-1001020

Force (nN) 05101520

Figu

re 3

.1. A

FM d

efle

ctio

n im

ages

of K

. ter

rige

na A

TCC

3352

7 tra

pped

in a

n Is

opor

e po

lyca

rbon

ate

mem

bran

e (im

age

size

2 µ

m×2

µm

; z-r

ange

: 75

nm) (

a) a

nd p

hysi

cally

ads

orbe

d on

pol

y-L-

lysi

ne c

oate

d gl

ass (

imag

e si

ze 3

µm

×3 µ

m; z

-ran

ge: 1

50 n

m) (

b) to

geth

er w

ith a

scan

ning

mic

rogr

aph

of

an A

FM ti

p co

ated

with

K. t

erri

gena

cel

ls (c

). Ex

ampl

es o

f for

ce-d

ista

nce

curv

es b

etw

een

K. t

erri

gena

and

silic

on n

itrid

e fo

r eac

h im

mob

iliza

tion

met

hod

are

pres

ente

d be

low

the

corr

espo

ndin

g im

ages

. Sol

id li

nes r

epre

sent

the

appr

oach

cur

ve, w

hile

the

dash

ed li

nes i

ndic

ate

the

retra

ctio

n cu

rve.


3.3 RESULTS Figure 3.1 shows AFM deflection images of K. terrigena immobilized by mechanical trapping in an Isopore polycarbonate membrane (Figure 3.1a), attached through electrostatic interactions on a positively charged glass slide (Figure 3.1b) as well as scanning electron micrograph of bacteria immobilized onto the silicon nitride AFM tip (Figure 3.1c). Examples of force-distance curves performed over the top of a trapped and physical adsorbed bacterium as well as between the bacteria-coated tip and a silicon nitride sheet are presented below the corresponding images.

At first sight, similarities between force-distance curves associated with mechanically trapped and physically adsorbed bacteria (Figures 3.1a and 3.1b) can be observed. Upon approach, a long range (about 500 – 800 nm) repulsive force is encountered, while in the examples given no adhesion is recorded upon retraction of the tip from the bacterial cell surface. In contrast, as the bacteria-coated tip approached a silicon nitride sheet, repulsion began at much shorter separation distance (about 15 nm) and a single adhesion peak is always present upon retraction.

Quantitative features of the force-distance curves have been summarized in Table 3.1. It is remarkably that intimate contact between the interacting surfaces is achieved for an applied force F0 that varies from 2.6 to 12 nN depending on the bacterial immobilization method used, whereas the decay length λ of the repulsive force upon approach ranged from 2.0 to 111 nm. As the AFM tip was retracted, adhesion forces were found in 15 % to 13 % of all the cases with an average attractive force of -0.26 and -0.5 nN at separation distances of 60 and 102 nm for mechanically trapped and physically adsorbed bacteria, respectively. However, retraction of bacteria immobilized onto an AFM tip away from a silicon nitride sheet always showed adhesion with an average attractive force of -35 nN at 78 nm separation distance. 3.4 DISCUSSION A proper interpretation of the force-distance curves between interacting surfaces in AFM requires bacterial immobilization that fully preserves the chemical and structural integrity of the cell surface. In this paper, the interaction forces between K. terrigena and silicon nitride are compared for three immobilization methods. Force-distance curves were different when bacteria were attached through fixation to the tip (Figure 3.1c) from those obtained for mechanically trapped or physically adsorbed bacteria (Figure 3.1a and 3.1b). For the latter two, qualitative similarities were found in force-distance curves, although for a bacterium immobilized through attachment to poly-L-lysine-treated glass stronger repulsive forces, occurring at larger separation distances, were measured upon approach (Table 3.1).

Mechanical trapping a single bacterium in a membrane filter with a pore size comparable with the dimensions of the cell does not require any chemical treatment or surface modification and the highest part of a trapped organism protrudes through the holes of the filter. Therewith, it can be easily probed with an AFM


Table 3.1. Characteristics of force-distance curves between K. terrigena ATCC33527 and silicon nitride for three different bacterial immobilization methods. F0 is the repulsive force at zero separation distance and λ the decay length of this repulsive force upon approach, while Fadh is the average adhesion force recorded upon retraction, together with the separation distance Dadh at which the adhesion force occurred. The percentage of force-distance curves for which adhesion upon retraction occurred is given, since not all force-distance curves showed adhesion upon retraction. All data are average values ± SD of 150 force-distance curves, taken over 5 to 10 different organisms while measuring on 10 different locations per organism.

Mechanical trapping Physical adsorption Bacteria-coated AFM tip

F0 (nN) 2.6 ± 1.7 12 ± 4 3.7 ± 0.5

λ (nm) 59 ± 52 111 ± 57 2.0 ± 0.5

% adhesion 15 13 100

Fadh (nN) -0.26 ± 0.05 -0.5 ± 0.2 -35 ± 2

Dadh (nm) 60 ± 8 102 ± 35 78 ± 13

under physiological conditions. In contrast, physical adsorption on a positively charged surface may stimulate the secretion of excess EPS by K. terrigena. The surface of K. terrigena adsorbed on a positively charged surface (Figure 3.1b) shows a similar morphology to the EPS substances previously scanned by Van der Aa & Dufrêne (2002). The surface presents stretchable coil-like structures which were depending on the scanning direction. A thicker negatively charged and highly hydrated EPS layer could account for the higher repulsion forces operating over larger distances observed upon approach of the AFM tip to such physically adsorbed bacteria. This is in line with observations by Razatos et al. (1998), reporting that an Escherichia coli mutant overproducing colanic acid in buffer, experienced greater repulsion upon approach of the AFM tip than the parent strain, which was attributed to the higher negative charge density of the capsular material produced.

The interaction force between a bacteria-coated AFM tip (Figure 3.1c) and silicon nitride sheets yields qualitatively and quantitatively distinct force-distance curves. Most notably upon approach, the distance over which repulsion is probed, is significantly reduced compared to both other methods. In addition, retraction of the bacteria-coated tip from the silicon nitride sheet always showed adhesion, whereas for the other two immobilization methods very weak adhesion forces (less than -0.5 nN) upon retraction was observed only in 13 % to 15 % of all force-distance curves recorded. The differences on adhesion forces could be readily attributed to the larger contact area probed by a bacteria-coated AFM tip. Assuming that five bacteria interact with the silicon nitride substratum an average adhesion force per bacterium of 7 nN can be calculated. The contact area for the other two immobilization method is


generally estimated based on the effective AFM tip radius thought to be ~250 nm. Therefore, average adhesion forces per bacterium of ~1 and 2 nN are found for mechanically trapped or physically adsorbed bacteria.

We envisage, that glutaraldehyde fixation of bacteria to a tip, stiffens the bacterium by cross-linking proteins and amino acids in the peptidoglycan layer with an impact on its adhesive properties. It has been found, for instance, that glutaraldehyde fixation caused yeast cells to become more hydrophobic (Bowen et al., 2001). However, Razatos et al. (1998) argued that glutaraldehyde treatment did not affect the adhesive properties of E. coli strains, because both contact angle and zeta potentials before and after glutaraldehyde treatment remained unchanged. In contrast, Burks et al. (2003) found that adhesion of these E. coli strains to glass was affected by glutaraldehyde treatment. Furthermore, AFM-based results showed that the addition of glutaraldehyde consistently increased the rigidity of the E. coli strains studied.

Table 3.2 summarizes the advantages and disadvantages of each AFM immobilization method evaluated in this study. In general, trapping bacterial cells in filters guarantees the physical and chemical integrity of the bacterial cell surface, whereas the applied vacuum needed to pull the cells into the holes could induce changes on mechanical cell surface properties. Also, adsorption of living cells onto positively charged surfaces may promote structural rearrangements in bacterial cell surface structure. Glutaraldehyde fixation of bacteria to the AFM tip clearly affects the chemical and structural integrity of the bacterial cell surface with a major impact on the interaction forces probed by AFM. Furthermore, complete coverage of the AFM tip by bacterial cells constitutes another problem. It is our experience that in three out of five cases the coverage is incomplete. See Figure 3.2.

In conclusion, the results from this study indicate that different immobilization methods of bacteria in AFM affect the qualitative and quantitative features of the force-distance curves between K. terrigena and silicon nitride. Mechanical trapping of single cells in a membrane filter is inferred to be the most suitable technique, as the two other methods evaluated change the chemical and structural integrity of the bacterial cell surface.

Figure 3.2. Electron micrographs of AFM silicon nitride tips after being coated with K. terrigena bacteria according to the procedure developed by Razatos et al. (1998). Bar marker represents 1 µm.


Dis

adva

ntag

es

♦ R

od-s

hape

d ba

cter

ia a

re d

iffic

ult t

o tra

p ♦

.Cel

ls m

ay b

e co

mpr

esse

d as

a re

sult

of a

pplie

d va

cuum

♦

EPS

may

acc

umul

ate

on th

e to

p of

a tr

appe

d ba

cter

ium

♦ C

hem

ical

trea

tmen

t of t

he su

bstra

tum

requ

ired

♦.P

hysi

co-c

hem

ical

pro

perti

es o

f bac

teria

l cel

l sur

face

pos

sibl

y af

fect

ed b

y th

e su

rfac

e m

odifi

catio

n ♦

.Imm

obili

zatio

n no

t alw

ays a

dequ

ate

for d

iffer

ent s

train

s

(Cam

esan

o et

al.,

200

0)

♦ L

ong

and

diff

icul

t pre

para

tion

proc

edur

e ♦

.Req

uire

s che

mic

al tr

eatm

ent o

f bot

h ba

cter

ial c

ell a

nd

su

bstra

tum

surf

ace

♦.P

hysi

co-c

hem

ical

and

mec

hani

cal p

rope

rties

of t

he

b

acte

rial c

ell s

urfa

ce c

hang

es

♦ T

he n

umbe

r of c

ells

inte

ract

ing

is u

nkno

wn,

as w

ell a

s th

eir

sp

atia

l orie

ntat

ion

whe

n in

tera

ctin

g w

ith th

e su

bstra

tum

♦.B

acte

ria-c

oate

d tip

s nee

d to

be

chec

ked

regu

larly

for

fu

ll co

vera

ge u

sing

ele

ctro

n m

icro

scop

y

Adva

ntag

es

♦ S

impl

e pr

epar

atio

n

♦ N

o ch

emic

al p

re-tr

eatm

ent o

f eith

er th

e tip

or b

acte

ria

♦

.Phy

sico

-che

mic

al p

rope

rties

of b

acte

rial c

ell

re

mai

n un

chan

ged

♦.E

xact

pos

ition

ing

of th

e tip

on

the

bact

eria

l cel

l

surf

ace

♦.C

onta

ct a

rea

can

be e

stim

ated

bas

ed o

n th

e

dim

ensi

ons o

f the

tip

♦ S

impl

e pr

epar

atio

n ♦

.Bac

teria

of d

iffer

ent s

hape

s and

dim

ensi

ons c

an b

e

stud

ied

♦.E

xact

pos

ition

ing

of th

e tip

on

the

bact

eria

l cel

l

surf

ace

♦.C

onta

ct a

rea

can

be e

stim

ated

bas

ed o

n th

e

dim

ensi

ons o

f the

tip

♦

Ver

satil

e ch

oice

of s

ubst

ratu

m

Mec

hani

cal t

rapp

ing

Phys

ical

ads

orpt

ion

Bac

teria

-coa

ted

tip

T

able

3.2

. Sum

mar

y of

adv

anta

ges

and

disa

dvan

tage

s as

soci

ated

to b

acte

rial i

mm

obili

zatio

n m

etho

d em

ploy

ed in

this

stud

y, i.

e. m

echa

nica

l tra

ppin

g of

bac

teria

in m

embr

ane

filte

rs, p

hysi

cal a

dsor

ptio

n of

bac

teria

to a

pos

itive

ly c

harg

ed su

rfac

e an

d ce

ll fix

atio

n to

the

tip u

sing

glu

tara

ldeh

yde.

4 On relations between microscopic and macroscopic physico-chemical properties of bacterial cell surfaces – an AFM study on Streptococcus mitis strains

Virginia Vadillo Rodríguez, Henk J. Busscher, Willem Norde , Joop de Vries and Henny C. van der Mei*

4.1 INTRODUCTION The bacterial cell wall is a critical structure, allowing organisms to selectively interact with their environment. Moreover, the cell wall imparts shape to the organisms, supports internal turgor pressure and acts as a selective barrier for

* Reproduced with permission of the American Chemical Society from Langmuir (2003) 19: 2372-2377


nutrients and metabolites. Chemically, the cell wall is constituted of a variety of spatially organized molecular structures, each with a specific function. All biological membranes, including those of the bacterial cell wall, are composed of a thin film of (glyco)lipids and proteins, held together mainly by non-covalent interactions. The composition of the outer cell wall varies considerably between Gram-positive and Gram-negative bacteria and may contain a variety of external structures, such as S-layers, capsules, fibrils, fimbriae and pili (Hancock, 1991). The combination of these structures determines the physico-chemical cell surface properties of a particular bacterial strain. These properties, however, are not fixed in space or time, but vary with environmental changes and as a result of mutation and various mechanisms of gene transfer between organisms (Savage & Fletcher, 1985).

Bacterial cell surface hydrophobicity and charge are commonly accepted as influential on bacterial interactions with their environment, but a generalized physico-chemical theory accounting for bacterial adhesion to substratum surfaces is still lacking (Bos et al., 1999). Cell surface hydrophobicity can be inferred from water contact angles on bacterial lawns (Busscher et al., 1994). For instance, water contact angles on oral streptococci can be as high as 103 degrees for Streptococcus mitis strains, but as low as 19 degrees for Streptococcus rattus (Van der Mei et al., 1998a). Extensive evaluations of a large number of bacterial strains and species have demonstrated that cell surface hydrophobicity is conveyed mostly to the cell surface through nitrogen containing groups, especially for oral streptococci (Van der Mei et al., 2000b).

Nearly all bacterial cell surfaces found in nature are negatively charged (Wilson et al., 2001). Hence, from an overall physico-chemical point of view, they are expected to be repelled by negatively charged substratum surfaces. However, Hayashi et al. (2001) and Poortinga et al. (2001) have both described that this electrostatic repulsion is often overestimated due to the neglect of bacterial cell surface softness. Soft, ion-penetrable cell surfaces experience less electrostatic repulsion than similarly charged, hard, ion-impenetrable surfaces, since their diffuse double layer charges are driven into the ion-penetrable cell walls causing an effective decrease in surface potential and, hence, electrostatic repulsion. Recently, Morisaki et al. (1999) explained adhesion of a negatively charged marine bacterium, Vibrio alginolyticus, onto a negatively charged substratum by considering the softness of the strains.

Although physico-chemical approaches based on overall cell surface hydrophobicity and charge density have explained adhesion to substrata of many bacterial strains and species, generalization is still impossible. Probably, this must be attributed in part to the fact that in such physico-chemical approaches bacterial cell surfaces have been considered as smooth, rigid and chemically homogeneous. Contact angles and surface charge densities are both macroscopic properties reflecting the cell surface chemistry and structure. Calculation of the interaction energy between a negatively charged bacterial cell surface and a negatively charged substratum surface using the DLVO-theory showed at low electrolyte concentration

– an AFM study on Streptococcus mitis strains 31

a shallow secondary minimum and a large potential energy barrier. If it was accounted for that only a minor number of positively charged sites existed on the cell surface, the energy barrier disappeared and highly adhesive conditions were revealed (Van Loosdrecht et al., 1990). In fact, the interaction between spirochetes and human erythrocytes has been described in terms of interactions between oppositely charged sites on the interacting surfaces, although both surfaces carry an overall negative charge on a macroscopic scale (Cowan et al., 1994). Obviously, a minor number of positively charged sites, while instrumental for adhesion, does hardly affect the macroscopic cell surface charge density.

Although macroscopic cell surface characterization adds to our understanding of bacterial adhesion, it has provided relatively little information at the molecular level of the forces governing the adhesion process. The introduction of the atomic force microscope and its application to biological surfaces (Dufrêne, 2000; Meagher & Griesser, 2002; Kidoaki & Matsuda, 2002) has offered new possibilities to obtain microscopic physico-chemical properties of bacterial cell surfaces. Recently, the turgor pressure of a spherical bacterium, Enterococcus hirae, in deionized water was derived from the indentation depth caused by an AFM tip and found to be between 4 and 6×105 Pa (Yao et al., 2002). Razatos et al. (1998) showed that the adhesion force between a silicon nitride AFM tip and Escherichia coli was affected by the length of lipopolysaccharide molecules on the cell surface and by the production of a capsular polysaccharide. Furthermore, it was discovered using AFM that E. coli JM109 and K12J62 have different surface morphologies dependent on environmental conditions, while lysozyme treatment led to the loss of surface rigidity and eventually to dramatic changes in bacterial shape (Bolshakova, et al., 2001).

Streptococcus mitis is a Gram-positive oral bacterial strain, that belongs to the primary colonizers of dental hard tissues and also adheres to mucous membranes,

Table 4.1. Macroscopic, physico-chemical cell surface properties of nine S. mitis strains, including their electrophoretic cell surface softness 1/λ and fixed charge density ρ (Vadillo-Rodríguez et al., 2002), their cell surface hydrophobicity by water contact angle θw (Van der Mei & Busscher, 1996) and elemental surface concentration ratio N/C by X-ray photoelectron spectroscopy (Van der Mei et al., 2000b).

S.mitis strains 1/λ (nm) ρ (106C m-3) θw (degrees) N/C ATCC9811 1.7 -2.2 68 0.110

ATCC33399 1.1 -3.5 56 0.106 244 1.8 -2.3 60 0.097 272 2.1 -1.6 54 0.131 357 1.0 -4.3 53 0.119 398 2.0 -2.7 59 0.116

BMS 1.2 -3.9 100 0.124 BA 2.5 -1.2 103 0.129 T9 1.7 -2.2 91 0.125


most notably the cheek and the tongue (Marsh & Martin, 1992). By comparison with other oral streptococci, S. mitis strains usually carry sparsely distributed but long fibrils (Cowan et al., 1992). A collection of nine S. mitis strains has been macroscopically characterized with regard to their cell surface hydrophobicities by water contact angles (Van der Mei & Busscher, 1996), and its surface charge properties using the soft-layer model as proposed by Ohshima (Vadillo-Rodríguez et al., 2002). X-ray photoelectron spectroscopy has been carried out to determine the chemical composition of the cell surface (Van der Mei et al., 2000b). A summary of the results obtained from these studies is presented in Table 4.1.

The aim of this chapter is to microscopically characterize the cell surfaces of these strains by atomic force microscopy and to find out if and how microscopic cell surface properties can be amalgamated into the macroscopic cell surface properties, previously determined. 4.2 MATERIALS AND METHODS 4.2.1 Bacterial strains, growth conditions and harvesting The nine S. mitis strains used in this study were all cultured in Todd Hewitt Broth (Oxoid, Basingstoke, UK). For each experiment, the strains were inoculated from blood agar in a batch culture. This culture was grown overnight and used to inoculate a second culture that was grown for 16 h in ambient air prior to harvesting. Bacteria were harvested by centrifugation (5 min at 10,000 g), washed twice with demineralized water and resuspended in water. To break up bacterial chains, cells were sonicated for 30 s at 30 W (Vibra Cell model 375, Sonics and Materials Inc., Danbury, CT, USA). Sonication was done intermittently while cooling in an ice/water bath. These conditions were found not to cause cell lysis in any strain. 4.2.2 Atomic force microscope Sample preparation. Bacterial cells were suspended in water to a concentration of 105 per ml after which 10 ml of the suspension was filtered through an Isopore polycarbonate membrane (Millipore) with a pore size of 0.8 µm (Kasas & Ikai, 1995). The pore size is chosen slightly smaller than the streptococcal dimensions to immobilize the bacteria by mechanical trapping. After filtration, the filter was carefully fixed with double-sided sticky tape on a sample glass and transferred to the AFM. AFM imaging and force-distance measurements. A Nanoscope III AFM (Digital Instruments, Santa Barbara, CA) operating in contact mode was used to image cells and to measure interaction forces. Measurements were taken at room temperature, either in deionized water or in a 0.1M KCl solution. “V”-shaped silicon nitride cantilevers from Park Scientific Instuments (Mountain View, CA) with a spring constant of 0.06 N m-1 and a probe curvature radius of ~50 nm, according to manufacturer specifications, were used.

Contact mode topographic images in deflection, height and friction modes were taken with an applied force maintained below 1 nN at a scan rate of ~2 Hz. Integral and proportional gains of the feedback loop were about two and three, respectively.


Individual force curves with z-displacements of 100 – 200 nm were collected at z-scan rates ≤ 1 Hz. The slope of the retraction force vs. distance curves in the region where probe and sample are in contact was used to convert the position sensitive detector voltage into cantilever deflection. The direct result of such a force measurement is the cantilever deflection d vs. position of the piezo z normal to the surface. To obtain a force-distance curve, d and z have to be converted into force and distance. The force F is obtained by multiplying the deflection d of the cantilever with its spring constant k, while the tip-sample separation D is calculated by subtracting the deflection d from the position z of the piezo.

Topographic images were recorded for at least ten bacterial cells of each S. mitis strain (either in deionized water or in a 0.1M KCl solution). To this end, the tip was positioned over the top of a trapped bacterium, scanning was stopped and ten force measurements were performed at randomly selected locations around the top for each bacterial cell studied. The selected force distance curves (see Figure 4.1 for an example) were analyzed in order to determine various characteristic parameters. First, the approach part of a force curve was fitted to a negative exponential F = F0 exp–D/Λ (4.1) in which F is the measured force at separation distance D, F0 is the force at zero separation distance and Λ is the characteristic decay length (separation distance over which F decays from F0 to F0/e). Secondly, the retraction part of a force curve usually showed various local attractive maxima. The magnitude of the adhesion forces, represented by these maxima, and the distances at which they occurred were quantitatively registered in histograms (see Figure 4.2 for an example), as well as the minimum Dmin and maximum Dmax, distances at which local maxima in adhesion forces occurred. Based on the relative prevalence of the local maxima in adhesion forces (expressed in percentages in the histograms, Figure 4.2), an average attractive force Fadh between the AFM tip and the cell surface of each S. mitis strain was calculated too. Forces with prevalence less than 2 % were neglected in this averaging process. 4.2.3 Statistical analysis In order to determine possible relationships between the microscopic properties obtained by force curve analysis and the macroscopic physico-chemical surface properties of the S. mitis strains, both sets of properties were submitted to a Pearson correlation test. Pearson’s correlation coefficient r reflects the strength and direction of the linear relationship between two variables. The outcome ranges from +1 to -1, where -1 is a perfect (inverse) correlation, 0 is no correlation and +1 is a perfect positive correlation.

In addition, in order to help the reader visualize the tendencies observed between our parameters, linear regression analyses between those parameters, showing high correlations in the Pearson analysis, were performed. Note, however, that strictly speaking the Pearson correlation test is different from linear regression analysis, as it constitutes only a rank testing.


Separation distance (nm)0 75 150 225 300

F(nN

)

-3

0

3

6

9

12

Dmin Dmax


F(nN

)

-3

0

3

6

9

12

Dmin Dmax

Figure 4.1. Force-distance curves for S. mitis 357 in water (a) and 0.1 M KCl (b). The solid lines represent the approach curve, while the dashed lines indicate the retraction curve. Minimum and maximum local maxima in adhesion force during retraction are marked by arrows.

Average values

Separation distance (nm)0 50 100 150 200 250 300

F (n

N)

-2.5

-2.0

-1.5

-1.0

-0.5

0.0

36 %

46 %

18 %

Average values

Separation distance (nm)0 50 100 150 200 250 300

F (n

N)

-1.0

-0.8

-0.6

-0.4

-0.2

0.0

7 %

67 % 26 %

Figure 4.2. Force distribution histogram for S. mitis 357 in water (a) and 0.1 M KCl (b), as derived from Figure 4.1, with the relative prevalence of each force indicated. 4.3 RESULTS Figure 4.1 presents an example of the interaction force between S. mitis 357 and a silicon nitride tip immersed in water and 0.1 M KCl solution, respectively, as a function of the separation distance between the tip and the bacterium. Interestingly, the approach curve is always repulsive and does not show attraction, not even at a short separation distance. Instead, the approach curve shows a gradually increasing repulsion between the tip and the cell surface that increases exponentially at close approach until contact at F0. Repulsive forces at contact as well as decay lengths are summarized in Table 4.2 for all nine strains. Repulsive forces at contact in water can either be larger (S. mitis ATCC9811, 357 and BA) or smaller (S.mitis 244, 398 and BMS) than in 0.1 M KCl, while for the strains S. mitis ATCC33399, 272 and T9 there was no statistically different between them. The decay lengths in water (“group average” 21 ± 10 nm) tend to be larger than in 0.1 M KCl (“group average” 15 ± 13 nm) for four out of the nine strains studied.


Tabl

e 4.

2.

Sum

mar

y of

qua

ntita

tive

data

der

ived

fro

m f

orce

-dis

tanc

e cu

rves

in w

ater

and

0.1

M K

Cl o

f ni

ne S

. miti

s st

rain

s, in

clud

ing

the

forc

e at

zer

o se

para

tion

dist

ance

Fo

and

its d

ecay

leng

th Λ

(bo

th f

rom

app

roac

h cu

rves

) an

d m

inim

um D

min a

nd m

axim

um D

max

dis

tanc

es a

t whi

ch lo

cal m

axim

a in

ad

hesi

on fo

rces

are

regi

ster

ed, t

oget

her w

ith th

e nu

mbe

r of l

ocal

max

ima

N a

nd th

e av

erag

e ad

hesi

on fo

rce

F adh

cal

cula

ted

(all

from

retra

ctio

n cu

rves

).

N 2 3 4 6 3 2 9 5 11

F adh

(nN

)

-0.5

± 0

.4

-1.4

± 0

.9

-1.8

± 1

.1

-1.0

± 0

.5

-0.7

± 0

.3

-1.5

± 0

.7

-1.1

± 0

.4

-1.4

± 0

.3

-2.2

± 1

.0

Dm

ax (n

m)

167 ±

33

170 ±

30

426 ±

29

534 ±

29

196 ±

43

120 ±3

5

867 ±

40

588 ±

23

1186

± 3

8

0.1

M K

Cl

Dm

in (n

m)

78 ±

28

14 ±

4

59 ±

28

29 ±

11

13 ±

0

36 ±

14

43 ±

10

36 ±

2

53 ±

19

N 6 2 4 6 3 2 5 4 9

F adh

(nN

)

-1.3

± 0

.6

-1.3

± 0

.8

-0.9

± 0

.4

-1.3

± 0

.7

-1.3

± 0

.3

-1.8

± 1

.0

-1.5

± 0

.8

-2.9

± 1

.2

-1.4

± 0

.8

Dm

ax (n

m)

585 ±4

0

98 ±

39

280 ±3

1

977 ±2

5

235 ±2

8

167 ±3

0

459 ±3

2

642 ±2

4

910 ±3

0

Retr

actio

n

wat

er

Dm

in (n

m)

64 ±

34

21 ±

11

17 ±

2

10 ±

0

37 ±

12

38 ±

25

50 ±

16

106 ±

24

43 ±

16

Λ (n

m)

50 ±

6

9 ±

2

13 ±

3

13 ±

6

10 ±

2

8 ±

1

10 ±

3

8 ±

1

11 ±

3

0.1

M K

Cl

F 0 (n

N)

2.1 ±

0.8

3.5 ±

0.7

2.8 ±

0.7

3.6 ±

1.4

1.1 ±

0.2

7.7 ±

1.0

1.6 ±

0.8

2.7 ±

0.6

4.8 ±

2.0

Λ (n

m)

33 ±

5

12 ±

2

10 ±

2

11 ±

2

20 ±

7

11 ±

3

33 ±

11

33 ±

8

25 ±

4

Appr

oach

wat

er

F 0 (n

N)

8.0 ±

1.0

3.4 ±

1.4

1.0 ±

0.4

4.7 ±

1.8

2.9 ±

0.5

3.2 ±

0.8

4.1 ±

1.5

4.5 ±

1.2

2.7 ±

0.7

S. m

itis

ATC

C98

11

ATC

C33

399

244

272

357

398

BM

S

BA

T9


Tabl

e 4.

3. P

ears

on c

orre

latio

n co

effic

ient

s be

twee

n m

icro

scop

ic c

ell s

urfa

ce c

hara

cter

istic

s fr

om A

FM a

nd m

acro

scop

ic c

ell s

urfa

ce c

hara

cter

istic

s in

wat

er o

r in

0.1M

KC

l for

nin

e S.

miti

s stra

ins.

Cor

rela

tion

coef

ficie

nts b

elow

0.5

hav

e be

en o

mitt

ed, w

hile

cor

rela

tions

mar

ked

in b

old

have

bee

n gr

aphi

cally

pre

sent

ed in

this

cha

pter

.

ρ (1

06 Cm

-3)

- - -

0.64

- - -

0.93

1

1/λ

(nm

)

- - - -

-0.5

9

- - 1 0.93

θ w (d

egre

es)

- -

0.73

-

-0.6

3

- 1 - -

N/C

- - -

0.67

-0.5

4

1 - - -

F adh

(nN

)

- -

-0.8

5

- 1 - - - -

Dm

ax (n

m)

- - - 1

-0.5

6

0.51

0.72

- -

Dm

in (n

m)

- - 1 0.66

-0.8

0

- - - -

Λ (n

m)

- 1 - - - - - - -

F 0 (n

N)

1 - - -

-0.5

1

- - - -

corr

elat

ion

coef

ficie

nts f

or A

FM a

naly

sis i

n w

ater

F 0 (n

N)

Λ (n

m)

Dm

in (n

m)

Dm

ax (n

m)

F adh

(nN

)

N/C

θ w (d

egre

es)

1/λ

(nm

)

ρ (1

06 Cm

-3)

corr

elat

ion

coef

ficie

nts f

or A

FM a

naly

sis i

n 0.

1M K

Cl


Upon retraction, multiple local maxima in adhesion forces can be observed. The separation distances at which these adhesion forces occur, their magnitude and relative prevalence as averaged per strain can be summarized in so-called distribution force histograms (see Figure 4.2, showing the distribution force histogram for S. mitis 357). The minimal and maximal distances at which local adhesion forces occur are presented in Table 4.2, together with the average adhesion force arising from a bacterial cell surface as calculated from the relative prevalence of all local adhesion forces. Adhesion forces tend to be slightly stronger in water than in 0.1 M KCl, but not for all S. mitis involved in this study. Some strains, like S. mitis ATCC33399 and 398 possess only two maxima in local adhesion forces at small separation distances, whereas other strains like T9 express nine local adhesion forces at separation distances up to 910 nm.

In order to determine possible correlations between the microscopic cell surface characteristics revealed by the AFM force curves, and previously published macroscopic cell surface properties (see Table 4.1), Pearson’s correlation coefficients r between all parameters were calculated (Table 4.3). In water, the minimum distance Dmin at which adhesion forces occur, correlates well with the average adhesion force Fadh (r = -0.85). The maximum distance Dmax at which adhesion occurs correlates with the amount of surface nitrogen as detected by XPS (r = 0.67) and also with the amount of fixed cell surface charge determined by electrophoresis analysis (r = 0.64). Cell surface hydrophobicity by water contact angles correlates with Dmin and the average adhesion force Fadh (r = 0.73 and -0.63, respectively). Remarkably, while cell surface hydrophobicity correlates best with Dmin in water, it correlates better with Dmax in 0.1 M KCl.

Dmax (nm)0 200 400 600 800 1000 1200

N/C

0.095

0.100

0.105

0.110

0.115

0.120

0.125

0.130

0.135

Figure 4.3. Nitrogen surface concentration ratio of S. mitis strains obtained by XPS as a function of the maximum distance Dmax at which a local maximum in adhesion force was registered by the AFM tip upon retraction in water.


Dmin(nm)0 20 40 60 80 100 120

wat

er c

onta

ct a

ngle

(deg

rees

)

50

60

70

80

90

100

110

Fadh (nN)-3.0 -2.5 -2.0 -1.5 -1.0 -0.5 0.0

wat

er c

onta

ct a

ngle

(deg

rees

)

50

60

70

80

90

100

110

Dmax (nm)0 200 400 600 800 1000 1200 1400

wat

er c

onta

ct a

ngle

(deg

rees

)

50

60

70

80

90

100

110

Figure 4.4. Water contact angle on lawns of S. mitis strains as a function of (a) the minimum distance Dmin at which a local maximum in adhesion force was registered by the AFM tip upon retraction in water (b) the average attractive force Fadh between the AFM tip and the cell surface of each S. mitis strain in water (c) the maximum distance Dmax at which a local maximum in adhesion force was registered by the AFM tip upon retraction in 0.1 M KCl.

a)

b)

c)

39– an AFM study on Streptococcus mitis strains

Fixe

d ch

arge

den

sity

(Cm

-3)

-4.5

-4.0

-3.5

-3.0

-2.5

-2.0

-1.5

-1.0

Dmax (nm)0 200 400 600 800 1000 1200

Figure 4.5. Fixed charge density of the bacterial cell wall of S. mitis strains as a function of the minimum distance Dmax at which a local maximum in adhesion force was registered by the AFM probe tip upon retraction in water.

Figures 4.3 to 4.5 present some of the correlations between different properties

as revealed by the Pearson correlation test. Figure 4.3 shows that the amount of macroscopically measured surface nitrogen influences, at least in water, the maximum distance at which adhesion forces occur in the retracting mode. In Figure 4.4a it can be seen that when the affinity of the cell surface for water is smallest (larger contact angle), the minimum distance over which adhesion forces in water were observed by the hydrophilic tip of the AFM is largest. Interestingly, in 0.1 M KCl, the water contact angle correlates best with Dmax (see Figure 4.4c). Although the water contact angle reflects the affinity of the surface for a hydrophilic medium as water, the average adhesion force experienced by the AFM tip upon retracting from the surface correlates inversely with the contact angle (see Figure 4.4b). Similar to the water contact angle, the fixed charge density correlates best with Dmax , as can be seen in Figure 4.5.

4.4 DISCUSSION The frequency of occurrence of surface structures with variable lengths on S. mitis strains is high and differs considerably between different isolates. In general, the density of fibrils on S. mitis strains has been described as sparse. Macroscopic cell surface properties of these strains have been extensively studied, but it is unknown how the microscopic properties of the cell surface, as determined e.g. by atomic force microscopy, make up the macroscopic properties of the cell surfaces.

Force-distance curves for the S. mitis isolates involved in this study show two common features, regardless of the environmental conditions: a repulsive force upon approach and multiple adhesion forces upon retraction. Thus, the interaction


is attractive but has to overcome an energy barrier upon approach. The repulsive interaction upon approach has a decay length that varies between 11 and 33 nm in water with associated contact forces of between 1.0 and 8.0 nN, dependent on the strain considered (see Table 4.2), while in 0.1 M KCl these ranges are between 8 and 50 nm and 1.1 and 7.7 nN, respectively. Note that although an electrostatic interaction between the silicon nitride tip and the bacterial cell surface was expected, the decay lengths observed are considerably larger than the Debye-Hückel lengths (9.6 and 0.96 nm in water and 0.1 M KCl, respectively). Electrosteric forces resulting from the presence of surface structures, with different length and charge, probably increase the range and magnitude of these repulsive forces. Previously, also Camesano and Logan (2000) concluded that the interaction between negatively charged bacteria and the silicon nitride tip of an atomic force microscope was dominated by electrosteric repulsion. Moreover, the energy that the AFM tip has to overcome in order to reach the bacterial surface, i.e. the area enclosed below the approach curve toward zero separation distance (contact point), can be calculated to be twice times as high in water ( ~20000 kT) than in 0.1M KCl (~10000 kT). For a sphere-flat plate configuration, the DVLO theory also predicts a higher energy barrier in water than in 0.1 M KCl, although the energy barriers calculated in water (~400 kT) and in 0.1 M KCl (~40 kT) do not correspond numerically with those calculated from the AFM approach curves. Recently, Ong et al. (1999) compared forces of interaction measured with the AFM to those of model predictions based on an extended-DVLO approach. They concluded that the inclusion of steric interactions in the theoretical models leads to an additional repulsive component for the wild-type Escherichia coli strain, expressing long LPS molecules on its cell surface. Accounting for the long appendages carried by S. mitis strains, this discrepancy between theory and experimental data maybe be due as well to steric interactions.

Upon retraction, multiple adhesion forces between the tip and the various S. mitis cell surfaces were observed. Hypothetically, these adhesion forces might be due to interactions between the tip and fibrils of different lengths on the cell surfaces. Moreover, the different lengths may represent different molecules and are not necessarily associated with a single polymer. Adhesion forces have been found to extend up to 1186 nm for S. mitis T9 in buffer solution. While these distances may appear large, they are of the order of magnitude of fibrillar lengths observed with electron microscopy after negative staining (Handley, 1990). Also, Frank and Belfort (1999) found adhesion forces between an AFM tip and Pseudomonas atlantica extending over 1200 nm, which they ascribed to the presence of extracellular polysaccharides. The above interpretation of local adhesion forces is supported by the observation that the number of local maxima in adhesion forces N (see Table 4.2) is not very sensitive to ionic strength, although in general the adhesion forces tend to be slightly stronger in water than in 0.1M KCl.

The relations between macroscopic and microscopic cell surface properties reported in Table 4.3 are not always evident. Amounts of macroscopically measured surface nitrogen clearly increase with the maximum distance over which


an adhesion force occurs in the retracting mode in water (Figure 4.3), as does the fixed charge density (Figure 4.5). Assuming that the nitrogen content reflects the length and density of the fibrils present on these strains and the charge is associated with components in these fibrils, the above relationships merely demonstrate that a thicker extracellular layer is required to accommodate those components. The water contact angle, however, results from short range molecular interactions: the more hydrophobic the inner dense layer of the cell surface (higher water contact angle), the stronger the adhesion force (Figure 4.4b) and the larger the minimum separation distance at which maximum adhesion is observed in water (Figure 4.4a). We hypothesize that the forces controlling the affinity of the bacterial cell surfaces for water arise largely from the presumably, more dense inner layer of the cell surface, represented by its thickness by Dmin. In 0.1M KCl, however, the water contact angle correlates best with the maximum separation distance over which an adhesion force occurs (Figure 4.4c), which is opposite to our expectations. Further, it was seen that even though the water contact angle reflects the affinity of the surface for a hydrophilic medium as water, the average adhesion force experienced by the AFM tip upon retracting from the surface correlates inversely with the contact angle. The high hydrophobicity of S. mitis surfaces in water could stimulate dehydration to account for the stronger adhesion forces probed by the hydrophilic AFM tip on more hydrophobic strains. This phenomena of hydrophobic dehydration of one of the two interacting surfaces has also been reported to be responsible for protein adsorption on solid surfaces. If the surfaces of the protein molecule and the sorbent are polar (hydrophilic), their hydration is favorable. In that case it is probable that some hydration water is retained between the sorbent surface and the adsorbed protein molecule. However, if (one of) the surfaces are (is) apolar, i.e. hydrophobic, dehydration would be a driving force for adhesion (Haynes & Norde, 1994).

In summary, this study is the first attempt to correlate microscopic and macroscopic bacterial cell surface properties. Although some of the relations between macroscopic and microscopic cell surface properties can be readily understood, full understanding of all relations observed require a better theory describing short-range molecular interactions than hitherto available. Moreover, this problem is more complicated, as the force-distance curves confirm that the cell surface is not homogeneous and that by consequence, water contact angles and charge densities reflect only average properties.

5 Relations between macroscopic and microscopic adhesion of Streptococcus mitis strains to surfaces Virginia Vadillo Rodríguez, Henk J. Busscher, Willem Norde, Joop de Vries and Henny C. van der Mei*

5.1 INTRODUCTION Bacterial adhesion to surfaces is ubiquitous and known to play an important role in a wide variety of environments, as on biomedical implants in the human body (Harkes et al., 1991), on food processing equipment (Visser & Jeurnink, 1997), on ship hulls (Cooksey & Wigglesworth-Cooksey, 1995), to surfaces in the oral cavity (Marsh & Martin, 1992) and wastewater plants (Characklis, 1973). However, the fundamental mechanisms governing bacterial adhesion are still poorly understood and have not been well defined. Initial bacterial attraction or repulsion to a surface

* Reproduced with permission of Society for General Microbiology from Microbiology (2004) 150: 1015-1022


44

may be described in terms of colloidal interactions. Thus, as bacteria move toward a substratum surface, the initial interaction between a bacterium and the surface is governed by long and medium range forces, primarily Lifshitz-Van der Waals and electrostatic forces. In addition, at close approach, other short-range, stereo-specific interactions may mediate irreversible adhesion of bacteria to a surface (Rijnaarts et al., 1995). All these interaction forces depend on physico-chemical properties of the substratum and the bacterial surfaces, such as hydrophobicity (Gannon et al., 1991), interfacial tensions (Busscher et al., 1984) and charge (Gannon et al., 1991). These properties are macroscopic in nature and represent an overall expression of the interactions taking place at a microscopic level. Previous studies have attempted to correlate bacterial adhesion with the macroscopic properties of both the substratum and the bacterial cell surfaces. In some cases, bacterial adhesion was related to the surface free energy of the bacteria and/or the substratum, whereas in others cases no relations could be detected (Absolom et al., 1983; Van Pelt et al., 1985). Nearly all bacterial cell surfaces found in nature are negatively charged and they are expected to be repelled by negatively charged substratum surfaces. However, even though experimental studies have shown the importance of electrostatic interactions in bacterial adhesion, discrepancies between experimental observations and theoretical expectations are frequently observed (Ong et al., 1999; Rijnaarts et al., 1995). For instance, adhesion of Escherichia coli to sludge flocs did not correlate with the bacterial zeta potential but with the fraction of positive charge present on the bacterial cell surface (Zita & Hermansson, 1997). Similarly, it was shown that despite small differences in DLVO (Derjaguin, Landau, Verwey, Overbeek) interaction energies, adhesion rates to glass differed strongly between fibrillated and non-fibrillated Streptococcus salivarius strains (Sjollema et al., 1990a). Therefore, it can be concluded that the DLVO theory, which accounts for long-range Lifshitz-Van der Waals and electrostatic interactions has only been marginally successful in describing adhesion of biological particles, such as bacteria, to surfaces. The same conclusion holds for the so-called extended DLVO theory (Van Oss et al., 1986), that also accounts for short-range acid-base interactions between a colloidal particle and a substratum surface.

Until recently, bacterial adhesion has been evaluated mainly by enumeration of the number of bacteria adhering to a surface (An & Friedman, 1997). The parallel plate flow chamber has turned out to be an effective tool for studying the bacterial adhesion to surfaces (McClaine & Ford, 2002). With the aid of phase-contrast microscopy, ultra-long working distance objectives and image analysis, bacterial adhesion can be monitored in real time and enumeration is instantaneous under the shear conditions of the actual experiments. The use of flow devices, however, is time-consuming and results provide no quantitative information on the magnitude of the interaction forces between bacteria and a surface.

Atomic force microscopy (AFM) provides new possibilities for probing the structural and physical properties of bacterial cells, as well as information on interaction forces involved in adhesion (Dufrêne, 2000; Dufrêne, 2002). Using topographic imaging, cell surface nanostructures (e.g. appendages and flagellae)

Adhesion properties of S. mitis

45

have been visualized, including changes of cell surface morphology during physiological processes. E. coli JM109 and K12J169, for instance, have been found to possess different morphologies depending on whether topographic images were taken in liquid or in air. Imaging in air revealed many topographic features that were missing for the cells imaged in liquid. The loss of topographic features in liquid was attributed to the dynamics of cellular filaments, that may collapse onto the wall surface creating a strain-specific topography of the surface. Furthermore, it was observed that lysozyme treatment led to the loss of surface rigidity and eventually to dramatic changes in bacterial shapes (Bolshakova et al., 2001). Force-distance curves provide complementary information on surface properties, including viscoelasticity and localized surface charge density and hydrophobicity. Interestingly, forces measured between the AFM tip and sulfate-reducing bacteria showed that both the adhesion forces at the cell-substratum periphery and cell-cell interface were higher than those measured in the center top of the bacterial cell. This has been suggested to be due to the accumulation of extracellular polymeric substances in interfacial regions (Fang et al., 2000). The role of hydrophobic interactions in bacterial adhesion at a microscopic level has been pointed out by Ong et al. (1999), measuring the interaction forces between Escherichia coli-coated AFM probes and solids of different surface hydrophobicity. It was shown that both attractive forces and cell adhesion were promoted by the hydrophobicity of the substratum surfaces.

Streptococcus mitis is one of the primary colonizers of surfaces in the oral cavity and colonizes both dental hard tissues as well as mucous membranes, most notably the cheeks and the tongue (Marsh & Martin, 1992). By comparison with other oral streptococci, S. mitis usually carries sparsely distributed but extremely long fibrils, as demonstrated by electron microscopy (Cowan et al., 1992). It is thought that bacterial surface appendages could exert localized attraction and consequently function to bridge a gap between substratum and the bacterial cell during adhesion (Smit et al., 1986).

The aim of the present study is to relate microscopic adhesion properties of S. mitis strains as derived from force-distance curves obtained using AFM to their macroscopic adhesion to surfaces in a parallel plate flow chamber.

5.1 MATERIALS AND METHODS

5.1.1 Bacterial strains, growth conditions and harvesting Nine S. mitis strains were used in this study, which were all cultured in Todd Hewitt Broth (Oxoid, Basingstoke, UK). For each experiment, the strains were inoculated from blood agar in a batch culture. This culture was used to inoculate a second culture that was grown for 16 h prior to harvesting. Bacteria were harvested by centrifugation (5 min at 10,000 g), washed twice with demineralized water and resuspended in water. To break up bacterial chains, cells were sonicated for 30 s at 30 W (Vibra Cell model 375, Sonics and Materials Inc., Danbury, CT, USA). Sonication was done intermittently while cooling in an ice/water bath.


46

5.1.2 Parallel plate flow chamber and data analysis The flow chamber (internal dimensions: 76×38×0.6 mm) and image analysis system have been described in detail previously (Busscher & Van der Mei, 1995). Images were taken from the bottom glass plate (55×38 mm) of the parallel flow chamber. The top plate of the chamber was also made of glass. Glass plates were cleaned by sonicating for 2 min in 2 % RBS35 surfactant solution in water (Omnilabo international BV, The Netherlands), rinsed thoroughly with tap water, dipped in methanol, and again rinsed with demineralized water. The flow chamber was cleaned with Extran (Merck, Germany) and thoroughly rinsed with water and demineralized water. Prior to each experiment, all tubes and the flow chamber were filled with 0.1M KCl solution, taking care to remove all air bubbles from the system. Once the system was filled, a bacterial suspension of 3×108 cells ml-1 in 0.1M KCl was allowed to flow through the system at a flow rate of 1.44 ml min-1, corresponding with a Reynolds number of 0.6 and a shear rate at the wall of the flow chamber of 10.6 s-1. Deposition was observed with a CCD-MXRI camera (High Technology) mounted on a phase-contrast microscope (Olympus BH-2) equipped with a 40x ultra-long-working distance lens (Olympus ULWD-CD Plan 40 PL). The camera was coupled to an image analyzer (TEA; Difa). The bacterial suspension was perfused through the system for 4 h with re-circulation at room temperature.

The total number of adhering bacteria per unit area n(t) was recorded as a function of time by image sequence analysis during 4 h and the affinity of an organism for the glass surface was expressed as the initial deposition rate j0, representing the initial increase of n(t) with time (see also Figure 5.1). Note that since the initial deposition rate is extracted only from the first, initial adhesion data, it represents the affinity of the organisms for the substratum surface without intervening influences of interactions between adhering bacteria. From the total number of adhering bacteria per unit area as function of time n(t) and the number of particles adhering over the full duration of an experiment n∞,, i.e. saturation adhesion, the so-called characteristic adhesion time τ was calculated using

n(t) = n∞ (1-e-t/τ) (5.1)

The characteristic adhesion time τ is determined by a combination of deposition, blocking and desorption, according to 1/τ = β + j0Ab (5.2) in which Ab represents the area on the substratum surface blocked by one adhering bacterium and β is the desorption rate coefficient (Bos et al., 1999). In order to derive the two unknowns Ab and β from Eq. 5.2, blocked areas for each particular strain were derived from the radial pair distribution g(r), as can be calculated from the spatial arrangement of the adhering bacteria (Sjollema & Busscher, 1990).


47

Time (s)

0 3000 6000 9000 12000 15000 18000

Adh

erin

g ba

cter

ia (1

06 cm-2

)

0

2

4

6

8

10

12

14

jo

S. mitis 357

Time (sec)

0 3000 6000 9000 12000 15000 18000

Adh

erin

g ba

cter

ia (1

06 cm-2

)

0

2

4

6

8

10

12

14

jo

S. mitis 244

Figure 5.1. Number of bacteria deposited as a function of time to a glass surface in 0.1M KCl as determined in a parallel plate flow chamber, for S. mitis 357 and S. mitis 244.


48

rela

tive

dens

ity g

(r)

1

rblocked

Ab

Figure 5.2. The relative density of adhering microorganisms, expressed as a radial pair distribution function g(r) as a function of the distance r between the adhering organisms The region where g(r) < 1 is associated with the blocked area Ab. Radial pair distribution functions describe the relative density of adhering organisms around a given center organism as a function of distance between the adhering organisms. Figure 5.2 shows an example of a radial distribution function, indicating the region for which g(r) < 1, i.e. Where the density of adhering organisms is less than average. Inserting the blocked areas Ab derived from the radial distribution functions, and the characteristic adhesion times τ from the measured kinetics, the desorption rate coefficients β can be directly calculated from Eq. 5.2.

All values given in this paper are the average of three experiments carried out with separately grown microorganisms. 5.1.3 Atomic force microscopy Bacterial cells were suspended in water to a concentration of 105 per ml after which 10 ml of the suspension was filtered through an Isopore polycarbonate membrane (Millipore) with a pore size of 0.8 µm (Dufrêne et al., 1999). The pore size was chosen slightly smaller than the streptococcal dimensions to immobilize the bacteria by mechanical trapping. After filtration, the filter was carefully fixed with double-sided sticky tape on a sample glass and transferred to the AFM. AFM measurements were made at room temperature under 0.1M KCl solution, using an optical lever microscope (Nanoscope III Digital Instruments) with an applied force maintained below 1 nN and a scan rate of ~2 Hz. “V”-shaped silicon nitride cantilevers from Park Scientific Instruments (Mountain View, CA) with a spring constant of 0.06 N m-1 and a probe curvature radius of ~50 nm, according to manufacturer specifications, were used. Individual force curves with z-displacements of 100 – 200 nm were collected at z-scan rates ≤ 1 Hz. The slope of the retraction force curves in the region where probe and sample are in contact was used to convert the voltage into cantilever deflection. The conversion of deflection data to force data was carried out as has been previously described by others (Dufrêne, 2000).


49

Topographic images were recorded for at least ten bacterial cells per culture of each S. mitis strain. To this end, the tip was positioned over the top of a trapped bacterium, scanning was stopped and ten force measurements were performed at randomly selected locations around the top for each bacterial cell studied. Force-distance curves for both approach and retract interactions were analyzed as follows: Approach. To quantify steric interactions between the AFM tip and cell surface polymers upon approach, a model developed for grafted polymers at relatively high surface coverage was used. The force per unit area between two parallel flat surfaces (Fst) with only one coated with polymer has been modeled following the work of Alexander (1977) and De Gennes (1987). To describe the force between a spherical AFM tip and a flat surface, the model was modified by Butt et al. (1999) by integrating the force per unit area over the tip surface to produce the interaction force

Fst = 50 kB T a L0 Γ 3/2 exp (-2π h / L0) (5.3)

where kB is the Boltzmann constant, T the temperature, a the tip radius, Γ the grafted polymer density per unit area, h the separation distance between the two surfaces and L0 the equilibrium thickness of the polymer layer. For these calculations the tip radius was assumed to be 250 nm (Drummond & Senden, 1994). Retraction. Retraction curves of all nine S. mitis strains showed various local attractive maxima. These attractive maxima and the distances at which they occurred were quantitatively registered in histograms (see Figure 5.3). Based on the prevalence of the local maxima in adhesion forces (expressed in percentages in the histograms), an average attractive force Fadh, average between the AFM tip and the cell surface of each S. mitis strain was calculated. Forces with prevalence less than 2 % were neglected in this averaging process. 5.1.4 Theory The results obtained in this study were explained on the basis of a combination of the extended DLVO (X-DLVO) theory and the steric model accounting for repulsion between bacterial polysaccharide molecules and the AFM tip, described above. The X-DLVO model accounts for Van der Waals, electrical double layer and acid-base interactions. The expressions quantifying these interactions as a function of separation distance can be found elsewhere (Van Oss, 1994) for different system geometries. The radius of the AFM tip was assumed to be small relative to the curvature of the bacterium, and therefore, the tip was approximated as a sphere and the bacterium as a flat surface (Johnson et al., 1995). A Hamaker constant of 10-20 J was chosen, consistent with earlier work on bacterial adhesion (Martin et al., 1996). Bacterial zeta potentials were recalculated from the surface fixed charge density previously obtained in a recent publication (Vadillo-Rodríguez et al., 2002) for all nine S. mitis by soft particle analysis according to Ohshima (Ohshima & Kondo, 1991), yielding an average surface potential of -1.6 mV. For

Sepa

ratio

n di

stan

ce (n

m)

010

0 20

0 30

040

050

060

0

F (nN)

-2.0

-1.5

-1.0

-0.50.0

50%

Sepa

ratio

n di

stan

ce (n

m)

0 10

020

030

040

050

060

0F (nN)

-2.0

-1.5

-1.0

-0.5

16%

6%14

%14

%

Sepa

ratio

n di

stan

ce (n

m)

010

0 20

0 30

040

050

060

0

F (nN) -2-1012345S.

miti

s 398

D ma

x

appr

oach

retra

ctio

n

0 10

020

030

040

050

060

0

F (nN) -2

-1 0 1 2 3 4 5

S. m

itis 2

72

Dm

ax

appr

oach

retra

ctio

n

Sepa

ratio

n di

stan

ce (n

m)

50%

50%

50

%

0.0

16%

34%

Figu

re 5

.3. R

etra

ctin

g fo

rce-

dist

ance

cur

ves

and

forc

e di

strib

utio

n hi

stog

ram

s fo

r S.

miti

s 3

98 a

nd S

. miti

s 27

2 in

0.1

M K

Cl.

Loca

l m

axim

a i

n a

dhes

ion

forc

e du

ring

retra

ctio

n ar

e m

arke

d by

arr

ows

for

the

max

imum

dis

tanc

e at

whi

ch t

hey

occu

r. Th

e r

elat

ive

pr

eval

ence

of e

ach

forc

e, a

s der

ived

from

forc

e di

stan

ce c

urve

s, is

indi

cate

d in

the

hist

ogra

ms.


51


F/a

(mN

/m)

-0.02

-0.01

0.00

0.01

0.02

StericVan der WaalsElectrostaticAcid-BaseX-DLVO

Steric

X-DLVO

Figure 5.4. The interaction force (divided by the radius of the AFM tip) between an AFM tip and a bacterial cell surface as a function of separation distance, based on extended DLVO theory and steric repulsions, as arising from average input values of the total collection of S. mitis isolates and a silicon nitride tip in 0.1 M KCl. silicon nitride tips, the zeta potential was estimated from electrophoretic mobility measurements on silicon nitride particles as a function of the pH (Harkes et al., 1991) and taken -2.0 mV in our calculations. Acid-base interactions were estimated based on electron-acceptor γ+ and electron-donor γ- parameters (Van Oss, 1994). These parameters were calculated for glass and bacteria by means of Young-Dupré equation after measuring contact angles with four different liquids (Cowan et al., 1992). The resulting values amount to γ+ = 1.5 mJ m-2, γ - = 56.3 mJ m-2 for glass and to average values γ+ = 2.1 mJ m-2 and γ - = 5.7 mJ m-2 for the S. mitis strains.

Figure 5.4 shows the result of this modeling, displaying the interaction forces as divided by the effective radius of the tip.

5.3 RESULTS Figure 5.1 presents examples of the deposition rates for S. mitis 357 and 244 to a glass surface in 0.1M KCl. The deposition behavior of all nine S. mitis strains followed an exponential rise in time to a maximum value, of which the initial linear part allowed calculation of the initial deposition rate j0. Blocked areas were determined from radial pair distribution functions and are summarized in Table 5.1, together with the initial deposition rates and the desorption rate coefficients β. As can be seen, S. mitis strains present varying adhesive abilities. For instance, the initial deposition rate j0 of S. mitis 357 is three times higher than that of S. mitis 244 and blocked areas Ab range from 0.7µm2 for S. mitis BA to 1.9 µm2 for S. mitis 357. Desorption rate coefficients are generally small (on the average 10-5 s-1) and are somewhat higher for S. mitis ATCC9811 than for the other bacterial strains.

Figure 5.3 shows two examples of the interaction forces measured between S. mitis 398 and S. mitis 272 and a silicon nitride AFM tip immersed in 0.1M KCl


52

Table 5.1. Initial deposition rates j0 as calculated from the initial linear increase of the number of adhering bacteria with time, blocked areas Ab determined from radial pair distribution functions and desorption rate coefficients β of S. mitis strains, suspended in 0.1M KCl and as measured in a parallel plate flow chamber on glass. ± indicates the standard deviation over three separate experiments with separately cultured organisms.

Parallel plate flow chamber

S.mitis strains j0 (cm-2 s-1) Ab(µm2) β (10-5s-1)

ATCC9811 542 ± 134 1.1 ± 0.2 10.9 ± 2.7 ATCC33399 439 ± 49 1.5 ± 1.2 2.5 ± 0.3

244 305 ± 29 1.1 ± 0.2 4.1 ± 1.2

272 443 ± 117 1.4 ± 0.4 1.8 ± 0.7

357 970 ± 159 1.9 ± 0.4 1.8 ± 0.4

398 827 ± 184 1.8 ± 0.2 4.8 ± 1.7

BMS 601 ± 111 1.3 ± 0.3 5.8 ± 2.2

BA 492 ± 85 0.7 ± 0.2 6.5 ± 1.8

T9 388 ± 76 1.0 ± 0.1 5.0 ± 0.2 Table 5.2. Summary of quantitative data derived from force-distance curves in 0.1M KCl of nine S. mitis strains, including equilibrium length L0 of the bacterial surface polymers, bacterial polymer density Γ and force at zero separation distance Fst (h=0) (from approach curves) and maximum Dmax distances at which local maxima in adhesion force occurs, together with the number of local maxima (N) and the average adhesion force Fadh, average calculated (from retraction curves).

AFM Approach Retraction

S.mitis strains

L0 (nm) Γ (1015m-2) Fst (h=0) (nN)

Dmax (nm) Fadh,average (nN)

N

ATCC9811 314 ± 37 3 ± 1 2.1 ± 0.8 167 ± 33 -0.5 ± 0.3 2

ATCC33399 57 ± 12 11 ± 2 3.5 ± 0.7 170 ± 30 -1.4 ± 0.8 3

244 82 ± 18 8 ± 2 2.8 ± 0.7 426 ± 29 -1.8 ±1.0 4

272 82 ± 37 9 ± 3 3.6 ± 1.4 534 ± 29 -1.0 ± 0.4 6

357 63 ± 12 5 ± 1 1.1 ± 0.2 196 ± 43 -0.7 ± 0.2 3

398 50 ± 6 21 ± 3 7.7 ± 1.0 120 ± 35 -1.5 ± 0.7 2

BMS 63 ± 18 6 ± 3 1.6 ± 0.8 867 ± 40 -1.1 ± 0.4 9

BA 50 ± 6 10 ± 2 2.7 ± 0.6 588 ± 23 -1.4 ± 0.3 5

T9 70 ± 18 12 ± 5 4.8 ± 2.0 1186 ± 38 -2.2 ± 0.9 11


53

as a function of the separation distance between the tip and the bacterium. Under our experimental conditions, all nine S. mitis strains present a repulsive force upon approach and multiple local maxima in attractive adhesion forces upon retraction. The repulsive forces upon approach indicate an equilibrium length L0 of the surface polymers between 50 nm up to 314 nm, depending on the strain considered (see Table 5.2). The polymer density Γ0, however, ranges from 3×10-3 nm-2 to 21×10-3 nm-2 (see also Table 5.2), in line with the sparse distribution of fibrils carried by S. mitis surfaces, as observed by electron microcopy (Cowan et al., 1992).

Upon retraction, the number of multiple local maxima in adhesion forces ranged from two at small separation distances for S. mitis 398, to eleven for S. mitis T9 at separation distances up to 1186 nm. The separation distances at which these adhesion forces occur, their magnitude and relative prevalence as averaged per strain are summarized in so-called distribution force histograms (see Figure 5.3, showing the distribution force histogram for S. mitis 398 and 272). The maximal distance Dmax at which local adhesion forces occurs together with the average adhesion force Fadh, average (arising from a bacterial cell surface as calculated from the relative prevalence of all local adhesion forces) and the number N of local maxima in adhesion forces are presented in Table 5.2. The average adhesion force Fadh, average ranges from -0.5 nN to -2.2 nN dependent on the strain considered. The large standard deviations associated to the values may be attributed the heterogeneous nature of the S. mitis cell surfaces.

Relations found between microscopic adhesion properties of S. mitis strains as derived from force-distance curves analysis and macroscopic adhesion in the parallel plate flow chamber are depicted in Figures 5.5 – 5.7. Figure 5.5 shows that the initial deposition rates j0 decrease as the force needed to achieve contact between the AFM tip and bacterial surface, i.e. Fst at zero separation distance, increases, indicating that the bacteria have to overcome a barrier before they can adhere. Note that S.mitis 398 constituted an exception to this behavior for reasons unknown and the strain has been omitted from this analysis. The radius of the area blocked rblocked by an adhering organism decreases in an almost linear fashion with the maximum distance Dmax over which the adhesion forces detected by the AFM tip upon retraction are operative. Note, in addition, that the order of magnitude of the radii of these blocked areas and the distance over which the adhesion forces act, are roughly similar (see Figure 5.6), while here S. mitis ATCC9811 constituted an exception. In Figure 5.7, it can be seen that unexpectedly, no relation could be found between the desorption rate coefficients β and the adhesion forces Fadh,

average. 5.4 DISCUSSION

Physico-chemical models to explain bacterial adhesion to solid substrata have only been partly successful for a limited number of strains, while usually a small number of strains have been involved in the verification of a model. The poor performance, so far, of physico-chemical models for microbial adhesion to surfaces


54

j0 (s-1cm-2)

0 200 400 600 800 1000 1200

F st (

h=0)

(nN

)

0

1

2

3

4

5

6

Figure 5.5. Force at zero separation distance Fst (h=0) between the AFM tip and the bacterial cell surface measured upon approach, as a function of the initial deposition rates j0 of eight S. mitis strains to glass, as determined in a parallel plate flow chamber in 0.1M KCl.

Dmax (nm)

0 200 400 600 800 1000 1200 1400

r bloc

ked (

nm)

0

500

1000

1500

2000

2500

Figure 5.6. The radius rblocked of the blocked areas Ab of nine S. mitis strains adhering on glass in 0.1 M KCl, as a function of the maximum distance Dmax at which a local maximum in adhesion force probed by the AFM tip upon retraction.


55

is due in part to the macroscopic approach of the system that is usually heterogeneous on a microscopic level. Equipment to measure physico-chemical properties of microbial cell surfaces with a microscopic resolution have for a long time impeded progress in the field. With the introduction of the Atomic force microscope, new possibilities for microscopic characterization of bacterial cell surfaces have become available. In this paper, we relate microscopic characterization of a large number of S. mitis strains, as probed by their interaction with the silicon nitride tip of the AFM, with their macroscopic adhesion to glass. High correlations between both sets of data were expected, since both glass and the silicon nitride tip are hydrophilic and negatively charged. At this stage it is important to note, that the relations presented in Figures 5.5 – 5.7 are obtained from a large collection of S. mitis strains. Therefore the relations depicted are expected to have a high degree of general validity, although not all strains behave fully according to the general trends we derive from our results.

The approach curves in the AFM measurements were used to assess steric interactions in adhesion of the organisms. The interaction forces given in Figure 5.4 demonstrate the relative importance of Lifshitz-Van der Waals, electrostatic, acid-base and steric terms for our collection of strains. X-DLVO interactions decay over a very short distance (~10 nm) and are therefore less influential on bacterial deposition to a substratum surface than steric forces that extend over much longer distances. Even though the X-DLVO does not show any energy barrier for adhesion, force-distance curves for all nine S. mitis strains always presented a repulsive force upon approach. Based on the theoretical predictions and on the distance observed at which this repulsive force becomes effective, we conclude that the force-distance curves for approach could be only described in terms of steric repulsion.

The steric interactions as obtained from AFM, play an important role in relating microscopic cell surface properties and macroscopic adhesion in the parallel plate

Fadh, average (nN)-2.5 -2.0 -1.5 -1.0 -0.5 0.0

β (1

0-5s-1

)

1

2

3

4

5

6

7

8

Figure 5.7. Desorption rate coefficients β of S. mitis strains in 0.1M KCl as a function of the average adhesion force Fadh,average between the AFM tip and the cell surface upon retraction.


56

flow chamber. According to X-DLVO theory and based on the analogy between the surface properties of the tip and the glass surface, macroscopic adhesion of S. mitis to glass substratum in the parallel plate flow chamber occurs under barrierless conditions (see Figure 5.4). However, the AFM tip detected a steric energy barrier that must be overcome in order to reach close contact with the bacterial surface. This energy barrier could be interpreted as an activation energy for the organisms to successfully deposit, as by the relation shown in Figure 5.5, the initial deposition rate increases as the force required to overcome the energy barrier by the AFM tip, decreases. These forces range between 1.1 and 7.7 nN for the nine S. mitis isolates employed in this study, and integrating Eq. 5.3 over the entire interaction range yields activation energies of ~2700 kT and 15000 kT, respectively. These activation energies are prohibitively high for spontaneous adhesion to occur for the entire collection of strains studied. Therefore, even though steric repulsion can be used to predict adhesion according to Figure 5.5, they are not fully responsible for adhesion at the separation distances at which they become operative.

The steric model is not only useful in explaining aspects of bacterial adhesion to macroscopic surfaces, but also to better understand the cell surface itself. L0 defines the equilibrium length of the bacterial surface polymers and under our specific ionic strength conditions amounts to several tens of nanometers, except for S. mitis ATCC9811. Other studies, also on fibrillated strains, reported equilibrium lengths L0 in the order of several hundreds of nanometers (Camesano & Logan, 2000). The present study was done, however, in a relatively high ionic strength solution, which may condense the electrically charged polymers on the cell surface and yields a relatively thin layer. Macroscopic as well as microscopic measurements at various ionic strength have provided evidence of a collapse of the fibrillated material at high ionic strength (Van der Mei et al., 1994; Van der Mei et al., 2000a).

The initial stages of macroscopic adhesion are governed mainly by interactions between the bacterial cell surfaces and the substratum surface, but in the more advanced stages, adhesion is in essence an interplay between interactions occurring between the substratum surface, a depositing bacterium and an already adhering one. It is interesting that this mechanism is confirmed by the combination of microscopic and macroscopic data shown Figure 5.6. As the distance over which the bacterium exerts adhesive forces extends, for instance through the extension of fibrils of different lengths (Van der Mei et al., 2000a), bacteria are brought closer together and the distance between adhering bacteria is reduced (Smit et al., 1986), yielding smaller blocked areas.

Desorption rate coefficients β are generally small for all S. mitis strains, on average in the order of ~10-5 s-1. The adhesion force measured by the AFM tip upon retraction was expected to be indicative for the desorption of bacteria in the parallel plate flow chamber. Figure 5.7 shows, however, that opposite to our expectation, no clear relation was found. Possibly, this has to do with the nature of the desorption process. Desorption in the parallel plate flow chamber takes place as a spontaneous process under the prevailing shear conditions, while in AFM the contact between bacterium and surface substratum is forced to break by application of an external force.


57

Summarizing, this study demonstrates that the repulsive force probed by AFM upon approach of the tip to a bacterial cell surface corresponds with a steric activation energy barrier, governing the rate of initial, macroscopic adhesion of the organisms to a glass surface. Moreover, the maximum distance at which attractive forces are probed by AFM upon retraction of the tip are influential on the area blocked by an adhering bacterium (i.e. the distance kept between adhering bacteria), while bacterial desorption could not be related with adhesive forces as probed by the AFM.

6 Atomic force microscopic corroboration of bond aging for adhesion of Streptococcus thermophilus to solid substrata Virginia Vadillo Rodríguez, Henk J. Busscher, Willem Norde, Joop de Vries and Henny C. van der Mei*

6.1 INTRODUCTION Microbial adhesion is at the basis of several medical (Dankert et al., 1986; Gristina et al., 1988), industrial (Bouman et al., 1982) and environmental problems (Cooksey & Wigglesworth-Cooksey, 1995). The initial stages of microbial adhesion are for a major part determined by mass transport of microorganisms towards a substratum surface, the efficiency of collisions between organisms and the substratum surface and the reversibility of the adhesion process. Microbial deposition, adhesion and desorption can best be studied in controlled flow devices

* Accepted for publication in Journal of Colloid and Interface Science


60

(Adamczyk, 1989) with in situ observation and real-time image analysis enabling the measurement of the deposition and adhesion as well as accurate information on the spatial arrangement of the adhering microorganisms and their residence time-dependent desorption (Sjollema et al., 1990b; Dabros & Van de Ven, 1982). Desorption implies that microbial adhesion is reversible. This is true for the very initial stages of the adhesion process, but overtime the adhesive bond between microorganisms and a substratum surface may strengthen, causing a prohibitive high activation energy for desorption. Alternatively, the adhesive bond may weaken overtime, leading to increased desorption rates.

A parallel plate flow chamber is a suitable device for studying the kinetics of microbial adhesion to surfaces, while modern image analysis techniques have made it possible to enumerate adhesion and residence time-dependent desorption simultaneously in situ (Meinders et al., 1992; Sjollema et al., 1989). The measured desorption will therefore be indicative for the reversibility of the process, although it may include a contribution from collisions between flowing and adhering organisms as occurring in natural systems (Van de Ven, 1989). Dabros & Van de Ven, (1982) proposed that desorption rate coefficients for a particle adsorbed at time τ and desorbing at time t, i.e. after residing on the surface for a time (t-τ), change exponentially from an initial value β0 to a final value β∞ during aging of the bond with a relaxation time 1/δ

β(t-τ) = β∞ - (β∞ - β0) e-δ (t-τ) (6.1)

Depending on whether the initial desorption rate coefficient β0 is larger or smaller than the final value β∞ , bonds are assumed to strengthen or weaken during aging.

Meinders et al. (1994) applied this model to analyze the residence time-dependent desorption of Streptococcus thermophilus B, a fouling organism encountered in dairy processing (Bouman et al., 1982), in a parallel plate flow chamber under moderate shear conditions of 50 s-1. Desorption rate coefficients decreased exponentially on glass with residence time (t-τ) from an initially high value β0 to an almost negligibly low value β∞. Initial desorption rate coefficients were higher for experiments carried out at pH 2 (30×10-4 s-1) than for experiments at pH 7 (23×10-4 s-1). The initial desorption rate coefficients decreased rapidly to about 10×10-6 s-1 with a relaxation time of about 48 and 59 s for pH 2 and 7, respectively. Consequently, it was concluded that the bond strength between S. thermophilus and glass substratum surfaces increased during aging, although this conclusion was never verified experimentally due to the absence of adequate instrumentation at that time.

Atomic force microscopy (AFM) provides a direct measure of the adhesive forces between microorganisms and surfaces. Utilizing a small tip mounted on a flexible cantilever, interaction forces between the tip and a bacterial surface can be recorded as the tip approaches, makes contact and retracts from the bacterial surface. Accordingly, AFM has been used extensively to study the initial interaction between bacteria and different substratum surfaces in aqueous solutions, including quantification of the contribution of surface polymers to bacterial

Adhesive bond strengthening by S. thermophilus

61

adhesion (Razatos et al., 1998; Camesano & Abu-Lail, 2002). Mechanical properties, such as turgor pressure and Young’s modulus of the bacterial cell surface, have also been calculated by means of AFM (Arnoldi et al., 2000; Velegol & Logan, 2002). In contrast, interfacial and adhesive forces originated during physical contact of the bacteria with the AFM tip are only sparsely found in literature (Abu-Lail & Camesano, 2003a; Lower et al., 2000).

The aim of this chapter is to analyze microscopically the time dependence of the bond strength between S. thermophilus B and the Si3Ni4 tip of an AFM, in order to present an independent verification of the conclusions drawn by Meinders et al. (1994) on aging of the bond between this organism and glass substratum surfaces. 6.2 MATERIALS AND METHODS

6.2.1 Bacterial strain and growth conditions S. thermophilus B was isolated from a heat exchanger plate in the downward section of a pasteurizer. Bacteria were grown overnight at 37 °C from a frozen stock in batch culture in M17 broth (OXOID), supplemented with 1 % saccharose. This culture was used to inoculate a second culture, which was grown for 20 h. After 20 h, cells were harvested by centrifugation at 5,000 g, washed twice in demineralized water and resuspended in 40 mM potassium phosphate solution with pH adjusted to 2 and 7 by addition of HCl or KOH. 6.2.2 Atomic force microscopy (AFM) Bacteria were suspended in water to a concentration of 105 per ml after which 10 ml of the suspension was filtered through an Isopore polycarbonate membrane (Millipore) with a pore size of 0.8 µm (Kasas & Ikai, 1995). The pore size was chosen slightly smaller than the streptococcal dimensions to immobilize the bacteria by mechanical trapping. After filtration, the filter was carefully fixed with double-sided sticky tape on a sample glass and transferred to the AFM. AFM measurements were made at room temperature under 40 mM potassium phosphate solution at pH 2 and 7, using an optical level microscope (Nanoscope III Digital Instrument). “V”-shaped silicon nitride cantilevers from Park Scientific Instruments (Mountain View, CA) with a spring constant of 0.06 N m-1 and a probe curvature of ∼ 50 nm, according to manufacturer specifications, were used. Individual force curves with z-displacements of 100 – 200 nm were collected over the top of a trapped bacterium at randomly selected locations. Retraction of the tip from the bacterial surface was carried out after different contact times between the AFM tip and bacterial cell surface, ranging from 1 to 200 s. The slope of the retraction force curves in the region where probe and sample are in contact were used to convert the voltage into cantilever deflection. The conversion of deflection data to force data was carried out as has been previously described by others (Dufrêne, 2000). The maximum adhesion peak Fadh, max upon retraction was recorded as function of contact time between the AFM tip and the bacterial cell surface. For each bacterial cell, two force-distance curves were measured for each given time, and five cells were examined for a given pH.


62

6.3 RESULTS AND DISCUSSION The initial stages of microbial adhesion are considered to be dominated by Lifshitz-Van der Waals and electrostatic interactions (Rutter & Vincent, 1980). Over a wide range of ionic strengths, the energy distance curve of the interaction between a bacterial cell and a surface of the same charge sign is characterized by the presence of a shallow, secondary minimum at separation distances of a few up to ten nm, an energy barrier closer to the surface and a deep, primary energy minimum at short separation distances (< 1 nm) (Norde & Lyklema, 1989). It is thought that after microorganisms are initially captured in a reversible fashion in the secondary energy minimum their extracellular polymeric substances (EPS) firmly anchor at the substratum surface thereby successfully surpassing the energy barrier (Busscher et al. , 1992).

The adhesive forces between a hydrophilic, negatively charged AFM tip and S. thermophilus B cell surface are shown in Figure 6.1 for pH 2 after a contact time of 10 and 130 s. As the tip retracts the bacterial surface, several adhesion events are observed before reaching a maximum value Fadh, max. After having overcome Fadh,

max, the tip eventually detaches stepwise reaching a null adhesion force between the tip and the bacterial cell surface at a relatively large separation distance (∼ 800 nm). Note that this does not necessarily indicate that the organisms are surrounded by an 800 nm thick EPS coat, as EPS chains may extend during retraction of the AFM tip.

Figure 6.2 shows the maximum adhesion peak Fadh, max observed upon retraction of the AFM tip from the bacterial cell surface as function of the contact time at pH 2 and pH 7. Bond strengthening occurs at both pH values yielding an increase in adhesive force over time. Although differences in initial and final desorption rate coefficients of S. thermophilus were reported for pH 2 and pH 7, these differences are not reflected in statistically significant way in the increases in adhesive force between the AFM tip and the cell surface (see Figure 6.2).

To our knowledge, this paper is the first one reporting experimental proof of aging of the bond between bacteria and a substratum surface and relating this bond aging to resident time-dependent desorption of organisms from a surface. Although glass is admittedly not similar in surface properties as Si3Ni4, of which the AFM tip is made, both surfaces are hydrophilic and negatively charged. As an important qualitative association between macroscopic resident time-dependent desorption and bond aging as measured in AFM, it is noted that the relaxation times of the resident time-dependent desorption are of the same order of magnitude as the time scales on which the adhesive bond strengthens according to the AFM measurements.

Meinders et al. (1994) calculated a bond strength energy of 16 to 17 kT per cell for S. thermophilus B on glass, by applying

βescape = (j0 a / D∞ c) (D∞ / a ∆h) exp(φm / kT) (6.2)

(where βescape is the escape rate coefficient, j0 the initial deposition rate, a the particle radius, D∞ the Stokes-Einstein diffusion coefficient, c the bacterial cell


63

(a)


0 200 400 600 800 1000 1200 1400

Forc

e (n

N)

-4

-2

0

2

4

6

8

10

Fadh, max

(a)

(b)


0 200 400 600 800 1000 1200 1400

Forc

e (n

N)

-4

-2

0

2

4

6

8

10

Fadh, max

Figure 6.1. Retraction line of a force-distance curve for S. thermophilus B in 40 mM potassium phosphate solution at pH 2 after a contact time of 10 s (a) and 130 s (b) between the AFM tip and the bacterial cell surface.


64

Time (s)

0 50 100 150 200 250

F adh,

max

(nN

)

-5

-4

-3

-2

-1

0pH 7pH 2

Figure 6.2. Maximum adhesion peak Fadh, max upon retraction as function of the contact time between the AFM tip and bacterial cell surface in 40 mM potassium phosphate solution at pH 2 (♦) and pH 7 (•). Error bars denote the SD in Fadh, max. concentration at the entrance of the flow chamber, ∆h the width of the energy minimum, φm depth of the energy minimum and kT the energy of thermal motion), as proposed by Xia et al. (1989), and inserting measured values for the initial deposition rates and a desorption rate coefficient βescape measured in the absence of particle flow. Alternatively, integration of the area under the maximum adhesion peak after 130 s of contact in the force-distance curves yields a bond strength energy of 104 – 105 kT between the tip and the organism. Comparing with the bond strength energy from Meinders et al. (1994), this would indicate that the bond between the tip and the bacterial cell surface is 103 – 104 fold stronger than between a bacterium and a macroscopic substratum surface. The bond strength energy per organism estimated by Meinders et al. (1994) suggests that desorption of S. thermophilus from glass, under the experimental conditions of the parallel plate flow chamber, takes place out of the secondary minimum of the energy distance curve according to the DLVO theory for colloidal stability (Bos et al., 1999). The experimentally decreasing desorption rate with increasing residence time of the bacterial cell near the surface may reflect the rate by which EPS bridge the separation between the cell and the surface and anchor to the latter. Irreversible anchoring thus occur as a result of multiple attachments of EPS to the substratum surface during bond aging, that stepwise detach in the parallel plate flow chamber until an organism is weakly attached. The final bond strength energy to overcome prior to escape from the surface and calculated from macroscopic desorption amounts 16 – 17 kT. Like after bond aging in the flow chamber, there is close contact between the AFM tip and the bacterial cell surface. This contact is forced


65

by pushing the organism into the primary minimum, and may become more intimate through multiple contacts between EPS and the AFM tip than in the flow chamber. Detachment of the AFM tip from the bacterial cell surface after close contact is reached by the application of an external force to the tip. Therefore, opposite to detachment in the flow chamber, detachment from the AFM tip is forced after stepwise disruption of multiple EPS contacts which would account for the huge bond strength energy estimated from the retracting force-distance curves.

In conclusion, this chapter presents qualitative associations between resident time-dependent bacterial desorption from a macroscopic glass surface with strengthening of the microscopic bond of a bacterium to the tip of an AFM. Quantitative considerations of the bond strength energy from force-distance curves and those derived from resident time-dependent desorption under flow, indicate that macroscopic desorption of bacteria takes place from the secondary minimum of the energy distance curve, whereas bacterial detachment from the AFM tip develops stepwise from the corresponding primary minimum. Bond strengthening thus seems to be a general result of multiple attachments of EPS to a substratum surface.

7 Dynamic cell surface hydrophobicity of lactobacillus strains with and without surface layer proteins Virginia Vadillo Rodríguez, Henk J. Busscher, Willem Norde, Joop de Vries and Henny C. van der Mei*

Several studies have shown that bacterial strains, such as lactobacilli, can protect the host against infection by invading uropathogens. The mechanism by which lactobacilli exert this protection is not fully understood, but adhesion is a commonly accepted prerequisite (Sanders, 1993). Several lactobacillus species possess a surface layer protein (SLP) anchored to the cell envelope. This surface layer consists of a (glyco-)protein, the so-called S-protein, which assembles into characteristic two-dimensional crystalline layers at the cell surface (Sára & Sleytr, 2000). The function of the S-layer on these organisms is unknown, but S-layers of lactobacilli are important in their adhesion to surfaces, as SLP conveys * Accepted for publication in Journal of Bacteriology


hydrophobicity to the lactobacillus cell surface (Van der Mei et al., 2003). Yet, adhesion of lactobacilli to surfaces often does not proceed according to expectations based on their cell surface hydrophobicity and hydrophobic strains do not always adhere best to hydrophobic substrata (Millsap et al., 1996), as outlined by surface thermodynamics (Absolom et al., 1983). This suggests cell surfaces of lactobacilli may adapt their cell surface hydrophobicity in response to environmental changes, like in pH or ionic strength.

Macroscopic bacterial cell surface hydrophobicity is commonly inferred from water contact angle measurements on bacteria deposited on membrane filters (Busscher et al., 1984). If water molecules have a greater preference to surround each other than to contact a bacterial cell surface, the surface appears hydrophobic and water droplets do not spread. If water molecules favor a microbial cell surface rather than each other the surface appears hydrophilic. Hydrophobic lactobacillus isolates with water contact angles above 100 degrees (L. acidophilus RC14) have been described, but also extremely hydrophilic ones with water contact angles of 19 degrees (L. casei 36) (Van der Mei et al., 1998a). Although cell surface hydrophobicity arises from interactions at the molecular level, hydrophobicity has never been assessed at the level of molecular cell surface components.

Atomic force microscopy (AFM) has emerged as a valuable tool for probing interaction forces at the molecular level with a high spatial resolution (Dufrêne, 2002). A sharp tip located at the free end of a flexible cantilever is approached and retracted from the surface under study. Interaction forces between the tip and the sample surface cause the cantilever to deflect. The deflection signal during the approach and retraction process is acquired to provide so-called force-distance curves (see Figure 7.1 as an example).

Separation distance (nm)0 50 100 150 200 250 300 350 400 450 500

F(nN

)

-2

-1

0

1

2

3

4

5

6

Fadh

Figure 7.1. Force-distance curve for L. acidophilus ATCC4356 interacting with an hydrophobic AFM tip at 10 mM KCl. The solid line represents the approach curve, while the dashed line indicates the retraction curve. The maximum adhesion force Fadh probed upon retraction is indicated on the graph.

Dynamic cell surface hydrophobicity of lactobacilli

69

In this chapter, the surfaces of L. acidophilus ATCC4356 and L. casei ATCC393 with and without SLP, respectively, have been probed with regard to their interaction forces with chemically functionalized AFM tips, i.e. terminated with hydrophobic (CH3) and hydrophilic (OH) groups. Experiments were done in a 10 mM and 100 mM KCl solution. The macroscopic cell surface hydrophobicity of the two strains has also been assessed by contact angle measurements with these low and high ionic strength solutions.

Bacterial strains were cultured in MRS (De Man, Rogosa Sharpe, Merck, Germany) at 37 °C in an atmosphere containing 5 % CO2. This culture was used to inoculate a second culture that was grown for 16 h prior harvesting. Bacteria were harvested by centrifugation (5 min at 10,000 g), washed twice with demineralized water and suspended in demineralized water, 10 or 100 mM KCl solution. Contact angle measurements were performed on bacterial lawns prepared by depositing about 50 layers of bacteria suspended in demineralized water on a cellulose acetate membrane filter (pore diameter 0.45 µm) (Van der Mei et al., 2003). For AFM experiments, bacteria were attached to a positively charged poly-L-lysine treated glass slide. “V”-shaped silicon nitride cantilevers with a spring constant of 0.06 N m-1 were functionalized by coating them with a thin layer of titanium and gold followed by their immersion in HS(CH2)11OH or HS(CH2)17CH3 solutions. Functionalized probes were always used immediately after preparation. AFM measurements were made at room temperature under 10 and 100 mM KCl solution using an optical level microscope (Nanoscope III Digital Instrument). An array of 32×32 force-distances curves were collected over the entire field of view, once a bacterium was imaged (see Figure 7.2a and 7.2a’ as an example). Adhesion maps were produced by taking the most negative force detected during the retraction curve (see Figure 7.1) and by plotting that value against x-y position of each force-distance curve (Figure 7.2b and 7.2b’). From the adhesion maps, a selected area of ∼800×800 nm2 over the top of each bacterium was used to generate an adhesion distribution histogram (Figure 7.2c and 7.2c’) from which an average adhesion force Fadh was calculated between functionalized AFM tips and the bacterial cell surfaces for each experimental condition studied. Three to five different organisms were studied in each particular case.

Adhesion maps indicated a heterogeneous surface distribution of interaction forces between the cell surfaces and functionalized tips for both strains, regardless of ionic strength (see the examples in Figure 7.2b and 7.2b’). Histograms showing the distribution of these interaction forces over the top of each bacterium are presented in Figures 7.2c and 7.2c’. The interaction forces detected by hydrophobic and hydrophilic AFM tips were averaged for each strain into an adhesion force Fadh and compared with contact angles measured with aqueous, low and high ionic strength solutions (see Table 7.1). In general, high interaction forces with a hydrophilic tip were found to coincide with low contact angles, whereas a cell surface with high contact angle showed the strongest interaction with a hydrophobic tip. In addition, both strains reversed their hydrophobic nature upon increasing the ionic strength from 10 to 100 mM. The lactobacillus strain with SLP

L. a

cido

philu

s ATC

C43

56 (S

LP)

-26

nN

(b)

(a)

3.5 µm

0

010203040506070-0.

23

-0.68

-1.51

-2.40

-3.35

-5.01

F adh

(nN

)

#events

(c)

0 nN

L. c

asei

ATC

C39

3 (n

o-SL

P)

05101520253035

-0.07

-0.33

-0.48

-0.77

-0.98

-1.33

-1.74

-2.70

-3.22

F adh

(nN

)

# events

(c’)

-3.5

nN

0 nN

(b

’)

(a’)

0 3 µm

F

igur

e 7.

2. A

rray

of 3

2×32

forc

e di

stan

ce c

urve

s ove

r the

AFM

fiel

d of

vie

w fo

r L. a

cido

philu

s ATC

C43

56 (S

LP) (

a) a

nd L

. cas

ei A

TCC

393

(no-

SLP)

(a

’) to

geth

er w

ith th

eir c

orre

spon

ding

adh

esio

n m

aps (

b an

d b’

) obt

aine

d us

ing

an h

ydro

phob

ic A

FM ti

p at

10

mM

KC

l. H

isto

gram

s (c

and

c’) s

how

the

d

istri

butio

n of

adh

esio

n fo

rces

ove

r a se

lect

ed a

rea

of a

bout

800

×800

nm

2 on

the

bact

eria

l cel

l sur

face

.

Dynamic cell surface hydrophobicity of lactobacilli 71

Table 7.1. Summary of contact angles with aqueous, low and high ionic strength solutions for two lactobacillus strains with and without SLP together with the average adhesion force Fadh as probed by hydrophobic and hydrophilic AFM tips. Contact angles represent a mean value of two independent sets of measurements. AFM data are representative of results obtained on three to five cells, using different probes and independent preparations. ± denote the standard deviation associated to the values calculated.

SLP strain no-SLP strain Ionic strength Ionic strength 10 mM 100 mM 10 mM 100 mM Contact angle (degrees) 76 ± 4 47 ± 3 32 ± 5 65 ± 5

Fadh, phobic (nN) -1.34 ± 0.19 -0.88 ± 0.13 -0.91 ± 0.09 -2.31 ± 0.45

Fadh, philic (nN) -0.11 ± 0.02 -2.14 ± 0.17 -1.96 ± 0.33 -0.47 ± 0.12

was found hydrophobic in 10 mM and became more hydrophilic in 100 mM, while the strain without SLP was hydrophilic in 10 mM and became hydrophobic in 100 mM.

The structure of the S-layer on L. acidophilus ATCC4356 is known to be composed of two sub-domains: an external N-terminal region showing predominantly hydrophobic amino acid residues and a C-terminal region, serving to attach the S-layer to the cell wall, which is mainly composed of positively charged hydrophilic residues (Smit et al., 2002). The dynamic cell surface hydrophobicity observed may be explained by a shrinkage of the S-layer due to reduced intra-molecular electrostatic repulsion at high ionic strength. Then, the inner hydrophilic region may become (partly) exposed at the aqueous periphery of the bacterial surface, rendering it more hydrophilic.

L. casei ATCC393 on the other hand, does not possess an S-layer. Yet, its cell surface shows dynamic hydrophobicity as well. X-ray photoelectron spectroscopy indicated that the surface of L. casei ATCC393 is rich in polysaccharides (Van der Mei et al., 2000b). At low ionic strength, this layer presents itself as a hydrophilic polyelectrolyte coating. At high ionic strength the polysaccharide layer is known to collapse and this evidently results in exposing a more hydrophobic surface.

It is interesting, that the dynamic behavior of the cell surface hydrophobicity of the lactobacilli was not only measurable macroscopically by contact angles on bacterial lawns, but also by AFM at a more microscopic level. Stronger interaction forces between the cell surfaces and hydrophobically or hydrophilically modified tips coincide with higher or lower contact angles with aqueous solutions. This is fully in line with surface thermodynamics, stating that hydrophobic surfaces favor interaction with hydrophobic surfaces. Analogously, hydrophilic surfaces show greater affinity for hydrophilic surfaces.

In conclusion, this study is the first to report the dynamic behavior of cell surfaces of lactobacilli with regard to their hydrophobicity in response to changes in environmental ionic strength. Dynamic cell surface hydrophobicity was


demonstrated both at a macroscopic and a more microscopic level by contact angle measurements and AFM, respectively. This dynamic behavior of bacterial cell surfaces upon changes in ionic strength offers to L. acidophilus ATCC4356 and L. casei ATCC393 a versatile mechanism to adhere to hydrophobic and hydrophilic surfaces in low and high ionic strength solutions, respectively.

8 Role of lactobacillus cell surface hydrophobicity as probed by AFM in adhesion to surfaces at low and high ionic strength Virginia Vadillo Rodríguez, Henk J. Busscher, Willem Norde, Joop de Vries and Henny C. van der Mei*

8.1 INTRODUCTION Lactobacilli are found widely distributed throughout nature and have been used for a long time in processing and preservation of food (Holzapfel et al., 1998). In addition, several studies have ascribed probiotic properties to certain lactobacillus strains, which are naturally found in the gastrointestinal or female urogenital tracts of both humans and animals (Herthelius et al., 1989; Sanders, 1993). In all

* Submitted to Colloids and Surfaces B: Biointerfaces


instances mentioned, the behavior of lactic acid bacteria is dependent on interfacial processes and hence on the cell surface physico-chemical properties.

The major surface-exposed protein in many lactobacillus species is located in the S-layer. The S-layer consists of one species of (glyco-)protein, the S-protein, which is assembled into characteristic two-dimensional crystalline layers at the cell surface. This assembly is an entropy-driven process during which individual S-protein monomers form multiple interactions with each other and with the underlying cell envelope (Beveridge, 1994). Although no generalized function is known for bacterial S-layers, their presence at the outermost cell surface suggests a role in adhesion.

Bacterial adhesion to surfaces is determined by an interplay of different physico-chemical surface properties, most notably hydrophobicity and electric charge of the surfaces of the bacterial cell and the substratum. Both the hydrophobicity and the electric charge are consequences of the chemical composition of the surfaces. It is known that SLP conveys hydrophobicity to the lactobacillus cell surface (Van der Mei et al., 2003). Yet, adhesion of lactobacilli to surfaces often does not proceed according to expectations based on their cell surface hydrophobicity and hydrophobic strains do not always adhere best to hydrophobic substrata (Millsap et al., 1996), as predicted by surface thermodynamics (Absolom et al., 1983). This suggests that cell surfaces of lactobacilli may adapt their hydrophobicity in response to environmental changes, like in pH or ionic strength (Vadillo-Rodríguez et al. 2004).

The interaction between bacteria and solid substrata is often described by the Derjaguin-Landau-Verwey-Overbeek (DLVO) theory for colloidal stability (Dejaguin & Landau, 1941; Verwey & Overbeek, 1948). The DLVO theory can be used to calculate the Gibbs energy of interaction between surfaces as a function of their separation distance. The total Gibbs energy of interaction is interpreted as the sum of attractive Lifshitz-Van der Waals interactions and an electrostatic contribution. In general, both bacteria and substratum surfaces are negatively charged and the electrostatic contribution is repulsive as a result of their overlapping electrical double layers (Marshall, 1976). Three different situations may be distinguished, depending on the electrolyte concentration of the surrounding medium. (i) At low electrolyte concentrations, when the double layers are extensive, a large Gibbs energy barrier has to be passed to reach close contact between the two surfaces. Hence, the bacterial cell is effectively repelled from the substratum surface. (ii) At intermediate electrolyte concentration, a (shallow) secondary minimum may be formed at some separation distance (typically 5 to 20 nm), where the organism may be captured in a reversible manner and (iii) at high electrolyte concentrations, the net interaction is attractive at all separation distances and results in a strong usually irreversible adhesion. In practice, the effect of hydrophobicity and electric charge on bacterial adhesion is generally studied by quantification of the number of attached organisms on substrata with different wettabilities and from different ionic strength suspensions (Millsap et al., 1996).

Experimentally, a parallel plate flow chamber is an extremely suitable device

Adhesion and interaction forces of lactobacilli 75

for monitoring the initial stages of bacterial adhesion in situ. Accurate measurements of the kinetics of the adhesion process can be performed and modern image analysis techniques have made it possible to enumerate adhesion and desorption simultaneously during deposition (Meinders et al., 1992; Sjollema et al., 1989). The measured desorption will therefore reflect the thermodynamic reversibility of the process, but may also include a contribution of collisions between flowing and adhering organisms as occurring in natural systems (Van de Ven, 1989).

Atomic force microscope (AFM) has emerged as a valuable tool for probing interaction forces as well as structural and physical properties of living microorganisms under physiological conditions. For instance, the role of hydrophobic interactions was pointed out by Ong et al. (1999) measuring the interaction forces between Escherichia coli-coated probes and solid substrata of different hydrophobicities. It was shown that both attractive forces and cell adhesion were promoted by the hydrophobicity of the substratum surfaces. Using nano-indentation measurements in aqueous solution, Yao et al. (2002) characterized the turgor pressure of several bacterial species, which were found to withstand an internal pressure of 4 to 6×105 Pa. It was also demonstrated in a recent paper that the elasticity of the cell wall of Saccharomyces cerevisiae varies significantly across the cell surface whereas a tenfold increase in Young’s modulus was found for the bud scar, in agreement with the accumulation of chitin in this region of the cell wall (Touhami et al., 2003). In addition, several studies have shown that quantitative force measurements can be obtained by functionalizing AFM probes with self-assembled monolayers (SAMs) of organic thiols terminated with selected terminal groups. Up to date, a variety of thin organic films, polymer surfaces and fungal spores have been characterized by using AFM tips functionalized with hydrophobically and hydrophilically terminated SAMs (Vezenove et al., 1997; Sinniah et al., 1996; Dufrêne, 2000). Recently, information on local isoelectric points have been reported for the yeast surface S. cerevisiae by studying its interaction with AFM probes modified with ionizable groups (Ahimou et al., 2002). Furthermore, it has been suggested that bacterial cell surface hydrophobicity or hydrophilicity can be assessed by comparison of the interaction forces between hydrophobically or hydrophilically modified AFM probes and a bacterial cell surface (Vadillo-Rodríguez et al., 2004).

The aim of this chapter is to investigate the role of lactobacillus cell surface hydrophobicity as probed by functionalized AFM tips, i.e. terminated with hydrophobic (–CH3) or hydrophilic (–OH) groups, on their adhesion to surfaces from low and high ionic strength suspensions. The adhesion of Lactobacillus acidophilus ATCC4356 and Lactobacillus casei ATCC393 with and without SLP, respectively, has been measured in a parallel plate flow chamber from low and high ionic strength suspensions to similarly functionalized glass surfaces and discussed against the background of the DLVO theory and the interaction forces measured through AFM.


8.2 MATERIALS AND METHODS 8.2.1 Bacterial strains, growth conditions and harvesting Two lactobacillus strains, L. acidophilus ATCC4356 and L. casei ATCC393 with and without S-layer respectively, were used in this study. Bacterial strains were streaked and grown for 24 h from frozen stock on MRS (De Man, Rogosa, Sharpe, Merck, Germany) agar plates and incubated at 37 ºC in an atmosphere containing 5 % CO2. Pre-cultures were grown under the same conditions by inoculating MRS broth with a colony from the agar plate. This culture was used to inoculate a second culture that was grown for 16 h prior to harvesting. Bacteria were harvested by centrifugation (5 min at 10,000 g), washed twice with demineralized water and suspended in 10 or 100 mM KCl buffer solution. 8.2.2 Substrata and tip modification Glass plates used in experiments were cleaned by sonicating for 2 min in 2 % RBS35 surfactant solution in water (Omnilabo International BV, The Netherlands), rinsed thoroughly with tap water, dipped in methanol, and again rinsed with demineralized water. Clean glass plates together with long/thin AFM “V”-shaped silicon nitride cantilevers from Digital Instruments Inc. (Santa Barbara, CA) were coated with electron beam thermal evaporation with a 4-nm-thick titanium layer followed by a 30-nm-thick gold layer. The coated substrata and cantilevers were immersed for 18 h in 1 mM solutions of HS(CH2)11OH and HS(CH2)17CH3 in ethanol and then rinsed with ethanol. Functionalized substrata and probes were always used immediately after preparation. In order to validate the quality of the surface modification, water contact angles were measured on functionalized glass surfaces. Water contact angles amounted to 10 – 20 degrees and 90 – 100 degrees for hydroxyl (–OH) and methyl (–CH3) terminated surfaces, respectively. 8.2.3 Parallel plate flow chamber and data analysis The parallel plate flow chamber (internal dimensions: 76×38×0.6 mm) and image analysis system have been described in detail previously (Busscher & Van der Mei, 1995). Images were taken from the bottom functionalized glass plate (55×38 mm) of the parallel plate flow chamber. The top plate of the chamber was made of glass and it was cleaned as described above. The flow chamber was cleaned with Extran (Merck, Germany) and thoroughly rinsed with water and demineralized water. Prior to each experiment, all tubes and the flow chamber were filled with 10 mM or 100 mM KCl solution, taking care to remove all air bubbles from the system. Once the system was filled, a bacterial suspension of 7.5×107 cells ml-1 in 10 mM or 100 mM KCl was allowed to flow through the system at a flow rate of 1.44 ml min-1, corresponding with a Reynolds number of 0.6 and a wall shear rate of 10.6 s-1. Deposition was observed with a CCD-MXRi camera (High technology) mounted on a phase-contrast microscope (Olympus BH-2) equipped with a 40x ultra-long-working distance lens (Olympus ULWD-CD Plan 40 PL). The camera was coupled to an image analyser (TEA; Difa). The bacterial suspension was perfused through the system for 4 h with re-circulation at room temperature.


The total number of adhering bacteria per unit area n(t) was recorded as a function of time by image sequence analysis during 4 h and the affinity of an organism for the glass surface was expressed as an initial deposition rate j0, representing the initial increase of n(t) with time (see Figure 8.1). Note that since the initial deposition rate is extracted only from the initial adhesion data, it represents the affinity of the organisms for the substratum surface without intervening influences of interactions between adhering bacteria. From the total number of adhering bacteria per unit area as function of time n(t) and the number of particles adhering over the full duration of an experiment n∞,, the so-called characteristic adhesion time τ was calculated using

n(t) = n∞ (1-e-t/τ) (8.1)

The characteristic adhesion time τ is determined by a combination of deposition, blocking and desorption, according to

1/τ = β + j0Ab (8.2)

in which Ab represents the area on the substratum surface blocked by one adhering bacterium and β is the desorption rate coefficient (Bos et al., 1999). In order to obtain the two unknowns Ab and β from Eq. 2, blocked areas were derived from the radial pair distribution g(r), as can be calculated from the spatial arrangement of the adhering bacteria (Sjollema & Busscher, 1990). Inserting the blocked areas Ab derived from the radial distribution functions, and the characteristic adhesion times τ from the measured kinetics, the desorption rate coefficients β can be directly calculated from Eq. 8.2.

All values given in this chapter are the average of three experiments carried out with separately grown microorganisms. 8.2.4 AFM Sample preparation. Bacterial cells were attached through electrostatic interactions to a glass plate made positively charged through adsorption of poly-L-lysine hydrobromide. The glass plates were first cleaned as previously described, after which a drop of 0.01 % (wt/vol) poly-L-lysine hydrobromide solution was added. After air-drying, the plate was rinsed with demineralized water and dipped into a bacterial suspension of concentration 105 per ml. After 15 min, the bacteria-coated glass was rinsed to remove loosely attached bacteria and transferred to the AFM. AFM measurements. AFM measurements were made at room temperature under 10 and 100 mM KCl solution using an optical level microscope (Nanoscope III Digital Instrument). An array of 32×32 force-distances curves with z-displacements of 100 – 200 nm at z-scan rates ≅10 Hz were collected over the entire field of view when a bacterium was imaged (see Figure 8.2a and 8.2a’ as an example). The slopes of the retraction force curves in the region where probe and sample are in contact were used to convert the voltage into cantilever deflection. The conversion of deflection into force was carried out as has been previously described by others (Dufrêne et


a)

a

Time (s)

0 2000 4000 6000 8000 10000 12000 14000 16000

Num

ber o

f bac

teria

(106 cm

-2)

0.0

0.5

1.0

1.5

2.0

2.5

3.0n ∞

j0

b)

B

Time (s)

0 2000 4000 6000 8000 10000 12000 14000 16000

Num

ber o

f bac

teria

(106 cm

-2)

0.0

0.5

1.0

1.5

2.0

2.5

3.0

n∞

j0

Figure 8.1 Number of bacteria deposited on hydrophilically functionalized glass as a function of time as determined in the parallel plate flow chamber for L. acidophilus ATCC4356 (with SLP) (a) and L. casei ATCC393 (no SLP) (b) in 10 mM KCl solution.


al., 2001) using a nominal spring constant for the functionalized tips of 0.21 N m-1, as determined by the Cleveland method (Cleveland et al., 1993). Force-distance curves taken over the top of each bacterium studied (see Figure 8.3 as an example) were analyzed in order to determine various characteristic parameters. Approach curves were fitted to an exponential function, where the interaction force F is described as F=F0 exp (-d/λ), in which F0 is the force at zero separation between the interacting surfaces, d the separation distance and λ the decay length of the interaction force F. The retracting curves were used to generate adhesion maps. Adhesion maps were produced by taking the most negative force detected during the retraction curve (Figure 8.3) as the value for adhesion and by plotting that value against x-y position of each force-distance curves (Figure 8.2b and 8.2b’). From the adhesion maps a selected area of ∼800×800 nm2 over the top of each bacterium was used to generate an adhesion distribution histogram (Figure 8.2c and 8.2c’) from which an average adhesion force Fadh was calculated between functionalized AFM tips and the bacterial cell surface for each experimental condition studied. Three to five different organisms were studied for each particular case. 8.2.5 DLVO theory The Gibbs energy of interaction G(h) between the bacterial surfaces and the functionalized glass substrata were calculated according to DLVO theory for colloidal stability using a sphere-plane geometry. Mathematical expressions quantifying the Lifshitz-Van der Waals and electrostatic interactions can be found elsewhere (Norde & Lyklema, 1989). A Hamaker constant of 6×10-21 J was used, consistent with earlier work on the interaction between lactobacilli and glass across water (Boonaert et al., 2001). Zeta potentials of functionalized glass were calculated from measured streaming potentials in a home-made parallel plate flow chamber (Van Wagenen & Andrade, 1980) yielding values of -29 mV and -24 mV for substrata terminated with –CH3 groups and -78 mV and -64 mV for substrata terminated with –OH groups, in 10 mM and 100 mM KCl solutions, respectively (original experimental values for the streaming potential of hydroxyl terminated substrata were re-calculated to account for the conductivity contribution of the underlying gold film based on Schweiss et al., 2001). Bacterial zeta potentials at low and high ionic strengths were derived from measured bacterial electrophoretic mobilities at 25 °C with a Lazer Zee Meter (PenKem, Bedford Hills, NY, USA) equipped with an image analysis option for tracking and zeta sizing using the Smoluchowski theory (Wit et al., 1997). The bacterial zeta potentials were -14 mV and -5 mV for L. acidophilus ATCC4356 and -7 mV and -2 mV for L. casei ATCC393, in 10 mM and 100 mM KCl solutions, respectively. 8.3 RESULTS Figure 8.1 shows the initial deposition rate of L. acidophilus ATCC4356 (a) and L. casei ATCC393 (b) on hydrophilically functionalized glass at 10 mM. The deposition behavior of both lactobacilli at all conditions studied followed an exponential rise in time to a maximum value n∞, of which the linear part allowed

0 nN

-3 n

N

SLP

stra

in

a)

0 µ

m

5

µm

b)

c)

01020304050

-1.40

-0.60

-0.26

-0.17

-0.11

-0.08

-0.04

0.00

Forc

e (n

N)

# events

0 nN

-3 n

N

no-S

LP st

rain

a’

)

0 µm

3 µm

b’)

c’)

0102030405060

-5.2

-4.1

-3.4

-2.6

-1.9

-1.2

-0.4

Forc

e (nN

)

# events

Fi

gure

8.2

Arr

ay o

f 32×

32 fo

rce

dist

ance

cur

ves o

ver t

he A

FM fi

eld

of v

iew

for L

. aci

doph

ilus A

TCC

4356

(with

SLP

) (a)

and

L. c

asei

ATC

C39

3 (n

o SL

P) (a

’) to

geth

er w

ith th

eir c

orre

spon

ding

adh

esio

n m

aps

(b a

nd b

’) o

btai

ned

usin

g a

hydr

ophi

lic A

FM ti

p in

10

mM

KC

l. H

isto

gram

s (c

an

d c’

) sho

w th

e di

strib

utio

n of

adh

esio

n fo

rces

ove

r a se

lect

ed a

rea

of a

bout

800

×800

nm

2 on

the

bact

eria

l cel

l sur

face

.


a)


0 50 100 150 200 250 300 350 400 450 500

Forc

e (n

N)

-3

0

3

6

9

12

15

Fadh

b)


0 50 100 150 200 250 300 350 400 450 500

Forc

e (n

N)

-3

0

3

6

9

12

15

Fadh

Figure 8.3. Example of force-distance curves for L. acidophilus ATCC4356 (with SLP) (a) and L. casei ATCC393 (no SLP) (b) interacting with a hydrophilic AFM tip in 10 mM KCl. The maximum adhesion force Fadh probed upon retraction is indicated in the graph.


a)

Gmax


0 10 20 30 40 50

G (k

T)

-30

0

30

60

90

120

150

Gsm

DGmax

b)

Gmax


0 10 20 30 40 50

G (k

T)

-30

-15

0

15

30

45

Gsm

DGmax

Figure 8.4. Gibbs energy curves as estimated based on the DLVO theory for colloidal stability for L. acidophilus ATCC4356 (with SLP) (a) and L. casei ATCC393 (no SLP) (b) strains interacting with hydrophilically functionalized glass in 10 mM KCl.


calculation of the initial deposition rate j0. The initial deposition rates together with the blocked areas as calculated from the radial pair distribution function (data not shown) were used to estimate the desorption rate coefficients β. The results of the flow experiments are summarized in Table 8.1. Generally, both lactobacilli show similar adhesion patterns at the two ionic strengths. At closer look, their initial deposition rate is observed to be lower when adhering to hydrophilically functionalized glass at 10 mM. The total number of adhering bacteria after 4 h of flow n∞ ranges from 2.2×106 to 6.0×106 cm-2 between the conditions studied and tends to be higher at elevated high ionic strength. The desorption rate coefficients β were only calculated for the strain lacking the SLP. Their values are small, albeit for the hydrophilic surface significantly higher than for the hydrophobic surface.

Figure 8.3 presents two examples of interactions forces measured between the strain possessing (a) and lacking (b) SLP and a hydrophilically functionalized AFM tip at 10 mM. The approach curve shows for both strains and all experimental conditions, a gradually increasing repulsion between the tip and the cell surface that increases exponentially at close approach until contact. Repulsive forces F0 at contact as well as the force decay lengths λ are summarized in Table 8.1 as calculated over a selected area of about 800×800 nm2 over the top of each bacterium studied. Interestingly, the strain with SLP tends to present higher values for both F0 and λ than the strain lacking SLP. Retraction of functionalized AFM tips from the bacterial surfaces always shows a local maximum in attractive adhesion force Fadh (Figure 8.3). Adhesion maps show a heterogeneous surface distribution of these adhesion forces for both strains regardless of ionic strength (see Figure 8.2b and 8.2b’). The interaction forces detected were averaged per strain into an adhesion force Fadh based on the histograms showing their surface distribution (see Figure 8.2c and 8.2c’). Values obtained for Fadh are presented in Table 8.1. As a general rule, adhesion forces Fadh are stronger for the combination of a hydrophobic bacterium/hydrophobic tip and a hydrophilic bacterium/hydrophilic tip regardless of ionic strength.

In order to compare our results with expectations based on the DLVO theory, the total interaction Gibbs energy as a function of the separation distance was plotted for all the bacterium/substratum combinations at the two ionic strengths investigated (see Figure 8.4a and 8.4b as example). At 10 mM, for all combinations, the energy curve shows a shallow, secondary interaction minimum at separation distances between the bacteria and a substratum of 15 to 21 nm (see Table 8.1 summarizing some characteristics of the Gibbs energy curves obtained). The depth of these secondary minima varies between -5 and -7 kT, yielding for each bacterial strain the least negative values when interacting with hydrophilic substrata. The interaction Gibbs energy shows a barrier ranging between 18 kT and 161 kT at separation distance varying between 2 – 5 nm. At 100 mM the net interaction is attractive at all separation distances and for all the combinations studied. Nevertheless, bacterial desorption is still observed at conditions of 100 mM ionic strength.


Table 8.1 Summary of quantitative data describing adhesion of a lactobacillus strain with or without SLP to hydrophobic and hydrophilic substrata at low and high ionic strength, including the initial deposition rate j0, total number of bacteria adhering after 4 h, n∞, desorption rate coefficients β, force at zero separation distance F0, decay length of the force λ (both from approach lines of AFM force-distance curves), and adhesion forces Fadh (from retraction lines). In addition, the Gibbs energy Gsm and location Dsm of the secondary minimum are given together with the position DGmax and the magnitude Gmax of the Gibbs energy barrier. A Hamaker constant of 6×10-21 J and a cell radius of 1700 nm were used for the Gibbs energy calculations.

L. acidophilus ATCC4356 L. casei ATCC393 Strain with SLP Strain without SLP

Ionic strength Ionic strength 10 mM 100 mM 10 mM 100 mM

Parallel Plate flow chamber j0, phobic (cm-2s-1) 571 ± 61 350 ± 59 433 ± 110 402 ± 37 j0, philic (cm-2s-1) 454 ± 22 532 ± 186 233 ± 68 427 ± 59 n∞, phobic (106) 4.0 ± 0.5 4.7 ± 0.5 4.0 ± 2.0 6.0 ± 0.2 n∞, philic (106) 2.9 ± 0.2 5.0 ± 0.6 2.2 ± 1.0 5.1 ± 0.6 β, phobic (10-5s-1) -a - a 7 ± 2 8 ± 1 β, philic (10-5s-1) - a - a 12 ± 2 11 ± 3

AFM

F0, phobic (nN) 24 ± 7 22 ± 8 14 ± 2 15 ± 1 F0, philic (nN) 29 ± 7 22 ± 1 15 ± 1 15 ± 3 λ, phobic (nm) 30 ± 10 34 ± 12 10 ± 1 11 ± 1 λ, philic (nm) 12 ± 2 18 ± 2 14 ± 1 9 ± 2 Fadh, phobic (nN) -1.3 ± 0.2 -0.9 ± 0.1 -0.9 ± 0.1 -2.3 ± 0.5 Fadh, philic (nN) -0.11 ± 0.02 -2.1 ± 0.2 -2.0 ± 0.3 -0.5 ± 0.1

DLVO

Gsm, phobic (kT) -6 -b -7 -b

Gsm, philic (kT) -5 -b -5.5 -b

-b -b

Dsm, phobic (nm) 18 -b 15 -b

Dsm, philic (nm) 21 -b 20 -b

-b -b

Gmax, phobic (kT) 95 -b 18 -b

Gmax, philic (kT) 161 -b 40 -b

-b -b

DGmax, phobic (kT) 2 -b 3.4 -b

DGmax, philic (kT) 3.4 -b 5.4 -b

a desorption rate coefficients β could not be estimated for these strains due to their long rod-shape, impeding calculation of the area blocked by an adhering organisms b at 100 mM the total Gibbs energy curve is attractive at all separation distance regardless of the strain considered


8.4 DISCUSSION The mechanism by which lactobacilli exert their protective functions in the intestinal and urogenital tract (Holzapfel et al., 1998; Herthelius et al., 1989; Sanders, 1993) is not yet known, but in order to offer their protection in a specific habitat it is essential that lactobacilli adhere well to surfaces. The presence of a regularly ordered, planar array of proteinaceous subunits (S-layer) at the outermost surface of some lactobacillus strains implies a likely role of S-layers in adhesion. S-layers of Lactobacillus subsp., for instance, have been shown to interact with receptors on host epithelial cells, thereby blocking receptor sites on the mucosal surfaces for the adherence of pathogenic species (Borinski & Holt, 1990).

It has been known that S-layers convey hydrophobicity to the lactobacilli cell surface. However, it was pointed out that strains with an SLP do not necessarily adhere better to hydrophobic substrata than strains without SLP (Van der Mei et al., 2003) and a reversal of the cell surface hydrophobicity was recorded for both L. acidophilus ATCC4356 and L. casei ATCC393 upon increasing the ionic strength. The lactobacillus strain with SLP was found hydrophobic in 10 mM and became more hydrophilic in 100 mM, while the strain without SLP was hydrophilic in 10 mM and became hydrophobic in 100 mM (Vadillo-Rodríguez et al., 2004). In a recent paper (Machado et al., 2004), it was observed that cultures of the strain L. casei ATCC393 grown under hyperosmotic conditions showed a significantly higher hydrophobicity than control cultures, which in turn, were rather hydrophilic. It was suggested that H3DG and AcylH3DG glycolipids present in the membrane envelope are related to the increased cell surface hydrophobicity induced by the external ionic strength of the medium. In addition, some L. acidophilus species have been reported as well to be sensitive to ionic strength upon changes in the permeability properties of the cell membrane, and hence, with a probable impact on the physico-chemical properties of the bacterial cell surfaces (Fernandez Murga et al., 1999). Therefore, these studies suggest that lactobacilli cell surfaces may adapt their cell surface hydrophobicity in response to environmental changes with a likely potential effect on their adhesion behavior.

In this study it is observed that the ability of L. acidophilus ATCC4356 and L. casei ATCC393 with and without SLP respectively, to adhere to hydrophobic and hydrophilic substrata is similar in low or high ionic strength media. Yet, at 10 mM ionic strength both strains showed lower initial deposition rates onto hydrophilic substrata. Considering that, according to DLVO theory, energy barriers as high as 161 kT up to 40 kT need to be overcome for L. acidophilus ATCC4356 and L. casei ATCC393, respectively in order to adhere in the primary minimum, it is very likely under those conditions the bacteria deposit in the secondary interaction minimum of the energy curve. The higher initial deposition rate found between these bacteria and the hydrophobic substratum at 10 mM could therefore be explained based on the deeper secondary minima (see Table 8.1). As well, the desorption rate coefficient β measured for the non-SLP strain is lower for the hydrophobic substratum than hydrophilic ones.


At 100 mM the DLVO theory predicts a net attractive force at all separation distances for all bacterium/substratum combinations studied. Therefore, primary minimum interaction between bacteria and substrata is expected. The data in Table 8.1 reveal that saturation values n∞ are indeed slightly higher at 100 mM. Moreover, the mass transport in the parallel plate flow chamber can be calculated by solving the convective-diffusion equation, which describes mass transport in terms of convection, diffusion and the interaction forces operating. Solving the convective-diffusion equation needs the assumption that attractive Lifshitz-Van der Waals forces between a particle and a substratum surface are counterbalanced by the hydrodynamic drag that a particle experiences when approaching a substratum surface, while electrostatic interactions are neglected (Bos et al., 1999). Accordingly, a theoretical deposition rate of 566 cm-2 s-1 was found for our lactobacillus strains assuming a bacterial hydrodynamic radius of 1700 nm. Considering that the strains studied are not spherical but rod-shaped bacteria, the actual theoretical deposition rate may slightly deviate from the value calculated. It seems that the deposition efficiency i.e. the probability that arrival of a bacterium to a substratum will actually result in deposition, is close to unity for nearly all combinations. Hence essentially all bacteria arriving at a substratum surface will deposit either in the secondary minimum (low ionic strength) or primary minimum (high ionic strength). It would be expected that bacteria have more difficulty to desorb from the primary minimum than from the secondary minimum. However, this expectation is not corroborated by the values obtained or the desorption probabilities β. Possibly, the desorption does not reflect thermodynamic stability but is primarily determined by collisions between flowing and adhering bacteria. It is also possible that the DLVO theory, as it does not take into account short range interactions, is not applicable to predict the interaction at short separation ( < 1nm).

Furthermore, it is observed that the magnitude of the forces detected upon approach of functionalized AFM tips to the bacterial cell surfaces as well as the distances over which such forces extend are not consistent with the DLVO model. AFM recorded a repulsive force F0 regardless of ionic strength. In AFM, the contact between the surfaces interacting is imposed and it is likely that they interact in the primary minimum. At low ionic strength a repulsive force between the bacterial cell surfaces and the AFM tips was expected. The total force needed for an organism to overcome the energy barriers predicted by DLVO at 10 mM can be calculated to amount to 10-11 N for both lactobacillus strains. This force is three orders of magnitude lower than the force recorded by AFM. In addition, the decay lengths associated to AFM forces extend over much larger distances than theoretically predicted. Therefore, it seems that AFM fails to detect DLVO forces. Note that the total force needed for the AFM to overcome in order to reach close contact with the bacterial cell surfaces is always larger for the strain with SLP than the strain without SLP. These observations lead to the hypothesis that the AFM tips likely probe mechanical cell surface properties upon approach, including electrosteric repulsions in addition to interaction forces. Consequently, the results obtained would reveal a harder coat for the strain having SLP, in agreement with


the proteinaceous crystalline structure known to be present at the outermost surface of these strains.

Upon retraction from the bacterial cell surfaces, the average adhesion force detected by the AFM tips is stronger for the combinations of hydrophobic bacterium/hydrophobic substratum and hydrophilic bacterium/hydrophilic substratum than for the hydrophobic/hydrophilic combinations. These adhesion strengths are not reflected in the macroscopic adhesion data (e.g. the desorption probability) neither at low nor high ionic strength. It is likely that the force applied when a bacterium comes into contact with a substratum as occurs in the AFM experiments influences its adhesion.

In summary, it is concluded that both L. acidophilus ATCC4356 and L. casei ATCC393 with and without SLP showed higher initial deposition rate to hydrophobic substrata at low ionic strength. At high ionic strength, a reversal of the cell surface hydrophobicity has been detected for both strains using functionalized AFM tips and no preferential macroscopic adhesion is recorded to either hydrophobic or hydrophilic substrata. This observation agrees with previous results on the adhesion of some lactobacillus strains which indicated that strains possessing a SLP do not always adhere better through hydrophobic interactions to hexadecane than strains without SLP (Van der Mei et al., 2003). Boonaert et al. (2001) experienced as well that, in water, the adhesion behavior of another hydrophilic no-SLP lactobacillus strain could not be explained based on the cell surface hydrophobicity, since bacteria adhered in higher numbers to hydrophobic polystyrene than to hydrophilic glass. In addition, AFM measurements failed to detect DLVO forces upon approach of the functionalized tips to the bacterial cell surfaces. Upon their retraction, stronger average adhesion forces were found for the combination hydrophobic bacterium/hydrophobic substratum and hydrophilic bacterium/hydrophilic substratum, independent of ionic strength. These strong adhesion forces are not reflected in the macroscopic observations on bacterial adhesion which suggest that the force applied when a bacterium comes into contact with a substratum influences its adhesion mode.

General Discussion

Bacteria are, with respect to their dimensions, colloidal particles, and hence the process of bacterial adhesion is commonly described by the Derjaguin-Landau-Verweij-Overbeek (DLVO) theory of colloidal stability. The DLVO theory estimates the interaction forces between the interacting surfaces based on Lifshitz-Van der Waals (LW) and electrostatic (EL) interactions and their decay length with separation distance. The most important parameter determining the LW interaction is the Hamaker constant, which is a material property. The value of the Hamaker constant is in most cases rather uncertain, yet for microbial cell surfaces it is currently being estimated from contact angles measurements on bacterial lawns. On the other hand, calculation of electrostatic interactions requires knowledge of the electrostatic surface potential of the interacting surfaces. The electrostatic surface potential of microbial cell surfaces cannot be measured directly but is commonly estimated from measured electrophoretic or electro-osmotic velocities observed during dynamic measurements in electric fields. Yet, application of DLVO theory to explain bacterial adhesion to solid substrata has only been successful for a


limited number of strains and species. In general, the degree of success of the DLVO theory to explain bacterial adhesion frequently decreases as the complexity of the cell surface of the organisms under consideration increases. It is important to realize that bacterial cell surfaces are both chemically and structurally more complex and heterogeneous than most inert colloidal particles. Therefore, a need exists for microscopic characterization of the physico-chemical properties of bacterial cell surfaces. Subsequently, a generalized physico-chemical theory to account for bacterial adhesion to substratum surfaces might become in reach.

The introduction of the atomic force microscope (AFM) and its application to biological surfaces has offered new possibilities to obtain microscopic physico-chemical properties of bacterial cell surfaces. Using AFM, the interaction force between a single bacterium and a solid substratum (such as an AFM probe) can be directly measured down to the nanoNewton range. Yet, interpretation of the force involved as well as distinction between its different components remains difficult. The main challenge in extracting tip-surface forces is the exact determination of the distance origin relative to the tip location. Bacterial deformation under the applied load or tip penetration through extracellular material could likely overestimate the distance range at which interaction forces are measured. The most current approach to interpret AFM force-distance curves is still to assume that the bacterium is linearly elastic and that all non-linearity displayed on the curves is a consequence of electrostatic and (electro)steric surface forces (Razatos et al., 1998; Ong et al., 1999). It is worthy to note that poor agreement has been found up to date between AFM force-distance curve interpretations and bacterial adhesion to solid substrata as macroscopically observed (Burks et al., 2003). In addition, attempts to fit DLVO theory with AFM data have mostly failed since AFM often predicts higher forces operating over larger distances than theoretically expected (Camesano & Logan, 2000).

This thesis represents a further attempt to correlate microscopic surface properties of bacterial species as derived from force-distance curve analyses to their macroscopic adhesion to solid substrata as directly observed under well-defined flow conditions. In order to formulate generic relations, bacterial species differing with respect to their outermost surface were considered along the subsequent studies, i.e. an entire collection of fibrillated Streptococcus mitis strains, an extracellular polymeric substances (EPS)-producing Streptococcus thermophilus B strain and two lactobacillus strains with and without a surface layer protein (SLP) anchored to their cell envelope. In general lines, it was observed that

a) approach lines of force-distance curves correspond with the rate of initial macroscopic adhesion of the organisms to a substratum

b) the maximum distance over which attractive forces are probed by AFM upon retraction of the tip are influential on the area blocked by an adhering bacterium on a substratum

c) bacterial desorption from solid substrata does not show a relation with adhesive forces as probed by AFM unless the contact time between bacterium and substratum is taken into account.

General Discussion 91

These three aspects will now subsequently be discussed. Approach lines of force-distance curves correspond with the rate of initial, macroscopic adhesion of the organisms to a substratum

Although the energy barrier detected upon approach of the AFM tip to the bacterial cell surface could be interpreted as an activation energy for the organisms to successfully deposit, the magnitude of the activation energies observed ranged between 2400 kT and 200000 kT (for S. mitis 244 in water and the lactobacillus strain with SLP in 100 mM KCl solution) among all strains considered. These activation energies are prohibitively high for their spontaneous adhesion observed on macroscopic substrata under laminar flow conditions. However, considering that most of the bacterial strains studied here bear extracellular material at their cell surface envelope, the following experimental situation could have been encountered during AFM:


← approach

Tip

defle

ctio

n (n

m)

(ii) (i)

(iii)

(v) (iv)

L0

(v)

(iv) (iii) (ii) (i)

H20 H20

→ bacterium bearing an extracellular coating → AFM tip

(i) At a long separation distance between the tip and the bacterial surface, the force experienced by the AFM tip is zero. (ii) As the tip approaches the bacterial surface, the cantilever may bend upwards due to a repulsive force. (iii) At a certain distance, the tip gets into direct contact with the extracellular material known to surround the cell envelope. After this first contact, the repulsion observed is due to a nonlinear deformation dominated by the initial penetration of the AFM tip through the


extracellular material. (iv) Subsequently, a few layers of water may need to be removed before intimate contact with the peptidoglycan layer is eventually reached. (v) Finally, deformation of the peptidoglycan layer could take place. This indicates that, the major part of the activation energy estimated belongs to a nonlinear deformation of the bacterial extracellular material. In the particular case of S. mitis strains, for instance, the non-linearity likely reveals an increased gradient on fibrillated material density towards the peptidoglycan layer. Additionally, the highest energy barriers detected for the lactobacillus strains correspond with a mechanically harder coat surrounding the envelope of these specific bacterial strains.

At high ionic strength, the DLVO theory for colloidal stability predicts for all S. mitis as well as for both lactobacillus strains studied a net attractive force at all separation distances between the bacterium and a substratum. Consequently and based on the previous interpretation of force-distance curves, the total force needed for the AFM tip to overcome in order to establish close contact with the bacterial cell surface would represent an estimation of the rigidity of the extracellular material surrounding the cell envelope whereas the distance range associated to such a force would make reference to the thickness of the extracellular material. Thickness values ranging from 50 to 314 nm were found for the polyelectrolyte layer surrounding S. mitis strains. Even though these strains are characterized as possessing extremely long fibrils at their cell surfaces, hydrodynamic measurements have revealed that long appendages of fibrillated strains collapse onto the cell surface upon increasing the ionic strength (Van der Mei et al., 2000a). Therefore, Figure 5.5 in Chapter 5 (showing that the rate of initial, macroscopic adhesion of S. mitis to a substratum increases as the force needed to achieve contact between the AFM tip and bacterial cell surface decreases) could be re-interpreted as follows:

a compressed and more rigid polymer layer results in higher bacterial adhesion. This phenomenon has been previously pointed out by Abu-Lail & Camesano (2003b) studying the role of biopolymer conformation on the adhesion of Pseudomonas putida KT2442. Additionally, the fibrillated Streptococcus salivarius HB has been reported to adhere in higher numbers (regardless of the substratum wettabilitty) as the ionic strength of the medium increases (Sjollema et al., 1990a). This observation likely corresponds once more with a compressed and more rigid surrounding polyelectrolyte layer. Results quantifying the macroscopic adhesion behavior of the lactobacillus strains at high ionic strength overlapped and hence, conclusions cannot be drawn to predict adhesion as function of the rigidity of their outermost surface. Nevertheless, AFM reveals a harder coat for the strain having SLP, in agreement with the proteinaceous crystalline structure known to be present at the outermost surface of these strains. Remarkably, the distances associated to the AFM approach lines of these bacteria are higher (56 -214 nm) than the ones associated to the fibrillated material collapsed onto the cell surfaces for several S. mitis strains. Both lactobacillus strains have been reported to be sensitive to hyperosmotic conditions regarding secretion of extracellular substances


(Machado et al., 2004; Fernandez Murga et al., 1999). A thick extracellular layer present as a result of the bacterial immobilization method used for immobilization in AFM, i.e. physical adsorption on positively charged glass, could account for the larger distances over which forces are recorded for the AFM tip upon approach to both lactobacillus strains. This phenomenon was pointed out in Chapter 3 comparing the interaction forces obtained between an EPS-producing Klebsiella terrigena strain and the silicon nitride tip of the AFM for three immobilization methods. Based on the results obtained, it was concluded that physical immobilization on a positively charged surface likely stimulates the secretion of EPS by K. terrigena.

At low ionic strength, it was shown that bacterial deposition probably takes place in the secondary minimum of the total Gibbs energy. Interestingly, in 10 mM KCl the AFM tip recorded a repulsion force regardless the strain studied. The presence of extracellular material implies that electrosteric repulsive interactions dominate the DLVO forces. Van der Mei et al. (2000a) have reported AFM characterization of a mechanically trapped ‘bald’ Streptococcus salivarius HBC12 in water. Approach lines were always repulsive (acting range ~21 nm) and no attraction was recorded at the position of the secondary minimum or at short separation distances. AFM likely does not allow to measure the distance dependence of attractive forces at short separation distances. When the gradient of the attractive forces exceeds the spring constant of the cantilever plus the gradient of possible repulsive forces, an instability may occur and at a certain distance the tip jumps onto contact with the surface. In addition, the attractive force predicted by the DLVO theory prior to reaching the secondary interaction minimum can be calculated from the slope of the distance dependence of the total Gibbs energy and amounts to 10-13 N for S. salivarius HBC12 interacting with hydrophilic, negatively charged glass across a low ionic strength solution. This force may be at least an order of magnitude lower than the force resolution of the AFM. Therefore, AFM likely fails to detect DLVO forces when bacterial cells are investigated.


0 10 20 30 40 50

G (k

T)

-40

-20

0

20

40

DLVOvdWEL

ξ bacterium= -8 mV; ξ substrata= -50 mV; H = 6×10-21 J; a = 500 nm


The maximum distance over which attractive forces are probed by AFM upon retraction of the tip are influential on the area blocked by an adhering bacterium on a substratum

This conclusion is based on the hypothesis that multiple adhesion forces upon retraction of the AFM tip from bacterial cell surfaces are due to multiple cell surface polymers adhering to and detaching from the AFM tip. This hypothesis is also supported in the literature (Abu-Lail & Camesano, 2003b) and here by the observation that the number of local maxima in adhesion forces was not very sensitive to ionic strength when the fibrillated S. mitis strains were investigated. In consequence, at the distance over which the bacterium exerts adhesive forces through the extension of fibrils of different lengths, it is likely that bacteria are brought closer together and the distance between adhering bacteria is reduced, yielding smaller blocked areas (shown on Figure 5.6, Chapter 5).

→ area blocked for an adhering bacterium → a fibrillated bacterium

This mechanism has been observed by AFM at a more molecular level for flexible rod-like proteins associated to microtubules found in the cytoplasm of eukaryotic cells. The projection domain of these proteins was identified as cross-bridges between microtubules based on the distance range of the forces probed by AFM (Mukhopadhyay & Hoh, 2001).

Bacterial desorption from solid substrata does not show a relation with adhesive forces as probed by AFM unless the contact time between bacterium and substratum is taken into account

Although the adhesion force measured by the AFM tip upon retraction was expected to be indicative for the desorption of bacteria in the parallel plate flow chamber, no clear relation could be found when S. mitis or lactobacillus strains were studied. It was hypothesized that this has to do with the nature of the desorption process. Desorption in the parallel plate flow chamber takes place as a spontaneous process under prevailing shear conditions, while in AFM the contact between the bacterium and surface substratum is forced to break by application of


an external force. Actually, it was shown in Chapter 6 that spontaneous desorption is related with the adhesion forces probed by the AFM if the contact time between bacterium and substratum is taking into account. AFM indicated that bond strengthening between the tip and Streptococcus thermophilus occurred within the same time as bond aging inferred from residence time-dependent desorption on a macroscopic level. In addition, comparison of the interaction energies derived from AFM and macroscopic desorption seems to indicate that bond strengthening arises as a result of multiple attachments of extracellular polymeric substances to a substratum surface. The fact that bridging controls desorption thus became evident.

The main forces controlling spontaneous bacterial desorption are thought to be Van der Waals attraction and bridging events taking place between cell surface macromolecules and the substratum. In order to investigate a possible role of Van der Waals forces to macroscopic spontaneous desorption β, β is plot vs. L0 i.e. the equilibrium length of the bacterial surface polymers as calculated applying the electrosteric model for the collection of S. mitis strains,

L0 (nm)

35 40 45 50 55 60 65 70 75

β (1

0-5s-1

)

1

2

3

4

5

6

7(103 º)

(100 º)

(91 º)

(60 º)

(59 º)

(56º )

(54 º) (53 º)

At first sight, it can be observed that the relation between β and L0 appears strongly influenced by the bacterial cell surface hydrophobicity (expressed between brackets in the graph) whereas 60 degrees appears to be a critical contact angle for strong desorption to occur. It has been reported that at high ionic strength steric interactions determined by cell surface macromolecules dominate adhesion (Rijnaarts et al., 1999). The steric interactions either promote adhesion by bridging or inhibit it by steric repulsion. The surfaces of the S. mitis strains may contain amphiphilic macromolecules with hydrophobic parts would induce steric repulsion on hydrophilic glass (but likely bridging on an hydrophobic substratum) while hydrophilic parts would likely promote bridging. Yet, as a general trend, both for more hydrophobic bacteria and more hydrophilic ones, desorption decreases with increasing thickness of the polyelectrolyte layer.


L0 β ↑

substratum substratum

L0β ↓

Accounting for Van der Waals forces, mainly arising from the body of the bacterial cell, the decrease of desorption β with the increase of L0 could be readily understood if the dependency of Van der Waals forces with the separation distance (in this case coinciding with L0 i.e. the thickness of the polyelectrolyte layer) is taking into account. In addition, note that L0 ranges between 40 and 71 nm. At such distances, Van der Waals forces are known to still play an important role.

Surface hydrophobicity has proven its value in bacterial adhesion studies (Rosenberg & Kjelleberg, 1986). Yet, it has also been observed that adhesion of certain strains to solid substrata often does not proceed according to expectations based on cell surface hydrophobicity and hydrophobic strains do not always adhere best to hydrophobic substrata, as predicted by surface thermodynamics (Van der Mei et al., 2003). Indeed, a definite role for bacterial and/or substratum hydrophobicity in adhesion has not been yet established and the term hydrophobicity remains still poorly defined.

In practice, the effect of cell surface hydrophobicity in adhesion is generally studied by quantification of the number of attached cells on substrata of different wettabilities. Chemical modification of AFM tips with organic thiols terminated with selected terminal groups has likely provided a new avenue to prove the enigmatic ‘hydrophobic interactions’. Accurate quantification of the role of hydrophobic interactions in adhesion processes would require the investigation of simpler systems than bacterial cell surfaces. Nevertheless, when bacterial cells are investigated, short-range ‘hydrophobic interactions’ could be expected to be recorded by hydrophobic and hydrophilic AFM tips either upon approach to or retract from the bacterial cell surface. Upon approach, a short-range hydration repulsion force is expected if both the interacting surfaces are hydrophilic. Hydrophilic surfaces bind a few layers of water molecules and their removal will lead to a repulsion force when the surfaces are forced into intimate contact (corresponds to (iv) in tip deflection vs. separation distance). If both surfaces are hydrophobic, attractive forces are expected. Again, AFM likely does not allow to measure the distance dependence of attractive forces at short separation distances. Upon retract, the average adhesion force detected by the AFM tips has been observed to be stronger for the combinations of a hydrophobic bacterium/hydrophobic substratum and hydrophilic bacterium/hydrophilic substratum than for the hydrophobic/hydrophilic combinations (Chapter 7 & 8).


The strength of hydrophilic interactions may reflect the complementarily and weak interactions between molecular groups on the surfaces interacting, i.e. hydrogen bonds, ionic interactions and ‘hydrophobic interactions’. The formation of each of these weak bonds contributes to a net decrease in the free energy of the system. ‘At short range’ the cumulative effect of many small binding forces could be enormous. Integration of the area under the maximum adhesion peak after contact in the force-distance curve has been reported to yield interaction energies of 104 – 105 kT between the tip and the EPS-producing K. terrigena (Chapter 3). In addition, the strength of hydrophobic interactions also result from the system achieving greater thermodynamic stability by minimizing the number of ordered water molecules, which would surround the hydrophobic surfaces in solution.

Based on related literature and on the results presented here it is concluded that AFM fails to detect long-range DLVO forces when bacterial cell surfaces are investigated whereas the major force contributions recorded are electrosteric, elastic, polymer extension and binding (examples of equations for these interactions using a sphere-plane geometry are presented in the Table below (Heinz & Hoh, 1999b)).

Approach

Retraction

→ Electrosteric

→ Polymer extension

→ Elastic

→ Binding (bond a

0.2 ≤ D/2L0 ≤ 0.9

Fst = 50kBTaLoΓ 3/2exp (-2πh/Lo)

F (δ) = 4E√a δ 3/2/ 3(1-ν)2

→ Ba→ AFM tip (apex approximated to an sphere)

a: radius of probe sphere E: elastic modulus h: separation distance kB: Boltzmann´s constant Lo: brush thickness in a good solvent L*: inverse Langevin function

l0 : monomer length N: number of unit in polymers T: absolute temperature x: elongation of polymer δ: indentation depth Γ: brush density per unit area

Λ: chτ0: recfrequτ: perruptuν: Po

F(x) = (kT/ l0)L* (x/Nl0)

ging)

F= (U-kTln(τ/τ0))/Λ

cterial surface (substratum)

aracteristic length of the bond iprocal of the natural bond

ency iod over which the bond will re isson ratio


Additionally, short range interactions such as hydrogen bonds, ionic interactions and ‘hydrophobic interactions’ might dominate at very short separation distances. Yet, distinction between the components of the total force involved requires accurate knowledge of the exact position of the tip relative to the bacterial cell surface. On any of the force-distance curves recorded for this thesis a clear transition between the different force regimes was observed. Perhaps, performing force-distance curves measurement at lower scan rate and plotting the mathematical gradient (‘rate of change’) of the tip deflection vs. the separation distance (instead of the deflection vs. separation distance) could give a better picture of the transitions taking place.

An accurate AFM microscopic characterization of physico-chemical properties of bacterial cell surfaces might benefit from the use of well-defined geometry probes (i.e. micro-spheres) with known physico-chemical properties. This would likely provide new insight for the development of theoretical models for colloid-like interactions. Additionally, systematic experiments need to be performed with strains differing in their outermost layer (e.g. fibrils, fimbriae, pili, capsules, S-layers). If different bacteria are compared there may be the need to introduce an empirical, strain-dependent factor to account for the surface structures of the cell.

In the near future, functionalization of AFM probes with biomolecules might allow mapping of the distribution of individual specific surface components and specific recognition forces which are known to play an important role in certain microbial adhesion processes. An important challenge will also be to combine force measurements with high resolution imaging on living cells. This will make it possible to correlate force interactions with morphological observations.

Summary

Application of physico-chemical models to explain microbial adhesion to surfaces has been successful for a limited number of strains and species (Chapter 1), owing to the macroscopic nature of the input data (i.e. zeta potentials and contact angles). Microorganisms have, from a physico-chemical point of view, generally been considered to be similar to inert polystyrene particles. However, microorganisms are not smooth particles and, in contrast to polystyrene particles, carry long, usually very thin surface structures protruding from the cell surface and radiating outwards into the surrounding liquid. Yet, the function of these surface structures in specific adhesion processes to different substratum surfaces as well as their influence on overall physico-chemical cell surface characteristics, remains to be identified for most bacterial strains. Therefore, relevant physico-chemical measurements on microbial cell surfaces require a microscopic resolution that can not be accomplished with most currently employed methods.

The introduction of the Atomic force microscope (AFM) and its application to biological surfaces has opened a new avenue to obtain microscopic, physico-chemical properties of bacterial cell surfaces. With a long-term goal of developing


a new theory for bacterial adhesion which will account for more complex interactions (e.g. charge heterogeneity, macromolecular bridging) that are not accounted for in DLVO theory, we initiated this thesis. The aim of this thesis was to reach a microscopic characterization of physico-chemical properties of bacterial cell surfaces as well as to find out if and how microscopic properties could be amalgamated into macroscopic cell surface properties and related to macroscopic bacterial adhesion on solid substrata.

A collection of nine Streptococcus mitis strains had been macroscopically characterized with regard to their cell surface hydrophobicities by water contact angles, and its surface chemical composition using X-ray photoelectron spectroscopy. In order to complete their macroscopic surface characterization, the electrophoretic softness and fixed charge density in the polyelectrolyte layer of S. mitis strains, usually carrying long sparsely distributed fibrils, were determined by the soft particle analysis using measured electrophoretic mobilities as a function of the ionic strength (Chapter 2). In general, S. mitis cell surfaces are electrophoretically soft with negative fixed charged density. Further, a comparison with surfaces of other bacterial strains that are reported to be soft indicates that the Ohshima theory does not provide information on the surface morphology causing the softness. The most likely reason is that the electro-osmotic flow occurs only in the very outer region of thick extracellular surface layers. Nevertheless, determining the surface softness was essential for proper characterization of the cell surface electrostatics.

For AFM experiments, bacterial cells need to be firmly anchored to a substratum surface in order to withstand the friction forces from the silicon nitride tip. In order to probe physico-chemical properties of bacterial cell surfaces under physiological conditions, it is required that immobilization does not affect the chemical and structural integrity of the cell surface. Different strategies for the immobilization of bacteria have been described in literature. Chapter 3 compares AFM interaction forces obtained between Klebsiella terrigena and silicon nitride for three commonly used immobilization methods, i.e. mechanical trapping of bacteria in membrane filters, physical adsorption of negatively charged bacteria to a positively charged surface and glutaraldehyde fixation of bacteria to the tip of the AFM. It was shown that different sample preparation techniques give rise to dissimilar interaction forces. Indeed, physical adsorption of bacterial cells on modified substrata may promote structural rearrangements in bacterial cell surface structures while glutaraldehyde treatment is thought to induce physico-chemical and mechanical changes on bacterial cell surfaces properties. In general, mechanical trapping of single bacterial cells in filters appears to be the most reliable method for immobilization.

After having established the most suitable bacterial immobilization method for AFM, a detailed analysis of the interaction forces between a silicon nitride AFM tip and the surface of the nine different oral bacterial S. mitis strains, was carried out. Chapter 4 presents a first attempt to correlate microscopic and macroscopic bacterial cell surface properties. Interestingly, microscopic features of force-

Summary 101

distance curves could be amalgamated in such a way, that relations between microscopic cell surface properties and macroscopic cell surface properties were obtained, even though these relations were not fully understood.

In Chapter 5, we attempted to correlate microscopic adhesion of S. mitis strains to the negatively charged, hydrophilic, silicon nitride tip of an AFM with macroscopic adhesion of the isolates to a negatively charged, hydrophilic glass in a parallel plate flow chamber. The repulsive force probed by AFM upon approach of the tip to a bacterial cell surface was found to correspond with an activation barrier, governing initial, macroscopic adhesion of the organisms to the glass surface. Moreover, maximum distances at which attractive forces are probed by the AFM upon retraction of the tip related with the area blocked by an adhering bacterium, i.e. the distance kept between adhering bacteria. Bacterial desorption could not be related with adhesive forces as probed by the AFM, possible due to the distinct nature of the desorption process occurring in the parallel plate flow chamber and the forced detachment in AFM.

Microbial desorption had also been studied in situ in controlled flow devices as a function of the organisms resident time on the surface (Meinders et al., 1994). It appeared that desorption of Streptococcus thermophilus decreased strongly within approximately 50 s after initial adhesion due to bond aging. In Chapter 6, bond aging between the S. thermophilus cell surface and the silicon nitride tip of an AFM was corroborated microscopically and related to the macroscopic, residence time-dependent desorption of the organism under flow. AFM indicated bond strengthening between the tip and the cell surface within 100 s of contact, which is in the same order of magnitude as bond aging inferred from residence-time dependent desorption. Additionally, comparison of the interaction energies derived from AFM and macroscopic desorption, seems to indicate that bond strengthening arises as a result of multiple attachments of extracellular polymeric substances to a substratum surface.

Bacterial surface hydrophobicity, generally measured by placing water droplets on carefully prepared and dried microbial lawns, is commonly accepted as influential on bacterial interaction with their environment. However, contact angles measured with liquid droplets on a microbial lawn are essentially representative of a fuzzy coat of cellular surface material, collapsed into a lawn. Therewith results are useful to interpret the long-range interactions between an organism and a substratum surface, but not necessarily for the interpretation of short-range interactions, which may be dominated by structural and chemical cell surface heterogeneities. In Chapter 7, the surfaces of L. acidophilus ATCC4356 and L. casei ATCC393 with and without SLP respectively, have been probed with regard to their interaction forces with chemically functionalized AFM tips, i.e. terminated with hydrophobic (CH3) and hydrophilic (OH) groups at low and high ionic strength solutions. The macroscopic cell surface hydrophobicity of the two strains was also assessed by contact angle measurements using the two ionic strength solutions investigated. In general, high interaction forces with a hydrophilic tip were found to coincide with low contact angles, whereas a cell surface with high


contact angle showed the strongest interaction with a hydrophobic tip. In addition, both strains reversed their hydrophobic nature upon increasing the ionic strength from 10 to 100 mM. It is interesting, that the dynamic behavior of the cell surface hydrophobicity of the lactobacilli was not only measurable macroscopically by contact angles on bacterial lawns, but also by AFM at a more microscopic level. This dynamic behavior of bacterial cell surfaces upon changes in ionic strength offers to L. acidophilus ATCC4356 and L. casei ATCC393 a versatile mechanism to adhere to hydrophobic and hydrophilic surfaces in low and high ionic strength solutions, respectively.

It is commonly found that adhesion of lactobacilli to solid substrata does not proceed according to expectations based on cell surface hydrophobicity. In Chapter 8 the role of cell surface hydrophobicity of the two lactobacillus strains with and without a SLP layer has been investigated with regard to their adhesion to hydrophobically or hydrophilically functionalized glass surfaces under well-defined flow conditions and in low and high ionic strength suspensions. Similarly, the interaction of the lactobacilli with similarly functionalized AFM tips was measured. In a low ionic strength suspension, both lactobacillus strains show higher initial deposition rates to hydrophobic glass than to hydrophilic glass, whereas in a high ionic strength suspension no clear influence of cell surface hydrophobicity on adhesion is observed. Independent of ionic strength, however, AFM detects stronger interaction forces when both bacteria and tip are hydrophobic or hydrophilic than when bacteria and tip have opposite hydrophobicities. These strong adhesion forces are not reflected in macroscopic observations on bacterial adhesion which suggest that the force applied when a bacterium comes into contact with a substratum influences its adhesion mode. In addition, the distance dependence of the total Gibbs energy of interaction could only be qualitatively correlated with bacterial deposition and desorption in the parallel plate flow chamber whereas AFM measurements failed to detect DLVO forces upon approach of the functionalized tips to the bacterial cell surfaces.

In the General Discussion, the main conclusions drawn in the thesis are further discussed, being:

(a) approach lines of force-distance curves correspond with the rate of initial, macroscopic adhesion of the organisms to a substratum,

(b) the maximum distance over which attractive forces are probed by AFM upon retraction of the tip are influential on the area blocked by an adhering bacterium on a substratum and

(c) bacterial desorption from solid substrata does not show a relation with adhesive forces as probed by AFM unless the contact time between bacterium and substratum is taken into account. In addition, some of the limitations of the system regarding accurate interpretation of force-distance curves as well as future perspectives are presented.

Samenvatting

Toepassen van fysisch-chemische modellen om hechting van micro- organismen aan oppervlakken te verklaren is meestal niet succesvol. Alleen voor een klein aantal stammen onder specifieke omstandigheden kunnen deze fysisch-chemische modellen de hechting aan een oppervlak beschrijven (Hoofdstuk1). Een van de redenen waarom deze modellen veelal in gebreke blijven is dat micro-organismen beschouwd worden als inerte polystyreen bolletjes, maar micro-organismen zijn geen gladde deeltjes en hebben vaak lange dunne oppervlakte structuren, die uitsteken vanaf het celoppervlak in de omliggende vloeistof. Deze celwand- structuren spelen hoogstwaarschijnlijk een belangrijke rol in de hechting van micro-organismen aan oppervlakken en zijn dan ook de reden dat de gangbare fysisch-chemische modellen slecht toepasbaar zijn. Om de fysisch-chemische eigenschappen van deze microstructuren te kunnen meten zullen er andere dan de traditionele technieken, zoals zeta potentiaal en randhoekmetingen, gebruikt moeten worden.

De introductie van de Atomic force microscoop (AFM) en de toepassing op biologische oppervlakken heeft de weg geopend om fysisch-chemische


eigenschappen op microscopisch niveau te kunnen meten. Het doel van dit proefschrift is een microscopische karakterisatie van de fysisch-chemische eigenschappen van bacterie-oppervlakken te geven, met als tweede doel deze microscopische eigenschappen te relateren aan macroscopische eigenschappen van een bacterie, zoals hun hechting aan vaste stof oppervlakken.

Een collectie van negen Streptococcus mitis stammen met lange dunne fibrillen aan hun oppervlak zijn in het verleden al macroscopisch gekarakteriseerd door het meten van hun hydrofobiciteit (water randhoekmetingen) en chemische samenstelling (X-ray photoelectron spectroscopy). Om hun macroscopische karakterisatie compleet te maken is ook de elektroforetische hardheid en ladingsverdeling gemeten in de polyelektrolietlaag rondom de cellen van S. mitis stammen. Dit is gedaan door het meten van de electroforetische mobiliteit als een functie van de zoutconcentratie (Hoofdstuk 2). De S. mitis cellen zijn elektroforetisch zacht en hebben een negatieve lading, maar meting van de electroforetische mobiliteit gaf geen verklaring op morphologisch niveau voor de gemeten zachtheid van het bacterie-oppervlak. De meest waarschijnlijke verklaring is dat de electro-osmotische vloeistofstroom alleen plaatsvindt in de buitenste lagen van de dikke extracellulaire laag aan het oppervlak.

Voor AFM experimenten is het van groot belang dat bacterie cellen stevig verankerd worden aan het substraat om de frictie-krachten met de silicium nitride tips te weerstaan. Om de fysisch-chemische eigenschappen van de bacterie oppervlakken onder fysiologische omstandigheden te kunnen meten, is het noodzakelijk dat de immobilisatie-methode geen invloed heeft op de chemische en structurele samenstelling van de bacterie. In de literatuur zijn verschillende immobilisatie-methoden beschreven. Hoofdstuk 3 vergelijkt de AFM interactiekrachten tussen Klebsiella terrigena en een silicium nitride tip voor drie veel gebruikte immobilisatie technieken, nl. het mechanisch vangen van een bacterie in een membraan filter, fysische adsorptie van de negatief geladen bacterie aan een positief geladen oppervlak en glutaardialdehyde fixatie van de bacterie aan de AFM-tip. Hieruit bleek dat verschillende preparatie technieken verschillende interactie krachten opleverden. Fysische adsorptie van bacterie-cellen op gemodificeerde oppervlakken kan structuurveranderingen van het bacterie oppervlak veroorzaken, terwijl glutaardialdehyde behandeling veranderingen teweeg brengt in de fysisch-chemische eigenschappen en ook mechanische veranderingen van het oppervlak veroorzaakt. Het mechanisch vangen van individuele bacteriën in een filter is de meest betrouwbare immobilisatie methode gebleken.

Na het vaststellen van de meest betrouwbare immobilisatie-methode voor AFM, is met deze methode een gedetailleerde studie gedaan naar de interactie-krachten tussen de silicium nitride tip en negen verschillende orale S. mitis stammen. In Hoofdstuk 4 wordt een eerste poging gedaan om microscopische en macroscopische eigenschappen van deze bacterie oppervlakken met elkaar te correleren. Verrassenderwijs konden de microscopische karakteristieken uit de kracht-afstand curves zo worden verklaard dat er relaties werden gevonden tussen

Samenvatting 105

de microscopische en de macroscopische oppervlakte-eigenschappen van de bacterie, ook al waren deze relaties niet altijd volledig te begrijpen.

In Hoofdstuk 5 hebben we geprobeerd een relatie te vinden tussen de microscopische hechting van S. mitis stammen aan de negatief geladen, hydrofiele silicium nitride tip van een AFM en de macroscopische hechting van de bacteriën aan een negatief geladen, hydrofiel glasoppervlak in de parallelle plaat stroomkamer. De afstotende kracht gemeten met AFM tijdens het naderen van de tip naar een bacterie oppervlak kwam overeen met de aktivatie drempel, die de snelheid van de initiële hechting van het organisme aan het glasoppervlak bepaalt. Ook is er een relatie gevonden tussen de maximum afstand waarover krachten werden gemeten tussen de AFM tip en de bacterie en de afstand tussen gehechte bacteriën op een oppervlak. Desorptie van gehechte bacteriën kon niet worden gerelateerd aan de aantrekkende krachten zoals die gevonden worden met AFM, waarschijnlijk omdat het desorptie proces, zoals dat in de parallelle plaat stroomkamer plaatsvindt, aanmerkelijk verschilt van het geforceerde lostrekken van de AFM tip.

De desorptie van micro-organismen is door een vorige promovendus in situ bestudeerd met gecontroleerde stroomsystemen als functie van de verblijftijd van het organisme op het oppervlak (Meinders et al., 1994). Het bleek dat de desorptie van Streptococcus thermophilus sterk afneemt in de eerste 50 s na de initiële hechting als gevolg van het versterken van de binding in de tijd. In Hoofdstuk 6, is dit versterkingsproces van de binding tussen het S. thermophilus oppervlak en de silicium nitride tip van de AFM microscopisch bestudeerd, en gerelateerd aan de macroscopische desorptie. De AFM geeft aan dat de binding tussen de tip en het bacterie oppervlak versterkt in de eerste 100 s na hechting. Dit is een zelfde tijdschaal als waarover macrocopische bindingsversterking plaatsvindt in de parallelle plaat stroomkamer. Het sterker worden van de binding bleek een gevolg te zijn van een toename in het aantal hechtingsplaatsen tussen bacterie en vaste stof oppervlak.

De hydrofobiciteit van het bacterie oppervlak, gewoonlijk gemeten door het plaatsen van een waterdruppel op een zorgvuldig bereide en gedeeltelijk gedroogde bacterielaag, is algemeen geaccepteerd als een eigenschap die van belang is voor bacteriële interactie met zijn omgeving. Deze zorgvuldig bereide en droge bacterielaag, is eigenlijk een gelachtige laag van willekeurig samengevouwen cellulair oppervlakte-materiaal. Daarom zijn deze resultaten alleen te gebruiken voor het interpreteren van lange afstand interacties tussen een organisme en een substraat-oppervlak, maar zijn de resultaten niet te gebruiken voor de korte afstand interacties, die gedomineerd worden door structurele en chemische heterogeniteiten. In Hoofdstuk 7 zijn met chemisch gemodificeerde AFM tips, met hydrofobe (CH3) en hydrofiele eindgroepen (OH), de interactie-krachten gemeten voor twee Lactobacillus acidophilus stammen met en zonder S(urface) L(ayer) P(rotein) bij een hoge en lage zoutconcentratie. De macroscopische hydrofobiciteit van beide stammen is ook gemeten door middel van randhoekmetingen bij beide gebruikte zoutconcentraties. Algemeen geldt dat grote interactie-krachten met een


hydrofiele tip corresponderen met lage randhoeken, terwijl een celoppervlak met een hoge randhoek de sterkste interactie geeft met een hydrofobe tip. Voor beide stammen bleek het hydrofobe karakter te veranderen naar meer hydrofiel of hydrophoob, wanneer de zoutconcentratie veranderde van 10 naar 100 mM. Het is interessant dat het dynamische gedrag van de hydrofobiciteit van het celoppervlak van de lactobacillen niet alleen macroscopisch meetbaar is met randhoeken, maar ook met de AFM op een microscopisch niveau. Dit dynamische gedrag van bacterie-oppervlakken onder invloed van de zoutconcentratie geeft deze stammen een mechanisme om te hechten aan zowel hydrofobe als hydrofiele oppervlakken, afhankelijk van de zoutconcentraties.

Algemeen is gevonden dat hechting van lactobacillen aan vaste oppervlakken niet gaat volgens verwachtingen op grond van de hydrofobiciteit van hun oppervlak. In Hoofdstuk 8 is de invloed van de hydrofobiciteit van beide lactobacillus stammen met en zonder S(urface) L(ayer) P(rotein) onderzocht op hun hechting aan hydrofoob en hydrofiel gemaakt glas, zowel bij hoge als lage zoutconcentraties. Ook is de interactie gemeten van de lactobacillen met hydrofobe en hydrofiele AFM tips. Bij een lage zoutconcentratie, hebben beide lactobacillus stammen een hogere initiële depositie snelheid op hydrofoob glas dan op hydrofiel glas, terwijl er bij een hoge zoutconcentratie geen duidelijke invloed is van de hydrofobiciteit op de hechting. Onafhankelijk van de zoutconcentratie, meten we met de AFM sterkere interactie-krachten als zowel de tip als de bacterie hydrofoob dan wel hydrofiel zijn, dan wanneer de bacterie en de tip een tegenovergestelde hydrofobiciteit hebben. Deze sterke hechtingskrachten zien we niet bij de macroscopische waarnemingen van bacteriële hechting, hetgeen doet vermoeden dat de toegepaste kracht waarmee een bacterie en een oppervlak in contact komen, de manier van hechten beïnvloedt. Daar komt ook nog bij dat de afstandsafhankelijkheid van de totale Gibbs energie bij hechting alleen kwantitatief gecorreleerd kan worden met bacteriele depositie en desorptie in de parallelle plaat stroomkamer terwijl met AFM metingen geen DLVO krachten konden worden gemeten tijdens het naderen van functionele tips naar het bacterie celoppervlak.

In de Algemene discussie worden de volgende drie belangrijke conclusies uit dit proefschrift verder besproken:

(a) De kracht-afstand curve tijdens naderen correleert met de snelheid van initiële macroscopische hechting van een micro-organisme aan een oppervlak,

(b) de maximale afstand waarover aantrekkende krachten worden gemeten met AFM bij terugtrekken van de tip beïnvloedt de onderlinge afstand tussen gehechte bacterien op een oppervlak en

(c) bacteriële desorptie van een oppervlak heeft geen relatie met de aantrekkende krachten zoals die gemeten met AFM, tenzij er rekening wordt gehouden met de tijd dat een bacterie en een oppervlak contact hebben.

Ook wordt er gesproken over de beperkingen van het AFM systeem wat betreft de precieze interpretatie van de kracht-afstand curves en toekomstige perspectieven worden gepresenteerd.

Resumen

El uso de modelos físico-químicos para explicar la adhesión microbiana sobre superficies ha tenido éxito para un número limitado de cepas y especies (Capítulo 1), debido en parte a la naturaleza macroscópica de los datos experimentales de que se dispone (es decir, potencial zeta y ángulos de contacto). En general, desde el punto de vista físico-químico, se han asimilado los microorganismos a partículas inertes de poliestireno. Ahora bien, los microorganismos no son partículas lisas, sino que, a diferencia de las de poliestireno, poseen ciertas estructuras finas y largas que se extienden desde su superficie hacía el liquido circundante. Sin embargo, hasta el momento, permanece sin identificarse para la mayoría de las cepas de bacterias la función específica realizada por tales estructuras en los procesos de adhesión microbiana, así como su influencia en las propiedades físico-químicas macroscópicas de las superficies celulares. En consecuencia, las medidas relevantes desde el punto de vista físico-químico de la superficie microbiana, precisan de una resolución microscópica que no puede lograrse con la mayoría de los métodos tradicionales. La introducción del Microscopio de Fuerza Atómica (AFM) y su


aplicación a las superficies biológicas ha abierto un nuevo camino para poder tener acceso a las propiedades físico-químicas microscópicas de las superficies bacterianas.

Iniciamos esta tesis con la intención, a largo plazo, de desarrollar una nueva teoría que permita explicar la adhesión microbiana a superficies, la cual habrá de tener en cuenta interacciones complejas (como pueden ser la heterogeneidad en las distribución superficial de cargas o los enlaces macromoleculares) que no han sido consideradas en la teoría de DLVO. Más concretamente, los objetivos de esta tesis han sido caracterizar microscópicamente las propiedades físico-químicas de las superficies bacterianas y dilucidar si tales propiedades microscópicas pueden amalgamarse -y cómo- en propiedades macroscópicas de la superficie celular y relacionarse con la adhesión observada macroscópicamente de las bacterias a los substratos sólidos.

Se ha completado la caracterización macroscópica de una colección de nueve cepas de Streptococcus mitis, de las que se disponía de su hidrofobicidad superficial, determinada mediante medidas de ángulos de contacto con agua, y su composición química superficial, por espectroscopia de rayos X de fotoelectrones (XPS), con la evaluación de su deformabilidad electroforética y su densidad de carga fija asociada a la capa de polielectrólitos de las cepas de S. mitis, generalmente portadores de largos fibrilos distribuidos de modo poco denso, utilizando el análisis de partículas deformables para las medidas de la movilidad electroforética en función de la fuerza iónica (Capítulo 2). En general, la superficie celular de S. mitis es electroforéticamente deformable, con densidad de carga fija negativa. Sin embargo, la comparación con otras cepas de bacterias que han sido descritas como deformables, indica que la teoría de Ohshima no permite obtener información sobre las características morfológicas de la superficie que provocan su deformabilidad. La razón más probable es que el flujo electroosmótico tiene lugar solo en la región más externa de las gruesas capas extracelulares. A pesar de ello, la determinación de la deformabilidad superficial es esencial para una correcta caracterización de la electrostática de la superficie celular.

Para realizar los experimentos de AFM es preciso anclar con suficiente fuerza las bacterias a un sustrato, de modo que sean capaces de resistir la fricción ejercida por la punta de nitruro de silicio. Pero para caracterizar las propiedades físico-químicas de la superficie de los microorganismos es preciso que dicha inmovilización no afecte ni a la integridad química ni estructural de la superficie celular. Existen varias estrategias, descritas en la bibliografía, para la inmovilización de bacterias. En el Capítulo 3 se presenta la comparación de la fuerzas de interacción determinadas por AFM entre Klebsiella terrigena y nitruro de silicio obtenidas empleando tres métodos comunes de inmovilización: bloqueo mecánico de las bacterias en filtros, adsorción física de bacterias cargadas negativamente sobre superficies cargadas positivamente y fijación de las bacterias con glutaraldeido a la punta de la sonda del microscopio AFM. Se ha puesto de manifiesto que métodos diferentes de preparación de las muestras conducen a fuerzas de interacción disimilares. La adsorción física de las bacterias sobre

Resumen 109

sustratos modificados puede provocar reordenamientos estructurales en la superficie celular, mientras que se considera que el empleo de glutaraldeido induce cambios en las propiedades físico-químicas y mecánicas de la superficie bacteriana. En general, el método de inmovilización mediante el bloqueo mecánico de bacterias individuales en filtros parece ser el más adecuado.

Una vez establecido el método más idóneo para la inmovilización de las bacterias de cara a su estudio por AFM, se ha realizado un análisis detallado de las fuerzas de interacción entre la punta de nitruro de silicio de la sonda del microscopio AFM y la superficie de nueve cepas de la bacteria S. mitis. El Capítulo 4 es un primer intento de correlacionar las propiedades microscópicas y macroscópicas de la superficie bacteriana. Es interesante el hecho de haber podido relacionar las características particulares de las curvas fuerza-distancia de modo que se ha logrado obtener relaciones entre las propiedades microscópicas y macroscópicas, aunque no se haya conseguido dar una justificación completa sobre el origen de estas relaciones.

En el Capítulo 5 hemos tratado de correlacionar la adhesión microscópica de las cepas de S. mitis a la punta, hidrófila y cargada negativamente, de nitruro de silicio de un AFM con su adhesión macroscópica a vidrio, hidrófilo y cargado negativamente, empleando una cámara de flujo plano-paralela. Se ha encontrado que la fuerza repulsiva observada al aproximar la sonda del AFM a la superficie de la célula se corresponde con una barrera de activación que gobierna la adhesión macroscópica inicial de los microorganismos a la superficie del vidrio. Además, se ha encontrado que la distancia máxima a la que se han detectado las fuerzas atractivas al alejar la sonda del AFM está relacionada con el área bloqueada por cada bacteria adherida, es decir, con la distancia entre bacterias adheridas. La desorción de las bacterias no ha sido posible relacionarla con la fuerza adhesiva determinada con AFM, posiblemente por las diferencias en la naturaleza de los procesos que se llevan a cabo en la cámara de flujo plano-paralela y la separación forzada en el AFM.

La desorción de microbios había sido estudiada in situ en función del tiempo de residencia de los microorganismos sobre la superficie en dispositivos de flujo controlado (Meinders et al., 1994), encontrándose una disminución muy importante de la desorción de Streptococcus thermophilus después de transcurridos aproximadamente 50 s de la adhesión inicial, proponiéndose que podría estar relacionada con el fortalecimiento del enlace al aumentar el tiempo de contacto entre la bacteria y el sustrato. En el Capítulo 6, se ha corroborado microscópicamente el efecto del tiempo de contacto sobre la intensidad del enlace entre la superficie de S. thermophilus y la punta de nitruro de silicio del AFM y se ha relacionado con la dependencia de la desorción en condiciones de flujo con el tiempo de residencia de las bacterias sobre la superficie. Los experimentos de AFM han puesto de manifiesto un fortalecimiento del enlace entre la sonda y la superficie de la célula tras 100 s de contacto, tiempo del mismo orden de magnitud que el inferido del estudio de la desorción en función del tiempo de residencia sobre la superficie. Además, la comparación entre las energías de interacción obtenidas por


AFM y de la desorción macroscópica parece indicar que el fortalecimiento del enlace está debido a la adhesión de múltiples polímeros extracelulares a la superficie del sustrato.

Esta generalmente aceptado que la hidrofobicidad de la superficie bacteriana, habitualmente determinada depositando gotas de agua sobre una capa de bacterias compactadas y secadas, influye en la interacción del microorganismo con su entorno. Sin embargo, el ángulo de contacto de las gotas de agua sobre esas capas de bacterias informa sobre las características de un recubrimiento desordenado de material de la superficie de las bacterias colapsadas en un cierto empaquetamiento. Aunque los resultados son útiles para interpretar las interacciones de largo alcance entre el microorganismo y la superficie del sustrato, no lo son necesariamente para la interpretación de las interacciones de corto alcance, que podrían estar controladas por las heterogeneidades estructurales y químicas de la superficie celular. En el Capítulo 7, se han estudiado las fuerzas de interacción mediante AFM de la superficie de L. acidophilus ATCC4356 y L. casei ATCC393 con y sin SLP, respectivamente, con sondas funcionalizadas, es decir terminadas con grupos hidrófobos (CH3) e hidrófilos (OH), en disoluciones con fuerza iónica baja y alta. La hidrofobicidad superficial macroscópica de ambas cepas se determinó mediante ángulos de contacto, empleando disoluciones con las dos fuerzas iónicas investigadas. En general, fuerzas de interacción altas con la sonda hidrófila coinciden con valores bajos del ángulo de contacto, mientras que las células que muestran ángulos de contacto altos presentan una interacción más fuerte con la sonda hidrófoba. Asimismo, ambas cepas invierten su naturaleza hidrófoba al aumentar la fuerza iónica de 10 a 100 mM. Es interesante el que el comportamiento dinámico de la hidrofobicidad superficial de los lactobacilos sea posible medirlo no solamente macroscópicamente mediante ángulos de contacto, sino también por AFM a nivel microscópico. El comportamiento dinámico de las superficies celulares de L. acidophilus ATCC4356 y L. casei ATCC393 frente a los cambios de la fuerza iónica confiere a estos microorganismos una versatilidad que les permite adherirse a superficies hidrófobas e hidrófilas en disoluciones con baja y alta fuerza iónica, respectivamente.

En general, la adhesión de los lactobacilos a sustratos sólidos no tiene lugar de acuerdo con las expectativas deducidas de la hidrofobicidad superficial. En el Capítulo 8 se ha investigado el papel que juega la hidrofobicidad superficial para dos cepas de lactobacilos con y sin capa de SLP en su adhesión a superficies de vidrio funcionalizado hidrofóbica o hidrofílicamente, en condiciones de flujo bien definidas de suspensiones con baja y alta fuerza iónica. Por similitud, se midió la interacción de los lactobacilos con sondas de AFM funcionalizadas del mismo modo. En las suspensiones de baja fuerza iónica, ambas cepas de lactobacilos presentan una adhesión inicial más alta al vidrio hidrófobo que al hidrófilo, mientras que con fuerzas iónicas altas no se ha observado una influencia clara de la hidrofobicidad superficial sobre la adhesión. Independientemente de la fuerza iónica, los experimentos de AFM indican que las fuerzas de interacción son más intensas cuando, tanto la bacteria como la sonda, son hidrófobas o hidrófilas que

Resumen 111

cuando tienen hidrofobicidades opuestas. Estas intensas fuerzas de adhesión no tienen reflejo en las observaciones de la adhesión macroscópica, lo que sugiere que la fuerza que actúa cuando una bacteria se pone en contacto con un sustrato afecta su modo de adhesión. Por otra parte, se ha podido correlacionar sólo cualitativamente la dependencia con la distancia de la energía de interacción total de Gibbs con la deposición y la desorción de las bacterias en la cámara de flujo plano-paralela, mientras que no se han detectado, mediante AFM, las fuerzas DLVO al aproximar las sondas funcionalizadas a la superficie de las bacterias.

En la Discusión General, se han discutido las conclusiones principales obtenidas de esta tesis que son:

(a) las curvas fuerza-distancia al aproximar la sonda a la superficie se corresponden con la razón con que se adhieren macroscópicamente los microorganismos a los sustratos.

(b) la distancia máxima a la cual se pueden detectar las fuerzas atractivas mediante la retracción de la sonda del AFM están relacionadas con el área que bloquea una bacteria adherida sobre un sustrato, y

(c) la desorción de bacterias de sustratos sólidos no muestra relación con las fuerzas adhesivas determinadas por AFM, excepto si se considera el tiempo de contacto entre las bacterias y el sustrato. Además, se presentan algunas de las limitaciones del sistema de cara a la interpretación precisa de las curvas fuerza-distancia, así como algunas perspectivas para futuros trabajos.

Reference list

Absolom, D.R., Lamberti, F.V., Policova, Z., Zingg, W., Van Oss, C.J., and Neumann, A.W. (1983) Surface thermodynamics of bacterial adhesion. Appl. Environ. Microbiol. 46: 90-97.

Abu-Lail, N.I., and Camesano, T.A. (2003a) Role of lipopolysaccharides in the adhesion, retention, and transport of Escherichia coli JM109. Environ. Sci. Technol. 37: 2173-2183.

Abu-Lail, N.I., and Camesano, T.A. (2003b) Role of ionic strength on the relationship of biopolymer conformation, DLVO contributions, and steric interactions to bioadhesion of Pseudomonas putida KT2442. Biomacromolecules 4: 1000-1012.

Adamczyk, Z. (1989) Particle Deposition from flowing suspensions. Coll. Surf. 39: 1-37. Alexander, S. (1977) Adsorption of chain molecules with a polar head. A scaling description. J. Phys.

II (Paris). 38: 983-987. Ahimou, F., Denis, F.A., Touhami, A., and Dufrêne, Y.F. (2002) Probing microbial cell surface

charges by atomic force microscopy. Langmuir 18: 9937-9941. An, Y.H., and Friedman, R.J. (1997) Laboratory methods for studies of bacterial adhesion. J.

Microbiol. Methods 30: 141-152. Amory, D.E., Genet, M.J., and Rouxhet, P.G. (1988) Application of XPS to the surface analysis of

yeast cells. Surf. Interf. Anal. 11: 478-486. Arnoldi, M., Fritz, M., Baurerlein, E., Radmacher, M., Sackmann, E., and Boulbitch A. (2000)

Bacterial turgor pressure can be measured by atomic force microscopy. Phys. Rev. E 62:1034-1043.


Beveridge, T.J. (1994) Bacterial S-layers. Curr. Opin. Struct. Biol. 4: 204-212. Boland, T., and Ratner, B.D. (1995) Direct measurement of hydrogen bonding in DNA

nucleotide bases by atomic force microscopy. Proc. Natl. Acad. Sci. USA 92: 5297-5301. Bolshakova, A.V., Kiselyova, O.I., Filonov, A.S., Frolova, O.Y., Lyubchenkoc, Y.L., and Yaminsky,

I.V. (2001) Comparative studies of bacteria with an atomic force microsocopy operating in different modes. Ultramicroscopy 86: 121-128.

Boonaert, C.J.P., Toniazzo, V., Mustin, C., Dufrêne, Y.F., and Rouxhet, P.G. (2002). Deformation of Lactococcus lactis surface in atomic force microscopy study. Colloids Surf. B Biointerfaces 23: 201-211.

Boonaert, C.J.P., Dufrêne, Y.F., Derclaye, S.R., and Rouxhet, P.G. (2001) Adhesion of Lactococcus lactis to model substrata: direct study of the interface. Colloids Surf. B: Biointerfaces 22: 171-178.

Borinski, R., and Holt, S.C. (1990) Surface characteristics of Wolinella recta ATCC 33238 and human clinical isolates: correlation of structure with function. Infect. Immun. 58: 2770-2776.

Bos, R., Van der Mei, H.C., and Busscher, H.J. (1999) Physico-chemistry of initial microbial adhesive interactions - its mechanisms and methods for study. FEMS Microbiol. Rev. 23: 179-230.

Bos, R., Van der Mei, H.C., and Busscher, H J. (1998) `Soft-particle' analysis of the electrophoretic mobility of a fibrillated and non-fibrillated oral streptococcal strain: Streptococcus salivarius. Biophys. Chem. 74: 251-255.

Bottomley, L.A., Coury, J.E., and First, P.N. (1996) Scanning probe microscopy. Anal. Chem. 68: 185-230.

Bouman, S., Lund, D.B., Driessen, F.M., and Schmidt, D.G. (1982) Growth of thermoresistant streptococci and deposition of milk constituents on plates of heat exchangers during long operation times. J. Food Prot. 45: 806-812.

Bowen, W.R., Lovitt, R.W., and Wright, C.J. (2001) Atomic force microscope study of the adhesion of Saccharomyces cerevisiae. J. Colloid Interface Sci. 237: 54-61.

Browing-Kelly, M.E., Wadu-Mesthrige, K., Hari, V., and Liu, G.Y. (1997) Atomic force microscopic study of specific antigen/antibody binding. Langmuir 13: 343-350.

Burks, G.A., Velegol, S.B., Paramonova, E., Lindenmuth, B.E, Feick, J.D., and Logan, B.E. (2003) Macroscopic and nanoscale measurements of the adhesion of bacteria with varying outer layer surface compostion. Langmuir 19: 2366-2371.

Busscher, H.J., and Van der Mei, H.C. (1995). Use of flow chamber devices and image analysis methods to study microbial adhesion. Methods in Enzymology 253: 455-477.

Busscher, H.J., Bialkowska-Hobrzanska, H., Reid, G., Van der Kuijl-Booij, M., and Van der Mei, H. C. (1994) Physicochemical characteristics of two pairs of coagulase-negative staphylococcal isolates with different plasmid profiles. Colloids Surf. B: Biointerfaces. 2: 73-82.

Busscher, H.J., Cowan, M.M., and Van der Mei, H.C. (1992). On the relative importance of specific and non-specific approaches to oral microbial adhesion. FEMS Microbiol. Rev. 88: 199-210.

Busscher, H.J., Handley, P.S., Rouxhet, P.G., Hesketh, L.M., and Van der Mei, H.C. (1991) In Microbial Surface Analysis: Structural and Physico-Chemical Methods Mozes, N.; Handley, P. S.; Busscher, H. J.; Rouxhet, P. G., Eds.; VCH Publishers Inc., New York.

Busscher, H.J., Weerkamp, A.H., Van der Mei, H.C., Van Pelt, A.W.J., De Jong, H.P., and Arends, J. (1984) Measurement of the surface free energy of bacterial cell surfaces and its relevance for adhesion. Appl. Environ. Microbiol. 48: 980-983.

Bustamante, C., and Rivetti, C. (1996) Visualizing protein-nucleic acid interactions on a large scale with the scanning force microscope. Annu. Rev. Biophys. Biomol. Struct. 25: 395-429.

Butt, H–J., Kappl, M., Mueller, H., and Raiteri, R. (1999) Steric forces measured with the atomic force microscope at various temperatures. Langmuir 15: 2559-2665.

Butt, H-J. (1992) Measuring local surface charge densities in electrolyte solutions with a scanning force microscope. Biophys. J. 63: 578-582.

Camesano, T.A., and Abu-Lail, N.I. (2002) Heterogeneity in bacterial surface polysaccharides, probed on a single-molecule basis. Biomacromolecules 3: 661-667.

Reference List 115

Camesano, T.A., and Logan, B.E. (2000) Probing bacterial electrosteric interactions using atomic force microscopy. Environ. Sci. Technol. 34: 3354-3362.

Camesano, T.A., Natan, M.J., and Logan, B.E. (2000) Observation of changes in bacterial cell morphology using tapping mode atomic force microscopy. Langmuir 16: 4563-4572.

Carstensen, E.L., and Marquis, R.E. (1968) Passive electrical properties of microorganisms: III. conductivity of isolated bacterial cell walls. Biophys. J. 8: 536-548.

Characklis, W.G. (1973) Attached microbial growths-I. Attachment and growth. Water Res. 7: 1113-1127.

Chilkoti, A., Boland, T., Ratner, B.D., and Stayton, P.S. (1995). The relationship between ligand-binding thermodynamics and protein-ligand interaction forces measured by atomic force microscopy. Biophys. J. 69: 2125-2130.

Cleveland, J.P., Manne, S., Bocek, D., and Hansma, P.K. (1993) A nondestructive method for determining the spring constant of cantilevers for scanning force microscopy. Rev. Sci. Instrum. 64: 403-405.

Cooksey, K.E., and Wigglesworth-Cooksey, B. (1995) Adhesion of bacteria and diatoms surfaces in the sea: a review. Aquat. Microbiol. Ecol. 9: 87-96.

Cowan, M.M., Mikx, F.H.M., and Busscher, H.J. (1994) Electrophoretic mobility and hemagglutination of Treponema denticola ATCC 33520. Colloids Surf. B: Biointerfaces 2: 407-410.

Cowan, M.M., Van der Mei, H.C., Rouxhet, P.G., and Busscher, H.J. (1992) Physico-chemical and structural properties of the surfaces of Peptostreptococcus micros and Streptococcus mitis as compared to those of mutans streptococci, Streptococcus sanguis and Streptococcus salivarius. J. Gen. Microbiol. 138: 2707-2714.

Dabros, T., and Van de Ven, T.G.M. (1982) Kinetics of coating by colloid particles. J. Colloid Interface Sci. 89: 232-244.

Dankert, J., Hogt, A.H. and Feijen, J. (1986) Biomedical polymers: bacterial adhesion, colonization, and infection. Biocompatibility 2: 219-301.

De Gennes, P.G. (1987) Polymers at an interface; a simplified view. Adv Colloid Interface Sci 27: 189-209.

Derjaguin, B.W., Landau, L. (1941) Theory of the stability of strongly charged lyophobic sols and of the adhesion of strongly charged particles in solutions of electrolytes. Acta Physiochim. URSS 14: 633-662.

Drummond, C.J., and Senden, T.J. (1994) Examination of the geometry of long-range tip––sample interaction in atomic force microscopy. Colloids Surf. A: Physicochem. Eng. Aspects 87: 217-234.

Dufrêne, Y.F. (2002) Atomic force microsocpy, a powerful tool in microbiology. J. Bacteriol. 184: 5205-5213.

Dufrêne, Y.F. (2001) Application of atomic force microscopy to microbial surfaces: from reconstituted cell surface layers to living cells. Micron. 32: 153-165.

Dufrêne, Y.F. (2000). Direct characterization of the physicochemical properties of fungal spores using functionalized AFM probes. Biophys. J. 78: 3286-3291.

Dufrêne, Y.F., Boonaert, C.J.P., Gerin, P.A., Asther, M., and Rouxhet, P.G. (1999) Direct probing of the surface ultrastructure and molecular interactions of dormant and germinating spores of Phanerochaete chrysosporium. J. Bacteriol. 181: 5350-5354.

Dufrêne, Y.F., Boonaert, C.J.P., Van der Mei, H.C., Busscher, H.J., and Rouxhet, P.G. (2001) Probing molecular interactions and mechanical properties of microbial cell surfaces by atomic force microscopy. Ultramicroscopy 86: 113-120.

Engel, A., Schoenenberger, C.-A., and Mueller, D.J. (1997) High resolution imaging of native biological sample surfaces using scanning probe microscopy. Curr. Opin. Struct. Biol. 7:279-284.

Fang, H.H.P., Chan, K-Y., and Xu, L-C. (2000) Quantification of bacterial adhesion forces using atomic force microscopy (AFM). J. Microbiol. Methods 40: 89-97.

Fernandez Murga, M.F., Bernik, D., Font de Valdez, G., and Disalvo, A.E. (1999) Permeability and stability of membranes formed by lipids extracted from Lactobacillus acidophilus grown at


different temperatures. Arch. Biochem. Biophys. 364: 115-121. Fives-Taylor, P.M., and Thompson, D.W. (1985) Surface properties of Streptococcus sanguis FW213

mutants nonadherent to saliva-coated hydroxyapatite. Infect. Immun. 47: 752-759. Florin, W.–L., Moy, V.T., and Gaub, H.E. (1994) Adhesive forces between individual ligand-receptor

pairs. Science 264: 415-417. Frank, B.P., and Belfort, G. (1999) Intermolecular forces between extracellular polysaccharides

measured using the atomic force microscope. Langmuir 13: 6234-6240. Gannon, J.T., Manilal, V.B., and Alexander, M. (1991) Relationships between cell surface properties

and transport of bacteria through soil. Appl. Environ. Microbiol. 57: 190-193. Gristina, A.G., Naylor, P., and Myrvik, Q. (1988) Infections from biomaterials and implants : a race

for the surface. Med. Prog. Technol. 14: 205-224. Hancock, I. C. (1991) In Microbial cell surface analysis – Structural and Physicochemical Methods;

Mozes, N.; Handley, P. S.; Busscher, H. J.; Rouxhet, P. G., Eds.; VCH Publishers Inc., New York.. Handley, P.S., Hesketh, L.M., and Moumena, R.A. (1991) Charged and hydrophobic groups are

localised in the short and long tuft fibrils on Streptococcus sanguis strains. Biofouling 4: 105-111. Handley, P.S. (1990) Structure, composition and function of surface structures on oral bacteria.

Biofouling 2: 239-264. Handley, P.S., Harty, D.W.S., Wyatt, J.E., Brown, C.R., Doran, J., and Gibbs, A.C.C. (1987) A

comparison of the adhesion, coaggregation and cell-surface hydrophobicity properties of fibrillar and fimbriate strains of Streptococcus salivarius. J. Gen. Microbiol.133: 3207-3217.

Hamers, R.J. (1996) Scanned probe microscopies in chemistry. J. Phys. Chem. 100: 13103-13120. Harkes, G., Feijen, J., and Dankert, J. (1991) Adhesion of Escherichia coli on to a series of

poly(methacrylates) differing in charge and hydrophobicity. Biomaterials 12: 853-860. Hayashi, H., Tsuneda, S., Hirata, A., and Sakasi, H. (2001) Soft particle analysis of bacterial cells and

its interpretation of cell adhesion behaviors in terms of DLVO theory. Colloids Surf. B: Biointerfaces 22: 149-157.

Haynes, C.A., and Norde, W. (1994) Globular proteins at solid/liquid interfaces. Colloids Surf. B: Biointerfaces 2: 517-566.

Heinz, W.F., and Hoh, J.H. (1999a) Relative surface charge mapping with the atomic force microscope. Biophys. J. 76: 528-538.

Heinz, W.F., and Hoh, J.H. (1999b) Spatially resolved force spectroscopy of biological surfaces using the atomic force microscopy. Trends Biotechnol. 17: 143-150.

Hermansson, M. (1999) The DLVO theory in microbial adhesion. Colloids Surf. B: Biointerfaces 14: 105-119.

Herthelius, M., Gorbach, S.L., Mollby, R., Nord, C.E., Pettersson, L., and Winberg, J. (1989) Elimination of vaginal colonization with Escherichia coli by administration of indigenous flora. Infect. Inmun. 57: 2447-2451.

Holzapfel, W.H., Haberer, P., Snel, J., Schillinger, U., and Huis in’t Veld, J.H.J. (1998) Overview of gut flora and probiotics. Int. J. Food Microbiol. 41: 85-101.

James, A.M. (1991) Charge properties of microbial cell surfaces. In Microbial Surface Analysis: Structural and Physico-Chemical Methods (ed. N. Mozes, P.S. Handley, H.J. Busscher, and P.G. Rouxhet), pp. 221-262, VCH Publishers Inc., New York.

Johnson, S. B., Drummond, C. J., Scales, P. J., and Nishimura, S. (1995) Comparison of techniques for measuring the electrical double layer properties of surfaces in aqueous solution: hexadecyltrimethylammonium bromide self-assembly. Langmuir 11: 2367-2375.

Kasas, S., and Ikai, A. (1995) A method for anchoring round shaped cells for atomic force microscope imaging. Biophys. J. 68: 1678-1680.

Kidoaki, S., and Matsuda, T. (2002) Mechanistic aspects of protein/material interactions probed by atomic force microscopy. Colloids Surf. B: Biointerfaces 23: 153-163.

Kiers, P.J.M., Bos, R., Van der Mei, H.C., and Busscher, H.J. (2001) The electrophoretic softness of the surface of Staphylococcus epidermidis cells grown in a liquid medium and on a solid agar. Microbiol. 147: 757-762.

Reference List 117

Lee, G.U., Chrisy, L.A., and Colton, R.J. (1994) Direct measurement of the forces between complementary strands of DNA. Science 266: 771-773.

Lower, S.K., Tadamnier, C.J., and Hochella, M.F. (2000) Measuring interfacial and adhesion forces between bacteria and mineral surfaces with biological force microscopy. Geochim. Cosmochim. Acta 64: 3133-3139.

Machado, M.C., Lopez, C.S., Heras, H., and Rivas, E.A. (2004) Osmotic response in Lactobacillus casei ATCC 393: biochemical and biophysical characteristic of membrane. Arch. Biochem. Biophys. 422: 61-70.

Marsh, P., and Martin, M. (1992) Oral Microbiology, 3rd edn. Chapman and Hall, London. Marshall, K.C., Pembrey, R., and Schneider, R.P. (1994) The relevance of X-ray photoelectron

spectrosocpy for analysis of microbial cell surfaces: a critical review. Colloids Surf. B: Biointerfaces 2: 371-369.

Marshall, K.C. (1976) Interfaces in microbial ecology. Harvard University Press, Cambridge, Mass. Martin, M.J., Logan, B.E., Johnson, W.P., Jewett, D.G., and Arnold, R.G. (1996) Scaling bacterial

filtration rates in different sized porous media. J. Environ. Eng. 122: 407-415. Matthyse, A.G., Holmes, K.V., and Gurlitz, R.W.G. (1981) Elaboration of cellulose fibrils by

Agrobacterium tumefaciens during attachment to carrot cells. J. Bacteriol. 145: 583-595. McClaine, J.W., and Ford, R.M. (2002) Characterizing the adhesion of motile and nonmotile

Escherichia coli to a glass surface using a parallel-plate flow chamber. Biotechnology Bioengineering 78: 179-189.

Meagher, L., and Griesser, H. J. (2002) Interactions between adsorbed lactoferrin layers measured directly with the atomic force microscope. Colloids Surf. B: Biointerfaces. 23: 125-140.

Meinders, J.M., Van der Mei, H.C., and Busscher H.J. (1994) Physicochemical aspects of deposition of Streptococcus thermophilus B to hydrophobic and hydrophilic substrata in a parallel plate flow chamber. J. Colloid Interface Sci. 164: 355-363.

Meinders, J.M., Noordmans, J., and Busscher, H.J. (1992) Simultaneous monitoring of the adsorption and the desorption of colloidal particles during deposition in a parallel plate flow chamber. J. Colloid Interface Sci. 152: 265-279.

Millsap, K.W., Reid, G., Van der Mei, H.C., and Busscher, H.J. (1996) Adhesion of Lactobacillus species in urine and phosphate buffer to silicone rubber and glass under flow. Biomaterials 18: 87-91.

Morisaki, H., Nagai, S., Ohshima, H., Ikemoto, E., and Kogure, K. (1999) The effect of motility and cell-surface polymers on bacterial attachment. Microbiol. 145: 2797-2802.

Mukhopadhyay, R., and Hoh, J.H. (2001) AFM force measurements on microtubule-associated proteins: the projection domain exerts a long-range repulsive force. FEBS Letters 505: 374-378.

Norde, W., and Lyklema, J. (1989) Protein adsorption and bacterial adhesion to solid surfaces: A colloid-chemical approach. Colloids Surfaces 38: 1-13.

Ohshima, H. (1995) Electrophoretic mobility of soft particles. Colloids Surf. A: Pysicochem Eng Aspects 103: 249-255.

Ohshima, H., and Kondo, T. (1991) On the electrophoretic mobility of biological cells. Biophysical Chemistry 39, 191-198.

Ong, Y. L., Razatos, A., Gerogiou, G., and Sharma, M.M. (1999) Adhesion between E. coli bacteria and biomaterial surfaces. Langmuir 15: 2719-2725.

Poortinga, A.T., Bos, R., and Busscher, H.J. (2001) Electrostatic interactions in the adhesion of an ion-penetrable and ion-impenetrable bacterial strain to glass. Colloids. Surf. B: Biointerfaces 20: 105-117.

Razatos, A., Ong, Y-L., Sharma, M.M., and Georgiou, G. (1998) Molecular determinants of bacterial adhesion monitored by atomic force microscopy. Pro.c Nat.l Acad. Sci. USA. 95: 11059-11064.

Rief, M., Gautel, M., Oesterhelt, F., Fernandez, J.M., and Gaub, H.E. (1997a) Reversible unfolding of individual titin Ig-domains by AFM. Science 276: 1109-1112.

Rief, M., Oesterhelt, F., Heymann, B., and Gaub, H.E. (1997b) Single molecule force spectroscopy on polysaccharides by AFM. Science 275:1295-1297.


Rijnaarts, H.H.M., Norde, W., Lyklema, J., and Zehnder, A.J.B. (1999) DLVO and steric contributions to bacterial deposition in media of different ionic strengths. Colloids Surf. B: Biointerfaces 14: 179-185

Rijnaarts, H.H.M., Norde, W., Bouwer, E.J., Lyklema, J., and Zehnder, A.J.B. (1995) Reversibility and mechanism of bacterial adhesion. Colloids Surf. B: Biointerfaces 4: 5-22.

Rosenberg, M., and Kjelleberg, S. (1986) Hydrophobic interactions: role in bacterial adhesion. Microbiol. Ecol. 9: 353-393.

Rouxhet, P.G., Mozes, N., Dengis, P.B., Dufrêne, Y.F., Gerin, P.A., and Genet, M.J. (1994) Application of X-ray photoelectron spectroscopy to microorganisms. Colloids Surf. B: Biointerfaces 2: 347-369.

Rutter, P.R., and Vincent, B. (1980) The adhesion of microorganisms to surfaces: physico-chemical aspects, Ellis Horwood Ltd. Chicester, England, pp. 79-91.

Sanders M. E. (1993) Effect of consumption of lactic cultures on human health. Adv Food Nutr Res 37: 67-130.

Sára M., and Sleytr, U.B. (2000) S-layer proteins. J Bacteriol 182: 859-868. Savage, D.B., and Fletcher, M. (1985) Bacterial Adhesion. Plenum Press: New York and London. Schweiss, R., Welzel, P.B., Werner, C.W., and Knoll, W. (2001) Dissociation of surface functional

groups and preferential adsorption of ions on self-assembled monolayers assessed by streaming potential and streaming current. Langmuir 17: 4304-4311.

Shao, Z., Mou, J., Czajkowsky, D.M., Yang, J., and Yuan, Y.J. (1996) Biological atomic force microscopy: what is achieved & what is needed. Adv. Phys. 45: 1-86.

Sinniah, S.K., Steel, A.B., Miller, C.J., and Reutt-Robey, J.E. (1996) Solvent exclusion and chemical contrast in scanning force microscopy. J. Am. Chem. Soc. 118: 8925-8931.

Sjollema, J., and Busscher, H.J. (1990) Deposition of polystyrene particles in a parallel flow cell. 2. Pair distribution functions between deposited particles. Colloids Surfaces 47: 337-352.

Sjollema, J., Van der Mei, H.C., Uyen, H.M.W., and Busscher, H.J. (1990a) The influence of collector and bacterial cell surface properties on the deposition of oral streptococci in a parallel plate flow cell. Adh. Sci. Technol. 4: 765-779.

Sjollema, J., Van der Mei, H.C., Uyen, H.M., and Busscher, H.J. (1990b) Direct observations of cooperative effects in oral streptococcal adhesion to glass by analysis of the spatial arrangement of adhering bacteria. FEMS Microbiol. Lett. 69: 263-269.

Sjollema, J., Busscher, H.J., and Weerkamp, J. (1989) Real-time enumeration of adhering microorganisms in a parallel plate flow cell using automated image analysis. J. Microbiol. Meth. 9: 73-78.

Smit, E., Jager, D., Martinez, B., Tielen, F.J., and Pouwels, P.H. (2002) Structural and functional analysis of the S-layer protein crystallisation domain of Lactobacillus acidophilus ATCC 4356: Evidence for protein-protein interaction of two subdomains. J. Mol. Biol. 324: 953-964.

Smit, G., Kijne, J.W., and Lugtenbert, B.J.J. (1986) Correlation between extracellular fibrils and attachment of Rhizobium leguminosarum to pea root hair tips. J. Bacteriol. 169: 4294-4301.

Tyler, B.J. (1997) XPS and SIMS studies of surfaces important in biofilm formation. Three case studies. Ann. N.Y. Acad. Sci. 831: 114-126.

Touhami, A., Nysten, B., and Dufrêne, Y. F. (2003) Nanoscale mapping of the elasticity of microbial cells by atomic force microscopy. Langmuir 19: 4539-4543.

Vadillo-Rodríguez, V., Busscher, H.J., Norde, W., De Vries, J., and Van der Mei, H.C. (2004) Dynamic cell surface hydrophobicity of Lactobacillus strains with and without surface layer protein. J. Bacteriol. manuscript in preparation.

Vadillo-Rodriguez, V., Busscher, H.J, Norde, W., and Van der Mei, H.C. (2002) Softness of the bacterial cell wall of Streptococcus mitis as probed my microelectrophoresis. Electrophoresis 23: 2007-2011.

Van der Aa, B.C., and Dufrêne, Y.F. (2002). In situ characterization of bacterial extracellular polymeric substances by AFM. Colloids Surf. B: Biointerfaces 23: 173-182.

Van der Mei, H.C., Van de Belt-Gritter, B., Pouwels, P.H., Martinez, B., and Busscher, H.J. (2003) Cell surface hydrophobicity is conveyed by S-layer proteins –a study in recombinant lactobacilli.

Reference List 119

Colloids Surf. B: Biointerfaces 28: 127-134. Van der Mei, H.C., Busscher, H.J., Bos, R., De Vries, J., Boonaert, C.J.P., and Dufrêne, Y.F. (2000a)

Direct probing by atomic force microscopy of the cell surface softness of a fibrillated and nonfibrillated oral streptoccal strain. Biophys. J. 78: 2668-2674.

Van der Mei, H.C., De Vries, J., and Busscher, H.J. (2000b) X-ray photoelectron spectroscopy for the study of microbial cell surfaces. Surf. Sci. Reports 39: 1-24.

Van der Mei, H.C., Bos, R., and Busscher, H.J. (1998a) A reference guide to microbial cell surface hydrophobicity based on contact angles. Colloids Surf. B: Biointerfaces 11: 213-221.

Van der Mei, H.C., Handley, P.S., Bos, R., and Busscher, H.J. (1998b) In Oral Biofilms and Plaque Control; Busscher, H.J.; Evans L.V., Eds.; Harwood Academic Publishers, Amsterdam.

Van der Mei, H.C., and Busscher, H.J. (1996). Detection by physico-chemical techniques of an amphiphilic surface component on Streptococcus mitis strains involved in non-electrostatic binding to surfaces. Eur. J. Oral Sci. 104: 48-55.

Van der Mei, H.C., Meinders, J.M., and Busscher, H.J. (1994) The influence of ionic strength and pH on diffusion of micro-organisms with different structural surface features. Microbiol. 140: 3413-3419.

Van der Mei, H.C., Léonard, A.J., Weerkamp, A.H., Rouxhet, P.G., and Busscher, H.J. (1988a) Surface properties of Streptococcus salivarius HB and nonfibrillar mutants: measurement of zeta potential and elemental composition with X-ray photoelectron spectroscopy. J. Bacteriol. 170: 2462-2466.

Van der Mei, H.C., Léonard, A.J., Weerkamp, A.H., Rouxhet, P.G., and Busscher, H.J. (1988b) Properties of oral streptococci relevant for adherence: Zeta potential, surface free energy and elemental composition. Colloids Surf. 32: 297-305.

Van der Mei, H.C., Weerkamp, A.H., and Busscher, H.J. (1987) Physico-chemical surface characteristics and adhesive properties of Streptococcus salivarius strains with defined cell surface structures. FEMS Microbiol. Lett. 40: 15-19.

Van de Ven, T.G.M. (1989) Effects of electrolytes, polymers and polyelectrolytes on particle deposition and detachment. Coll. Surf. 39: 107-126.

Van der Wal, A., Minor, M., Norde, W., Zehnder, A.J.B., and Lyklema, J. (1997) The electrokinetic potential of bacterial cells. Langmuir 13: 165-171.

Van Loosdrecht, M.C., Norde W., and Zehnder A.J. (1990) Physical chemical description of bacterial adhesion. J. Biomater. Appl. 5: 91-106.

Van Loosdrecht, M.C.M., Lyklema, J., Norde, W., and Zehnder, A.J.B. (1989) Bacterial adhesion: A physicochemical approach. Microb. Ecol. 17: 1-15.

Van Oss, C.J. (1995) Hydrophobicity of biosurfaces - origin, quantitative determination and interaction energies. Colloids Surf. B: Biointerfaces 5: 91-110.

Van Oss, C. J. (1994) Interfacial forces in Aqueous Media, New York, NY: Marcel Dekker. Van Oss, C.J., Chaudhury, M.K., and Good, R.J. (1987) Monopolar surfaces. Adv. Coll. Interface Sci.

28: 35-64. Van Oss, C.J., Good, R.J., and Chaudhury, M. (1986) The role of Van der Waals forces and hydrogen

bonds in hydrophobic interactions between biopolymers and low energy surfaces. J. Colloids Interface Sci. 111: 378-390.

Van Pelt, A.W., Weerkamp, A.H., Uyen, M.H., Busscher, H.J., De Jong, H.P., and Arends, J. (1985) Adhesion of Streptococcus sanguis CH3 to polymers with different surface free energies. Appl Environ. Microbiol. 49: 1270-1275.

Van Wagenen, R.J., and Andrade. J.D. (1980) Flat plate streaming potential investigations: hydrodynamics and electrokinetic equivalency. J. Colloid Interface Sci. 76: 305-314.

Velegol, S.B., and Logan, B.E. (2002) Contributions of bacterial surface polymers, electrostatics, and cell elasticity to the shape of AFM force curves. Langmuir 18: 5256-5262.

Verwey, E.J.W., and Overbeek, J.G. (1948) Theory of the stability of lyophobic colloids. Elsevier, Amsterdam, 1948.

Visser, J., and Jeurnink, T.J.M. (1997). Fouling of heat exchangers in the dairy industry. Experimental Thermal Fluid Science 14: 407-424.


Vezenov, D.V., Noy, A., Rozsnyai, L.F., and Lieber, C.M. (1997) Force titrations and ionization state sensitive imaging of functional groups in aqueous solutions by chemical force microscopy. J. Am. Chem. Soc. 119: 2006-2015.

Weerkamp, A.H., and McBride, B.C. (1980) Characterization of the adherence properties of Streptococcus salivarius. Infect Immun 29: 459-468.

Wilson, W.W., Wade, M.M., Holman, C., and Champlin, R.F. (2001) Status of methods for assessing bacterial cell surface charge properties based on zeta potential measurements. J. Microbiol. Methods. 43: 153-164.

Wit, P.J., Noordmans, J., and Busscher, H. J. (1997) Tracking of colloidal particles using microscopic image sequence analysis. Application to particulate microelectrophoresis and particle deposition. Colloids. Surf. A: Physicochem. Eng. Aspects 125: 85-92.

Xia, Z., Woo, L., and Van de Ven, T.G.M. (1989) Microrheological aspects of adhesion of Escherichia coli on glass. Biorheology 26: 359-375.

Yao, X., Walter, S., Burke, S., Steward, S., Jericho, M.H., Pink, D., Hunter, R., and Beveridge, T.J. (2002) Atomic force microcopy and theoretical considerations of surface properties and turgor pressures of bacteria. Colloids. Surf. B: Biointerfaces. 23: 213-230.

Zita, A., and Hermansson, M. (1997) Effects of bacterial cell surface structures and hydrophobicity on attachment to activated sludge flocs. Appl. Environ. Microbiol. 63: 1168-1170.

Acknowledgements

Todo empieza cuando María José Nuevo recuerda mis intenciones de volver al extranjero y María Luisa González me hace saber sobre la posibilidad de escribir una tesis doctoral en Groningen (¿quien sabía donde estaba Groningen en aquel entonces?!). Os agradezco vuestro interés de entonces y de ahora. Además, a Maria Luisa le debo la entera traducción del “Summary” a español, que en verdad, nunca habría existido si ella no le hubiera dedicado su tiempo y paciencia. Muchas gracias Marisa.

Dear Prof. Henk Busscher, a difficult person to deal with as you always said (although about me). What could I say that you don´t know?… You have created the scientific person I am considered to be. Your trust in the work we developed stimulated and enabled me to overcome the doubts, fears and lingering start of the unknown, which evolved into this project. I understand now (and only now) that your vision and quiet confidence kept me on the right track all the while. It remains a unique experience for me. Not least, I much appreciate your capacity to rise above any problem I caused.

Dear Prof. Henny van der Mei, thank you for your level-headed approach to this project throughout the entire course of the four years. You brought stability, patience and all the microbiology knowledge that I certainly lacked. Apart from that, you were able to provide invaluable hands-on expertise in the lab. In particular, I would like to thank you for your availability to my concerns, whether related to my project or extra-curricular.

Dear Prof. Willem Norde. I dare to say (although it may sound far too poetic) that you have been the warm wind from the south that always blew through the lab far too infrequent. I want you to know that your generosity with your limited time was always compensated by your expertise. Somehow, you remain to me one of the reluctant legends of science.

To you, all three my promoters, I am indebted to you for educating my mind in the scientific realm and providing me the controlled freedom I needed.

Dear Joop, you taught me what I know about the AFM. More importantly, I thank you for your unfailing support and accepting me for who I am without any pre-judgment. Of course, I do not forget that you gave me the space to satisfy one of my addictions. Despite the downside of it, the benefit is that gave me the chance to get to know you further. I want you to know that your writing the programs for analyzing the data, providing the Dutch summary, ensuring that the thesis gets to the printer and long long etcetera, makes you highly responsible for any success which may be associated to this book.

Dear Flo, you were the family away from home. I love the fact we were always able to find humor in the absurdities of life as a foreign PhD student, which indeed was a tremendous source of energy and inspiration. Nothing would have been the same without you around. I thank you for your understanding, company, patience, respect and willingness to experience new situations. I knew you were always there and that’s priceless Flo.

Dear Jerome, with you I have shared my most sincere thoughts, fears and happiness. To all, you brought your independent opinion and experience, providing the stairs to different points of view and intellectual growth. For many moments, your understanding annihilated the loneliness and impotence that life brings from time to time without previous appointment, that is a great feat.

124

Dear Luc, allias `the suave marques de Groningen´. I need to say that your forthright and principled nature is rare and to be cherished. I love to see how you always danced to your own tune and I considered myself privileged to have been exposed to your particular sense of humor and company. Cannot wait to hear that great music album you have with in you.

If anybody ever sees Joop, Flo, Jerome or Luc, and they asked for a drink, please buy them hundred. It`ll be my round.

Dewi, you were probably one of the nicest roommates one could ever hope for. I highly appreciate your tolerance and patience toward the existing cultural differences, making them a virtue and not necessarily a fault.

To Ietse Stokroos, I am grateful to you for providing the invaluable expertise in coating AFM tips and plates, the electron micrographs and, importantly, the flexible and accommodating nature which you displayed all through the time. .

Dear Ina and Ellen, thank you for your kindness and help which started at the very beginning and last till the end.

Karim and Geert. Muchas gracias por el cuidado y atención que siempre me habéis brindado. Quiero dejar claro que el hecho de que no nos hayamos visto demasiado durante los últimos años no tiene nada que ver con el olvido.

Norel, I have seen that I could count on you in my hour of need. I believe we were always able to enjoy oneanother´s company despite our different value system in life. It`s not to be taken lightly.

Ward, I know I never treated you in the way that you deserved. I want you to know I truly enjoy your originally in mind and spirit.

To Chris, Tjar, Anna, René, Geert. Thank for the your sociability, interest and enjoyable conversations. Chris, I did not forget all your lessons in getting me started in the microbiology lab.

Albert, Bart, and Wim (or Bin!!) the wonderful institution of after-hours AIO borrels left with you…all too soon!!

Queridas Graciela y Sonia, representáis las contadas amistades que han sobrevivido la distancia y diferencias. No puedo sentirme más que agradecida por vuestro continuo interés y apoyo en todo lo que he empezado a lo largo de los últimos años. Se que poseéis la habilidad para convertir en realidad vuestras esperanzas y sueños... girls, the future is ours!!

Caro Michele, non hai mai fallito a celebrare onestamente la mia felicité e mio successi, questo non he a cosa da prendere leggermente. So che gli ultimi anni non ti ho contattato tanto quanto deva, non vuele dire che non penso a te, affato, tu sempre sei nel mio cuore. Ancora mi piace quando ridi ogni volta che pronuncio nel modo sbagliato “luccello”, “lorzaino”…..

A mis queridos padres y hermano, a quien les dedico este libro. El apoyo incondicional que he recibido de vosotros a lo largo de todos los años que recuerdo, apoyo tanto asociado a mi vida como a mi carrera, me da el valor necesario para convertirme en la persona que espero ser. Me habéis ofrecido la libertad de viajar y ver el mundo, y aunque la distancia que nos separa duela en ocasiones, quiero que sepáis que solamente conozco un lugar llamado ´hogar´y que está y siempre estará donde quiera que vosotros os

125

halléis. Estoy orgullosa de la educación que me habéis otorgado y en verdad es el mejor legado que nunca pude esperar. Espero que os deis cuenta que quien soy y lo que hasta ahora he conseguido es debido a vosotros. Os quiero mucho.

Notes

Stellingen

behorende bij het proefschrift


Virginia Vadillo Rodríguez

Groningen, 1 september 2004

1. The soft particle model as forwarded by Ohshima is not conclusive with respect to the surface structures causing the softness (this thesis, Chapter 2).

2. It is a delicate task to amalgamate microscopic

properties derived from AFM measurements into macroscopic cell surface properties (this thesis, Chapter 1).

3. The maximum distance over which attractive forces are probed by AFM upon retraction of the tip are influential on the area blocked by an adhering bacterium on a substratum (this thesis, Chapter 5).

4. Bacterial desorption from solid substrata is related with adhesive forces as probed by AFM, provided the contact time between the bacterium and substratum is taken into account (this thesis, General Discussion)

5. The force applied when a bacterium comes into contact with a substratum influences is adhesion mode (this thesis, Chapter 8).

6. AFM fails to detect long-range DLVO forces when bacterial cells are investigated (this thesis, Chapter 8)

7. An empirical, strain-dependent factor to account for the surface structures on a bacterial cell should be considered when developing theoretical model to describe bacterial adhesion.

8. The way we communicate to people determines our quality of life.

9. Think faster to live longer.

10. Time within a man’s life time is not a straight line.

11. Understanding is an exponential process, that can sometimes have a large characteristic time τ.

12. Often we do not know what we are looking for until we find it, also in science.

13. Those who believe “Er gaat niets boven Groningen” should try the south of Spain.

14. In love or life not getting what you want is sometimes a wonderful stroke of luck.

University of Groningen Macroscopic and microscopic ...

Documents