The application of atomic force microscopy to topographical studies

Planta (2000) 211: 641±647

The application of atomic force microscopy to topographical studies and force measurements on the secreted adhesive of the green alga Enteromorpha J. A. Callow1, S. A. Crawford2, M. J. Higgins2, P. Mulvaney3, R. Wetherbee2 1

School of Biosciences, University of Birmingham, Birmingham, B15 2TT, UK School of Botany, University of Melbourne, Parkville 3052, Victoria, Australia School of Chemistry, University of Melbourne, Parkville 3052, Victoria, Australia

2 3

Received: 22 February 2000 / Accepted: 20 April 2000

Abstract. Atomic force microscopy (AFM) enables the topographical structure of cells and biological materials to be resolved under natural (physiological) conditions, without ®xation and dehydration artefacts associated with imaging methods in vacuo. It also provides a means of measuring interaction forces and the mechanical properties of biomaterials. In the present study, AFM has been applied for the ®rst time to the study of the mechanical properties of a natural adhesive produced by a green plant cell. Swimming spores of the green alga Enteromorpha linza (L.) J. Ag. (7±10 lm) secrete an adhesive glycoprotein which provides ®rm anchorage to the substratum. Imaging of the adhesive in its hydrated state revealed a swollen gel-like pad, approximately 1 lm thick, surrounding the spore body. Force measurements revealed that freshly released adhesive has an adhesion strength of 173 1.7 mN m)1 (mean SE; n 90) with a maximum value for a single adhesion force curve of 458 mN m)1. The adhesive had a compressibility (equivalent to Young's modulus) of 0.54 ´ 106 0.05 ´ 106 N m)2 (mean SE; n 30). Within minutes of release the adhesive underwent a progressive `curing' process with a 65% reduction in mean adhesive strength within an hour of settlement, which was also re¯ected in a reduction in the average length of the adhesive polymer strands (polymer extension) and a 10-fold increase in Young's modulus. Measurements on the spore surface itself revealed considerably lower adhesion-strength values but higher polymer-extension values than the adhesive pad, which may re¯ect the deposition of dierent polymers on this surface as a new cell wall is formed. The study demonstrates the value of AFM to the imaging of plant cells in the absence of ®xation and dehydration artefacts and to the characterisation of the mechanical properties of plant glycoproteins that have potential utility as adhesives. Abbreviation: AFM atomic force microscopy Correspondence to: J. A. Callow; E-mail: [email protected]; Fax: +44-121-4145925

Key words: Atomic force microscopy ± Bioadhesion ± Biofouling ± Enteromorpha ± Glycoprotein (adhesive) ± Secretion (glycoprotein)

Introduction Atomic force microscopy (AFM), introduced by Binnig et al. (1986), relies on the sensitive detection, by laser, of de¯ections to a small, cantilever-mounted tip, which occur in response to intermolecular forces as the tip is raster-scanned across a surface. In addition to producing high-resolution topographical information about materials in their native state, AFM in force mode also lends itself to measurements of fundamental mechanical properties of materials such as adhesive strength, compressibility and elasticity. Several groups have reported measurements of the elasticity and adhesive strength of single polymer strands (Rief et al. 1997 and refs therein; Chatelier et al. 1998). Rief et al. (1997) showed that polymer deformation can be described by a Langevin function, and that segment elasticity is dominated by twists in bond angles. Whilst the technique is extensively used by surface scientists, application by biologists is still relatively limited (for reviews, see Shao and Yang 1995; Nagao and Dvorak 1998; Vinckier and Semenza 1998) and the value of some studies is limited by use of non-hydrated methods. With the more recent development of ¯uid cells, AFM has been used to study adhesion forces of biological materials in solution, for example, cell-adhesion proteoglycans of marine sponges (Dammer et al. 1995), the lustrin A adhesive that binds nacre tablets in the abalone shell (Smith et al. 1999), and whole cells of Escherichia coli (Ong et al. 1999). A detailed understanding of the mechanical properties of adhesive polymers at plant cell surfaces is currently lacking. In this context the green alga Enteromorpha provides a useful model system for studying adhesion to substrata. The motile spores of this alga attach to a range of substrata through a mechanism that

642

J. A. Callow et al.: AFM studies on Enteromorpha spore adhesive

involves the secretion of a preformed adhesive glycoprotein. The attached spore develops a cell wall and the resultant sporeling ultimately develops into a mature, sessile plant. Although recent biochemical and immunological studies in our laboratory have revealed certain of the molecular properties of materials contributing to the adhesive (Stanley et al. 1999; Callow et al. 2000), the physical state and mechanical properties of the secreted adhesive polymers are unknown. The released adhesive appears to be ®brillar in character when viewed by transmission (Evans and Christie 1970) and scanning electron microscopy (Stanley et al. 1999). However, the likelihood of ®xation, dehydration and other sample preparation artefacts requires caution in the interpretation of images obtained by these methods. Atomic force microscopy, on the other hand, permits the non-invasive imaging of specimens in their natural, hydrated state, without extensive sample preparation. It is of both fundamental and practical interest, e.g. in the design of materials resistant to attachment and in the development of bioadhesive materials, to understand the forces that determine the adhesion characteristics of organisms that exhibit strong adhesion to substrata. In the present paper we report the results of the ®rst study to apply AFM in both imaging and force modes to a green plant cell. The general topographical features of the adhesive pad of Enteromorpha spores in their native state have been determined and force curves have been used to compute adhesion-strength values of secreted adhesive as it cures over time.

images collected. In some cases spores were re-scanned at 90° to the original scan. To measure the mechanical properties of the spore adhesive, the atomic force microscope was used in force mode, whereby the de¯ection of the cantilever tip is measured as it is brought vertically towards (i.e. in the z direction) and then pulled away from a single spot on the sample surface. Initial attempts to do this by making a crude topographic image in contact mode, then switching to forcecurve mode, gave unsatisfactory results because the method was slow and often resulted in spores detaching due to lateral forces between the tip and spore. This precluded collection of reliable force-curve measurements on freshly settled spores at very early time points. Therefore a dierent method (`force-curve mapping') was used to obtain force-curve measurements on single spores within minutes of settling. Having brought a spore and the cantilever tip (`short fat' tips with a spring constant of 0.58 Nm)1 were used for force measurements) into close proximity with each other, characteristic force curves were sampled for the polystyrene surface. The cantilever was then moved closer to the spore in 0.5-lm increments in the X direction, monitoring the shape of the force curves continuously. Contact with the adhesive pad could be readily recognised as a change in sample height and by the shift in the appearance of force curves from the `hard' non-compressible, nonadhesive character of the polystyrene surface, to the `soft', compressible nature of the adhesive pad, with its characteristic adhesive `pull-os'. The cantilever was then advanced a further 1 lm onto the adhesive pad and a series of 10 force curves exhibiting adhesion events was taken. Adhesive properties were sampled at two adjacent positions at each time-point by moving the cantilever by 0.5 lm in the Y direction, before returning to the starting position. This procedure was repeated at dierent time-points, after each of which a short imaging scan was performed to check that the tip was still contacting the adhesive pad. At the end of each experiment, force curves were carried out on the polystyrene surface to check that the tip had not become contaminated. For calculation of mechanical properties, raw output data collected as tip de¯ection (Volts) versus piezo displacement (Z, nm) were transformed into force (mN) versus distance (nm) curves using the Dimforce programme (P.G. Hartley, CSIRO Molecular Science, Clayton, Australia) based on the methods described by Ducker et al. (1992) and Hartley (1999), using a value for the tip diameter of 0.1 lm (Biggs and Mulvaney 1994). `Compressibility' (N m)2) was calculated as the gradient of the slope on the extending portion of the force-displacement curve between the point of contact and the region of constant compliance. Àdhesion-strength' (mN m)1) and `polymer-extension' (nm) parameters were calculated from the retracting portion of the force-displacement curve.

Materials and methods Algal materials Reproductive thalli of Enteromorpha linza (L.) J. Ag. were collected from Port Melbourne (37°50¢31¢¢S 144°55¢54¢¢E) Victoria, Australia and zoospores were released as described by Callow et al. (1997). Released spores were allowed to settle, in darkness, onto 35-mmdiameter polystyrene Petri dishes. After a settling period of 5± 30 min, unsettled spores were removed and the resulting bio®lms were washed in ®lter-sterilised seawater (FSW). Settled spores were examined in FSW by AFM immediately, or after incubation for various times in the light (15±17 lmol photons m)2 s)1) at 15 °C.

Atomic force microscopy of hydrated specimens Imaging and force measurements were performed at room temperature (20 °C) on a Dimension 3100 atomic force microscope (Digital Instruments, Santa Barbara, Calif., USA) with a Nanoscope III controller. Dishes containing settled spores were secured to the stage of the atomic force microscope by means of a small metal disc secured by epoxy adhesive or double-sided Sellotape to the underside of the dish. Samples were imaged in FSW in the ¯uid cell in contact mode using `long thin' DNPS silicon nitride tips (Digital Instruments) with a measured spring constant of 0.06 Nm)1 (Cleveland et al. 1993; Sader et al. 1995, 1999). Settled spores were located with the optical objective and manually brought into the scanning region of the cantilever tip. The tip was brought into contact with the surface and scanning commenced using a scan size of 30±50 lm, a scan rate of 1 Hz and a z-range (height) of 7 lm. Once contact had been made with a spore the scan rate was reduced to 0.5 Hz and height and de¯ection

Scanning electron microscopy Spores settled onto glass coverslips were ®xed overnight in 5% glutaraldehyde in seawater, and washed three times in seawater before being transferred to 1% OsO4 in cacodylate buer (pH 7.2) for 1 h. Samples were dehydrated in a graded ethanol series. Dehydrated spores were then dried with a Bal-Tec CPD 030 critical-point drier using ethanol as the intermediary agent. Coverslips with attached spores were mounted onto scanning electron microscopy stubs with carbon sticky tabs and an electrical conduction bridge formed between stubs and samples with Electrodog high-conductivity paint (Acheson, Plymouth, UK). The samples were coated with platinum in a Dynavac Xenosput sputtercoater, and viewed in a Phillips XL-30 ®eld-emission scanning electron microscope operated at 2.0 kV.

Results Imaging of settled Enteromorpha zoospores by AFM in contact mode presented two technical problems. First,


643

dislodge the spores, typically with their surrounding adhesive. For freshly settled spores it proved to be impossible to scan a whole spore plus the adhesive material without dislodging it, and consequently the earliest time at which a whole spore could be imaged at low resolution was 4.5 h after settling. The spore body itself was beyond the dynamic range of the atomic-forcemicroscope drive to reveal any structure but the surrounding adhesive was successfully imaged as an extensive, swollen, gel-like pad, some 20 lm ´ 15 lm in dimension (Fig. 1). Cross-section analysis of height images revealed the pad to be approximately 1 lm thick (data not shown). The validity of the general shape and topography was con®rmed by scanning specimens at 90° to the original scan but some surface-features, notably lines at the edges of the pad, were revealed as tip artefacts since such features were reversed in position when scan data were collected in `retrace' mode. Imaging of the adhesive of young spores at higher levels

Fig. 1. An AFM contact-mode image of a settled zoospore of Enteromorpha 4.5 h after settling, showing the adhesive pad (ap) surrounding the original spore (s) which is beyond the dynamic range. The striations at the edge of the adhesive pad (arrows) are tip artefacts. The bulbous structure (arrowhead) may be a bacterium trapped under the adhesive

the adhesive material released, especially by freshly adhered spores, appeared to be extremely soft and therefore prone to smearing by the scanning tip. Second, a settled spore is approx. 5±7 lm in diameter and therefore presents a substantial obstacle to the movement of the cantilever tip. This not only generates the possibility of tip convolution artefacts due to the extreme size range presented, but also creates a resistance to the tip which was often suciently great to

Fig. 2. Scanning electron micrograph of a ®xed, critical-point-dried spore of Enteromorpha, 5 h after settling, displaying adhesive (arrows)

Fig. 3a±c. A sample of AFM approach-retract cycles presented as uncorrected piezo displacement-photodiode voltage plots, illustrating dierent types of adhesion pro®le. The examples are taken from measurements on spores at dierent times after settlement. The x-axis shows the vertical movement of the coupled cantilever/piezo (nm), and the y-axis the bending of the cantilever (V) as detected by the split photodiode. The regions of interest shown against the approaching (or extending) and retracting curves are (i) a baseline region where the tip and sample are separated; (ii) a region where tip and sample start to interact and the sample starts to be compressed; (iii) the region of `constant compliance' where the tip and sample movement are coupled; (iv) a hysteretic region of the retracting curve caused by an adhesion force (v) which is overcome as the tip-sample bonds break (vi) and the tip returns to the baseline (vii)

644

of resolution was impossible due to smearing, and use of the atomic force microscope in `¯uid tapping' mode did not improve imaging. The general topography of the adhesive pad imaged by AFM is in marked contrast to the shrunken, ®brillar appearance shown by scanning electron microscopy (Fig. 2). Surface force measurements were made on the adhesive pad using the atomic force microscope in force mode. In successive approach-retract cycles on all samples, at all time-points, some variation was observed in the characteristics of the force curves and this is illustrated in Fig. 3 in the form of raw instrument output (i.e. photodiode voltage vs. piezo-displacement). Although not all approach-retract cycles resulted in an adhesion peak, most retraction curves showed the hysteresis typical of tip-sample adhesion eects. Simple adhesion events with a single `pull-o' (Fig. 3a) were observed but more typical were saw-toothed curves with multiple pull-os (Fig. 3b). Some force curves showed a retardation of the adhesion peak (Fig. 3c). To compute adhesive forces, photodiode voltagepiezo displacement curves were converted to normalised force-distance curves which take account of the tip radius and spring constant of the cantilever. An example of a normalised force-distance curve is shown in Fig. 4, together with an indication of how values for adhesion strength and polymer extension were determined. The maximum adhesion `pull-o' observed from an individual retraction curve was 458 mN m)1. An average value for adhesive strength of 173 17 mN m)1 (mean SE) was determined from over 90 force curves taken from the adhesion pads of 3 separate spores within 35 min of settlement. To follow the change in adhesive properties with time, samples of adhesion curves were taken from three positions on the adhesive pad surface at intervals over a period of 5.5 h, the ®rst sample being taken after only 13 min settlement. Two-way analysis of variance (ANOVA) revealed no signi®cant dierences between sample positions (P > 0.05, F 1.85). The time-course (Fig. 5) revealed a rapid, 65% reduction in adhesion strength within the ®rst hour of settlement, the value thereafter reaching a fairly constant level. Curing pro®les consistent with this general pattern were obtained for other spores in separate experiments. The values for polymer extension showed a signi®cant increase in the ®rst hour after settlement before subsequently declining as the adhesive cured (Fig. 6). Analysis of the same batch of spores after overnight incubation (three separate spores measured with a new tip) revealed no further reduction in adhesion-strength values. However, there was a 10fold increase in polymer-extension values to an average of 2.1 lm, compared with only 0.19 lm at 5.5 h after settlement. To compare adhesive properties of the adhesive pad with those of the spore surface, in separate experiments the cantilever tip was placed directly over the spore body, visual location being con®rmed by the height of the cantilever above the surface. The spore surface was much less adhesive than the adhesive pad, values for freshly settled spores being


Fig. 4. A corrected force-distance retract curve derived from the curve shown in Fig. 3b to illustrate the method of calculating values for maximum pull-o adhesion strength and polymer extension

Fig. 5. Changes in mean adhesion strength with time after settlement for the adhesive pad (d) and the spore surface (s), obtained in two separate experiments on dierent Enteromorpha spores. At each time point, samples of 30 retraction curves were taken and adhesionstrength values were computed for the maximum `pull-o' (Fig. 4). Values 95% con®dence limits

only 25% of those of the pad (Fig. 5). Although adhesion initially reduced with time after settlement, the general trend was towards a slight, but signi®cant increase in adhesiveness rather than a reduction. A similar trend was noted for polymer-extension values (Fig. 6). Adhesive curing was also re¯ected in changes in Young's modulus (E), calculated from normalised approach curves as the gradient of the slope between the point of contact and the region of constant compliance (Fig. 7). For freshly settled spores, approach curves revealed a soft, compressible material with a slope of shallow gradient averaging 0.54 ´ 106 0.05 ´ 106 N m)2 (mean SE; n 30) (Fig. 8). The value of E increased 10-fold within 30 min as the adhesive became less compressible, or stier over time, reaching an average value of 5.06 ´ 106 0.17 ´ 106 N m)2 (mean SE).


645

Fig. 6. Changes in mean polymer extension with time after settlement for the adhesive pad (d) and the spore surface (s), obtained from two separate experiments on dierent Enteromorpha spores. Polymer-extension values were calculated as shown in Fig. 4. Each time point represents the mean of 30 retraction curves 95% con®dence limits

Fig. 8. Changes in mean values for Young's modulus (E) for the adhesive pad after settlement. Each point is the average of 30 dierent approach curves 95% con®dence limits

Discussion All sessile, marine organisms that have a dispersal phase in their life histories require an adhesive strategy that allows them to form a strong and permanent bond to a range of substrates, under water, over a wide range of temperatures, salinities and conditions of turbulence. Whilst much progress has been made in characterising the molecular and certain material properties of the protein glues used by marine invertebrates (Vreeland et al. 1998), there is little comparative information on adhesives used by marine plants. The present paper, together with recent cell and molecular studies (Stanley et al. 1999; Callow et al. 2000), attempts to address this by studying the general morphology and material properties of the Enteromorpha spore adhesive by AFM. Because of the extreme height of Enteromorpha spores (in AFM terms), the softness of the adhesive, and the propensity for the spores to detach when exposed to lateral forces of the tip, the value of AFM in imaging settled Enteromorpha spores, without intro-

Fig. 7. A normalised force-distance approach curve derived from the raw output shown in Fig. 3c, to illustrate the method of computing Young's modulus (E)

ducing imaging artefacts, was rather limited. Nevertheless, studies of the general topography of the adhesive pad amply revealed the limitations of previous images based largely on scanning electron microscopy, providing a totally dierent view of the general topography of the secreted adhesive. Rather than a reticulate network of ®brils the adhesive in its native, hydrated state is a swollen pad-like structure, without any obvious ®brillar structure at higher resolution. The asymmetrical shape of the adhesive pad, also observed in scanning-electronmicroscopy images (Stanley et al. 1999) may be derived from a non-radial pattern of adhesive discharge as the pyriform spores settle onto a surface at dierent angles. From the image illustrated in Fig. 1, the volume of an adhesive pad may be roughly estimated as 225 lm3, assuming an average pad thickness of 1 lm. Adhesive is released from a spore as it settles, by the discharge of the glycoprotein contents of a large number of adhesive vesicles from the anterior region. From previously published micrographs (Evans and Christie 1970) the volume of the roughly cone-shaped apical (non-chloroplast-containing) region of a swimming spore containing these vesicles is approximately 7.5 lm3. If the proportion of that cone occupied by vesicles is estimated as 10%, then this gives a volume for the condensed contents of the vesicles of 0.75 lm3. Comparison of this with the estimated volume of the adhesive pad gives a `swelling factor' of the order 300-fold. Even allowing for the possibility of substantial errors in these various estimates, this indicates that the process of adhesion involves a considerable expansion of the condensed vesicle contents following their discharge. In successive approach-retract force curve cycles some variation in adhesion pro®le was observed. For statistical and steric reasons (Dammer et al. 1995) some cycles failed to reveal tip-adhesive interactions but most cycles demonstrated the hysteresis typical of adhesive events. Curves with simple `pull-os' were obtained, representing the breaking of single polymer strands, but a more common pro®le was the `saw-toothed' curve representing multiple pull-os. This may be attributed to

646


the breaking of successive adhesive polymer links, although similar patterns obtained by pulling the elastic protein titin have been ascribed to the sequential unfolding of modular adhesive domains of single molecules (Marsalek et al. 1999). In other instances the adhesive peak was retarded until the tip was 500±600 nm above the surface, indicating the presence of very elastic polymers with relatively weak intramolecular adhesive interactions. Similar eects were observed for adhesion proteoglycans of sponges (Dammer et al. 1995). Some of the variations in the retraction pro®les may represent a response to the adhesive molecules being continually pulled and broken during successive cycles. Another possibility is that the adhesive pad contains several polymer types, some of which are adhesive and exert strong in¯uences on the tip, others being more elastic in character. How `sticky' is the Enteromorpha spore adhesive compared with other biological glues? Adhesive-strength values derived from the force plots take account of the tip radius of the atomic force microscope and are expressed as N m)1. This is equivalent to J m)2, which is therefore an estimate of energy per unit area of surface, or surface energy (c), the energy required to separate two bodies. The work required to separate dissimilar bodies is termed adhesion and the value of c for many solids and liquids is approximately 20 mN m)1 (20 mJ m)2). The adhesive pad of Enteromorpha, when freshly released, gave a maximum value for adhesive strength of 458 mN m)1 (i.e. from an individual retraction curve) and an average adhesive strength of 173 mN m)1 (determined from 90 retraction curves and three separate spores). These values are considerably greater than the »30 mN m)1 reported for E. coli lipopolysaccharide by Ong et al. (1999). However, the determination of absolute values for adhesive strength (and compressibility/ elasticity) obtained by AFM, are complicated by considerations of the area of contact between tip and sample. In the present paper the tip of the atomic force microscope has been approximated to a sphere of nominal radius 0.1 lm. The choice of 0.1 lm is based upon calibration work reported earlier by Biggs and Mulvaney (1994) for the Digital Instruments V-shaped cantilevers. More-accurate values can be determined by probing the sample with a silica sphere of known radius glued to the tip. This approach was not followed because the dimensions of available spheres were approximately the same as those of the Enteromorpha spore itself. Also, the greater contact area of a sphere is less predictable for a settled spore of uneven height than would be the case for surface ®lms used in precise surface chemistry studies. Considerations of tip geometry and chemistry do not aect the interpretation of relative changes in mechanical properties with time since the same tip is used throughout an experiment. However, considerations of surface area of contact are unimportant if the atomic force microscope is measuring the breaking of individual adhesive polymer strands, and others report the adhesive properties of biological glues determined by AFM in terms of àdhesion forces', i.e. without regard for considerations

of tip radius. Thus, Dufrene et al. (1999) reported `strong' adhesion forces of 9 nN for the spore adhesive of the fungus Phanerochaete chrysosporium. Smith et al. (1999) estimate the breaking force of the abalone nacre adhesive to be of the order of 1 nN and Dammer et al. (1995) estimate binding strengths of sponge cell-adhesion glycoproteins to be of the order 0.1 nN. If the adhesive-strength values in the present paper are converted to adhesive forces by multiplying by the nominal tip radius, then comparable measurements for freshly released Enteromorpha spore adhesive would be 17 1.7 nN (mean SE), the highest value for an individual force curve being 46 nN. These values suggest a very strong adhesive. The substantial decrease in adhesion-strength values with time, and the approximately parallel reduction in polymer-extension values, suggests that the adhesive undergoes a rapid `curing' process in which the adhesive polymers become progressively cross-linked and therefore less extensible with time. Christie et al. (1970) referred to a progressive `hardening' of the adhesive with time and this was substantiated by Callow et al. (2000) who showed that attached spores become progressively less sensitive to detachment after treatment with proteolytic enzymes. Further, Stanley et al. (1999) showed that the presumptive adhesive polymer, a self-aggregating, N-linked 110-kDa glycoprotein (under denaturing conditions), was less extractable by buer or detergent (SDS) 1 h after settlement. Evans and Christie (1970) suggested, on the basis of transmission electron microscopy, that the spore surface was also covered in adhesive. However, in the present study, comparisons of adhesiveness and polymer extension between the adhesive pad and the spore surface with time, revealed dierences suggesting that the two surfaces have dierent properties. After settlement, a new cell wall is synthesised by the spore over a period of several hours and it is possible that the low adhesion values, but higher polymer-extension values re¯ect the deposition of less sticky, but more-extensible wall polymers. The material properties of the spore adhesive in terms of `compressibility' were obtained by measuring the slope of corrected approach curves as the tip de¯ects following contact with the sample. A hard sample gives rise to a de¯ection of the cantilever which is linear with piezo travel while a soft, compressible or elastic material deforms in response to the applied force, resulting in a non-linear relationship. The slope of the normalised curve is: E DF =RDD with the units of Nm)2 or force per unit area. Young's modulus (E), often called the èlastic modulus', describes the mechanical resistance of a material while undergoing elongation or compression, also has dimensions of force per unit area. Atomic-force-microscopy force curves have been used to examine the viscoelastic properties of relatively few biological materials but the value of »5 ´ 105 N m)2 for the freshly-released Enteromorpha spore adhesive is of the same order as a 20% solution of


647

gelatin (Mahon 1999), while the 10-fold greater value for cured spore adhesive is similar to that of natural rubber (Pocius 1997). The approach outlined in this paper opens up the prospect of more detailed studies on the adhesion properties of cells and their interaction with substrates, at nanoscale levels of resolution. In the context of algal biofouling, for example, it is possible to envisage probing adhesive strength by AFM using microscope tips that have been derivatised with materials with dierent physico-chemical properties, and coatings such as `foul-release' silicone elastomers. This approach will complement traditional quantitative settlement studies (e.g. Callow et al. 1997) and adhesion-strength testing under ¯ow-shear stress (e.g. Schultz et al. 2000).

Ducker WA, Senden TJ, Pashley RM (1992) Measurement of forces in liquids using a force microscope. Langmuir 8: 1831± 1836 Dufrene YF, Boonaert CJP, Gern PA, Asther M, Rouxhet PG (1999) Direct probing of the surface ultrastructure and molecular interactions of dormant and germinating spores of Phanerochaete chrysosporium. J Bacteriol 181: 5350±5354 Evans LV, Christie AO (1970) Studies on the ship-fouling alga Enteromorpha. I. Aspects of the ®ne-structure and biochemistry of swimming and newly settled zoospores. Ann Bot 34: 451± 466 Hartley PG (1999) Measurement of colloidal interactions using the atomic force microscope. In: Farinato R & Dubin P (eds) Colloid-polymer interactions: from fundamentals to practice. Wiley, New York, pp 253±286 Mahon J (1999) Shear and compressive moduli of pectin gels. MEngSci thesis, University of Melbourne Marszalek PE, Lu H, Carrian-Vasquez M, Oberhauser AF, Schulter K, Fernandez JM (1999) Mechanical unfolding of titin modules. Nature 402: 100±103 Nagao E, Dvorak JA (1998) An integrated approach to the study of living cells by atomic force microscopy. J Microsc 191: 8±19 Ong Y-L, Razatos A, Georgiou G, Sharma MM (1999) Adhesion forces between E. coli bacteria and biomaterials surfaces. Langmuir 15: 2719±2725 Pocius AV (1997) Adhesion and adhesives technology. Hanser Publishers, Munich Vienna New York Rief M, Oesterhelt F, Heymann B, Gaub HE (1997) Single molecule force spectroscopy on polysaccharides by atomic force microscopy. Science 275: 1295±1297 Sader JE, Larson I, Mulvaney P, White LR (1995) Method for the calibration of atomic force microscope cantilevers. Rev Sci Instr 66: 3789±3798 Sader JE, Chon JWM, Mulvaney P (1999) Calibration of rectangular atomic force microscope cantilevers. Rev Sci Instr 70: 3967±3969 Shao Z, Yang J (1995) Progress in high resolution atomic force microscopy in biology. Quart Rev Biophys 28: 195±251 Smith BL, Schaer TE, Viani M, Thompson JB, Frederick NA, Kind J, Belcher, Stucky GD, Morse DE, Hansma PK (1999) Molecular mechanistic origin of the toughness of natural adhesives, ®bres and composites. Nature 399: 761±763 Schultz MP, Finlay JA, Callow ME, Callow JA (2000) A turbulent channel ¯ow apparatus for the determination of the adhesion strength of microfouling organisms. Biofouling, in press Stanley MS, Callow ME, Callow, JA (1999) Monoclonal antibodies to adhesive cell coat glycoproteins secreted by zoospores of the green alga Enteromorpha. Planta 210: 61±71 Vinckier A, Semenza G (1998) Measuring elasticity of biological materials by atomic force microscopy. FEBS Lett 430: 12±16 Vreeland V, Waite JH, Epstein L (1998) Polyphenols and oxidases in substratum adhesion by marine algae and mussels. J Phycol 34: 1±8

This study was carried out with ®nancial support to J.A.C from the University of Melbourne (EC Dyason Universitas 21 Fellowship) and the Royal Society. P.M. and R.W. were supported by the Australian Research Council.

References Biggs S, Mulvaney P (1994) Measurement of forces between gold surfaces in water by atomic force microscopy. J Chem Phys 100: 8501±8505 Binnig G, Quate CF, Gerber Ch (1986) Atomic force microscope. Phys Rev Lett 56: 930±933 Callow ME, Callow JA, Pickett-Heaps JD, Wetherbee R (1997) Primary adhesion of Enteromorpha (Chlorophyta, Ulvales) propagules: quantitative settlement studies and video microscopy. J Phycol 33: 938±947 Callow JA, Stanley MS, Wetherbee R, Callow ME (2000) Cellular and molecular approaches to understanding primary adhesion in Enteromorpha: an overview. Biofouling, in press Chatelier X, Senden TJ, Joanny J-F, di Meglio J-M (1998) Detachment of a single polyelectrolyte chain adsorbed on a charged surface. Europhys Lett 41: 303±308 Christie AO, Evans LV, Shaw M (1970) Studies on the ship-fouling alga Enteromorpha, II. The eect of certain enzymes on the adhesion of zoospores. Ann Bot 34: 467±482 Cleveland JP, Manne S, Bocek D, Hansma PK (1993) A nondestructive method for determining the spring constants of cantilevers for scanning force microscopy. Rev Sci Instr 64: 403±406 Dammer U, Popescu O, Wagner P, Anselmetti D, Guntherodt H-J, Misevic GN (1995) Binding strength between cell adhesion proteoglycans measured by atomic force microscopy. Science 267: 1173±1175