Colloids and Surfaces A: Physicochemical and Engineering Aspects 149 (1999) 481–489
Molecular adsorption vs. cell adhesion at an electrified aqueous interface S. Kovacˇ, V. Svetlicˇic´ *, V. Zˇutic´ Center for Marine and Environmental Research, Ru2er Bosˇkovic´ Institute, PO Box 1016, 10 001 Zagreb, Croatia Received 7 October 1997; accepted 11 May 1998
Abstract Cell adhesion in the presence of surface-active polymers has been studied at the dynamic liquid interface of the dropping mercury electrode. Dextrans with average molecular weights of 70 000 and 500 000 were selected as surfaceactive polymers and marine nanoflagellate Dunaliella tertiolecta as model cell. The electrochemical response at the dropping mercury electrode measures directly the contribution of each component in the process of the initial monolayer formation. Competition between adsorption of molecules and adhesion of cells was discerned by analyzing two parameters: the frequency of cell adhesion and the time for monolayer film formation at the electrode surface. Adhering cells and adsorbed polymers form separate domains in the monolayer. © 1999 Elsevier Science B.V. All rights reserved. Keywords: Biofilm formation; Biopolymer adsorption; Cell adhesion; Dropping mercury electrode; Dunaliella tertiolecta
1. Introduction Mixtures of dissolved molecules and microbial cells are typical of a natural environment such as seawater. Studies of dynamics of initial monolayer formation are essential for understanding and control of processes in natural aquatic systems, for industrial applications, and in biomaterials where the most extensive studies have been performed [1]. Adsorption of dissolved organic molecules has been extensively studied at natural and model surfaces in the aquatic environment [2,3], while the effect of adsorbed molecules on cell adhesion at an aqueous interface is not well understood. Adsorption of an initial molecular layer is gen* Corresponding author. Fax: +385 1 46 80 242; e-mail:
[email protected]
erally considered as a prerequisite for cell adhesion in the process of biofilm formation [1,4]. There are also vast data in the literature on the inhibitive effects of adsorbed molecules [5–7]. Moreover, it is not clearly resolved when cell adhesion takes place through molecular contact with the substrate in an aqueous environment [8,9]. Application of electrodes to probe cell adhesion is still scarce [10,11], although the electrode surface can easily be controlled in terms of surface charge and interfacial energy in a precise manner by an applied potential for a given liquid phase composition. Probing cell adhesion in the presence of surfaceactive polymers at a freshly exposed dynamic liquid interface of the dropping mercury electrode (DME) [12–14] is a subject of this study. The DME is particularly suited as a probe for colloidal [15,16 ] and heterodispersed [17–22] systems,
0927-7757/99/$ – see front matter © 1999 Elsevier Science B.V. All rights reserved. PII S0 9 2 7- 7 7 5 7 ( 9 8 ) 0 05 1 9 - 6
482
S. Kovacˇ et al. / Colloids Surfaces A: Physicochem. Eng. Aspects 149 (1999) 481–489
owing to its continuously renewable surface and efficient transport of solutes and particles by convective streaming [23].
process and the representative behavior can be determined only by analyzing a large set of data collected under identical experimental conditions.
2. The electrochemical method The electrochemical method was used, that can directly differentiate molecular adsorption from adhesion of surface-active particles [17]. The method is basically chronoamperometry of dissolved oxygen at potentials of current maximum (‘‘polarographic maximum’’) [24,25] which is caused by gradients of interfacial tension at the fast dropping mercury electrode. The adsorption of organic molecules causes a decrease in the surface tension gradients at the mercury electrode/solution interface, and consequently a suppression of convective streaming. In the chronoamperometric curve, the adsorption manifests as a gradual decrease in oxygen reduction current [24]. On the other hand, surfaceactive particles, such as organic droplets or cells in aqueous environments, can be characterized through their attachment signals [20,22,25]. The signals appear as sharp perturbations in I–t curves of oxygen reduction due to a localized drop of interfacial tension caused by molecular contact. The amplitude of attachment signals reflects the particle size, and the frequency of signals corresponds to the concentration of reactive particles. The potential range of attachment signals appearance serves to estimate the energy of adhesion. When the oxygen reduction current decreases to the diffusion-limited value, it is assumed that the electrode surface is fully covered with organic molecules [24]. The dynamics of film formation (full monolayer, surface coverage, h=1) can be followed by measuring the time, t, when current drops to the diffusion-limited value (I ). t is then d the time needed for a monolayer formation at the growing interface. It has to be pointed out that the dropping mercury electrode is also a model liquid interface. It has a fast growing renewable surface, and an experiment can be repeated many times at will. This is an important aspect of the method since the arrival of cells at the interface is a stochastic
3. Experimental 3.1. Cell culture The axenic cell culture of marine nanoflagellate Dunaliella tertiolecta Butcher (strain CCMP 1320) was obtained from Provasoli-Guillard Center for Culture of Marine Phytoplankton, Bigelow Laboratory for Ocean Sciences. Phytoplankton cells were grown in F-2 medium [26 ] in batch cultures at ambient conditions. Components of the medium were added to seawater filtered through a Gelman filter 0.22 mm. Cell counts were made using a haemacytometer. Cells were harvested after approximately 8 days of growth. The majority of growth medium (#90%) was removed after a mild centrifugation (1500g, 5 min). The maximum cell density in stock suspension was 1010/l. The electrochemical measurements were made within 3 h after preparation of a stock suspension. Aliquots of stock cell suspension were added to organic-free electrolyte or dextran solutions of given concentration immediately before electrochemical measurement.
3.2. Dextran solutions Non-polar dextran of average molecular weight 70 000 (D-70) (Serva, research grade), and dextran sulphate, sodium salt (Merck, for laboratory use) of average molecular weight 500 000 (D-500*) were used as models of dissolved molecules. The concentration of stock solution for D-70 was 300 mg/l, and for D-500* 625 mg/l. Aliquots of stock solution were added to 20 ml of organic-free electrolyte solution. The aqueous solutions contained 0.1 M NaCl and 5 mM NaHCO to main3 tain pH at 8. Water used for preparation of the solutions was ultra-pure MilliQ water.
S. Kovacˇ et al. / Colloids Surfaces A: Physicochem. Eng. Aspects 149 (1999) 481–489
483
3.3. Electrochemical measurement
4.1. Electrochemical response in dextran solutions
The electrochemical measurements were performed in a standard Methrom vessel with 20 ml volumes of cell suspension, solutions of organic molecules or their mixtures. The measured samples were air saturated and the vessel was open to air throughout the experiments. A dropping mercury electrode (drop life 2.0 s, flow rate 6.03 mg/s, maximum surface area 4.57 mm2) was used in electrochemical measurement with 0.1 M Ag/AgCl electrode as reference in a three-electrode configuration. Measurements were done in 0.1 M NaCl electrolyte solution at 20°C for the following reasons: (1) the literature contains much physical and chemical information about the mercury/aqueous interface [27]; and (2) streaming maximum of oxygen reduction is better pronounced than at ionic strength of seawater. The measurements were performed using a PAR 174A polarographic analyzer. Current–time (I–t) curves at a constant potential were recorded and stored using a Nicolet 3091 digital oscilloscope connected to a PC. Because of the stochastic nature of the processes in cell suspensions, at least 30 current–time curves were recorded in each sample. Results of electrochemical measurement are presented as mean values obtained by analyzing 15 or 30 I–t curves. The I–t curves presented later in Figs. 2–4 were selected as most representative for a given suspension.
Water soluble dextrans (a-1,6-glucose polymer) were chosen as surface-active biopolymers. Dextrans are known for their specific adsorption at the mercury electrode in a range of positive but also negative surface charges [28]. Adsorption is fast and controlled by mass transport from the solution. In this study we used dextran sulphate of molecular weight 500 000 (D-500*) and neutral dextran of molecular weight 70 000 (D-70). Fig. 1 compares I–t curves for oxygen reduction in the presence of increasing concentrations of D-500*: 50, 100, 200, and 625 mg/l as indicated on the corresponding curves. The thin line (curve 0) represents the response in the organic-free electrolyte. At dextran concentration of 50 mg/l suppression of oxygen reduction current is significant, but current is not suppressed to the diffusion-limited value, I , during the drop life (2 s). At dextran d concentration 100 mg/l the current drops to I at d time t =1.3 s, while at dextran concentration of 1
4. Results Chronoamperometric curves were recorded at constant potential, −400 mV, where the current of oxygen reduction is the most sensitive to adsorption and adhesion of organic constituents. At this potential the hydrophobic mercury surface bears a positive charge (surface charge density s=+3.8 mC/cm2), and the interfacial tension is close to the maximum value (electrocapillary maximum). Under such conditions at the interface, hydrophobic and electrostatic attractions are expected to prevail in adhesion of cells, as well as in the adsorption of organic solutes.
Fig. 1. Overlaid current–time curves for oxygen reduction at −400 mV in 0.1 M NaCl aqueous solution in the presence of 50, 100, 200 and 625 mg/l D-500*. Curve 0 is the I–t curve in organic-free electrolyte. Time t for a monolayer formation at the growing interface is also indicated for concentrations >50 mg/l.
484
S. Kovacˇ et al. / Colloids Surfaces A: Physicochem. Eng. Aspects 149 (1999) 481–489
625 mg/l the current is suppressed to the diffusionlimited value almost during the whole drop lifetime (t =50 ms). The experimental parameter t 3 will be used in the analysis of dynamics of monolayer film formation.
4.2. The electrochemical response of cells As the model unicellular organism, we used Dunaliella tertiolecta, a marine nanoflagellate without the cell wall. The organism is simple to grow in batch culture and forms stable cell suspensions. The cell size of 6–10 mm and flexibility of the cell membrane are features of choice to obtain characteristic electrical signals for attachment of single cells. The arrival of a cell and its adhesion at the mercury interface leaves a fingerprint on the I–t curve of oxygen reduction (Fig. 2) in the form of an asymmetrical spike (sharp increase and slower decay), termed the attachment signal. The sharp rise in current reflects a fast establishment of molecular contact between a cell and the electrode, and the following slower decay of current reflects the subsequent spreading of cell material at the interface. The random appearance of attachment signals suggests a stochastic process. The average number of attachment signals is proportional to the cell density in the sample. Since the attachment signal for each cell adhesion has a finite duration, without tailing (as in the case of oil droplets [18]), we concluded that the spread cell material forms a compact patch at the free electrode surface. The attachment signals have been investigated separately [29,30] in oxygen-free suspensions in order to characterize the adhesion of single cells. In the oxygen-free solution, current–time transients, the attachment signals, are caused only by the double layer charge displacement from the contact area at the electrode [18]. This allows characterization of the nature and surface area of the contact between the cell and the metal interface. The displacement current (I ) at a given time D t is proportional to the contact area (A) and an initial charge density at the mercury surface
Fig. 2. Typical I–t curves with attachment signals of single cells recorded in dispersion of D. tertiolecta, cell density 5×107/l. The thin line compares the oxygen reduction current in organicfree electrolyte before addition of cell suspension (curve 0) and diffusion-limited current (I ). d
(s ) [18]: Hg dA I =− s D dt Hg The amount of displaced charge (q ) is obtained D by integration of displacement current (I ) from D the initial contact at time t to the time t when 0 m
S. Kovacˇ et al. / Colloids Surfaces A: Physicochem. Eng. Aspects 149 (1999) 481–489
485
maximum contact area was attained [19]:
P
t q = m I dt=−A s D f Hg D t0 Direct contact between the attached cell and the electrode surface was demonstrated, and the contact area after spreading at the positively charged electrode has been determined as (0.8–2.0)×10−4 cm2/cell, which is by two orders of magnitude larger than the original cell. Fig. 2 shows selected I–t curves of oxygen reduction in suspension of Dunaliella tertiolecta with cell density of 5×107/ml. Each I–t curve is unique and reflects the distribution of particles and molecules around the electrochemical probe at a given time interval. Attachment signals have duration 60–200 ms and an amplitude of 0.6–2.2 mA. Each attachment signal ends in a distinct suppression of oxygen reduction current. With the chosen cell density the average frequency of attachment signals is 10±2.1 during 2 s of drop life. The sequence of attachments on each mercury drop results in a stepwise attenuation of streaming maximum towards the maximum suppression (I ), which d corresponds to h=1. In contrast to I–t curves recorded in dextran solutions ( Fig. 1), where t is perfectly reproducible on each I–t curve, the time for complete film formation in Fig. 2 differs for each I–t transient. The average value for 30 curves analyzed is t=1.50±0.32 s. Notice that the suppression of oxygen streaming maximum in Fig. 2 is partly due to the presence of dissolved surfaceactive material inherent to the cell culture [31]. The adsorption of the dissolved surface-active material causes surface coverage of h=0.25. 4.3. Electrochemical response of cells in dextran solutions We present two sets of experimental I–t curves in the presence of D-500* that best illustrate the effect of dissolved dextran molecules in the mixture (Figs. 3 and 4). The thin line is the I–t curve recorded before addition of D. tertiolecta cells. The attachment signals are apparent even in the presence of high concentration of adsorbable
Fig. 3. Selection of I–t curves recorded in suspension of D. tertiolecta cells (5×107/l ) with 50 mg/l of D-500*. The thin line is the I–t curve for dextran solution prior to addition of cell suspension.
organic polymer. The frequency of their appearance decreases with increase of dextran concentration. The time for complete film formation is considerably shorter than in the cell suspension or dextran solution alone [Fig. 5(a)]. Signals appear during the whole drop lifetime, although they are less frequent after t. The attachment frequency
486
S. Kovacˇ et al. / Colloids Surfaces A: Physicochem. Eng. Aspects 149 (1999) 481–489
Fig. 4. Selection of I–t curves recorded in suspension of D. tertiolecta cells (5×107/l ) with 100 mg/l of D-500*. The thin line is the I–t curve recorded in the absence of D. tertiolecta cells.
(N/drop life) and film formation time (t) were analyzed on 15 I–t curves for each concentration of D-500*. The relationship between cell attachment frequency and increasing concentration of D-500* is presented in Fig. 5(b). For constant cell density (5×107/l ), the average attachment frequency decreases exponentially (10±2 to 1.6±1.2) with increasing dextran concentration (10 to 300 mg/l ).
Fig. 5. (a) Relationship of film formation time t with increase in concentration of D-500* before ( · · · ) and after addition of 5×107/l D. tertiolecta cells (———). Bars denote standard deviation. (b) Effect of D-500* concentration on frequency of cell attachment at DME (N/drop life). Cell density 5×107/l.
Cell adhesion reduces the film formation time for D-500* concentrations up to 300 mg/l. This is clear evidence for the mixed film formation consisting of spread cellular material and adsorbed dextran molecules. At higher dextran concentrations the rate of film formation is fast, leaving only a small fraction of the electrode surface available for cell adhesion. Thus, the molecular
S. Kovacˇ et al. / Colloids Surfaces A: Physicochem. Eng. Aspects 149 (1999) 481–489
adsorption and cell adhesion are concurrent processes. To confirm this interpretation we studied cell adhesion in the presence of smaller dextran molecules (D-70) that are more efficient in film formation [Fig. 6(a)]. The film formation time in dextran solutions was not affected by the presence of D. tertiolecta cells for D-70 concentrations
487
150 mg/l. The average attachment frequency decreased from 9±1.4 to 0.5±0.5 with increasing dextran concentration from 5 to 200 mg/l [Fig. 6(b)]. As adhesion of cells results in compact patches of spread material and the rest of the electrode surface is covered with randomly adsorbed dextran molecules [17], we expect that the resulting film has a mosaic-like structure.
5. Discussion
Fig. 6. (a) Dependence of film formation time t on concentration of D-70 before ( · · · ) and after addition of 5×107/l D. tertiolecta cells (———). Bars denote standard deviation. (b) Effect of D-70 concentration on frequency of cell attachment at DME (N/drop life). Cell density 5×107/l.
Adsorption of dissolved organic molecules and adhesion of phytoplankton cells at the mercury electrode/solution interface results in a coverage of the electrode surface with organic material that displaces counter ions and water molecules from the interface (schematic presentation in Fig. 7). Suppression of oxygen reduction current is proportional to the surface coverage in both cases. The electrochemical method enables a distinction between cell adhesion and polymer adsorption. Upon impact of D. tertiolecta cells with the interface, a fast deformation occurs, as in the case of oil droplets, followed by rupture of the fluid membrane and spreading of cell content onto the mercury surface, experimentally detected as a wellresolved attachment signal. The experimentally determined surface area covered by cell material is 100 times larger than the cross-section of the original cell. Adsorption of dextran molecules is fast and controlled by mass transport from the solution. There are two regimes of transport, before and after the time t when the oxygen reduction current drops to the diffusion-limited value. t is the time needed for monolayer formation at the growing interface. For t
t, transport of dissolved oxygen and dextran is diffusion-controlled while arrival of cells remains a stochastic process. For the series of I–t curves, the parameter t has a range of values [Figs. 5(a) and 6(a)] both in suspension of cells and in mixtures with dextran, due to random arrival and attachment of cells to the interface. We recorded the appearance of
488
S. Kovacˇ et al. / Colloids Surfaces A: Physicochem. Eng. Aspects 149 (1999) 481–489
Fig. 7. Schematic presentation of interaction between a cell, biopolymer molecule and positively charged mercury electrode in aqueous electrolyte solution whereby displacement of the electrode double layer charge takes place.
attachment signals, i.e. cell adhesion, for t>t, although a full surface coverage is assumed to be attained already at t. This unexpected finding can be interpreted as a competition between cell adhesion and molecular adsorption for the fraction of a fresh surface area formed by the drop growth. The frequency of cell attachment decreases exponentially with increase in dextran concentration [Figs. 5(b) and 6(b)] for two reasons: increase in the rate of dextran film formation leaves less surface available for cell adhesion but also reduces the convective streaming and the cell transport. Competition between adsorption of molecules and adhesion of cells in solutions of two different dextrans can be discerned by analyzing the time for complete film formation and frequency of cell adhesion occurring after t. In the presence of lower molecular weight dextran, D-70 attachment signals are absent at t>t if t∏0.5 s. In the presence of higher molecular weight, D-500* attachment signals are detected during the whole drop lifetime, even for shortest t measured (625 mg/l, t=0.05 s). We concluded that smaller molecules (D-70) are more efficient in covering the mercury surface (faster film formation and/or a more compact film) [32], and that cell adhesion is less probable in competition with adsorption of molecules than in solutions of D-500*. D-500* molecules diffuse more slowly and seem to form a less compact D-500* film at the interface. In the mixture of cells and biomolecules, we demonstrated a composite film formation by simultaneous adsorption and adhesion processes. Spread cells and adsorbed polymers presumably form mosaic-like structures as adhesion of cells results in compact patches of spread material and
the rest of the electrode surface is covered with randomly adsorbed dextran molecules [17]. The electrochemical response at the dropping mercury electrode measures directly the contribution of each component in the process of initial monolayer formation. We are currently expanding this study using solutions of non-ionic detergents and proteins as well as suspensions of marine bacteria [33].
References [1] W.G. Characklis, K.C. Marshall, in: W.G. Characklis, K.C. Marshal (Eds.), Biofilms, Wiley, New York, 1990, chapter 1. [2] W. Stumm, Chemistry of the Solid–Water Interface, Processes at the Mineral–Water and Particle–Water Interface in Natural Systems, Wiley, New York, 1992. [3] W. Stumm, J.J. Morgan, Aquatic Chemistry, Chemical Equilibria and Rates in Natural Waters, Environmental Science and Tehnology, Wiley-Interscience Series of Texts and Monographs, Wiley, New York, 1996. [4] K.E. Cooksley, B. Wigglesworth-Cooksley, Aquat. Microb. Ecol. 9 (1995) 87. [5] N.F. Owens, D. Gingell, P.R. Ritter, J. Cell Sci. 87 (1987) 667. [6 ] J.H. Paul, W.G. Jeffrey, Can. J. Microbiol. 31 (1985) 224. [7] E.A. Vogler, Colloids Surf. 42 (1989) 233. [8] A. Trommler, D. Gingell, H. Wolf, Biophys. J. 48 (1985) 835. [9] H.H.M. Rijnaarts, W. Norde, E.J. Bouwer, J. Lyklema, A.J.B. Zehnder, Colloids Surf. B: Biointerfaces 4 (1995) 5. [10] D. Gingell, J.A. Fornes, Nature 256 (1975) 211. [11] A.S. Gordon, S.M. Gerchakov, L.R. Udey, Can. J. Microbiol. 27 (1981) 698. [12] A.L. Bard, L.R. Faulkner, Electrochemical Methods, Fundamentals and Applications, Wiley, New York, 1980, pp. 146–158. [13] J. Heyrovsky´, J. Ku˚ta, Principles of Polarography, Czechoslovak Academy of Science, Prague, 1965.
S. Kovacˇ et al. / Colloids Surfaces A: Physicochem. Eng. Aspects 149 (1999) 481–489 [14] A.W. Adamson, Physical Chemistry of Surfaces, Wiley, New York, 1982, pp. 212–217. [15] V. Zˇutic´, E. Nicolas, P. Gerard, L. Gierst, J. Electroanal. Chem. 44 (1973) 107. [16 ] M. Heyrovsky´, J. Jirkovsky´, Langmuir 11 (1995) 4288. [17] V. Zˇutic´, V. Svetlicˇic´, J. Tomaic´, Pure Appl. Chem. 62 (1990) 2269. [18] V. Zˇutic´, S. Kovacˇ, J. Tomaic´, V. Svetlicˇic´, J. Electroanal. Chem. 349 (1993) 173. [19] N. Ivosˇevic´, J. Tomaic´, V. Zˇutic´, Langmuir 10 (1994) 2415. [20] V. Svetlicˇic´, N. Ivosˇevic´, V. Zˇutic´, D. Fuks, Croat. Chem. Acta 70 (1997) 141. [21] N. Ivosˇevic´, V. Zˇutic´, Croat. Chem. Acta 70 (1997) 167. [22] D. Bizzoto, J. Lypkowski, Progr. Colloid Polym. Sci. 103 (1997) 201. [23] V.G. Levich, Physiochemical Hydrodynamics, PrenticeHall, Engelwood Cliffs, NJ, 1962. [24] R.G. Barradas, F. Kimmerle, J. Electroanal. Chem. 11 (1966) 163. [25] S. Kovac´, Heterocoalescence of organic particles and mercury electrode in aqueous electrolyte solution and seawater, M.Sc. Thesis, University of Zagreb, Zagreb, 1993 (in Croatian).
489
[26 ] R.R.L. Guillard, J.H. Ryter, Com. J. Microbiol. 8 (1962) 229. [27] J. Lyklema, R. Parsons, Electrical Properties of Interfaces. Compilation of Data on the Electrical Double Layer on Mercury Electrodes, Office of Standard Reference Data, National Bureau of Standards, Department of Commerce, Washington, DC, 1983. [28] B. Malfroy, J.A. Reynaud, Anal. Biochem. 84 (1988) 1. [29] N. Ivosˇevic´, Organic particles in the sea: interaction with mercury electrode, M.Sc. Thesis, University of Zagreb, Zagreb, 1995 (in Croatian). [30] N. Ivosˇevic´, S. Kovac´, V. Zˇutic´, R. Lewin, 9th Int. Conf. on Surface and Colloid Science, Sofia, 6–12 July 1997, Book of Abstracts, p. 351. ´ osovic´, E. Marcˇenko, N. Bihari, F. Krsˇinic´, [31] V. Zˇutic´, B. C Mar. Chem. 10 (1981) 505. [32] R.A. Frazier, M.C. Davies, G. Matthijs, C.J. Roberts, E. Schacht, S.J.B. Tendler, P.M. Williams, Langmuir 13 (1997) 4795. [33] S. Kovac´, V. Zˇutic´, V. Svetlicˇic´, Cell adhesion at the dropping mercury electrode, in: European Research Conference on Natural Waters and Water Technology, San Feliu de Guixols, Spain, 4–9 October 1997.