TWO TYPES OF LAMBDA POLYPEPTIDE CHAINS INHUMAN IMMUNOGLOBULINS BASED ON AN AMINO ACID

SUBSTITUTION AT POSITION 190

BY ETTORE APPELLA AND DANIEL EIN

LABORATORY OF BIOLOGY, AND IMMUNOLOGY BRANCH, NATIONAL CANCER INSTITUTE,

NATIONAL INSTITUTES OF HEALTH

Communicated by C. B. Anfinsen, March 30, 1967

Studies in this and in other laboratories have been concerned with the chemicalcomposition of immunoglobulin light polypeptide chains and with the genetic sig-nificance of the chemical evidence. Many recent investigations have been devotedto the comparison of amino acid sequences of Bence-Jones proteins of manl-3 andmouse.4-6 These polypeptide chains are about 214 residues in length and consistof variable and invariable halves of about equal length. In man, as in mouse, theinvariant portion comprises the 107 amino acids of the COOH-terminal region.There are two classes of light polypeptide chains of man, designated kappa andlambda.Human lambda Bence-Jones proteins share no antigenic determinants with kappa

polypeptide chains and there are no tryptic peptides in common.7 Amino acid se-quence analysis, however, has shown a large degree of homology (about 44%)8between kappa and lambda chains. Lambda Bence-Jones proteins share ten com-mon peptides. To date, these ten peptides appeared identical in all lambda Bence-Jones proteins.Two antigenic forms of lambda polypeptide chains, tentatively designated Oz

(+) and Oz (-), have been distinguished by a rabbit antiserum produced to alambda-type Bence-Jones protein.9 The present investigations were undertaken todefine a structural basis for the antigenic specificity. This communication reportsthe existence among lambda-type Bence-Jones proteins of a single amino acid sub-stitution in the carboxyl terminal half of the molecule at position 190. A good cor-relation exists between the immunological subtype and the structural variation asdemonstrated by peptide analysis.

Materials and Methods.-Bence-Jones proteins were initially isolated from urine by 80% am-monium sulphate precipitation. Proteins Ru and Ni were reprecipitated twice with 50% am-monium sulphate. Other proteins were further purified by DEAE-chromatography (phosphatebuffer pH 8.0, gradient elution), Sephadex G-100 gel filtration (0.2 M Tris-HCl buffer pH 8.0plus 0.2 M NaCl), Geon-Pevikon block electrophoresis (veronal buffer pH 8.6 A 0.6), or anycombination of these as was necessary. Immunoelectrophoresis and Ouchterlony analysis at con-centrations exceeding 15 mg/ml showed that all preparations were homogeneous by these criteria.Polyacrylamide gel electrophoresis (pH 8.8)10 of the Ru and Ni Bence-Jones proteins showed a

single component in each case.Reduction of the proteins was accomplished by incubation in 10 M urea and 0.01 M Dithio-

threitol under N2 for 3 or 4 hr at 37°C. After reduction, iodoacetic acid was slowly added to afinal concentration of 0.022 M maintaining a pH of 8.2 by addition of 0.2 N NaOH. The reac-tion was terminated after 15 min by the addition of sufficient 0-mercaptoethanol to bring the finalconcentration to 0.025 M. The carboxymethylated product was dialyzed against water andlyophilized.

Performic-acid oxidation of the enzyme was performed according to the method of Moore"but without the use of HBr as a reducing agent.Trypsin (2 X crystallized, lot 6241, Worthington) was treated with -(1-tosylamido-2 phenyl)

1449

1450 BIOCHEMISTRY: APPELLA AND EIN PROC. N. A. S.

ethyl chloromethyl ketone to eliminate residual chymotryptic activity.12 Digestion was (carriedout with an enzyme-substrate ratio of 1:100 in 0.5% ammonium bicarbonate at 370C for about12 hr after which samples were frozen and lyophilized. The lyophilized digest was dissolved illdilute ammonia water and applied to a Whatman 1o1. 3 filter paper inl a concentration of 1 mg/cm.

Electrophoretic separation was performed in the first dimension under Varsol with the follow-ing conditions: (a) pyridine-acetic acid buffer at pH 6.5 at 3000 volts for 60 min; (b) pyridine-acetic acid buffer at pH 3.6 at 2800 volts for 45 min; and (c) formic-acetic acid buffer at pH 1.9at 2800 volts for 45 min. Descending chromatography in the second dimension for a period of18 hr was performed with butanol-acetic acid-water (4:1:5).

Peptides were detected on a paper strip with a ninhydrin spray reagent" and the starch-iodinereagent for peptide bonds.'4 Arginine-containing peptides were located by the use of phenan-threnequinone reagent,"5 tyrosine- or histidine-containing peptides with the Pauly"6 reagent.Peptides were eluted from paper with 0.06 N ammonia or hydrochloric acid at room temperature ina closed chamber. Samples of suitable sizewere hydrolyzed with constant boiling HCl (at 1100Cfor 24 hr) and were analyzed in a Spinco model 644 B amino acid analyzer which was equipped foraccelerated sensitive analysis.'7

Digestion with chymotrypsin was carried out with an enzyme-substrate ratio of 1:50 in 0.8%ammonium bicarbonate for 12 hr, after which the digest was frozen and lyophilized. Electro-phoretic and chromatographic separation of the peptides were performed as described above.Digestion with carboxypeptidases A and B (Worthington, lot 1630, and lot 61 B) was carriedout with 5 ,g of enzymes in 0.2 ml dilute bicarbonate buffer for 3-5 hr at 370C. The reactionwas terminated by addition of 0.5 ml of Na-citrate buffer pH 2.2 and the peptide sample appliedto the amino acid analyzer.

Amino-terminal residues of tryptic and chymotryptic peptides were identified by the Dansylmethod of Gray and Hartley.'8 Dansyl-chloride (1-dimethyl-amino-naphthalene-5-siilphonylchloride, DNS-Cl) was coupled to the samples at 37CC, and the product was hydrolyzed withconstant boiling HCl. The resulting a-DNS amino acids were identified by high-voltage electro-phoresis at pH 4.55 except for DNS-serine, DNS-proline, and DNS-alanine, and these aminoacids may be separated by electrophoresis under Varsol in formic-acetic acid buffer at pH 1.9.

Results.-Studies were done with 12 lambda Bence-Jones proteins of which sevenwere Oz (+) by immunochemical tests and five were Oz (-). Two proteins, Ni(Oz +) and Ru (Oz -) were initially investigated in detail.Tryptic digests of the Ni and Ru proteins were compared after chromatography

in butanol-acetic acid-water and electrophoresis in pyridine-acetic acid-water atpH 3.6. In each digest, ten common peptides were resolved, as previously de-fined.7 Nine of these were apparently identical in the two digests; the only sig-nificant difference in the ten common spots being peptide A15. Peptide A15 movedfarther toward the cathode in the Ni than in the Ru digest, and tests with the phe-nanthrenequinone reagent revealed that A15 from Ni was arginine negative whileA15 from Ru was arginine positive. This difference is well shown in Figure 1,which is a 1:1 mixture of tryptic digests from the Ni and Ru proteins.

Peptide A15 was purified in larger quantity by high-voltage electrophoresis ofthe tryptic digest at pH 6.5 followed by electrophoresis at pH 3.6. Measurementof the amino acid composition (Table 1) showed that the A15 peptide varied in asingle amino acid, Ni having lysine and Ru having arginine. By dansyl analysesfor N-terminal residues and the known specificity of trypsin, the sequence of aminoacids of the peptide A15 is determined to be:

Ser-His-Lys (Ni)Ser-His-Arg (Ru)

Chymotryptic digestion of both Ni and Ru proteins released a basic peptide

VOL. 57, 1967 BIOCHEMISTRY: APPELLA AND EIN 1451

Of~~~~~~~~~~ ....A

'A~~~~~FIG. 1.-Peptide map of mixed tryptic digest of Ni and Ru lambda Bence-Jones proteins. See

text for details.

(shown to be residues 187 through 192, Fig. 2) w~hich stained for histidine in bothproteins but for arginine only in the Ru protein. The N-terminal residue of thisbasic chymotryptic peptide was identical for both proteins. The sequence of aminoacids in this basic chymotryptic peptide from the Ni protein was determined by thedansyl method, by carboxypeptidase-A and -B digestion and by tryptic digestion.As shown in Table 2 and Figure 2 the sequence of amino acids of the basic chymo-tryptic peptide from Ni as compared to that of Ru and other proteins19 is:

Lys-Ser-His-Lys-Ser-Tyr (Ni)Lys(Ser,His,Arg,Ser,Tyr) (Ru)

The amino acid composition and sequence of a number of tryptic and chymotrypicpeptides of two lambda Bence-Jones proteins has been reported by Milstein"and Putnam.8 The peptide Al15 has now been shown to be linked in sequence to theC-terminal portion of the protein. These studies further indicate that the arginineof peptide A15 (as in Ru in this study) occupies position 190. There has been noprior report, however, of lysine in position 190, as found here for Ni.

1452 BIOCHEMISTRY: APPELLA AND EIN PROC. N. A. S.

TABLE 1AMINO ACID COMPOSITION OF PEPTIDES FROM TRYPTIC DIGEST OF CARBOXYMETHYLATED

LAMBDA BENCFJONES PROTEINS NI AND RuPeptides*

Amino - Al-, 1,lTzA1M, A4 -A3-acid Ni Ru Ni Ru Ni Ru Ni Ru Ni Ru

Lys ... ... 1.0 1.0 1.0 ... 1.0 1.0 1.0 1.0H~is ... ... 1.0 1.1 1.0 1.2 ... ... ... ...

Arg .. .. .. ... ... 1 .0 ... ... ... ..CMC 0.8 0.8 0.9 0.9 ... ... ... ... ...

Asp .. .. ..... ... ... ... ... 1.3 1.3Thr 1.9 1.9 2.0 1.8 ... ... 3.0 2.8 ...

Ser 1.1 1.0 2.6 2.6 0.8 1.2 1.2 1.1 2.2 2.2Glu 1.1 1.0 3.0 3.0 ... ... 1.0 1.1Pro 0.8 1.3 ... ... ... ... 1.2 1.2 1.2 1.1Gly .. ... 1.0 1.1 ... ... 1.1 1.1 .. ..Ala 1 .0 1.1 ... ... ... ... 1.1 1.1 1.1 1.2Val 0.8 1.0 2.2 1.9 ... ... 0.9 0.9 0.8 1.0Tyr ... 1.0 0.8 ... ... ... ... ...Yield(%) 56 20 14 5 12 29 30 28 33 35NH2-ter-minalresiduet Ser Ser Ser Ser Thr Thr Ala Ala Ala Ala* The nomenclature follows that of Putnam7 except for peptide T2 which was named by Milstein.19 The

figures given are the molar ratios of each amino acid relative to lysine or arginine.t The NH2-terminal amino acid was determined by the Dansyl method (see text for details).

The amino acid compositions and N-terminal residues of four additional commonpeptides were also investigated and found to be the same for the two proteins (Niand Ru) (Table 1). The results of analyses of these four other common peptidesagreed with those previously reported.7. 8

A systematic study of ten additional lambda-type Bence-Jones proteins wasundertaken. Some of these proteins were carboxymethylated with iodoacetic acidand digested with trypsin while others were digested in the native state. The tryp-

190 195(Oz)Lys

X Chains LysSerHis -SerTyr Ser Cys Glx Val Thr HisArg

ValK Chains GlutLys His Lys- *Tyr Ala Cys GluuValThr His-

Leu(Inv)

200 205 210X Chains Glx Gly ---- ---- Ser Thr Val Glu Lys Thr Val AlaProThr

K Chains Gln Gly Leu Ser Ser Pro Val Thr Lys Ser Phe Asn Arg Gly

X Chains Glu Cys Ser (carboxyl end)

K Chains Glu Cys (carboxyl end)

FIG. 2.-Comparison of the partial amino acid sequences of human kappa and lambda Bence-Jones proteins. Data for the lambda Bence-Jones proteins derived from Milstein"' and Titaniet al.8 The kappa chain sequence is derived from Hilschmann and Craig,' Titani et al.,2 and Mil-stein.3

VOL. 57, 1967 BIOCHEMISTRY: APPELLA AND EIN 1453

TABLE 2AMINO ACID COMPOSITION AND SEQUENCE OF CHYMOTRYPTIC PEPTIDE FROM DIGEST OF

HUMAN LAMBDA CHAIN BENCE-JONES NIAmino Acid Composition of

Method Recovered Material Amino Acid Sequence DeducedAcid hydrolysis Lys His Ser Tyr

2.0 0.8 1.8 0.8CarboxypeptidaseA* 0 0 0.8 1.0 ( ) Ser Tyr

Carboxypepti-dasesAandB* 0.8 0.3 0.8 1.0 ( ) His Lys Ser Tyr

Dansyl-Edmandegradation ++ + Lys Ser His Lys Ser Tyr* Molar recovery of liberated amino acids.

tic digest was then separated by electrophoresis at pH 6.5 and peptide A15 fur-ther purified by electrophoresis at pH 3.6. Six of these proteins had lysine in pep-tide A15, four had arginine (Table 3). The Bence-Jones proteins were also classi-fied for the Oz antigenic specificity (Table 3). In each case the presence of lysinein peptide A15 was associated with the Oz (+) antigen and the presence of argininewith an Oz (-) reaction.Discussion.-Two forms of lambda Bence-Jones proteins can be defined, based

on amino acid variations at position 190. A single substitution-lysine for ar-ginine-has been demonstrated, which correlates with serological subgroups oflambda polypeptide chains tentatively designated Oz (+) and Oz (-). The Ozantigen is detected by a precipitin reaction and has been found in IgG myelomaproteins, in light chains from pooled normal IgG, and in whole normal IgG mole-cules.9 Seven Oz (+) Bence-Jones proteins have lysine at position 190, whilefive Oz (-) proteins have arginine.These findings complement previous observations with kappa polypeptide chains.

Kappa chains are divisible into two groups based on a single substitution of valinefor leucine at position 191. This substitution corresponds with the Inv (a) (leu-cine) and Inv (b) (valine) specificities which represent genetic polymorphism ofkappa polypeptide chains." 3, 21 Inv factors are found only on kappa type Bence-Jones proteins and on kappa light chains of IgG, IgM, and IgA molecules. TheInV factors are generally considered to be controlled by alleles at a single locus, soit may be that single amino acid substitutions are responsible for the serologicalspecificity of allelic variants of immunoglobulins.Evidence from studies with hemoglobin variants indicates that single amino

substitutions can give rise to antigenic differences.20 They can be detected by anti-

TABLE 3Oz TYPES AND A15 AMINO ACID COMPOSITION

Amino AcidsProtein Oz Lys Arg His SerOz + 1.0 ... 0.9 0.7Re + 1.0 ... 0.8 1.0Ge + 1.0 ... 0.9 1.0Cof + 1.0 ... 1.1 0.7O'N + 1.0 ... 0.9 0.6Go + 1.0 ... 0.9 0.9Ri - ... 1.0 1.1 1.1Cor ... 1.0 1.0 1.1BJB ... 1.0 0.7 1.0Ca ... 1.0 0.9 1.0

1454 BIOCHEMISTRY: APPELLA AND EIN PROC. N. A. S.

a-chain and anti-fl-chain sera using complement-fixation tests. These antisera arecapable of recognizing mutational differences involving a single amino acid substi-tution. A substitution which causes a charge difference could alter the three-di-mensional structure of a protein and thereby its antigenic characteristics.

It is also possible that the Oz and InV serological specificities are associated withbut not caused by the single amino acid substitutions described for the invariantregion. Amino acid changes in the variable region may be coordinate with in-variant region changes thus accounting for antigenic differences. The relationshipof these serological specificities to the existence of multiple substitutions presentin the variable portion remains to be elucidated. It may be that conformationalalterations induced by variable-region amino acid changes have little effect on theantigenic structure of the invariant region. A structural barrier between the twoportions of the chain may exist, so that any change in the variable region would notaffect the invariant region and thereby impair its recognition of the heavy chain.Not enough data about the amino acid sequences of kappa and lambda light

polypeptide chains are available to test these hypotheses. Studies are in progress todetermine whether the Oz types of lambda polypeptide chains, which appearanalogous to the Inv factors of kappa polypeptide chains, do indeed represent he-reditable polymorphism.Sumnmary.-Two forms of lambda Bence-Jones proteins, differing in a single

amino acid in the invariant region, have been demonstrated. These polypeptidechains occur in two antigenic types, Oz (+) and Oz (-). Of 12 proteins studied, theseven Oz (+) proteins have lysine at position 190 and five Oz (-) proteins have ar-ginine at the same position. This amino acid interchange resembles the inter-change at position 191 of kappa polypeptide chains which is associated with theallotypic Inv factors.

The authors wish to thank Dr. John L. Fahey for his encouragement and helpful advice.1 Hilschmann, N., and L. C. Craig, these PROCEEDINGS, 53, 1403 (1965).2 Titani, K., E. Whitley, Jr., L. Avogardo, and F. W. Putnam, Science, 149, 1090 (1965).Milstein, C., Nature, 209, 370 (1966).

4 Hood, L. E., W. R. Gray, and W. J. Dreyer, these PROCEEDINGS, 55, 826 (1966).6 Perham, R., E. Appella, and M. Potter, Science, 154, 391 (1966).6 Gray, W. R., W. J. Dreyer, and L. E. Hood, Science, 155, 465 (1967).7 Putnam, F. W., and C. W. Easley, J. Biol. Chem., 240, 1626 (1965).8 Titani, K., M. Wikler, and F. W. Putnam, Science, 155, 828 (1967).9 Ein, D., and J. L. Fahey, in press.10 Davis, B. J., Proc. N.Y. Acad. Sci., 121, 404 (1964).11 Moore, S., J. Biol. Chem., 238, 235 (1963).12 Koska, V., and F. A. Carpenter, J. Biol. Chem., 239, 1799 (1964).13 Levy, A. L., and D. Chung, Anal. Chem., 25, 396 (1963).14 Rydon, H. N., and P. W. G. Smith, Nature, 169, 922 (1952).15 Yamada, S., and H. A. Itano, Biochim. Biophys. Acta, 130, 540 (1966).16 Block, R. J., E. L. Durrum, and G. Zweig, A Manual of Paper Chromatography and paper

Electrophoresis (New York: Academic Press, Inc. 1958).17 Hubbard, R. W., Biochim. Biophys. Res. Commun., 19, 679 (1965).18 Gray, W. R., and B. S. Hartley, Biochem. J., 89, 379 (1963).29 Milstein, C., J. Mol. Biol., 21, 203 (1966).20 Reichlin, M., E. Bucci, C. Fronticelli, J. Wyman, E. Antonini, C. Ioppolo, and A. Rossi-

Fanelli, J. Mol. Biol., 17, 8 (1966).21 Baglioni, C., L. Alescio-Zonta, D. Cioli, and A. Carbonara, Science, 152, 1517 (1966).