Protein purity was assessed by SDS 8-18% PAGE (ExcelGel, GE Healthcare) heavily overloaded with samples run under reducing

conditions and stained with Brilliant blue R350. Native protein integrity, absence of aggregation and dissociation were demonstrated this website by native, non‐denatured 3-8% gradient PAGE in Tris acetate (NuPAGE Novex, Invitrogen), and by size exclusion chromatography of 0.02 mg samples in a volume of 100 μL on a 10 × 30 cm Superdex 200 column equilibrated and eluted at 0.5 mL/min with 10 mM Tris, 140 mM NaCl, pH 8.0 for SAP and 10 mM Tris, 140 mM NaCl, 2 mM CaCl2, pH 8.0 for CRP. The integrity of the protomers of SAP and CRP was verified by electrospray ionization mass spectrometry (ESIMS). After buffer exchange into pure water 2-4 μL samples were diluted 1/10 with a 50%MeCN/49.9%H2O/0.1%HCOOH v/v/v mixture and infused into the electrospray 5-FU source of a Quattro II triple quadrupole mass spectrometer (Micromass) under the following conditions: ES positive ion mode, 2.49 s scan with 0.11 s interscan delay, mass range m/z700–2750, cone voltage ramp 17–116 V, capillary at 3 kV. The concentrations of the specific proteins were confirmed by specific immunoassays for human CRP ( Eda et al., 1998 and Erlandsen and Randers, 2000) and SAP ( Nelson et al., 1991) respectively. Functional

integrity of the proteins for specific ligand recognition in vitro was established by their complete, strictly calcium dependent binding to phosphoethanolamine-Sepharose ( Hawkins et al., 1991). The authentic native state of the SAP preparation and its functional integrity for localization to amyloid deposits were investigated in vivo in normal healthy C57BL/6 mice and C57BL/6 mice in which AA amyloidosis had been induced by repeated injection of casein ( Hawkins et al., 1988a and Hawkins et al., 1991), in comparison with a highly purified non‐GMP batch of human SAP.

SAP was trace radiolabeled with 125I as previously described ( Hawkins et al., 1988a and Hawkins et al., 1991). Unlabeled non‐GMP SAP was spiked with labeled GMP SAP at approximately 0.3 μg (100,000 cpm) per mg. Normal healthy adult female C57BL/6 mice received 1 mg of the spiked SAP by IV injection and were then Cyclin-dependent kinase 3 bled at intervals thereafter for assay of total human SAP by electroimmunoassay and counting to estimate clearance of the labeled GMP human SAP. Two groups of AA amyloidotic mice received 0.3 μg tracer doses of either GMP or non‐GMP 125I‐SAP by IV injection. After 24 h they were bled out, killed and radioactivity was determined in the spleen and liver, which contain the amyloid deposits in this model. Pro‐inflammatory effects of the preparations in vivo were sought in wild type adult female C57BL/6 mice weighing ~ 20 g each, which were pre‐bled 48 h before testing to provide individual baseline values of the sensitive murine acute phase reactants, SAP ( Pepys et al., 1979a) and serum amyloid A protein (SAA), and then given 720 μg per mouse of each human protein IV (~ 30 mg/kg).