A 25,000 molecular weight protein constituent of human amyloid fibrils related to amyloid protein AA

Review: amyloidogenesis-unquestioned answers and unanswered questions

Current assumptions and conclusions in several active areas of amyloid research are examined to see how consistent the data from chosen in vitro and in vivo model systems are with clinical and anatomic observations. These areas include the assembly of amyloid-like fibrils in vitro, the nucleation phenomenon, amyloid fibril structure in vivo and in vitro, common structural components of the amyloids, and the regression of tissue amyloid and proteolysis of amyloid proteins. Divergences and congruencies are highlighted, which in turn suggests caution in the interpretation of present data, greater collaboration and communication among investigators, and, additional areas and techniques for investigation.

Immunohistochemical identification and crossreactions of amyloid-A fibril protein in man and eleven other species

Antisera were prepared in rabbits, sheep or chicken against purified amyloid fibril protein AA from man, mouse, stone marten, dog, cow and hamster. These antisera were tested by immunodiffusion against all purified antigens and applied to tissue sections containing amyloid from man, mouse, hamster, guinea pig, rabbit, cat, dog, mink, stone marten, pine marten, cow and horse. The binding of the antibodies to amyloid in tissue sections was assessed by the indirect immunoperoxidase method. The strongest reactions in the immunodiffusion and immunohistochemical methods were found between amyloid deposits of members of a given species and an antibody raised against protein AA from the same species. In contrast to the lack of cross-reactivity in immunodiffusion (except in the mouse-man relationship), extensive cross-reactions were observed immunohistochemically in phylogenetically related species, e.g. between stone marten, pine marten and mink, or between hamster and mouse. However, cross-reactions were also observed in combinations such as man-mouse, man-dog, man-cat, mouse-horse, and dog-cow. In addition, individual antisera showed variations in immunohistochemical reactivity with amyloid deposits of different members of one given species. Moreover, antisera prepared in rabbits reacted more restrictedly than those prepared in sheep, while rabbit antisera against any AA-protein did not react with rabbit amyloid. Finally, the widest degree of cross-reactivity including almost all mammalian species investigated was observed with a chicken antiserum to human amyloid AA protein.

A second component in bovine AA amyloid fibrils not identical with protein AA is essential for AA amyloid fibrillogenesis

Amyloid fibrils were isolated from the renal papillae and glomeruli of cows with spontaneous AA amyloidosis. The fibrils were solubilized by treatment with guanidine hydrochloride (Gu HCl) and subjected to gel filtration on Sephacryl S-200. Two other fractions were obtained beside the void volume and the AA fractions. Reaggregation studies were performed by dialysing the fractions, separately or in combinations, against Gu-HCl-free solutions. Protein AA alone (about 10 kd) appeared not to precipitate. The other fractions alone and the combinations of fractions tested formed precipitates. The precipitates containing all fractions (including protein AA) or protein AA plus a fraction containing a 19- and a 23-kd protein revealed congophilic green birefringent fibrillar material. Dialysis against acidic and calcium-containing solutions gave the best results. Amyloid fibril-like material was visible on electron microscopic examination. The amino acid composition of the 19 + 23-kd material appeared to be slightly different from protein AA and evidently unlike SAP. On immunofluorescence-absorbance studies the 19 + 23-kd material appeared evidently unlike protein AA and SAP. From these findings it is concluded that for spontaneous formation of AA amyloid fibrils other non-AA proteins are necessary.

Partial amino acid sequence homology between an heredofamilial amyloid protein and human plasma prealbumin

Amyloid fibril protein has been isolated from the tissues of a patient of Swedish ancestry with autosomal dominant heredofamilial amyloidosis. After solubilization in guanidine HCl, a significant amount of the protein was contained in a homogeneous low molecular weight fraction. Molecular weight of approximately 14,000, amino acid analysis, double immunodiffusion analysis and immunoelectrophoresis all supported this material being a prealbumin-related protein. Automated sequence analysis gave a mixture of amino acids at each step, suggesting an heterogeneous NH2-terminus. After cleavage of the protein with cyanogen bromide, a homogeneous peptide was obtained with the sequence Val-Val-Val-Leu-Asp-Ala-Val-Arg-Gly-Thr-Pro- corresponding in 9 of the 11 positions analyzed with the known sequence of human prealbumin, starting with position 14. Antiserum raised to the amyloid protein reacted with normal human prealbumin. After absorption with normal human serum, this antiserum continued to detect a determinant in the amyloid patient's serum, suggesting an abnormal serum prealbumin, which may be the precursor of the fibril protein in this type of heredo-familial amyloidosis. Indirect immunohistochemical studies on kidney tissue from the patient with amyloidosis showed marked staining with anti-prealbumin and anti-heredofamilial amyloid protein, but not with anti-AA or anti-kappa antisera. No genetic association between this family and amyloidosis and Portuguese families with familial amyloid polyneuropathy is known.

Kinetics of serum amyloid protein A in casein-induced murine amyloidosis

Serum amyloid protein A (SAA), the precursor of secondary amyloid protein, is elevated in chronic diseases which are associated with an increased incidence of amyloid. However, SAA is also elevated in acute bacterial and viral infections and somes forms of cancer. The murine model of casein-induced amyloidosis was studied to determine the relationship between SAA production and amyloid deposition. SAA levels measured by radioimmunoassay were found to be as high as 200 times the normal level in CBA/J mice receiving daily parenteral casein. After a single injection of casein the SAA level was elevated by 3h and peaked by 12-18 h. Similar levels were found in casein-treated A/J mice, a strain less susceptible to the induction of amyloid. Parenterally administered bovine serum albumin, which has low potential for amyloid induction, gave SAA levels in CBA/J and A/J mice comparable to casein treatment. These data show that, while SAA levels are elevated during chronic antigenic stimulation, there are other factors involved in amyloid formation. These factors may include alterations in the degradation of SAA by the reticuloendothelial system caused by substances such as casein. Nude (athymic) mice were shown to attain high levels of SAA after receiving casein parenterally. Therefore, thymus-derived lymphocytes are not necessary for the synthesis of SAA.

Amyloid-related serum protein SAA in endotoxin-induced amyloidosis of the mink

The concentration of the amyloid AA-related serum protein (SAA) was markedly increased after endotoxin injections in mink, mouse, rabbit, and man. It was particularly studied during the development of endotoxin-induced amyloidosis of the mink. Protein SAA was markedly elevated in all mink 24 h after the first endotoxin injection but had fallen to relatively low levels before the onset of amyloidosis at about 4 wk. These results are consistent with SAA being a circulating precursor of the amyloid fibril protein, AA. However, both proteins may be derived by proteolysis from a common precursor.