Amino Acid Sequence of an Amyloid Fibril Protein of Unknown Origin

TIIE JOURNAL OF BIOLOGICAL CHEMISTRY Vol. 247, No. 17,Issue of September 10,~~. 5653-5655, 1972 Printed

Communication

Amino

Acid

Fibril

Protein

Sequence

of an Amyloid

of Unknown

Origin

(Received for publication, DANIEL JUDITH

EIN,* SHIGERU KIMURA,~ MAGNOTTA,* AND GEORGE

May 10, 1972)

WILLIAM D. G. GLENNER~

TERRY,*

From the Immunology Brunch, National Cancer Institute,* an? the Laboratory of Experimental Pathology, National Institute of Arthritis and Metabolic Diseases,1 ATational Institutes of Health, Bethesda, Maryland 20014

SUMMARY

Recent investigations of the nature of amyloid fibril proteins have shown that many of these proteins derive from immunoglobulin light polypeptide chains. Immunochemical analyses have demonstrated antigenic cross-reactivity between human amyloid fibril proteins and either human K or X Bence-Jones proteills, i.e. light polypeptide chains (1). NHt-terminal amino acid sequence analysis of two amyloid proteins has shown them to have homogeneous sequences which are highly homologous to the NH*-terminal sequence of human K Bence-Jones proteins (2). Finally, certain Hence-Jones proteins can be enzymatically digested in vi!ro under conditions comparable to those found in viva to form fibrils bearing the typical tinctorial and ultrastructural features of native amyloidfibrils (3). These findings further confirm the close relation between amyloid fibril proteins and immunoglobulill proteins, primarily the light polypeptide chains. Having established that some amyloid proteins are identical with light polypeptide chains, it was reasonable to ask whether all amyloid fibril proteins are immunoglobulin-associated (3, 4). One amyloid fibril protein which demonstrated properties usually associated with amylojd, such as fibrillar structure with ppleated sheet conformation (5), had a highly unusual amino acid composition lacking valine, proline, threonine, and half-cystine from it,s total of over 40 amino acids (6). No comparable length of any known immunoglobulin polypeptide chain lacks all of these amino acids. This suggests that the Amyloid IV protein might be derived from a previously unrecognized immunoglobulin or from an immunoglobulin polypeptide chain containing an internal deletion. Alternatively, the Amyloid IV protein might have an origin other than from immunoglobulins. Consistent with this possibility was the finding that the sequence of the first 15 amino 5653

U.S.A.

acids of this protein was different from any immunoglobulin sequence reported in the literature (7). Identical NHt-termina1 sequences have been obtained from a variety of amyloid fibril In order to define more clearly proteins of human (8,9) 0rigin.l the nature of this class of “non-immunoglobulin” amyloid fibril proteins, the complete sequence of Amyloid IV has been determined and shown to be nonhomologous with the sequences of known immunoglobulin polypeptide chains. MATERIALS

AND

METHODS

Amyloid IV protein was purified initially by column chromatography in 5 M guanidine hydrochloride and 1 w acetic acid as previously described (10). This protein was then rechromatographed on a column (1.5 x 90 cm) using P-30 (Bio-Rad Laboratories, Richmond, Calif.) equilibrated withO. N ammonium bicarbonate, pH 8.7, and the major protein fractions were pooled and lyophilized. The protein was purified to homogeneity as assessed by sodium dodecyl sulfate-polyacrylamide disc gelelectrophoresis at pH 7.1. The amino acid composition and peptide maps of this protein have been reported (10) as has the sequence of the NHzterminal 15 amino acids determined with the Beckman model 890 automatic amino acid Sequencer (7). For the present study, automatic amino acid sequencing was repeated using the same procedures with 10 mg of protein. Tryptic (TPCK2-treated) and chymotryptic digests (Worthington) of Amyloid IV protein were prepared using 1 yc solutions of protein in 0.1 x ammonium bicarbonate, pH 8.7, at an enzyme to substrate ratio of 1:lOO. Digestions were performed at 37” for 5 hours and stopped by freezing and lyophilization of the digests. Peptide maps were prepared by the method of Katz et al. (11). These maps were sprayed with a dilute ninhydrin spray and heated in an 80” oven for up to 10 min. Resulting ninhydrinAmino acid positive spots were cut out and eluted with water. compositions of peptides were performed by hydrolyzing portions of peptides for 24 hours at 110” and examining the hydrolysates by high voltage electrophoresis on paper (12). Appropriate peptides were hydrolyzed under similar conditions and analyzed on the Beckman model 121 amino acid analyzer. Tryptophan-containing peptides were detected by Ehrlich’s reagent (10). Identical peptides from up to four maps were then pooled for amino acid sequence determination. Amino acid sequencing of peptides was performed by a modification of the dansyl method (13). Identification of dansyl-amino acid derivatives was accomplished by thin layer chromatography on polyamide sheets. The position occupied by tryptophan was assumed to be that position in the tryptophancontaining tryptic peptides for which no dansyl-amino acid could be found. For the corresponding chymotryptic peptide tryptophan was assigned to the COOH-terminal residue because of the specificity of the enzyme and because all other positions were 1 G. G. Glenner, R. Anders, R. Guyer, and W. D. Terry, manuscript in Drenaration. ;T