PHYSICOCHEMICAL CHARACTERIZATION OF METAL-BINDING PROTEINS USING HPLC-ICP-MS, HPLC-MA-AAS AND ELECTROSPRAY-MS

A size exclusion HPLC method was developed and interfaced with ICP-MS detection for determining the metal profiles of commercially available rabbit liver metallothioneins (MT) and metallothionein-like proteins (MLP) extracted from fresh water mussels and hemolyzed osprey blood. T he redox state of the cysteine residues was indirectly evaluated via a cadmium saturation approach in the presence or absence of a reducing agent, followed by HPLC-microatomization (MA)-AAS and HPLC-ICP-MS anal yses. An electrospray-MS protocol was also developed to accurately mea sure the molecular weight of rabbit MT isoform II. Nanogram quantities of Cd-MT/MLP were poorly chromatographed on silica based supports. A copolymeric styrene-divinylbenzene size exclusion support provided a s ymmetrical peak (rabbit MT standard) and linear HPLC-MA-AAS calibratio n curves [r=0.9988; from the LOD (27 ng, as protein) to about 300 x LO D], indicating negligible losses of Cd during the chromatography of tr ace quantities. Go-injection of Cd2+ saturated samples with beta-merca ptoethanol (BMSH) was essential to repress Cd2+-support interactions w hich otherwise induced an undesirable metal-affinity retention mechani sm. In the presence of added Cd2+, 22 mmol/L BMSH did not significantl y compete for Cd2+ specifically bound to MT, while preventing non-spec ific binding to non-thiolic complexing sites. Crude mussel and osprey blood MLP extracts (in cold, deoxygenated Tris-HCl buffer) were obtain ed by ultracentrifugation (145,000 g) and thermocoagulation/centrifuga tion, respectively. Incubation with BMSH was prerequisite to obtain a maximum saturation of mussel and osprey blood MLP by Cd2+, even for sa mples conserved (-80 degrees C) in the presence of BMSH (22 mmol/L). T hese observations indicated that a major proportion of the cysteine re sidues present in these MLP were oxidized. The assumption of a fully r educed MT/MLP pool binding metals in a definite stoichiometry has been the basis of several quantitative metal binding assays involving the saturation of the thiolic complexing sites with a metallic marker (Ag, Cd2+, or Hg2+). Since thiolic agents may interfere, the metal satura tion protocols do not include a reducing step to ensure that all cyste ines in a MT/MLP extract are available for co-ordination. Given that v ariations in the redox state of crude MT/MLP extracts may compromise t he accuracy of metal saturation assays, it is concluded that the prepa ration of reference samples certified for total metallothionein conten t would be desirable.