Current Concepts of the Formation and Com position of Amyloid

Transcription

1 ANNALS OF CLIN IC A L AND LABORATORY SCIEN CE, Vol. 5, No , Institute for Clinical Science Current Concepts of the Formation and Com position of Amyloid GEORGE G. GLENNER, M.D. Laboratory of Experimental Pathology, National Institute of Arthritis and Metabolic Diseases, National Institutes of Health, Bethesda,M D ABSTRACT Immunochemical and chemical studies of purified amyloid fibril protein from a wide variety of human tissues reveals that amyloid fibrils may derive from immunoglobulin proteins most frequently in the primary disease and can be classified as of kappa- or lambda-type. In addition a protein of unknown origin is the source of amyloid fibrils in another group of cases usually of secondary type. The creation of fibrils with all the characteristics of amyloid fibrils has been accomplished by proteolysis of some Bence Jones proteins. This indicates that amyloid fibrils may be formed by proteolytic digestion of circulating light polypeptide chains of immunoglobulin proteins. Introduction Amyloidosis is a disease that has attracted the attention and investigative efforts of medical researchers for well over a century. Controversy as to its pathogenesis and chemical composition date from the coinage by Virchow of the term amyloid to designate the starch-like quality of iodine staining of the deposits40 and the findings shortly thereafter of the proteinacious character of these deposits by Friedreich and Kekule.10 The subsequent literature on the nature of amyloid has been abundant and conflicting.23 Distinguishing Properties of Amyloid Deposits Certain properties of the amyloid substance have been sufficiently characteristic so as to distinguish these deposits from those found in other disease processes: (1) Congo red staining and green polarization birefrin- 257 gence,25 (2) chemical evidence of tryptophan as a constant constituent,4,11,15 (3) the electron microscopic definition of a fibrillar component,1,2,36 (4) relative resistance of this fibrillar component to proteolytic digestion,35 (5) crystallographic evidence of the presence of a /3-pleated sheet confirmation in amyloid fibril concentrates5 and (6) insolubility of the fibrillar component in the usual aqueous solvents.11,20 Evidence that the fibrillar component represents the predominant distinguishing feature of the amyloid substance can be derived from the aforementioned properties. In addition, there is evidence that the most generally accepted characterizing feature of the deposits, the Congo red polarization birefringence, is dependent upon the /3-pleated sheet conformation of the fibrils.12,39 Although the /3-pleated sheet conformation has been found as a partial characteristic of an increasingly wider number of naturally occuring proteins, the /3-structure

2 258 GLENNER as the major characteristic of a mammalian protein, as seen in amyloid fibrils, has not been previously described. The fact that this conformation is not an artifact of preparation was shown by examination of native, undessicated preparations,5 a procedure frequently used in crystallographic analysis.9 Amyloid Fibril Concentrates Early attempts to purify the fibrillar material were by the use of differential centrifugation in which a top layer concentrate of amyloid fibrils was obtained after saline homogenization and low speed centrifugation. Modifying this method of fibril concentation, Pras et al30 obtained, from a top layer preparation, a suspension of fibrils in distilled water and were able to produce a concentrated fibril precipitate from this water soluble preparaton by salt addition. The latter procedure was modified by us and we obtained from the aqueous fibril suspension on ultracentrifugation at 100,000 x g for one hour a pellet containing a concentrate of amyloid fibrils.11 In an occasional case it was found that the amyloid fibrils were concentrated not in the top layer, but in a lower fraction of the pellet of the initial saline homogenate.20 Fibril Protein Purification The object of our study was to purify the major protein of the highly concentrated preparations of amyloid fibrils from the tissue of different patients in which amyloid deposits constituted the major component of the tissue bulk as estimated by Congo red polarization birefringence. Column chromatography by gel filtration of amyloid fibrils of aqueous neutral or basic (ph 11) extracts of amyloid deposits indicated by the elution of Congo red birefringent fibrils in the excluded (void) volume that they could not be fractionated and purified using these aqueous solvents.20 Our early work indicated, however, that solubilization of the majority of the amyloid fibril concentrate could be accomplished in the presence of 6 M guanidine-hcl11 and that the further addition of a reducing agent, dithiothreitol, invariably permitted almost a complete solubilization of all the protein constituents of amyloid fibril concentrates from all tissue sources investigated. It was not possible to reproduce with other systems (including 8 M urea or 1 percent sodium dodecyl sulfate [SDS]) the solvent capacity found with 6 M guanidine. Denaturation of the amyloid concentrate had been shown to cause a loss of both Congo red birefringence and fibril morphology and a distortion of the characteristic x-ray diffraction pattern of the native fibril. It was considered reasonable, therefore, to isolate the major protein from the concentrate of amyloid fibrils with the assumption that this major protein was derived solely from the fibrillar component. Utilizing this solubilization procedure and gel filtration column chromatography, it was possible to obtain the major protein of the amyloid fibril concentrate from tissues of different patients in good yield.20 It was found that gel filtration on Sepharose 4 B permitted fractionation of a small proportion of high molecular weight protein and permitted the isolation of a lower molecular weight major peak containing the majority (over 50 percent) of the protein of the concentrate. Subsequent gel filtration of this major fraction on Sephadex G-100 permitted further purification of the major protein component from minor lower molecular weight materials.20 At each step in the fractionation procedure the isolated protein was monitored by SDS disc gel electrophoresis in conjunction with tryptophan staining of the gel. Analysis of Major Amyloid Fibril Protein Chemical and physical analyses of the major proteins obtained from 10 tissues of six different patients were performed. The significant findings from these analytical

3 AMYLOID FORMATION AND COMPOSITION 259 studies of the major fibril protein indicated that amyloid fibril proteins purified from different tissues of the same patient were identical both in amino acid, molecular weight, amino-terminal group analysis and on peptide mapping However, the amyloid fibril protein from different patients was shown to differ markedly. Nonetheless, certain similar underlying characteristics of these proteins were evident, i.e. a high dicarboxylic and short chain amino acid content and the invariable presence of tryptophan. Further data indicated the presence only of unreactive or aspartic aminoterminal groups, molecular weights ranging from 5,000 to 18,300 and the presence of two or more major proteins in several preparations.20 An exceptional protein was amyloid IV, which was found to have an arginine amino-terminus and four absent amino acids including half cystine. Antibodies were obtained against the purified amyloid protein from one patient and these antibodies after absorption with normal human serum, liver and spleen were found to react with a component in the serum obtained from the same patient, but failed to react with the serum of other amyloidotic patients.15 Based on the aforementioned data, it was concluded that the majority of amyloid fibrils studied probably represented the tissue deposition of a segment of circulating immunoglobulin protein. The molecular weight ranges found by us indicated that in most cases, the fibrils were composed of the amino-terminal variable fragment of either the light or heavy polypeptide chain.15,20 Proof of Immunoglobulin Origin of Amyloid Fibril Protein Our findings were consistent with the evidence presented by Osserman and associates26,27,28 during the past ten years that in most cases a relationship existed between amyloid deposits and immunoglobulin light chains. Since it was postulated by us that most amyloid fibrils represented primarily the amino-terminal variable fragment of an immunoglobulin protein, it was decided to test this postulate directly and conclusively. The two proteins having aspartic amino-terminal amino acids lent themselves to automatic sequencing using the Beckman 890 amino acid sequencer. The results of the initial sequencing14 and further sequence analysis to position 35 and 3616 showed the variable region of these proteins were homologous with Ker, a kappa light chain of V. Since one of these patients was clinically defined as having primary or idiopathic amyloidosis and the other was known to have amyloidosis secondary to epidermolysis bullosa and subsequent pulmonary tuberculosis, it was apparent that amyloid fibrils in both a primary and secondary (infectious) case could originate from immunoglobulin proteins. Since the molecular weights of these proteins were 7,500 and 18,300, respectively, they represented primarily the variable portion of a light chain protein. Antibodies obtained against the aminoterminal unreactive amyloid proteins were tested against highly purified Bence Jones proteins in immunodiffusion.22 The unabsorbed amyloid antibodies reacted with a limited number of lambda Bence Jones proteins. For example, the antibody to amyloid VI reacted with seven of 23 lambda Bence Jones proteins and, when tested with amyloid VI, the amyloid protein was shown to have antigenic determinants in common with this group of lambda polypeptide light chains. These studies showed that, as in the case of Bence J ones proteins, these amyloid fibril proteins could be classified as of kappaor lambda-type. Creation of Amyloid Fibrils from Bence Jones Proteins If, indeed, amyloid proteins are derived primarily from the variable fragment of immunoglobulin light chain, it should be possible to obtain from the variable segment of a light chain (Bence Jones protein) a fibril having the characteristics of amyloid fibrils.

4 2 6 0 G L E N N E R Using the technique described by Solomon, et al33 to obtain the variable fragment of a light chain on peptic digestion at ph 3.5 at 37 C, a precipitate with some but not all Bence Jones proteins was obtained by us following a 5 hour incubation. Proof that this precipitate represented the aminoterminal variable fragment of the Bence Jones protein was afforded by molecular weight studies, peptide mapping (which indicated absence of major Constant region peptides) and sequence analysis (which showed the fibril protein to be homologous with the variable region of the parent lambda Bence Jones protein). This precipitate had all the features that characterize amyloid deposits including, (1) green polarization birefringence after Congo red staining, (2) fibrils similar in many respects to those of amyloid fibrils and (3) a /8-pleated sheet pattern indistinguishable from that of amyloid fibrils.13 It is apparent, therefore, that regardless of clinical classification, i.e. primary or secondary, amyloid fibrils may consist of immunoglobulin protein and that they can be created from Bence Jones proteins. The origin of amyloid fibrils associated with overt multiple myeloma or plasma cell dyscrasia has also been established. The amyloid fibril and urinary Bence Jones protein of one such patient are identical by limited amino acid sequence, carboxyl-terminal analysis and peptide map criteria.38 This demonstrates that at least in some patients with amyloidosis, plasma cell dyscrasia and Bence Jones proteinuria, the amyloid fibril is derived from the homogeneous light chain. In a further case of a patient with amyloidosis without overt Bence Jones proteinuria or the presence of a distinguishable M-component in his serum, an intact light chain with antigenic determinants in common with the isolated amyloid fibril protein of the patient has been isolated from the serum.17 No detectable whole myeloma protein having the same idiotypic specificity as the amyloid fibril protein could be demonstrated in either the serum or urine of the patient. These findings show that (1) under certain circumstances both whole light chains and their variable region fragments may become amyloid fibrils and (2) in some cases of amyloidosis unassociated with detectable M-components in the serum or Bence Jones proteinuria, the amyloid fibril can be derived from all or a portion of a homogeneous light polypeptide chain of an immunoglobulin protein. Further chemical analyses has shown that not all amyloid fibrils are derived from immunoglobulin proteins. Our definition of amyloid is, after all, based on certain derivative properties. It is a material deposited in tissues which exhibits Congo red polarization birefringence, has a fibrillar appearance by electron microscopy and a characteristic x-ray diffraction pattern indicating a /3- pleated sheet conformation.5 Under appropriate conditions, proteins other than light chains could assume these same properties. An anti-parallel chain /3-pleated sheet conformation has been found by infrared studies in the Fd fragment of immunoglobulin heavy chains.39 Studies of amino acid polymers show that poly-l-lysine in its (3- conformation9 has many of the properties of amyloid fibrils12 except for some differences in the spacing of the polypeptide chains of the /3-sheet. Partial amino acid sequence determinations of amyloid fibril protein IV from a patient with classic secondary amyloidosis7 and an identical amino acid sequence of the fibril protein from a patient with classic primary disease17 do not correspond, as of this writing, to that of any sequenced protein. Thus, some cases of amyloidosis may result from tissue deposition of portions of immunoglobulins other than light chains or of proteins other than immunoglobulins. Pathogenesis of Amyloidosis The fact that amyloid fibrils could be created by us from some Bence Jones proteins at a physiologic temperature in the

5 AMYLOID FORMATION AND COMPOSITION 261 presence of a proteolytic enzyme having an acidic ph optimum13 suggests that one possible pathogenetic mechanism for amyloid formation may be by means of intralysosomal catheptic digestion6 of light polypeptide chains of immunoglobulins. This mechanism is supported by the frequent close spatial relationship between amyloid deposits and cells of the macrophage system32,37 and by the electron microscopic observations of fibrils within plasmalemmal invaginations and membrane-bound vesicles of macrophages.19,34 However, not all Bence Jones proteins when digested by proteolytic enzymes form amyloid fibrils. Therefore, one must consider the possibility that there exist amyloidogenic amino acid sequences. Following proteolysis, light polypeptide chains having these sequences may form aggregates of what are primarily variable region fragments. This concept is supported by the greater association of amyloidosis and multiple myeloma in those cases with lambda Bence Jones proteinuria,29 the greater proportion of lambda- to kappa-type amyloid fibril proteins reported by us,17 and the evidence that following proteolysis primarily lambda Bence Jones proteins are degraded to form amyloid fibrils.13 Since the Bence Jones proteins that were digested to create amyloid fibrils were derived from patients without known amyloidosis, it is possible that proteolysis per se is not a necessary and sufficient condition for fibril formation and that in vivo factors of as yet unknown types may play a role in this disease process. Of considerable interest is the homogeneity of the portion of immunoglobulin protein deposited in different tissues of the same patient as amyloid fibrils. Both the type and homogeneity of these proteins may be dependent upon a selection process occurring at the level of either immunoglobulin-synthesizing cell, immunoglobulin-degrading phagocytic cells or both. In addition to previous mechanism for the formation of amyloid fibrils, one must also consider the possibility that a physical change affecting circulating immunoglobulin proteins could also produce fibrils from a whole light polypeptide chain in the absence of proteolysis. That physical treatment can produce this effect has been demonstrated by infrared and x-ray diffraction studies with some but not all light chains studies.39 Osserman and coworkers28 have provided evidence to indicate that amyloid-related Bence Jones proteins have a greater affinity to normal tissues than do Bence Jones proteins unrelated to amyloidosis. This would imply a chemical characteristic of amyloid-related Bence J ones proteins residing in the variable region of the light chain. Amyloid fibril deposition is probably dependent upon an equilibrium in which the rates of amyloid fibril formation and chronicity of the stimulus affect the rate of removal of the fibrils by continuous proteolytic action. The evidence that, following removal of an antigenic stimulus, renal glomerular deposits are more resistant to mobilization than hepatic and splenic deposits24 tends to support this concept. Thus differences in tissue localization may indicate changes in the equilibrium caused by removal of the stimulus for immunoglobulin synthesis of an amyloidogenic light chain or an increase in degradation of amyloid fibrils owing to other as yet unknown factors. Distinctions between primary and secondary amyloidosis based on relative differences in tissue localization may be explained in part by such a mechanism. Of primary importance, however, is the finding of immunoglobulin fragments in primary as well as secondary amyloidosis. This emphasizes the importance of ruling out chronic inflammatory disease and investigating each case for the possible presence of exogenous or endogenous3 antigenic stimuli. Such stimuli are known to affect the levels of free light chains,8,18 and their removal would, therefore, increase the rate of mobilization of amyloid deposits of immunoglobulin origin.

6 262 GLENNER References 1. CAESAR, R.: Elektronenmikroscopische Befunde am Amyloid. Nova Acta Leopoldina 31:87-97, COHEN, A. S. and CALKINS, E.: Electron microscopic observations on a fibrous component in amyloid of diverse origins. Nature 183: , COIMBRA, A. and ANDRADE, C.: Familial amyloid polyneuropathy: an electron microscope study of the peripheral nerve in five cases. I. Interstitial changes. Brain , COOPER, J. H.: An evaluation of current methods for the diagnostic histochemistry of amyloid. J. Clin. Path. 22: , EANES, E. D. and GLENNER, G. G.: X-ray diffraction studies on amyloid filaments. J. Histochem. Cytochem. 16: , EHRENREICH, B. A. and COHN, Z. A.: The uptake and digestion of iodinated human serum albumin by macrophages in vitro. J. Exp. Med. 126: , EIN, D., KIMURA, S., and GLENNER, G. G.: An amyloid fibril protein of unknown origin: partial amino-acid sequence analysis. Biochem. Biophys. Res. Commun. 46: , EPSTEIN, W. V. and TAN, M.: Increase of L-chain proteins in the sera of patients with systemic lupus erythematosus and the synovial fluids of patients with peripheral rheumatoid arthritis. Arthritis Rheum. 9: , FASMAN, G. D.: Factors responsible for conformation stability. Poly-a-Amino Acids, Fasman, G. D., ed. Marcel Dekker, New York, p. 499, FRIEDREICH, N. and KEKULE, A.: Zur amyloidfrage. Virchow Arch. Path. Anat. 16:50-75, GLENNER, G. G., CUATRECASAS, P., ISERSKY, C., BLADEN, H. A., and EANES, E. D.: Physical and chemical properties of amyloid fibers. II. Isolation of a unique protein constituting the major component from human splenic amyloid fibril concentrates. J. Histochem. Cytochem. 17: , GLENNER, G. G., EANES, E. D., and PAGE, D. L.: The relation of the properties of Congo redstained amyloid fibrils to the /3-conformation. J. Histochem. Cytochem. 20: GLENNER, G. G., EIN, D., EANES, E. D., BLADEN, H. A., TERRY, W., and PAGE, D.: The creation of amyloid fibrils from Bence Jones proteins in uiiro. Science 174: , GLENNER, G. G., HARBAUGH, J., OHMS, J. I., H ARAD A, M., and CUATRECASAS, P.: An amyloid protein: the amino-terminal variable fragment of an immunoglobulin light chain. Biochem. Biophys. Res. Commun. 41: , GLENNER, G. G., HARADA, M., ISERSKY, C., CUATRECASAS, P., PAGE, D and KEISER, H. R.: Human amyloid protein: diversity and uniformity. Biochem. Biophys. Res. Commun. 41: , GLENNER, G. G., TERRY, W., HARADA, M., ISERSKY, C., and PAGE, D.: Amyloid fibril proteins: proof of homology with immunoglobulin light chains by sequence analysis. Science 171: , GLENNER, G. G., TERRY, W. D., and ISERSKY, C.: Amyloidosis: Its Nature and Pathogenesis. Seminars Hemat. 10:65-86, GORDON, D. A., EISEN, A. Z., and VAUGHAN, J. H.: Studies on urinary 7 -globulins in patients with rheumatoid arthritis. Arthritis Rheum. 9: , GUEFT, B. and GHIDONI, J. J.: The site of formation and ultrastructure of amyloid. Amer. J. Path. 43: , HARADA, M., ISERSKY, C., CUATRECASAS, P., PAGE, D BLADEN, H. A., EANES, E. D., KEISER, H. R., and GLENNER, G. G.: Human amyloid protein: chemical variability and homogeneity. J. Histochem. Cytochem. 19:1-15, HEEFNER, W. A. and SORENSON, G. D.: Experimental amyloidosis. I. Light and electron microscopic observations of spleen and lymph nodes. Lab. Invest. 11: , ISERSKY, C., EIN, D., PAGE, D. L., HARADA, M., and GLENNER, G. G.: Immunochemical crossreactions of human amyloid proteins with immunoglobulin light polypeptide chains. J. Immunol., in press. 23. LETTERER, E.: History and development of amyloid research. Amyloidosis. Mandema, E., Ruinen, L., Schölten, J. H. and Cohen, A. S., eds. Excerpta Medica, Amsterdam, pp. 3-9, LOW ENSTEIN, J. and GALLO, G.: Remission of the nephrotic syndrome in renal amyloidosis. New Eng. J. Med. 282: , MISSMAHL, H. P. and HARTWIG, M.: Polarizationoptische Untersuchungen an der Amyloidsubstanz. Virchow Arch. Path. Anat. 324: , OSSERMAN, E. F.: The plasmacytic dyscrasias. Amer. J. Med. 31: , OSSERMAN, E. F., TALAL, N and TAKATSUKI, K.: Amyloidosis: tissue proteinosis: gammaloidosis. Ann. Intern. Med. 55: , OSSERMAN, E. F., TAKATSUKI, K., and TALAL, N.: The pathogenesis of amyloidosis. Seminars Hemat. 1:3-86, PICK, A. I. and OSSERMAN, E. F.: Amyloidosis associated with plasma cell dyscrasias. Amyloidosis. Mandema, E., Ruinen, L., Schölten, J. H. and Cohen, A. S., eds. Excerpta Medica, Amsterdam, pp , PRAS, M., ZUCKER-FRANKLIN, D., RIMON, A., and FRANKLIN, E. C.: Physical, chemical and ultrastructural studies of watersoluble human amyloid fibrils. J. Exp. Med. 130: , REIMANN, H. A. and EKLUND, C. M.: Longcontinued vaccine therapy as a cause of amyloidosis. Amer. J. Med. Sei. 190:88-92, SMETANA, H.: The relation of the reticulo-endothelial system to the formation of amyloid. J. Exp. Med. 45: , SOLOMON, A. and McLAUGHLIN, C. L.: Bence Jones proteins and light chains of immunoglobulins.

7 AMYLOID FORMATION AND COMPOSITION I. Formation and characterization of aminoterminal (variant) and carboxyl-terminal (constant) halves. J. Biol. Chem. 244: , SORENSON, G. D. and BARI, W. A.: Murine amyloid deposits and cellular relationships. Amyloidosis. Mandema, E., Ruinen, L., Scholten, J. H. and Cohen, A. S., eds. Excerpta Medica, Amsterdam, pp , SORENSON, G. D. and BININGTON, B.: Resistance of murine amyloid fibrils to proteolytic enzymes. Fed. Proc , SPIRO, D.: The structural basis of proteinuria in man. Electron microscopic studies of renal biopsy specimens from patients with lipid nephrosis, amyloidosis, and subacute and chronic glomerulonephritis. Amer. J. Path. 35:47-73, TEILUM, G..- Periodic acid-schiff-positive reticuloendothelial cells producing glycoprotein. Functional significance during formation of amyloid. Amer. J. Pathol. 32: , TERRY, W. D., PAGE, D. L., KIMMURA, S., ISOBE, T., OSSERMAN, E. F., and GLENNER, G. G.: Structural identity of Bence Jones and amyloid fibril proteins in a patient with plasma cell dyscrasia and amyloidosis. J. Clin. Invest. 52: , TERMINE, J. D., EANES, E. D., EIN, D and GLENNER, G. G.: Infrared spectroscopy of human amyloid fibrils and immunoglobulin proteins. Biopolymers J J : , VIRCHOW, R.: Cellular Pathology. J. B. Lippincott and Co., Philadelphia, p. 409, 1863.