Adenoviral expression of murine serum amyloid A proteins to study amyloid fibrillogenesis

Serum amyloid A (SAA) proteins are one of the most inducible acute-phase reactants and are precursors of secondary amyloidosis. In the mouse, SAA(1) and SAA(2) are induced in approximately equal quantities in response to amyloid induction models. These two isotypes differ in only 9 of 103 amino acid residues; however, only SAA(2) is selectively deposited into amyloid fibrils. SAA expression in the CE/J mouse species is an exception in that gene duplication did not occur and the CE/J variant is a hybrid molecule sharing features of SAA(1) and SAA(2). However, even though it is more closely related to SAA(2) it is not deposited as amyloid fibrils, We have developed an adenoviral vector system to overexpress SAA proteins in cell culture to determine the ability of these proteins to form amyloid fibrils, and to study the structural features in relation to amyloid formation. Both the SAA(2) and CE/J SAA proteins were synthesized in large quantities and purified to homogeneity. Electron microscopic analysis of the SAA proteins revealed that the SAA(2) protein was capable of forming amyloid fibrils, whereas the CE/J SAA was incapable. Radiolabelled SAAs were associated with normal or acute-phase high-density lipoproteins (HDLs); we examined them for their clearance from the circulation. In normal mice, SAA(2) had a half-life of 70 min and CE/J SAA had a half-life of 120 min; however, in amyloid mice 50 % of the SAA(2) cleared in 55 min, compared with 135 min for the CE/J protein. When the SAA proteins were associated with acute-phase HDLs, SAA(2) clearance was decreased to 60 min in normal mice compared with 30 min in amyloidogenic mice. Both normal and acute-phase HDLs were capable of depositing SAA(2) into preformed amyloid fibrils, whereas the CE/J protein did not become associated with amyloid fibrils. This established approach opens the doors for large-scale SAA production and for the examination of specific amino acids involved in the fibrillogenic capability of the SAA(2) molecule in vitro and in vivo.