Viscous Boundary Lubrication of Hydrophobic Surfaces by Mucin

Nov 28, 2008 - found that mucin facilitates lubrication between hydrophobic PDMS ... The observed boundary lubrication behavior of mucin was found to ...
0 downloads 0 Views 1MB Size
Langmuir 2009, 25, 2313-2321

2313

Viscous Boundary Lubrication of Hydrophobic Surfaces by Mucin Gleb E. Yakubov,*,† James McColl,‡ Jeroen H. H. Bongaerts,†,§ and Jeremy J. Ramsden‡ UnileVer Corporate Research, Colworth Science Park, Bedfordshire MK44 1LQ, U.K., Department of Materials, Cranfield UniVersity, Bedfordshire MK43 0AL, U.K., and SKF Engineering and Research Centre, KelVinbaan 16, 3439 MT Nieuwegein, The Netherlands ReceiVed June 13, 2008. ReVised Manuscript ReceiVed NoVember 28, 2008 The lubricating behavior of the weakly charged short-side-chain glycoprotein mucin “Orthana” (Mw ) 0.55 MDa) has been investigated between hydrophobic and hydrophilic PDMS substrates using soft-contact tribometry. It was found that mucin facilitates lubrication between hydrophobic PDMS surfaces, leading to a 10-fold reduction in boundary friction coefficient for rough surfaces. The presence of mucin also results in a shift of the mixed lubrication regime to lower entrainment speeds. The observed boundary lubrication behavior of mucin was found to depend on the bulk concentration, and we linked this to the structure and dynamics of the adsorbed mucin films, which are assessed using optical waveguide light spectroscopy. We observe a composite structure of the adsorbed mucin layer, with its internal structure governed by entanglement. The film thickness of this adsorbed layer increases with concentration, while the boundary friction coefficient for rough surfaces was found to be inversely proportional to the thickness of the adsorbed film. This link between lubrication and structure of the film is consistent with a viscous boundary lubrication mechanism, i.e., a thicker adsorbed film, at a given sliding speed, results in a lower local shear rate and, hence, in a lower local shear stress. The estimated local viscosities of the adsorbed layer, derived from the friction measurements and the polymer layer density, are in agreement with each other.

1. Introduction Movement is a survival prerequisite for almost all living organisms. Efficient lubrication of moving internal and external surfaces is vital to minimize wear and loss of energy during motion. Examples are numerous and diverse: mammals lubricate their joints using synovial fluid and coat their respiratory airways, gastro-intestinal, ocular, and many other surfaces with a mucosal layer, while cephalopods utilize a mucus-coated foot on which they move.1 Biolubrication is also crucial for processing of foods in the mouth, which has a large bearing on oral health, mastication, swallowing, and mouthfeel.2,3 For example, an astringent mouthfeel is believed to be caused by the interaction of polyphenols with the salivary film on oral surfaces, resulting in a aggregation of salivary proteins at those surfaces, which in turn reduces oral lubrication.2-5 Adsorption of highly hydrated hydrophilic and amphiphilic polymers onto surfaces reduces friction, most notably between hydrophobic surfaces.6-9 A natural example of such an amphiphilic polymer is lubricin,10 a mucin-like glycoprotein with a high content of sulfonated sugars and sialic acid residues. Another much-studied example is hyaluronic acid (HA),11-15 * Towhomcorrespondenceshouldbeaddressed.E-mail:[email protected]. † Unilever Corporate Research. ‡ Cranfield University. § SKF Engineering and Research Centre.

(1) Roussel, P.; Delmotte, P. Curr. Org. Chem. 2004, 8, 413–437. (2) de Wijk, R. A.; Prinz, J. F. J. Texture Stud. 2006, 37, 413–427. (3) Noble, A. C. Chem. Taste: Mechanisms, BehaViors, Mimics 2002, 825, 192–201. (4) de Wijk, R. A.; Prinz, J. F. Food Qual. Pref. 2005, 16, 121–129. (5) Breslin, P. A. S.; Gilmore, M. M.; Beauchamp, G. K.; Green, B. G. Chem. Sens. 1993, 18, 405–417. (6) Jung, Y. C.; Bhushan, B. Nanotechnology 2006, 17, 4970–4980. (7) Andrienko, D.; Dunweg, B.; Vinogradova, O. I. J. Chem. Phys. 2003, 119, 13106–13112. (8) Zhu, Y. X.; Granick, S. Langmuir 2002, 18, 10058–10063. (9) Cassin, G.; Heinrich, E.; Spikes, H. A. Tribol. Lett. 2001, 11, 95–102. (10) Zappone, B.; Ruths, M.; Greene, G. W.; Jay, G. D.; Israelachvili, J. N. Biophys. J. 2007, 92, 1693–1708. (11) Benz, M.; Chen, N. H.; Israelachvili, J. J. Biomed. Mater. Res., Part A 2004, 71A, 6–15.

one of the main components of synovial fluids. In both cases the lubricating properties are thought to originate from dense brushtype coatings and/or high charge densities. Attempts have been made to understand the mechanism of biolubrication at the molecular scale. Adsorbed biopolymers and (glyco-)proteins form a barrier against bare surface-surface interaction through steric, brush-brush, and/or electrostatic repulsion. The ubiquity of charged biopolymers at biological interfaces might imply that polyelectrolye polymers facilitate lubrication across the majority of biological lubricating systems.16,12,11 However, the creation of such a barrier is not sufficient to ensure efficient boundary lubrication as it does not necessarily provide an efficient sliding mechanism.17 It has been proposed that hydrated salt ionsstypically accumulated as the counterions to the polyelectrolyte coatingssact as molecular ball bearings if they are in some way ‘trapped’ close to the surface and cannot be squeezed out. Hydrated ions, analogous to normal ball bearings, are thought to withstand compression forces acting on the surfaces, while at the same time providing a very efficient sliding mechanism.18-22 Mucin is one of the important constituents of saliva, and it is often assumed that it is the principle molecule contributing to oral lubrication.23-27 Mucins are high molecular weight gly(12) Zhu, Y. X.; Granick, S. Macromolecules 2003, 36, 973–976. (13) Tadmor, R.; Chen, N. H.; Israelachvili, J. Macromolecules 2003, 36, 9519–9526. (14) Tadmor, R.; Chen, N. H.; Israelachvili, J. N. J. Biomed. Mater. Res. 2002, 61, 514–523. (15) Jay, G. D.; Lane, B. P.; Sokoloff, L. Connect. Tissue Res. 1992, 28, 245–255. (16) Raviv, U.; Giasson, S.; Kampf, N.; Gohy, J. F.; Jerome, R.; Klein, J. Nature 2003, 425, 163–165. (17) Feiler, A. A.; Bergstrom, L.; Rutland, M. W. Langmuir 2008, 24, 2274– 2276. (18) Raviv, U.; Klein, J. Science 2002, 297, 1540–1543. (19) Raviv, U.; Laurat, P.; Klein, J. J. Chem. Phys. 2002, 116, 5167–5172. (20) Raviv, U.; Perkin, S.; Laurat, P.; Klein, J. Langmuir 2004, 20, 5322– 5332. (21) Zhu, Y. X.; Granick, S. Phys. ReV. Lett. 2004, 93, 096101. (22) Briscoe, W. H.; Titmuss, S.; Tiberg, F.; Thomas, R. K.; McGillivray, D. J.; Klein, J. Nature 2006, 444, 191–194. (23) Lindh, L.; Glantz, P. O.; Carlstedt, I.; Wickstrom, C.; Arnebrant, T. Colloids Surf., B 2002, 25, 139–146.

10.1021/la8018666 CCC: $40.75  2009 American Chemical Society Published on Web 01/15/2009

2314 Langmuir, Vol. 25, No. 4, 2009

coproteins and are ubiquitous in most organisms.1 They play a range of key roles, from nonspecific agents of the immune system1,28,29 to lubrication facilitators.30,24 The latter is aided by their ability to adsorb onto surfaces of practically any chemical nature.31,32 For example, there are two types of oral cavity surfaces: the ‘hard’ hydrophilic enamel surfaces of the teeth33 and the ‘soft’ hydrophobic (in the absence of a saliva coating) surfaces of gums, tongue, and palate.34 It is worth noting that the soft surfaces have a wide range of surface structures, roughness, and levels of keratinization, with the tongue (keratinized) having almost millimeter-sized taste buds, while gums (keratinized) and buccal (nonkeratinized) surfaces are characterized by micrometer-scale roughness.35 Recently Lee at al.30 have investigated the effect of pH and ionic strength on the adsorbing and lubricating properties of porcine gastric mucin. Although mucin’s negative charge density increases with increasing pH, they found that this does not result in enhanced lubricating properties. For mucin solutions at pH 7, when porcine gastric mucin (PGM) is highly charged, no lowering of the boundary lubrication coefficient (in comparison with the absence of mucin) was observed, as might have been expected following the polyelectrolyte lubrication model. Efficient boundary lubrication was observed at pH 2 though, when PGM is neutral. This observation suggests that the polyelectrolyte model may not be universal, and other lubrication mechanisms related to adsorption energy, conformation, and dynamics (including wear) of the adsorbed film may be common and important in biological systems. For example, the high-molecular-weight MUC5B mucins that are reported to be responsible for saliva lubrication36,24 exist in a variety of glycoforms, dominated by less charged subpopulations.37 Therefore, one can suggest that oral lubrication may originate not from charged polyelectrolytes, but rather from structural and local rheological properties of the adsorbed layer of mucins and their complexes with lowermolecular-weight proteins. In this paper, we focus on the weakly charged, short-sidechain, heavily glycosylated mucin, porcine gastric mucin “Orthana”, which has been well characterized recently.38,39 We investigate the effects of bulk concentration, ionic strength, and hydrophobicity of the underlying solid substrate on the softcontact lubricating properties of mucin using ball-on-disk (24) Berg, C. H.; Lindh, L.; Arnebrant, T. Biofouling 2004, 20, 65–70. (25) Bongaerts, J. H. H.; Rossetti, D.; Stokes, J. R. Tribol. Lett. 2007, 27, 277–287. (26) Stokes, J. R.; Davies, G. A. Biorheology 2007, 44, 141–160. (27) Feiler, A. A.; Sahlholm, A.; Sandberg, T.; Caldwell, K. D. J. Colloid Interface Sci. 2007, 315, 475–481. (28) Bansil, R.; Stanley, E.; Lamont, J. T. Annu. ReV. Physiol. 1995, 57, 635– 657. (29) Bansil, R.; Turner, B. S. Curr. Opin. Colloid Interface Sci. 2006, 11, 164–170. (30) Lee, S.; Muller, M.; Rezwan, K.; Spencer, N. D. Langmuir 2005, 21, 8344–8353. (31) Yakubov, G. E.; Papagiannopoulos, A.; Rat, E.; Waigh, T. A. Biomacromolecules 2007, 8, 3791–3799. (32) Arvidsson, A.; Lofgren, C. D.; Christersson, C. E.; Glantz, P. O.; Wennerberg, A. Biofouling 2004, 20, 181–188. (33) Berg, I. C. H.; Elofsson, U. M.; Joiner, A.; Malmsten, M.; Arnebrant, T. Biofouling 2001, 17, 173–187. (34) Ranc, H.; Elkhyat, A.; Servais, C.; Mac-Mary, S.; Launay, B.; Humbert, P. Colloids Surf., A 2006, 276, 155–161. (35) Newland, J. R.; Meiller, T. F.; Wynn, R. L.; Crossley, H. L. Oral Soft Tissue Diseases; Lexi-Comp: Hudson, OH, 2005. (36) Cardenas, M.; Elofsson, U.; Lindh, L. Biomacromolecules 2007, 8, 1149– 1156. (37) Thornton, D. J.; Khan, N.; Mehrotra, R.; Howard, M.; Veerman, E.; Packer, N. H.; Sheehan, J. K. Glycobiology 1999, 9, 293–302. (38) Yakubov, G. E.; Papagiannopoulos, A.; Rat, E.; Easton, R. L.; Waigh, T. A. Biomacromolecules 2007, 8, 3467–3477. (39) Di Cola, E.; Yakubov, G. E.; Waigh, T. A. Biomacromolecules 2008, 9, 3216-3222.

YakuboV et al.

tribometry. The use of soft surfaces makes the results relevant to the biolubrication of soft tissue. Many previous studies40,24 have utilized hard surfaces such as silica or mica, which results in either very high contact pressures (of the order of GPa), or very low loads and, possibly irrelevant, low sliding speeds. Only a few reported tribological studies30,25 have employed soft surfaces. The surfaces used in the present work do not mimic all the features of any specific biological surface, i.e., they are not meant as a true biomimetic substrate. Instead, they provide a generic model of soft biological contacts, and we use them here to investigate the influence of surface hydrophobicity, as well as surface roughness, on biolubrication. Furthermore, we correlate the lubricating properties of the adsorbed mucin layers with the structure and dynamics of the adsorbed film, assessed using optical waveguide lightmode spectroscopy (OWLS).

2. Experimental Section 2.1. Biopolymers and Other Chemicals. Pharmaceutical-grade PGM, molecular weight 546 kDa, was purchased from A/S Orthana Kemisk Fabrik (Kastrup, Denmark). “Orthana” mucin is used in a saliva substitute formulation “Saliva Orthana”. It has been found to adopt a dumbbell conformation in the bulk,38 with monomer dimensions of Rg ≈ 26 nm, dimers of Rg ≈ 52 nm, and with hydrodynamic radius of unglycosylated globular domains of Rhbead ≈ 10 nm.38,39 The commercial preparation was extensively dialyzed to remove all salts and other low-molecular-weight additives and finally lyophilized and stored for use as required.38 All solutions were made by dissolving weighed portions of the lyophilized material in ultrapure water (see below). The sample was shaken for 2 h and subsequently filtered through a Sartorius “Minisart” filter (200 nm pore size). The solutions with concentrations between 0.1 and 100 mg/mL were prepared and used immediately after preparation. The resulting mucin solutions were found to be Newtonian in a shear rate range of 1-105 s-1 (see Supporting Information for details of rheological characterization performed). Water was purified using a commercial system comprising an SG reverse osmosis precleaning unit and a Barnstead NANOpure Diamond unit equipped with semiconductor-grade ion-exchange resins, an ultrafilter (0.2 µm) and a UV oxidation chamber. The purified water had a resistivity 18.2 of MΩ · cm. 2.2. Tribological Measurements. Tribological measurements were carried out on a Mini Traction Machine (MTM) (PCS Instruments Ltd., UK). This technique has been described in previous studies.41 Briefly, friction forces between a disk and a loaded ball (applied normal force L) are measured as a function of speed using a force transducer on the ball. Ball and disk are both driven by separate motors so that the slide-to-roll ratio (SRR), defined as

SRR )

Vball - Vdisk (Vball + Vdisk) ⁄ 2

(1)

can be arbitrarily adjusted. The velocities determine the entrainment speed U, defined as

U ) (Vball + Vdisk) ⁄ 2

(2)

For each entrainment speed the measurements were performed using both conditions Vball > Vdisk and Vball < Vdisk, thus keeping the SRR constant. The average of the two measurements was taken, enabling evaluation of friction force (Ff) free from offset errors in the lateral force measurement and also removing the contribution from the (40) Raviv, U.; Tadmor, R.; Klein, J. J. Phys. Chem. B 2001, 105, 8125–8134. (41) Bongaerts, J. H. H.; Fourtouni, K.; Stokes, J. R. Tribol. Int. 2007, 40, 1531–1542.

Viscous Boundary Lubrication of Hydrophobic Surfaces

Langmuir, Vol. 25, No. 4, 2009 2315

Table 1. Contact Angles Measured on Smooth PDMS Surfaces for Water, Mucin Solution, and Air Bubble Immersed Either into Water or Mucin Solution surface

liquid phase

hydrophobic PDMS hydrophobic PDMS hydrophilic PDMS hydrophilic PDMS hydrophobic PDMS surface modified with an adsorbed layer of mucin* hydrophobic PDMS hydrophobic PDMS

deionized water 30 mg/mL mucin solution deionized water 30 mg/mL mucin solution deionized water air bubble in water air bubble in 30 mg/mL mucin solution

static contact angle 106.1 ( 3 103.6 ( 2 47.6 ( 6 43.3 ( 4 94.7 ( 2 75.7 ( 2 104 ( 3

* The hydrophobic PDMS surfaces were immersed into 30 mg/mL mucin solution for 30 min, subsequently rinsed twice with deionized water, and then dried using a N2 jet.

rolling friction (see refs 42-44 for more details). The measured friction force was used to calculate the friction coefficient µ, defined as

µ ) Ff⁄L

(3)

In a typical experiment the friction coefficient µ was measured five times for every combination of U, load L (typically 1 N) and SRR (typically 50%), starting at high U and then reducing it in a stepwise manner. The ball and disk were immersed in the polymer test solution for at least 30 min before starting each measurement and kept in the solution during measurement. Each tribopair was only used for one liquid and then discarded. For low polymer concentrations 0.4 no concentration-dependency was observed in the mixed regime, which we attribute to the adsorbed film thickness being much smaller than the effective gap between the surfaces, i.e., δ , h ≈ rs. 3.3. Contact Angle Effects on Tribological Behavior of Mucin. Previously it has been shown that for nonionic surfactant systems48 and for lubricants with a range of surface contact angles,41 a lateral Uη shift is due to changes in the water contact angle upon adsorption of surfactant on hydrophobic surfaces such as PDMS. The adsorption of a monolayer of surfactant onto the hydrophobic surface renders it more hydrophilic, causing the lubricant (water) to be more easily entrained into the rubbing contact. This results in an extension of the elastohydrodynamic regime to lower values of Uη and a lateral shift of the mixed regime to lower values of Uη. Figure 2a and b shows that mucin solutions also result in a lateral shift of the Stribeck curves to lower values of Uη. Figure 2b demonstrates that even for the lowest mucin concentration, a significant lateral shift of the mixed regime is observed, while boundary lubrication is comparable to that of water. We propose that the observed Uη shift of the mixed regime is related to the changes in the receding contact angle of water on PDMS. The contact angle is influenced by adsorption of the amphiphilic mucin glyco-proteins, similar to surfactants. In addition, our experiments indicate that the effects of contact angle dominate in the mixed lubrication regime when h ≈ rs . δ. We have performed receding contact angle measurements using an air bubble immersed either into water or 30 mg/mL mucin solution. This measurement is more relevant than liquid drops in air measurements, since our surfaces are always immersed in fluid. For a hydrophobic smooth PDMS surface, we see that the air contact angle is higher for mucin solution (∼104°) than for water (∼74°), a difference of about ∼30°. This suggests that the

Viscous Boundary Lubrication of Hydrophobic Surfaces

Figure 5. Stribeck curves at room temperature for a rough hydrophilic tribopair at various mucin bulk concentrations.

surface energy required to create a substrate-air interface (∆WSLV ) γ(1 + cos θ)) is ∼48 mJ/m2 less for the mucin solution. Though this change in contact angle does not render hydrophobic surfaces truly hydrophilic, it would appear to suffice as an explanation for the lateral shift of the Stribeck curve and the extension of the elastohydrodynamic regime to lower values of Uη. Interestingly, mucin solutions do not spread well over ‘virgin’ hydrophobic surfaces (Table 1). Results of contact angle measurements of 30 mg/mL mucin solution in air demonstrate that only a minor decrease (∼6°) in water contact angle is observed with an adsorbed mucin layer. To further investigate the influence of surface contact angle we have examined the lubrication of mucin solutions between hydrophilic PDMS surfaces. The advancing contact angles of a 30 mg/mL mucin solution and water on hydrophilic PDMS are very similar (Table 1), and both are significantly lower than the receding contact angle of mucin on hydrophobic PDMS. OWLS data on HL surfaces may not be representative due to the presence of Ti4+ ions on the surface, resulting in much higher adsorption than for HL PDMS. However, ellipsometry experiments on HB and HL silica surfaces indicate that the total adsorbed amount is ∼4 times lower for HL than for HB surfaces. Independent QCM experiments utilizing HB and HL PDMS-coated quartz crystals yield, upon exposing them to mucin solutions, a frequency shift (which corresponds to adsorbed mass) that is ∼10 times lower for HL than for HB PDMS (manuscript in preparation). The corresponding friction coefficients on hydrophilic rough surfaces are presented in Figure 5. Within experimental accuracy all curves overlap, indicating there is no added lubricating effect of adsorbed mucin over bare hydrophilic PDMS surfaces and that viscous boundary lubrication does not seem to occur. We attribute this to the low adsorbed mass of mucin ‘Orthana’ on HL PDMS substrates so that effects related to the adsorbed film are insignificant, i.e., the condition h ≈ rs . δ holds over entire Uη range. Also, it is possible that no full coverage of the mucin layer on HL surfaces is obtained and lubrication can be simply governed by the properties of bare HL PDMS. 3.4. Effects of Applied Load and Slide-to-Roll Ratio on Mucin Lubrication. Equations 4 and 5 implicitly suggest that for viscous boundary lubrication friction coefficient should increase linearly with sliding speed (µ ≈ Vs), which is not observed here; we observe a boundary friction plateau, i.e., µ ≈ const. In order to investigate this further, we have performed experiments with a varying slide-to-roll ratio, which enables more detailed examination of the complex flow in the tribological contact and at the same time expands the range of sliding speeds for given values of load and entrainment speed, following Vs ) SRRU.

Langmuir, Vol. 25, No. 4, 2009 2319

Figure 6. Boundary friction coefficient for rough hydrophobic surfaces versus sliding-to-roll ratio. In the inset the same data are presented as friction coefficient versus sliding speed (Vs ) SRRU) for speeds below 5 mm/s. Lines represent power law fits to the data and are given to guide the eye.

Figure 7. Boundary friction coefficient for rough hydrophobic surfaces versus applied load for various mucin bulk concentrations and entrainment speeds.

Figure 6 presents the sliding friction coefficients at various entrainment speeds U < 5 mm/s and varying SRR from 0% (pure rolling) to 200% (pure sliding) for two representative mucin concentrations 1 and 30 mg/mL (load 1 N). One can clearly see (in the insert of Figure 6) that the sliding friction coefficient increases with speed at very low sliding speeds, roughly below Vs ) 1 mm/s (∼SRR < 30%) and then levels off, especially for the 1 mg/mL solution. The observed dependency of friction coefficient on sliding speed and SRR is weaker than linear (slope