The practical limitations encountered with high-resolution structural techniques such as isotope labelling, namely the size of the protein/colloid system and the size of the protein itself, mean that methods such as CD and vibrational spectroscopies remain attractive options. These low-resolution techniques provide a global estimation of the secondary structures (a-helix, ^-sheet or random coil) without specifying the local region involved in the conformational transition. The latter techniques are suitable for studying proteins at liquid-solid interfaces in more or less restricted conditions. For CD spectroscopy, limitations arise from the significant scattering of the wavelength of interest due to the size of the sorbent particles. The use of silica nanoparticles of diameter not exceeding 20 nm has been developed specifically for CD experiments (Billsten et al. 1998). Adsorption of globular proteins of different structural stabilities has been studied by CD on charged silica nanoparticles (Giacomelli and Norde 2001; Kondo and Fukuda 1998; Kondo et al. 1991). These studies have shown that the largest conforma-tional changes are observed for the adsorption of the less stable proteins such as BSA, myoglobin or haemoglobin. In contrast, the more rigid proteins such as cytochrome c, immunoglobulin G (IgG) and ribonuclease (RNase), which are characterised by their low adiabatic compressibility, are less unfolded by contact with the hydrophilic surface (Table 2).
Partially hydrophobic supports could be used for CD experiments by modification of silica particles by polymer silane coupling agents (Kondo et al. 1996) or by the use of Teflon latex nanoparticles (Giacomelli and Norde 2003; Zoungrana et al. 1997). Recent CD measurements of protein adsorption on silica nanoparticles of various diameters combined with
NMR have shown that the amount and the secondary structures of the adsorbed protein depend on the particle curvature (Lundqvist et al. 2004).
FTIR spectroscopy reveals not only changes in the global secondary structures ofproteins (Byler and Susi 1986; Surewicz et al. 1993), but also allows determination of the average solvation states by hydrogen-deuterium (H/D) isotope exchange, as detailed later on (Mitchell et al. 1988; Muga et al. 1991) and can quantify the protonation states of the aspartate and glutamate side chains (Chirgadze et al. 1975). Transmission-FTIR spectroscopy is still used for direct comparisons between a protein in solution and the same protein adsorbed on various types of solid particles such as chromatographic supports (Boulkanz et al. 1997; Soderquist and Walton 1980) or clay minerals (Revault et al. 2005; Servagent-Noinville et al. 2000). The only limitation is a particle size not exceeding 10 ^m. Conformational changes could be correlated with chromatographic retention times in the case of serum albumin and interferons (Boulkanz et al. 1995; Pantazaki et al. 1998).
A strategy for studying both adsorption kinetics and protein structural transition by FTIR is to use attenuated total reflection (ATR) mode at the liquid-planar solid interface (Ball and Jones 1995; Chittur 1998; van Straaten and Peppas 1991). In order to obtain a large variety of sorbent phases, many modification of the ATR crystal plate could be obtained either by deposition of lipids (Sharp et al. 2002) or polymers (Green et al. 1999; Lenk et al. 1989; Sukhishvili and Granick 1999) or by chemical grafting of SAMs (Cheng et al. 1994; Noinville et al. 2003). Adsorption kinetics maybe monitored by measuring the time-dependent increase in the amide band in a flow system or in stationary conditions (Chittur et al. 1986; Jeon et al. 1992). The earliest ATR-FTIR experiments were devoted to the real-time adsorption kinetics of proteins of interest to biomaterials including im-munoassays, protein chromatography, biosensors and biocatalysis (Cheng et al. 1994; Fu et al. 1993; Lenk et al. 1989; Müller et al. 1997; Ong et al. 1994). For instance, adsorption of IgG onto different supports was largely studied with respect to both the adsorbed amount and the structure to ensure antigen binding (Giacomelli et al. 1999a). The Y-shaped IgG molecules are composed of two Fab segments and one Fc domain. The Fab segments containing the binding sites show a different stability from the whole IgG molecules. The less structurally stable Fc fragment is more readily adsorbed than the Fab fragments, thus affecting the orientation of the adsorbed IgG (Buijs et al. 1996).
Was this article helpful?