Characterization of Amyloid Proteins AA and SAA as Apolipoproteins of High Density Lipoprotein (HDL)

An AA-like protein with a molecular weight of 8600 complexed to high-density lipoprotein (HDL) was demonstrated in several acute-phase sera with high levels of SAA. The protein 'apo AA' (to distinguish it from tissue AA) was isolated by elution from sodium dodecyl sulphate (SDS)-polyacrylamide gel, and showed antigenic identity with purified tissue protein AA in double immunodiffusion. Normal HDL was shown to bind purified tissue AA in vitro. When the in vitro-associated HDL-AA complexes were given intravenously to mice during induction of amyloidosis, human AA was incorporated in the amyloid fibrils. Both apo AI and apo AII were shown to displace SAA from acute phase HDL when added to HDL-SAA complexes in vitro. This might be of importance in amyloidogenesis, as the liver and the small intestine, which are the main sites for AI and AII synthesis, are also sites of early amyloid deposition.     

Serum amyloid A apolipoprotein (apo SAA). Implications in inflammation and in lipoprotein modifications

The measurement of serum amyloid A apolipoprotein (apo SAA) during acute phase inflammation offers a high interest because of its specificity, sensitivity and early increase of its levels, compared to other acute phase proteins. Furthermore apo SAA is transported in serum in association with lipoproteins, in particular with their denser subpopulation, HDL3 thus inducing their modification. The decrease in Lp AI:AII concentrations in inflammatory diseases is the consequence of the decrease in HDL3. In general the HDL3 composition was changed with a displacement of apo AI by SAA. Another interest to this protein is its relationship with amyloidosis. Apo SAA is the presumed precursor of amyloid A protein, which can be deposited in various tissues, leading to secondary amyloidosis.     

Is the serum amyloid A protein in acute phase plasma high density lipoprotein the precursor of AA amyloid fibrils?

Serum amyloid A protein (SAA), an apolipoprotein of high density lipoprotein (HDL), is generally considered to be the precursor of AA protein, which forms the fibrils in reactive systemic amyloidosis in man and animals. This view is based on amino acid sequence identity between AA and the amino-terminal portion of SAA. However, in extensive and well-controlled studies of experimentally induced murine AA amyloidosis, we were unable to demonstrate a direct precursor-product relationship between SAA, in SAA-rich HDL preparations from acute phase or amyloidotic mouse or human serum, and AA protein in the amyloid deposits. This raises the possibility that SAA in its usual form, as an apolipoprotein of HDL synthesized during the acute phase response, may not be the major precursor of AA fibrils. The amyloidogenic forms of circulating SAA molecules may not be isolated during the preparation of HDL. Alternatively, particularly in the light of recent evidence that SAA mRNA is expressed in many different tissues throughout the body of appropriately stimulated animals, amyloidogenic SAA may be derived from sources other than the liver cells in which SAA-rich HDL is synthesized.     

Purification and Characterization of Hamster Serum Amyloid A Protein (SAA) by Cholesteryl Hemisuccinate Affinity Chromatography

High-density lipoprotein (HDL) apolipoprotein was separated from hamster serum by cholesteryl hemisuccinate affinity chromatography (CHAC) in comparison with the density-gradient ultracentrifugation (DGUC). The apolipoprotein recovery from serum by CHAC was 70% and by DGUC 80%. This disadvantage is compensated for by the ease of purification by CHAC, a method particularly suited for the processing of large amounts of serum. From the acute-phase HDL CHAC fraction, apo SAA was isolated by gel filtration. Using isoelectrofocusing, two-dimensional gel electrophoresis, and titration curve, four isotypes of hamster apo SAA were identified and characterized. In the acute-phase serum, one of the isotypes was predominant (apo SAA1). In serum of amyloidotic animals, the relative contribution of apo SAA1 was considerably lower, suggesting selective removal of the latter during amyloidogenesis and possibly its deposition in hamster AA amyloid. Furthermore, the affinity chromatography method was modified with gradient elution of affinity-bound material. By this method HDL apolipoprotein was separated into three subclasses. Apo SAA was shown to associate with two different subclasses. In acute-phase serum most of the apo SAA1 was found in the subclass with the lowest affinity for the cholesteryl beads, whereas the latter was depressed in amyloidotic serum, suggesting that the amyloidogenicity of a particular apo SAA isotype is determined by its cholesteryl-binding properties.     

Degradation of Amyloid Proteins by Different Serine Proteases

A protein extract was obtained from normal human serum by adsorption to unsubstituted Sepharose 4B. This extract contained one or several enzymes with SAA and AA degrading capacity. The optimal pH for degradation of SAA was about 7.3. On fractionation of the enzyme extract on Sephadex G-160, the active component was eluted in the V0 peak. The V0 fraction, which on double immunodiffusion analysis was found to contain alpha 2-macroglobulin, was also active against synthetic substrates used to determine the activity of thrombin and plasma kallikrein. Gel filtration under dissociating conditions and molecular weight estimation further indicated the presence of those enzymes in the preparation. Several serine proteases which are known to be inhibited by alpha 2-macroglobulin possessed AA and SAA degrading activity. On degradation of SAA, an intermediate split product with molecular weight similar to AA was formed. Kallikrein, plasmin and elastase were also able to degrade intact amyloid fibrils suspended in phosphate-buffered saline.     

Transformation of Amyloid Precursor SAA to Protein AA and Incorporation in Amyloid Fibrils in Vivo

Experimental amyloidosis was induced in mice by intraperitoneal injections of endotoxin (lipopolysaccharide (LPS)). In addition to LPS, a group of mice received high-density lipoprotein (HDL)-SAA complexes isolated from human acute-phase serum, whereas a group of control mice received saline in addition to LPS. Isolated amyloid fibrils from the mice given HDL-SAA contained human AA protein, as shown by immunodiffusion, immunoblot, and enzyme-linked immunosorbent assay techniques, in addition to mouse AA. In contrast, amyloid from the control mice contained exclusively AA of mouse origin. Thus, the experiments provided solid evidence that SAA is the precursor for amyloid fibril protein AA.     

Metabolic disorders of lipoproteins--influences of compositional changes of lipoproteins upon their metabolic behavior

Compositional changes of apoproteins and lipids in lipoproteins influence their affinities for receptors and enzymes. Decrease of apo C proteins and increase of apo E in chylomicron and very low density lipoproteins (VLDL) during their catabolism might promote the binding to remnant receptor. On the other hand, the affinity for lipoprotein lipase (LPL) gradually decreases and that for hepatic lipase increases. However, the responsiveness of VLDL to LPL might be under the control of triglyceride (TG)/surface component ratios but not of the apoprotein ratios in ordinary circumstances judging from the results of the releases of fatty acids from VLDL by LPL in vitro. Responses of VLDL from diabetic patients to LPL significantly decreased compared with those from non-diabetic subjects. Glycation of VLDL in vitro impaired their responses to LPL. Therefore, delayed catabolism of VLDL in diabetes might partially depend upon glycation of VLDL besides the decreased LPL activity. Low density lipoproteins (LDL), apoproteins of which consist mostly of apo B protein and had a low TG level, showed a high affinity to the LDL receptor. However, LDL from hypertriglyceridemic subjects, in which the TG contents was increased, had a low affinity to the receptor. Since high density lipoproteins (HDL) from patients in acute phases contain a large amount of serum amyloid A protein (SAA), the percentages of apo A proteins markedly decreased. When SAA-rich HDL were incubated with leucocytes, SAA were degraded rapidly, although other apoproteins remained to be unchanged. Therefore, such HDL become unstable, and this might induce low HDL levels in the acute phase.     

Expression of mouse apolipoprotein SAA1.1 in CE/J mice: isoform-specific effects on amyloidogenesis

Amyloid A (AA) amyloid deposition in mice is dependent upon isoform-specific effects of the serum amyloid A (SAA) protein. In type A mice, SAA1.1 and SAA2.1 are the major apolipoprotein-SAA isoforms found on high-density lipoproteins. During inflammation, both isoforms are increased 1000-fold, but only SAA1.1 is selectively deposited into amyloid fibrils. Previous studies showed that the CE/J mouse strain is resistant to amyloid induction. This resistance is not due to a deficiency in SAA synthesis, but is probably related to the unusual SAA isoform present. The CE/J mouse has a single acute-phase SAA protein (SAA2.2), which is a composite of the SAA1.1 and SAA2.1, with an amino terminus similar to the nonamyloidogenic SAA2.1. Recently, genetic experiments suggested that the SAA2.2 isoform might provide protection from amyloid deposition. To determine the amyloidogenic potential of the CE/J mouse, we generated SAA adenoviral vectors to express the various isoforms in vitro and in vivo. Purified recombinant SAA proteins demonstrated that SAA1.1 was fibrillogenic in vitro, whereas SAA2.2 was unable to form fibrils. Incubation of increasing concentrations of the nonamyloidogenic SAA2.2 protein with the amyloidogenic SAA1.1 did not inhibit the fibrillogenic nature of SAA1.1, or alter its ability to form extensive fibrils. Injection of the mouse SAA1.1 or SAA2.2 adenoviral vectors into mice resulted in isoform-specific expression of the SAA proteins. Amyloid induction after viral expression of the SAA1.1 protein resulted in the deposition of amyloid fibrils in the CE/J mouse, whereas SAA2.2 expression had no effect. Similar expression of the SAA2.2 protein in C57BL/6 mice did not alter amyloid deposition. These data demonstrate that the failure of the CE/J mouse to deposit amyloid is due to the structural inability of the SAA2.2 to form amyloid fibrils. This mouse provides a unique system to test the amyloidogenic potential of altered SAA proteins and to determine the important structural features of the protein.     

Characterization of Bovine Amyloid Proteins SAA and AA

The bovine serum amyloid A (SAA) and tissue amyloid A (AA) proteins were isolated and characterized. SAA was isolated from acute phase high density lipoprotein (HDL) of a cow suffering from acute mastitis, and was identified by amino acid sequence analysis. No AA-like protein was found in complex with HDL in serum. Amyloid fibrils isolated from a bovine kidney contained a 9 kDa AA protein and a considerable amount of a 14 kDa protein. Amino acid sequence analysis showed that the largest protein probably represents undegraded SAA. This is an interesting observation which confirms previous works indicating that SAA can be incorporated in the amyloid fibrils without a prior degradation to AA. The partial amino acid sequences of bovine SAA and AA were strikingly homologous to the sequences of corresponding proteins in man and other species.     

No significant influence of serum amyloid A1 genotypes on serum lipids or HDL clearance

Serum amyloid A1(SAA1), the major acute phase isotype of SAA protein family, consists of three common allelic variants(SAA1.1, SAA1.3 and SAA1.5) in the Japanese population. We have recently reported that subjects with the SAA1.5 allele have higher plasma SAA concentrations than those without it, a phenomenon probably due to the delayed catabolism of the isotype SAA1.5. Since SAA is present in high density lipoprotein(HDL), this study assessed whether SAA genotype influenced the serum lipid study by altering HDL metabolism. In a total of 279 healthy adults, no difference was noted in their total cholesterol, HDL-cholesterol or triglyceride concentrations among six genotype groups. Plasma clearance of human apolipoprotein AI(apoAI) was studied in mice by giving HDL reconstituted with each recombinant human SAA1 isotype. The apoAI clearance did not differ among each of the SAA1 isotype-conjugated HDLs. Moreover, the changes in content of SAA in HDL also did not alter the apoAI clearance. These results suggest that SAA1 may not play an active role in plasma HDL metabolism.     

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.     

The Primary Structure of Rabbit Serum Amyloid A Protein Isolated from Acute Phase Serum

Serum amyloid A (SAA) protein, a sensitive acute phase protein and the precursor of protein AA in secondary amyloid, was purified from pooled acute phase rabbit serum using two different methods: isolation of protein SAA directly by octyl-Sepharose chromatography of total serum, and dissociation and isolation of apoSAA from acute phase high density lipoprotein (HDL). The protein SAA fraction obtained was further purified using gel filtration and ion exchange chromatography. Rabbit protein SAA has 104 amino acid residues, like human SAA, and has a partially blocked N terminus. The highly conserved region from position 33 to position 63 found in SAA from all species studied was confirmed also in rabbit SAA. No microheterogeneities were observed. The amino acid sequence showed extensive N-terminal homology with the rabbit amyloid A protein, except for the microheterogeneity in position 12 in protein AA. It also showed identical amino acid sequence with that deduced from the rabbit cDNA clone pSAA 55. Complete homologies were found with clone SAA 2, except for positions 22 and 78, clone SA8-1, except for positions 22 and 79 and clone SA7-3, except for position 22. This pSAA 55/SA7-3/SA8-l/SAA2-like protein was the only SAA isotype found both in total serum and in the HDL fraction. Isotypes corresponding to other SAA-like genes could not be found in this pool of acute phase rabbit sera.     

Apolipoprotein A-I-derived amyloid in atherosclerosis. Its association with plasma levels of apolipoprotein A-I and cholesterol

Wild-type apolipoprotein A-I (apo A-I)-derived amyloid commonly occurs in atherosclerotic plaques. To clarify apo A-I amyloid formation, plasma levels of apo A-I and cholesterol were related to the presence of amyloid in atherosclerotic plaques in 15 patients with peripheral atherosclerosis, subjected to arterial reconstruction. Plasma levels of apo A-I and high-density lipoprotein (HDL) cholesterol were slightly higher in patients with apo A-I-derived amyloid than in those without, but the difference was not significant. Levels of low-density lipoprotein cholesterol and total cholesterol were significantly higher in the group with amyloid. High concentrations of apo A-I in the arterial intima are probably of greater importance to amyloid formation than high plasma levels of the protein. During atherosclerosis, the acute phase reactant serum amyloid A may displace apo A-I from HDL, leading to increased concentration of lipid-free apo A-I in the intima and conformational changes of apo A-I, which make it more fibrillogenic. Some forms of amyloid fibrils have been shown to be cytotoxic. Apo A-I-derived amyloid is possibly a pathogenically important factor in atherosclerosis.     

A simple separative method for measuring serum amyloid A in high-density-lipoprotein and low-density-lipoprotein fractions using the phosphotungstic acid-Mg2+ precipitation procedure

We developed a simple separative method for measuring serum amyloid A (SAA) in both high-density-lipoprotein (HDL) and low-density-lipoprotein (LDL) fractions. It was devised using the SAA agglutination method and phosphotungstic acid-Mg2+ precipitation procedure for evaluating HDL-cholesterol (HDL-C). The new method is also able to detect amyloid A (AA) in each fraction with precision. The results of both the present method and the method using SAA agglutination and the dextran sulfate-Mg2+ precipitation procedure showed a strong correlation when used to measure the level of SAA in the LDL fraction of patients (r = 0.997; p < 0.0001). Reference intervals in normal healthy subjects (n=75) ranged from 0.5 to 4.7 microg/ml in the HDL fraction and from 0.1 to 1.9 microg/ml in the LDL fraction. SAA in the LDL fraction of subjects with hyperlipidemia was significantly higher than in normal subjects and subjects with normal lipidemia. SAA in the HDL fraction and total sera of subjects with hyperlipidemia was significantly higher than in normal subjects; however, it was not higher than in patients with normal lipidemia. The present methods for detecting SAA, especially in the LDL fraction, might benefit from analyzing patho-physiological events in various lipid disorders.     

The role of fibronectin in the development of experimental amyloidosis. Evidence of immunohistochemical codistribution and binding property with serum amyloid protein A.

Azocasein-induced amyloid A (AA) amyloidosis in CBA/K1Jms mice was investigated to elucidate a preference of serum amyloid A (SAA) deposition in the spleen. By indirect immunofluorescence using anti-SAA/AA antibodies the initial deposition of SAA/AA was recognized in the marginal zone of spleen at 20 days after azocasein injection. Indirect immunofluorescence using anti-fibronectin antibodies also showed meshwork positivity in the corresponding area more intensely than that in controls. Immunoelectron microscopy using anti-SAA/AA revealed the presence of positively stained flocculent materials on cell surfaces of macrophages in the marginal area in addition to amyloid fibril. The tissue fibronectin rapidly increased in the spleen and maintained 10 times more than that of controls until the 20th day. Binding assay of SAA on frozen sections revealed the presence of SAA-binding substances in the perifollicular area. Affinity chromatographic assay showed fibronectin have a binding capacity to SAA1 and SAA2. By binding assay on the microtiter plate, SAA had more affinity to fibronectin than those of heparan sulfate, collagen type I, or serum amyloid P component. These results indicate that fibronectin plays an important role in the development of amyloidosis by working as a linking protein between SAA and the cell surface of macrophages.     

Immunolocalization of serum amyloid A and AA amyloid in lysosomes in murine monocytoid cells: Confocal and immunogold electron microscopic studies

Murine AA amyloid (AA) protein represents the amino-terminal two-third portion of SAA₂, one of the isoforms of serum amyloid A. Whether plasma membrane-bound or lysosomal enzymes in activated murine monocytoid cells degrade SAA₂ to generate amyloidogenic AA-like peptides is not clearly understood, although AA has been localized in the lysosomes. Here we show, using confocal and immunogold microscopy (IEM), that both SAA and AA localize in lysosomes of activated monocytoid cells from amyloidotic mice. Rabbit anti-mouse AA IgG (RAA) and two monoclonal antibodies against murine lysosome-associated membrane proteins (LAMP-1 and LAMP-2) were used to immunolocalize SAA/AA and lysosomes, respectively. Confocal analysis co-localized both anti-RAA and anti-LAMP-1/LAMP-2 reactivities in the perikaryal organelles which by IEM proved to be electron-dense lysosomes. LAMP-1/LAMP-2-specific gold particles were also localized on lysosomal and perikaryal AA. The results suggest sequestration of SAA into the lysosomes. Since monocytoid cells are not known to phagocytose native amyloid fibrils, our results implicate lysosomes in AA formation.     

Humoral proinflammatory cytokine and SAA generation profiles and spatio-temporal relationship between SAA and lysosomal cathepsin B and D in murine splenic monocytoid cells during AA amyloidosis

Evidence shows that tissue macrophages (MPhis), in mice undergoing AA amyloidosis, endocytose acute-phase humoral serum amyloid A (SAA) and traffic it to lysosomes where it is degraded. Incomplete degradation of SAA leads to intracellular nascent AA fibril formation. In vitro, cathepsin (Cat) B is known to generate amyloidogenic SAA derivatives, whereas Cat D generates non-amyloidogenic SAA derivatives, and interferon (IFN-gamma)-treated MPhis show selective increase in Cat B concentration, a factor conducive to AA amyloidogenesis. To understand the cumulative effect of these factors in AA amyloidosis, humoral levels of SAA, IFN-gamma, tumour necrosis factor (TNF-alpha) and granulocyte-macrophage colony-stimulating factor were determined in azocasein (AZC)-treated CD-1 mice. We correlated these responses with the spatio-temporal distribution of SAA, Cat B- and Cat D-immunoreactive splenic reticuloendothelial (RE) cells. AZC-treated CD-1 mice similar to that of A/J mice showed partial amyloid resistance; their peak humoral IFN-gamma and SAA responses overlapped during the pre-amyloid phase. Unexpectedly, Cat D immunoreactivity (IR), instead of Cat B IR, was predominant in the splenic RE cells, indicating an apparent lack of causal relationship between IFN-gamma-mediated increase in Cat B expression. Partial amyloid resistance in CD-1 mice, probably a genetic trait, may be linked to high levels of Cat D expression, causing a delay in nascent AA fibril formation.     

Candidate genes involved in cardiovascular risk factors by a family-based association study on the island of Kosrae,Federated States of Micronesia

Altered plasma levels of lipids and lipoproteins, obesity, hypertension, and diabetes are major risk factors for atherosclerotic cardiovascular disease. To identify genes that affect these traits and disorders, we looked for association between markers in candidate genes (apolipoprotein AII (apo AII), apolipoprotein AI-CIII-AIV gene cluster (apo AI-CIII-AIV), apolipoprotein E (apo E), cholesteryl ester transfer protein (CETP), cholesterol 7alpha-hydroxylase (CYP7a), hepatic lipase (HL), and microsomal triglyceride transfer protein (MTP)) and known risk factors (triglycerides (Tg), total cholesterol (TC), apolipoprotein AI (apo AI), apolipoprotein AII (apo AII), apolipoprotein B (apo B), body mass index (BMI), blood pressure (BP), leptin, and fasting blood sugar (FBS) levels.) A total of 1,102 individuals from the Pacific island of Kosrae were genotyped for the following markers: Apo AII/MspI, Apo CIII/SstI, Apo AI/XmnI, Apo E/HhaI, CETP/TaqIB, CYP7a/BsaI, HL/DraI, and MTP/HhpI. After testing for population stratification, family-based association analysis was carried out. Novel associations found were: 1) the apo AII/MspI with apo AI and BP levels, 2) the CYP7a/BsaI with apo AI and BMI levels. We also confirmed the following associations: 1) the apo AII/MspI with Tg level; 2) the apo CIII/SstI with Tg, TC, and apo B levels; 3) the Apo E/HhaI E2, E3, and E4 alleles with TC, apo AI, and apo B levels; and 4) the CETP/TaqIB with apo AI level. We further confirmed the connection between the apo AII gene and Tg level by a nonparametric linkage analysis. We therefore conclude that many of these candidate genes may play a significant role in susceptibility to heart disease.     

Rabbit Serum Amyloid Protein A: Expression and Primary Structure deduced from cDNA Sequences

Serum amyloid A protein (SAA), the precursor of amyloid protein A (AA) in deposits of secondary amyloidosis, is an acute phase plasma apolipoprotein produced by hepatocytes. The primary structure of SAA demonstrates high interspecies homology. Several isoforms exist in individual species, probably with different amyloidogenic potential. The nucleotide sequences of two different rabbit serum amyloid A cDNA clones have been analysed, one (corresponding to SAA1) 569 base pairs (bp) long and the other (corresponding to SAA2) 513 bp long. Their deduced amino acid sequences differ at five amino acid positions, four of which are located in the NH2-terminal region of the protein. The deduced amino acid sequence of SAA2 corresponds to rabbit protein AA previously described except for one amino acid in position 22. Eighteen hours after turpentine stimulation, rabbit SAA mRNA is abundant in liver, while lower levels are present in spleen. None of the other extrahepatic organs studied showed any SAA mRNA expression. A third mRNA species (1.9 kb) hybridizing with a single-stranded RNA probe transcribed from the rabbit SAA cDNA, was identified. SAA1 and SAA2 mRNA were found in approximately equal amounts in turpentine-stimulated rabbit liver, but seem to be coordinately decreased after repeated inflammatory stimulation.     

Effect of acute inflammation on rat apolipoprotein mRNA levels

Hybridization studies using specific cDNA have been used to determine the mRNA levels for rat apolipoproteins AI, AII, AIV, and E in extracts of rat liver and intestine. The ratios of intestinal mRNA/liver mRNA for apolipoprotein AI (apo AI), apo AIV, and apo E were 1.3, 1.7, and 0.1, respectively. Apo AII mRNA was detected in the liver but not in the intestine. The mRNA levels for apo AII and apo AIV in rat liver decreased during inflammation to minimums of 40% and 25% of normal, respectively. The mRNA levels for apo AIV in the intestine, apo E in the liver and for apo AI in both the liver and intestine did not change significantly during inflammation. The time course for the decrease in the hepatic mRNA levels for apo AIV was similar to those previously observed for the negative acute-phase proteins albumin and transthyretin. The serum levels for apo AIV were not affected by inflammation.