Cathepsin D activity protects against development of type AA amyloid fibrils.

BACKGROUND: The extracellular, fibrillar deposits of reactive (secondary) amyloidosis are composed of amyloid A (AA) protein, a proteolytically derived fragment of the acute phase protein serum amyloid A (SAA). While complete degradation of SAA precludes amyloid formation, limited cleavage which generates AA protein is considered part of the pathogenic mechanism. MATERIALS AND METHODS: In this study, we investigated SAA degradation by lysosomal enzymes cathepsins B, D, and K, and assessed the impact of cathepsin activity on AA amyloid formation in a cell culture model using peripheral blood mononuclear cells from healthy volunteers. RESULTS: Lysates of human mononuclear cells were capable of degrading SAA. Degradation was significantly reduced by inhibition of cathepsin D with pepstatin A. Inhibition of cathepsin B or cathepsin K, however, had no effect. The SAA fragment pattern generated by mononuclear cell lysates was similar to that produced by incubating SAA with purified human cathepsin D. Consistent with in vitro findings, amyloid formation in human monocyte cultures was increased by 43% when cathepsin D was inhibited, but remained unaffected by inhibition of cathepsin B or cathepsin K. CONCLUSION: These data provide evidence that cathepsin D but not cathepsin B or cathepsin K is physiologically important in SAA degradation and hence in preventing SAA from accumulating and serving as precursor of AA amyloid fibrils.

Lovastatin inhibits formation of AA amyloid.

Amyloid A (AA) amyloidosis is a severe complication of many chronic inflammatory disorders, including the hereditary periodic fever syndromes. However, in one of these periodic fever syndromes, the hyper IgD and periodic fever syndrome, amyloidosis is rare despite vigorous, recurring inflammation. This hereditary syndrome is caused by mutations in the gene coding for mevalonate kinase, an enzyme of the isoprenoid pathway. In this study, we used a cell culture system with human monocytes to show that inhibition of the isoprenoid pathway inhibits amyloidogenesis. Inhibition of the isoprenoid pathway by lovastatin resulted in a dose-dependent reduction of amyloid formed [53% at 10 microM (P=0.01)] compared with mononuclear cells that are exposed only to serum AA. The inhibitory effects of lovastatin are reversible by addition of farnesol but not geranylgeraniol. Farnesyl transferase inhibition also inhibited amyloidogenesis. These results implicate that the isoprenoid metabolism could be a potential target for prevention and treatment of AA amyloidosis.

A new human hereditary amyloidosis: the result of a stop-codon mutation in the apolipoprotein AII gene.

Hereditary systemic amyloidosis may be caused by mutations in a number of plasma proteins including transthyretin, apolipoprotein AI, fibrinogen Aalpha-chain, lysozyme, and gelsolin. Each type of amyloidosis is inherited as an autosomal dominant disease and is associated with a structurally altered protein that aggregates to form amyloid fibrils. Here we report that the amyloid protein in a family with previously uncharacterized hereditary renal amyloidosis is apolipoprotein AII (apoAII) with a 21-residue peptide extension on the carboxyl terminus. Sequence analysis of the apoAII gene of affected individuals showed heterozygosity for a single base substitution in the apoAII stop codon. The mutation results in extension of translation to the next in-frame stop codon 60 nucleotides downstream and is predicted to give a 21-residue C-terminal extension of the apoAII protein identical to that found in the amyloid. This mutation produces a novel BstNI restriction site that can be used to identify individuals with this gene by restriction fragment length polymorphism analysis. This is the first report of apoAII amyloid in humans and the first mutation identified in apoAII protein. Amyloid fibril formation from apoAII suggests that this lipoprotein, which is predicted to have an amphipathic helical structure, must undergo a transition to a beta-pleated sheet by a mechanism shared by other lipoproteins that form amyloid.

Binding, trafficking and accumulation of serum amyloid A in peritoneal macrophages.

Murine serum amyloid A1.1 (SAA1.1) has been conjugated with the fluorophore Texas Red (TxR), and its interaction with peritoneal macrophages has been visualized by scanning confocal microscopy. Binding of TxR-SAA to cell surfaces was inhibited by an excess of unlabelled SAA indicating the involvement of saturable receptors. Internalized TxR-SAA was seen initially as small punctate signals which in some cells evolved into a fine fluorescent network, a pattern typical of tubular endosomes. Colocalization of TxR-SAA with Cy5-labelled low density lipoprotein (LDL) but not with Oregon Green-labelled transferrin suggested that SAA trafficked through endosomes and lysosomes for degradation rather than through recycling compartments. Consistent with this catabolic pathway, macrophages loaded with TxR-SAA lost fluorescence within several days after being shifted to a fluorophore-free medium. In sharp contrast to this, cells maintained under amyloid-forming conditions, i.e. in the presence of unlabelled SAA and amyloid-enhancing factor (AEF) before and after treatment with TxR-SAA, remained brightly fluorescent over the course of 5 days. Immunocytochemistry verified the accumulation of SAA within macrophages. These findings support the hypothesis that a decreased catabolism of internalized SAA plays a role in AA amyloid pathogenesis.

Expression of SAA and amyloidogenesis in congenic mice of CE/J and C57BL/6 strains.

Inbred strains of mice display different susceptibilities to experimental induction of reactive amyloidosis. CBA/J, C57BL/6, and ICR are among the most susceptible strains, while CE/J mice appear to be totally resistant. In contrast to amyloidogenic strains which express two major acute phase serum amyloid A proteins (SAA1 and SAA2), CE/J mice produce only a single isoform, designated SAA2.2. Studies indicate that CE/J x CBA/J hybrid mice expressing both SAA2.2 and SAA1.1/SAA2.1 are amyloid resistant, and this has led to the hypothesis that SAA2.2 may protect mice against amyloid formation even in the presence of fibrillogenic SAA1.1. We have tested this hypothesis in mice derived from CE/J and C57BL/6 strains. CE/J mice were mated with C57BL/6 mice to produce F1 hybrids. Congenic mice were then produced by backcrossing each successive generation to C57BL/6 mice. Representative mice from F2 and F3 generations were analyzed to determine SAA genotype and susceptibility to amyloid induction by repeated casein injections. All F2 and F3 mice examined, including those which carried the SAA2.2 gene, developed extensive splenic AA amyloid. Expression of SAA2.2 in mice testing positive for the SAA2.2 gene was confirmed by sequence analysis of HDL-associated SAA proteins. These results demonstrate that the presence of SAA2.2 is not sufficient to protect CE/J x C57BL/6 hybrid mice from amyloid development. This is consistent with our observation that macrophage cultures, derived from C57BL/6, CBA/J, or CE/J mice, undergo amyloid deposition when treated with SAA1.1 alone or with equal amounts of SAA1.1 and SAA2.2, but show no deposition when treated solely with SAA2.2. We conclude from these studies that while SAA2.2 is non-fibrillogenic, its physical presence is not sufficient for protection against amyloid formation.

Protein aging hypothesis of Alzheimer disease.

Alzheimer disease (AD), the most common form of aging-related neurodegenerative disorders, is associated with formation of fibrillar deposits of amyloid beta-protein (Abeta). While the direct involvement of Abeta in AD has been well documented, the relations between Abeta production, amyloid formation, and neurodegeneration remain unknown. We propose that AD is initiated by a protein aging-related structural transformation in soluble Abeta. We hypothesize that spontaneous chemical modification of aspartyl residues in Abeta to transient succinimide induces a non-native conformation in a fraction of soluble Abeta, rendering it amyloidogenic and neurotoxic. Conformationally altered Abeta is characterized by increased stability in solution and the presence of a non-native beta-turn that determines folding of Abeta in solution and the structure of Abeta subunits incorporated into amyloid fibrils. While the soluble 'non-native' Abeta is both the factor triggering the neurodegenerative cascade and the precursor of amyloid plaques, these two events result from interaction of Abeta with different sets of cellular components and need not coincide in space and time. Extensive literature data and experimental evidence are provided in support of this hypothesis.

A cell culture system for the study of amyloid pathogenesis. Amyloid formation by peritoneal macrophages cultured with recombinant serum amyloid A.

A murine macrophage culture system that is both easy to employ and amenable to manipulation has been developed to study the cellular processes involved in AA amyloid formation. Amyloid deposition, as identified by Congo red-positive, green birefringent material, is achieved by providing cultures with recombinant serum amyloid A2 (rSAA2), a defined, readily produced, and highly amyloidogenic protein. In contrast to fibril formation, which can occur in vitro with very high concentrations of SAA and low pH, amyloid deposition in culture is dependent on metabolically active macrophages maintained in neutral pH medium containing rSAA2 at a concentration typical of that seen in acute phase serum. Although amyloid-enhancing factor is not required, its addition to culture medium results in larger and more numerous amyloid deposits. Amyloid formation in culture is accompanied by C-terminal processing of SAA and the generation of an 8.5-kd fragment analogous to amyloid A protein produced in vivo. Consistent with the possibility that impaired catabolism of SAA plays a role in AA amyloid pathogenesis, treatment of macrophages with pepstatin, an aspartic protease inhibitor, results in increased amyloid deposition. Finally, the amyloidogenicity exhibited by SAA proteins in macrophage cultures parallels that seen in vivo, eg, SAA2 is highly amyloidogenic, whereas CE/J SAA is nonamyloidogenic. The macrophage culture model presented here offers a new approach to the study of AA amyloid pathogenesis.

Differential plasma clearance of murine acute-phase serum amyloid A proteins SAA1 and SAA2.

Serum amyloid A (SAA) proteins SAA1 and SAA2 are prominent acute-phase reactants which circulate in association with the high-density-lipoprotein (HDL) fraction of plasma. Plasma levels of SAA1 and SAA2 increase dramatically, by as much as 1000-fold, within 24 h of tissue injury and then rapidly decrease with cessation of the inflammatory stimulus, suggesting that SAA clearance and/or catabolism is important to the re-establishment of homoeostasis. In this context, aberrant SAA catabolism has long been considered a potential factor in the pathogenesis of reactive amyloidosis. To initiate studies aimed at understanding the differential regulation of SAA metabolism, we have produced 35S-labelled murine SAA1 and SAA2 in Escherichia coli, bound them individually to HDL, and then compared the plasma-clearance characteristics of SAA1 and SAA2 under normal and acute-phase conditions. When bound to normal HDL, SAA2 [half-life (t1/2) = 30 min] was cleared significantly faster than SAA1 (t1/2 = 75 min). Clearance of SAA1 and SAA2 was significantly slower when each was bound to acute-phase HDL as opposed to normal HDL, when clearance rates were determined in acute-phase mice versus normal mice, and when normal HDL was remodelled to contain both recombinant isotypes rather than just one of the isotypes. Thus it appears that an increased amount of SAA on HDL, or possibly the combined presence of both isotypes on HDL, is associated with a prolongation in the plasma half-life of SAA.

Accelerated amyloid deposition in mice treated with the aspartic protease inhibitor, pepstatin.

The development of amyloidotic diseases is believed to be determined in large part by the structure and metabolism of the amyloid subunit protein. The amino-terminal region of serum amyloid A (SAA), the subunit precursor protein in reactive amyloidosis, appears to confer fibrillogenic potential. Here we present data consistent with the hypothesis that amyloid A fibrillogenesis is favored when proteolysis of the amino-terminal region of SAA is impaired. Murine tissue extracts were found to contain pepstatin-inhibitable protease activity that cleaved mouse SAA2 between Glu8 and Ala9. Tissues obtained from mice that had been treated with pepstatin for 3 days lacked this activity. To investigate a possible relationship between inhibition of aspartic proteases and amyloidogenesis, mice were treated with pepstatin while concurrently undergoing a standard amyloid induction protocol (repeated casein injections). Pepstatin-treated mice showed amyloid deposition significantly sooner than the control group, which had received only casein. During the preamyloidotic phase, pepstatin-treated mice had higher concentrations of SAA in serum and spleen than control mice. In addition, clearance of injected 125I-labeled SAA from plasma was significantly delayed. Based on these findings, it is reasonable to postulate that inhibition of aspartic protease activity can lead to an accumulation of amino-terminally intact SAA molecules and thereby accelerate amyloid fibril formation.

Metabolism of amyloid proteins.

Metabolic processing of amyloid precursor proteins is an important factor in the genesis of practically all forms of amyloidosis. Of the three major forms of systemic amyloidosis, reactive amyloid (amyloid A protein; AA) formation shows the most consistent role of partial proteolysis of serum amyloid A (SAA) to AA proteins which form fibrils. Immunoglobulin amyloidosis is also usually associated with C-terminal degradation of the fibril precursor light chain protein. Although it is commonly thought that transthyretin amyloidosis is associated with fibril formation from the tetrameric circulating plasma transthyretin, chemical analyses of transthyretin fibril deposits show significant fragmentation of the fibril protein constituents. In addition, it has been documented that proteolytic fragments are the fibril subunit proteins in gelsolin, cystatin C. Alzheimer's beta-amyloid precursor protein and apolipoprotein AI (apoAI) amyloidoses. Notable exceptions to the role of proteolysis in amyloid fibril formation would appear to be the lysozyme and beta 2-microglobulin amyloidoses. Few studies have examined the metabolism of amyloid-forming proteins. Perhaps the best data are on apoAI, which show decreased plasma residence time for the amyloidogenic Gly26Arg apoAI (1.8 d vs. normal 4.5 d). Similarly, preliminary data show increased clearance of Val30Met transthyretin when compared with the wild-type protein (18 h vs. 26 h). Also, biosynthetically 35S-labelled SAA proteins reconstituted with HDL show increased plasma clearance of murine SAA2, the amyloid fibril subunit protein, when compared with murine SAA1. Few data are available on metabolism of amyloid immunoglobulin light chain proteins, but it has been shown that radiolabelled Bence-Jones proteins are cleared very rapidly from the circulation. A better understanding of the metabolism of precursor proteins in each of the amyloid deposition diseases will give insight into the mechanisms of fibril formation and pathogenesis of amyloidosis.

Genes encoding human serum amyloid A proteins SAA1 and SAA2 are located 18 kb apart in opposite transcriptional orientations.

The three expressed genes of the human serum amyloid A gene family (SAA1, SAA2 and SAA4) have been isolated in four contiguous clones selected from a chromosome 11 cosmid library. Analysis of these clones revealed that SAA1 and SAA2 are located 18 kb apart in opposite transcriptional orientations, while SAA4 lies 11 kb downstream from SAA2 in the same orientation: 3'(SAA1)5'-18 kb-5'(SAA2)3'-11 kb-5'(SAA4)3'. A fifth SAA clone isolated from this library was noncontiguous with the other four and contained the SAA3 pseudogene.

In vitro degradation of serum amyloid A by cathepsin D and other acid proteases: possible protection against amyloid fibril formation.

The effects of acid proteases on degradation of serum amyloid A protein (SAA) were investigated in vitro. Human recombinant SAA1 (rSAA1), when incubated with human spleen extracts at pH 3.2, was degraded in the amino-terminal portion of the molecule. This reaction was inhibited by an acid protease inhibitor, pepstatin. The degraded SAA molecules lacking nine or more amino-terminal residues, when exposed to in vitro fibril-forming conditions, failed to form Congo red positive precipitates and did not show amyloid fibril-like structure by electron microscopy. This suggests that the amino-terminal portion of SAA is essential for fibril formation. Cathepsin D, one of the lysosomal enzymes, also initiated degradation of rSAA1 at the amino-terminus. Cathepsin D immunoreactivity was detected in marginal areas of amyloid deposits in spleens from patients with reactive amyloidosis. These findings suggest that cathepsin D or similar acid proteases may be involved in SAA catabolism and may protect against amyloid formation.

Characterization of amyloid A protein in human secondary amyloidosis: the predominant deposition of serum amyloid A1.

Serum amyloid A protein (SAA) is the plasma precursor for amyloid A protein (AA), the subunit protein in amyloid deposits of secondary or reactive amyloidosis. Several forms of acute phase SAA have been identified in human plasma. To elucidate whether one of these forms of SAA predominates in the formation of AA amyloid deposits, the amino acid sequence of the subunit protein in six cases of reactive amyloidosis was investigated. Minimal heterogeneity was present at the N-terminus as all samples started with residue 1, 2, or 3 of SAA. The C-terminus, however, was more heterogeneous with the AA protein in each case terminating at multiple sites from residue 58 to 84 of SAA. Since less than 20% of the AA protein in each case contained sequence past residue 67 of SAA, the sequence and recovery of tryptic peptides containing residues 52, 57, and 60 where human SAA1 and 2 differ was used to determine the relative amounts of SAA1 and 2 present. One sample contained only SAA1 sequence, four contained approx. 11% or less of SAA2 sequence, and the sixth contained 24-33% of SAA2 sequence. Thus, while five of the six AA samples contained both SAA1 and 2, the predominant form in all cases was SAA1. In three of the six cases, the protein defensin was isolated along with the AA protein from the fibrils. This may suggest neutrophil involvement in SAA processing to AA fibrils.

Cathepsin B generates the most common form of amyloid A (76 residues) as a degradation product from serum amyloid A.

Amyloid A protein (AA), the chief constituent of reactive amyloid deposits, is derived from serum amyloid A (SAA) and most commonly corresponds to the amino-terminal 76 residues (AA76). Digestion of recombinant human SAA1 with a lysosomal thiol protease, cathepsin B, and analysis of the products by SDS-PAGE and amino-terminal sequencing revealed that AA76 was generated as a minor and transient degradation product. Digestion with neutrophil elastase generated intermediates different from AA76. This finding suggests that cathepsin B may play an important role in amyloid fibrilogenesis by converting SAA to AA.

Fibril formation from recombinant human serum amyloid A.

Three isotypes of human serum amyloid A (SAA), SAA1, SAA2 beta, and SAA4 were expressed at high levels in Escherichia coli (E. coli) using a pET vector expression system. Each SAA cDNA was ligated to the vector pET-21a(+) and transformed into E. coli, strain BL21(DE3)pLysS. Expression conditions required high concentrations of antibiotics in order to obtain a high ratio of synthesized SAA to total E. coli proteins. Each recombinant SAA (rSAA) was purified by molecular sieve chromatography followed by chromatofocusing or hydrophobic interaction chromatography. The yield of purified protein was 5-10 mg per 11 of culture. When subjected to in vitro fibril forming conditions, rSAA1 formed amyloid-like fibrils confirmed by Congo red staining and electron microscopy. In contrast, rSAA2 beta and rSAA4 showed negative Congo red staining and curvilinear or flattened fibrillar structures on electron microscopy. This suggests that SAA1 has greater potential for forming amyloid fibrils than either SAA2 beta or SAA4.

Recombinant murine serum amyloid A from baculovirus-infected insect cells: purification and characterization.

Serum amyloid A (SAA) is an extremely sensitive acute-phase reactant and precursor to the subunit protein in reactive amyloid deposits. Although the mouse has long served as an informative experimental model, both the function of SAA and the pathogenic mechanism of amyloid formation remain unknown. The production of SAA by a heterologous system was pursued as means of generating readily-renewable amounts of SAA of defined sequence. Murine SAA2 has been expressed in and purified from baculovirus-infected insect cells. Using the transfer vector pBlueBac, SAA2 cDNA was cloned into baculovirus DNA such that expression was under the control of the polyhedrin promoter. Lysates prepared from infected cells contained three amyloid A-immunoreactive forms which accumulated intracellularly over a three day periods. The form having the lowest relative molecular mass, 12.5 kDa, co-migrated in SDS-polyacrylamide gels with the SAA2 present in murine acute-phase serum. Recombinant SAA2 was purified by Sepharose CL-6B chromatography followed by chromatofocusing between pH 8 and pH 5. Amino-terminal sequencing of the purified 12.5 kDa sample confirmed the first 20 residues of mature murine SAA2. After incubation with normal mouse serum, purified recombinant SAA2 fractionated exclusively with lipoprotein complexes, suggesting that it was bound to HDL. Based on this observation, we believe that recombinant SAA can serve as a suitable substitute for the native protein in physiologically relevant studies.