Membrane-Grafted Hyaluronan Films: A Well-Defined Model System of Glycoconjugate Cell Coats Ralf P. Richter,* ,² Kai K. Hock, ² Jeffrey Burkhartsmeyer, ² Heike Boehm, ² Pit Bingen, ² Guoliang Wang, § Nicole F. Steinmetz, ‡ David J. Evans, ‡ and Joachim P. Spatz ² Department of New Materials and Biosystems, Max-Planck-Institute for Metals Research, Heisenbergstrasse 3, 70569 Stuttgart, Germany, Department of Biophysical Chemistry, The UniVersity of Heidelberg, INF 253, 69120 Heidelberg, Germany, the Department of Biological Chemistry, John Innes Centre, Colney, Norwich NR4 7UH, United Kingdom, and Q-Sense AB, Redegatan 13, 42677 Va ¨stra Fro ¨lunda, Sweden Received December 7, 2006; E-mail: ralf.richter@urz.uni-heidelberg.de Hyaluronan (HA), ubiquitous in the extracellular space, is intriguing by virtue of its simplicity. 1 With a persistence length of ∼4 nm and a contour length of up to a few micrometers, this linear polymer of identical disaccharide units forms an extended random coil in solution. 2 An important biological feature is its attachment to the surface of living cells 3 and the interaction with diverse hyaluronan binding proteins. 1,4 The soft and hydrated pericellular coats thereby created play key roles in the general protection of the cell and in the communication with its environment. Only little is currently known about the supramolecular structure of these assemblies and how it relates to properties and functions. The predominant role of water and the low degree of order in these matrices renders direct structural investigations difficult and, notwithstanding progress in the field, 1,4,5 the in vivo investigation of these coats remains a challenge. For a thorough characterization of the relationship between structure and function, it is desirable to move from living cells with their complex dynamics to model systems with tunable complexity. Here we present a simple method to create hyaluronan films that are well-defined in their molecular attachment to the substrate, a supported lipid bilayer (SLB). The confinement to a solid support makes these model coats accessible to characterization with a range of surface-sensitive techniques that are not easily applicable on living cells. By employing such techniques, we investigate the films’ thickness, grafting density, permeability, and mechanical properties. SLBs not only provide tunable densities of adhesion sites 6 together with a background of very low unspecific binding, 7 but they also mimic the biological environment of hyaluronan. 3 We incorporated a fraction of 5% biotinylated lipids into small unilamellar vesicles (SUVs). Their exposure to glass or silica results in the formation of SLBs 8 that provide enough functional sites for stable immobilization of a monolayer of streptavidin (SA) 9 (Figure 1A,B). SLB-formation as well as coverage with SA was directly tracked and controlled by quartz crystal microbalance with dis- sipation monitoring (QCM-D) (Figure 1C). Exposure of end-biotinylated HA led to immediate binding, as observed by QCM-D. Strong shifts in dissipation, D, accompanied by relatively small shifts in resonance frequency, f, are characteristic of the formation of a very soft and highly hydrated layer such as expected for hyaluronan (Figure 1C). The strong dependence of the QCM-D response on the molecular weight, M w , of HA is remarkable: much stronger responses in f and D for short-chained hyaluronan (HA50), as compared to long-chained HA (HA1000), provide a first indication that the HA-film density increases strongly as M w decreases. Nonbiotinylated HA did not generate any QCM-D response. The thickness, h, of the polysaccharide films and the average distance between neighboring HA-anchoring sites, s, were deter- mined by colloidal probe reflection interference contrast microscopy (RICM) 10 and reflectometry, respectively (Table 1, Supporting Information). A comparison of the results with the molecules’ radii of gyration, 2 R g , and contour lengths, L c , is instructive. The maximal anchoring densities generated with our grafting-to approach were sufficient to induce significant chain stretching for all hyaluronan lengths: while HA1000 stretched to more than five times its radius ² Max-Planck-Institute and Heidelberg University. ‡ John Innes Centre. § Q-Sense. Figure 1. Immobilization of hyaluronan (HA) on a supported lipid bilayer. (A,B) Schematic illustration of the immobilization strategy (not drawn to scale). A solid-supported lipid bilayer (SLB) is formed by the exposure of small unilamellar vesicles (SUVs), containing a fraction of biotinylated lipids, to a silica or glass surface. Hyaluronan, biotinylated at its reducing end, is immobilized on the SLB via streptavidin (SA). The molecular conformation of hyaluronan is dependent on the grafting density, giving rise to a mushroomlike state (A) or a brushlike state (B). (C) Kinetics of film formation, as monitored in real time by QCM-D. All steps in the immobilization scheme can be tracked. The two-phase behavior together with the final changes in frequency and dissipation, Δf )-25 Hz and ΔD < 0.3 × 10 -6 , upon exposure of SUVs are characteristic for the formation of an SLB of good quality. 8 An additional frequency shift of -28 Hz and small changes in dissipation upon addition of SA confirm the formation of a protein monolayer. 9 The strong increase in dissipation upon HA-binding reflects the highly hydrated and viscoelastic state of the forming film. Published on Web 04/05/2007 5306 9 J. AM. CHEM. SOC. 2007, 129, 5306-5307 10.1021/ja068768s CCC: $37.00 © 2007 American Chemical Society