Human myeloma light chains with increased molecular weight: High frequency among λ chains

0161-589o~s3;o4n397-i I $03.00!0 Q 1983 Pergamon Press Ltd.

,~~olr~ulrrr lnmu,~hg~ Vol. 20, No. 4, pp. 397-407, 1983 Printed in Gretit Britain.

HUMAN MYELOMA LIGHT CHAINS WITH MOLECULAR WEIGHT: HIGH FREQUENCY CHAINS*

INCREASED AMONG 2

JEA~-~~~~RE BOUVET, JACQUES PILLOT and PANAYOTIS LIACQPQU~OS Laboratoire d’immunologie Microbienne. Institut Pasteur, Paris, France: Institut d’Immuno-biologie, H&pita1 Broussais, Paris, France.

and

Abstract-The discovery of a human myeloma protein comprising a K L-chain with an increased mol. wt of 30,000) (Bouvet et ul.. 1980) prompted investigations on the incidence of such heavier L-chains among other human myeloma proteins. In LOS samples examined, 34 were found to have L-chains heavier than normal (23,00~24,~0~, ranging from 25,000 up to 31,OM), and five of lighter mol. wt (21,OOt?-22,000). These mol. wt abnormalities were detected by electrophoresis in sodium dodecyl sulfate lo’.;, polyacrylamide gels (SDS-PAGE) after reduction with 2-mercaptoethanol. The mol. wt of three of the heavier ti or i. chains was also estimated by filtration through a Sephadex GlOO column and by sedimentation equilibrium. All three methods indicated a mol. wt increase of about 1%2%; as compared with the usual mol. wt. The distribution of the high mol. wt chains among all L-chains examined was found to be 1i out of 62 K chains (17.7”,,) and 23 out of 43 2 chains (53”,;) (P < 0.001). A preferential association of such L-chains with H-chains producing multiple bands in SDS-PAGE (P K 0.01) and an association between multiple L-chain and multiple H-chain band (P < 0.05) were also observed. In contrast, no abnormal L-chain was found in immunoglobulins from normal subjects. Spontaneous degradation of the normal H-chains sometimes yielded fragments of 30,~Omol. wt. These fragments were easily distingtlishable from abnormal L-chains. The nature of extra mol. wt in heavy L-chains was investigated for the presence of carbohydrate moiety. Four large and three normal size L-chains were examined for amino-sugar and sialic acid content. A small amount (one residue per molecule) of amino-sugar was detected only in two normal and two heavy L-chains. whereas sialic acid was only found in the heaviest {27,0~30,~~ L-chains (Lh) and in small percentage (one or two residues per molecule). Total sugar estimation in one Lh chain indicated a proportion not exceeding three or four residues per L-chain (mol. wt 1,000) and this is insufficient to explain the 1%25%) (3,600-6,000) mol. wt increase. It is therefore possible that, at least in some heavy myeloma L-chains, an additional peptide is expressed. Whatever the nature of the increase it would be of interest to elucidate whether this is a marker of malignant process or of an intermediate step of normal lg synthesis.

The mol. wt of immunoglobulin L-chains has been estimated to be 23,000 (Small & Lamm, 1966; Filitt~-Wurmser et al., 1970; Dorrington & Rochey, 1970). Except in rare cases with defective L-chains (Lewis et al., 1968; Isobe & Osserman, 1974) human myeloma proteins were found to have structurally normal L-chains, although Virella & Coelho (1974) described an apparent increase in mol. wt estimation of myeloma i chains, when compared * This work was supported by research grants from the Institut National de la Sante et de la Recherche Medicale (INSERM) ATP No. 597891 and CRL No. 821004. f Abbreviations used in this paper: BJ. Bence-Jones protems: Kh, kappa heavy; Km. kappa medium: Kn, kappa normal; ~1. kappa light: Ih, lambda heavy; lm, lambda medium; in, lambda normal; II, lambda light; 2-ME, 2-mercaptoethanol; SDS-PAGE, Sodium dodecyi sulfatepolyacrylamide gel electrophoresis. 397

with K chains by SDS-PAGE? technique. In a previous study {Bouvet et al., 1980, 1983), we reported a Ich chain of 30,000, isolated from a myeloma patient (DA) having three M-components (IgG, Kh, IgG h-n, IgA An) simultaneously present in the serum. Subsequently, Spira et al. (1980) and Preud’homme et al. (1980) also reported human myeloma L-chains with increased mol. wt. More recently, Savvidou et at. (1981a, 1981b) studied another case in which mol. wt increase was explained by an abnormal glycosylation of the L-chain. In the present study, we analysed 105 human M-components to determine the relative frequency of the increased apparent mol. wt of L-chains in SDS-PAGE, and we confirmed by other methods the increased mol. wt of DAk-h and two other Lh chains. Sugar content was found to be very low, with less than four residues per L-chain.

41ATERIALS

AND

RIETHOI~S

Set-a or urine from 105 myeloma patients were selected for the presence of M-components in amounts able to be purified in a onestep procedure. IgG myelotnas were therefore Favored (94 cases), but some Bencc-Jones (fout cases), IgA (two cases), IgG + IgA (two cases) myelotnas. as well as IgM macroglobulinemi~~s (three cases). were also included. Benign gatnmapathies and the majority of IgA myelotnas were not used since they needed large amounts of serutn and multi-step purification procedures. Ten set-a frotn normal patients and commercial Cohn fraction II were used as controls. Pwjjicatiorz

of IgG

r117rlB.1

pmtc3irl.s

IgG and BJ proteins were purified from serum and urine respectively by using DE 52 ion-exchange chromatography in tube tech nique. Samples (1 ml) were dial& against 0.015 !\/I phosphate buffer pH 8. and mixed twice with two volutnes of DE 53 equilibrated in the same buffer. Tubes were rotated one ht at room temp. and centrifuged. The supernatant contained only IgG or BJ proteins. other serum proteins being rctaincd by the DE 53.

IgM was prepared by a method based on the euglobulinic properties of many tnonoclonal IgM (Bouvet et N/. manuscript in preparation).

IgA was purified by the method of Fine bi Steinbuch ( 1970). using successive ammonium sulfate and caproic acid precipitations. followed by DE 52 ion-exchange chromatography.

The apparent mol. wts were estimated by the method of Weber & Osborn (1969). with reduced protein mol. wt markers (Pharmacia) and normal human IgG. After reduction with I-ME. all the samples were dissolved in urea-SDS buffer. heated 30 min at 60’ C, and run in IO?,, slab gels prepared according to the system of Laemmli (1970). According to their apparent mol. wt the L-chains were tertned as Ln (normal), Lh (high). Lm (medium) and LI (light).

Purification of IgG from patient\ iZm. A\ and BE were performed b) ammonium ~~11liltc precipitation followed b! ion-exchange chromatogt-aph! on DE-52 column equilibrated in 0.00 .21 phosphate huffcr. Proteins were eluted step-u& using diffcrcnt molaritics of phosphate bull‘er and eluates \vcre analysod. after reduction. by SI)S PA<;E. IgG + IgA paraprotein~ pre5ctit simnltaneouslq in DA and Di palicnt\ wcrc prcparcd as alread) dcscribcd for the DA patient (Bottvet c’l ~71..19X0). I-‘urthcr puritication of the t\\n DA IgG was obtained b!; several passapcs OII DE-53 columns equilibrated in 0.0015 !21 phosphate buffer. and eluted w;ith 0.005 .2f and 0.01 !2,1of the same bulfer (Boujct c’l trl.. 19S2).

H- and L-chains Deere separutcd using difl’crent techniques depending on the purpose of the esperimcnt. For mol. \vt determination, proteins were reduced with Z-ME and the chain\ sepal-atcd on a Sephadex G 100 column equtltbrated in l”1’ acetic acid (F1cischman (11 (I/.. 1961 ). For Lhe amino-sugar determinations and analysis of hydrophobic pcptidcs. the IgC; mol. eculcs kvere reduced with dilhiothreitol iii gtt:tnidine buffer and chains wcrc scparatcd on ;I Scphade\ G 100 column equilibrated with 6 .!I urea I Y acctic acid. For neutral sugar dotcltnination. Z-ME reduced chains were separated on a P60 Biogcl col~~n~tt Lvhich doch not contain glucidic components. cquilibraretl \vilh I \ acetic acid.

Reduced and alkylated L-chains from DA IgG (oh and 7,~) were mixed and filtered simultaneousl) through ;I 150 x 1.j cm Scphadcl GIOO colutnn ecluilibratcd with 1 .L’ acetic acid which has been previousl!, calibrated with reduced protein markers (Pharmacia). Rcspcctive elution volumes of the two L-chains were determined by SDS PAGE analysis of elutcd fraction. The elution volumes of the isolated Lh chains from patients Av and BE were also dctcrtnined by the same method. Sedimentation equilibrium was performed by Dr. A. Fritsch (Institut Pasteur. Paris) on a Spinco-Beckman model E analytical centrifuge according to Ihe high hpeed method 01’

Increased

mol. 11t in Myeloma

Yphantis (1964). An An-H was used with 12 rnm Kel-F centerpieces. and three samples (DA Kh, DA Kn, and normal L-chains. or. BE/h. Av Kh and normal L) were run simultaneously at 40.000 revimin and 20 C. Concentrations were measured with absorption optics at 2X0 nm and scanner equipment. Purified L-chains were dissolved in 1 1%’acetic acid and thoroughly dialysed against the same solvent.

The neutral sugar content of L-chains was determined by the artthrone reaction (Scott & Melvin. 1953). ;V-acetyl-osamine and osamine content was performed after hydrolysis with 4 ,‘I’ HCl for 4 hr _ at 100 C and IV-acetylation, using the method of Reissig ef 01. (1955) as modified by Lev-vy & McAllall (1959). The inffuence of sialic acid content on SDS-PAGE mol. wt estimation was determined by comparing migrations before and after treatment of purified M-components (1 mg,‘ml) with 0.05 U of I’ihr.io cl~olrrlrc~ neuraminidase for 24 hr at 37 C.

Purified L-chains from 11 tnyeloma proteins and from normal IgG (10 mg in 1 ml of 0.126 M atnmoiiium bicarbonate) were digested with trypsin (1 mg:ml dissolved in lo-’ .V Hc’l). Incubation was performed for 4 hr at 37 C’. and 100 /iI of trypsin solution was added every hr. After digestion, L-chains were centrifuged at 10.000 g for 15 min and examined for the presence of an insoluble pellet due to hydrophobic peptides.

RESULTS

mation of the mol. wt of the so-called 30.000 mol. wt marker in several commercial batches. Since the purpose of our study was to establish the frequency of abnormal L-chains. we used as mol. wt marker normal L-chains (2~,000 mol. wt) and est~~blished that the socalled 30,000 mol. wt band was in fact 25.000 mol. wt. In addition, new batches of Pharmacia markers showed a regular mol. wt of 30.000 for carbonic anhydrase and confirmed our previous estimations. The apparent mol. wt of L-chains frotn normal sera from commercial preparations was homogeneous (Fig. 1). No accessory bands were found. even after overloading the gel; in particular, the 30.000 mol. wt region did not show any band. although many faint unexpected bands were found in the 40.~~0~~0,00~ mol. wt region, and between 55,000 (H-cll~ti?l) and 100.000 mol. wt. The lo@ I 10,000 mol. wt band was found to occur frequently in some samples from pure polyclonal or monoclonai IgG. when not heated, and was interpreted as dimeric H-chain.

The mol. wt of L-chains from 66 myeloma patients was close to that of normal L-chains (Ln) (Table 1). In 62 of these sera. only one band was found both in L-chain and in H-chain regions by SDS-PAGE. Obvious abnormalities were found in the apparent mol. wt or in the number of chain bands from the other 39 samples (Table 2). In I1 of these cases the largest L-chain was estimated to more than 27,000 mol. wt (Lh chains) and in 23 cases it was estimated between 25.000 and 26,000 (Lm chains) (Fig. 2). In f‘lve cases the apparent mol. wt was smaller than expected (Ll chains).

Since the myeloma sera contained high concentrations of paraproteins, one-step preparations were contaminated by only negligible atnounts of normal IgG. as shown by zone electrophoresis. However, albumin was sometimes found as a trace contaminant.

Total

When estimated by comparison with protein tnarkers (Fig. l), in SDS-PAGE, the mol. wt of reduced normal L-chains was 28.000. This apparent mol. wt contrasted to the established 11101.wt of 21,000 and was due to a wrong esti-

3’)“)

L-C‘hninb

Lll Lm Ln

5 6 47

6 17 I9

LI

4 62

4:

myelomar

II 33 66” \ 104

hIn one cast of this group. t\rio Ln chains could he distinguished into the 3.000 mol. w~23,OOO mol. wt arc’:!.

400

JEAN-PIERRE

HOUVET.

JACQUES

PILLOT

and PANAYOTIS

LIACOPOULOS

a b

d

Fig. I. SDSPACiE analy\ls of normal polqclonal human reduced IgG in IO”,,gel. A. sl~ghtl> overloaded gel with 3 mg.‘ml. 2.ME treated IpG (heated samples): H. heavily ovel-loaded gel with IO mgml. 2-ME treated IgG (not heated): M. reduced mol. wt markers: (a). 94.000; (b). 67.000; (cl 33,000: and (d) carbonic anhydrase. Note the lack of L-chains of mol. wt cqual to or higher than that of(d). A and R were two samples prepared separately from the Ig preparation: note the presence in H of H-chain dimcra

2

3

4-m 67-

-Ir 23Fig. 2. SDS PAGE analysis of reduced M-components form three patients. I = reduced normal IgG. 2 = BJ protein of patient Du. 3 = same patient serum IgG (Lm chain), note that urinar) and \cruz L-chams are slightly different. 4 = Lh + Lm + Ln chains (arrows) in IgGi paraprotein preparation from patient Ma. 5 : Lm + Ln chains from patlent RI (l,gG~l.

Increased

mol. wt in Myeloma

The simultaneous presence of more than one band in the L-chain region was found in 19 cases. Size abnormality was found to be independent of the class of associated H-chain. From the five BJ proteins which were examined, three were Ln, one Lm and one Lh. All these BJ preparations reacted only with anti-Lchain anti-sera. As shown in Table 1, Lh or Lm chains were present in only 11 out of 62~ chains, but in 23 out of 43 i chains (P < 0.001). Differences in distribution were also found between myelomas of BJ type or with one H-chain band, and those with two or three H-chain bands. In this second group, five out of 11 contained a Lh chain, in contrast with six of 94 in the first group (P < 0.05) (Table 2). There was also an association between L-chain mol. wt heterogeneity and H-chain multiple banding in SDS-PAGE since of 19 cases with more than one L-chain band, seven had more than one H-chain band, whereas of 86 cases with one L-chain, only four had more than one H-chain band (P < 0.01). H-chain origirz gf’.sonw30,000 ,fiom normal IgG

mol. wt bands

One possibility for the presence of the 30,000 mol. wt band was degradation of H-chain. After keeping several days at room temp., sample preparations of normal IgG, in SDS urea buffer, showed modified migrations when analysed by SDSPAGE. Under these conditions, the H-chain band decreased, and a dis-

Table 2. Occurrence of multlple L-chains in human myelomas. and their association with single or multiple H-cham banding in SDS&PAGE analysis H chains L chains Lhh 2Lh Lh + Lm Lh + Ln Lh + Lm + Ln ?Lh + Ln Lh + 2Ln Lh + Ln + LI Lm Lm + Ln Lm + 2Ln Ln 2Ln LI Total

Oorl”

2

3

Total

2

II

I I I 2

I

I 1 : 14 5

::I

1 ~

2

-1 -

1

1 62

3

1 5 94

’ Bence-Jones proteins. ‘For Lh. Lm, Ln and Ll definition

23

7

5 105

4

see Table

nil a

Fig. 3. SDS-PAGE analysis of stored reduced normal IgG. Arrows indicate the presence of extra H-chain bands, absent in fresh preparations. (M, mol. wt markers).

Crete accessory band of 30,000 mol. wt appeared concomitantly while the normal L-chain band of 23,000 mol. wt remained unchanged (Fig. 3). In contrast, when fresh preparations were examined, normal L-chains were not associated with this 30,000 mol. wt band. In serum Am, after a 3-year storage at 4°C under aseptic conditions, the electrophoretic pattern on cellogel was modified. Whereas the original serum sample showed one paraprotein band (IgG K), two bands appeared after storage. Purification of the stored sample that the 0.005 M phosphate buffer eluted one band that contained pure L-chains; the other band was eluted with 0.015 M and contained part H-chain only reacting with anti-Fc serum, but not with anti-F(ab’)‘. These degradative products of H-chains produced 45,000 and 30,000 mol. wt bands after reduction. When the purified H-chain sample was stored at room temp. in urea-SDS solution, only one 30,000 mol. wt band appeared.

66

->

1.

401

L-Chains

Analysis

qf BE M-components

BE serum showed only one band when analysed by zone electrophoresis (Fig. 4). However, after DE-52 ion-exchange chromatography, it

was possible to isolate throc different molecular species: 1gG i.n, IgG im. and lgG i.h. the latter being the most abundant. The clcctrophoretic mobilities of the three proteins were , but clearly, difkrent when purified plTp-

arations were examined (Fig. 4A) though they were fused in the original serum (Fig. 4~). SDS PAGE analysis of each purified sample s110wed dilferences in L-chains and in H-chains mol. wt (Fig. 4B). Anti-L-chain sera reacted with IpG i.h preparations free of any other Ln chain.

+

D_i patielit contained three M-cornpoliellts present in the serum. From these M-components, two (I& ~1 and IgA ki) had normal Ln chains: however. the minor component. i&G xh. had L-chains of 30,000 mol. wt. SDSPA(;E analysis of the reduced IgG preparation after separation on a Scphadex Gl()() column showed that the ~1-i chain clearly differed from fin chain in terms of exclusion volumes (Fig. 51.

possibility was that the apparent excess ol wt or Lh in SDS PAGE could be due to an ~~~?~~ol.~~al eiectropl~oretic behaviour of these proteins. Therefore. two other methods. namely gel filtration and sedimentation equilibrium. were used for studying mol. wt of three different Lh chains (DA /- 15.6”,, (3,hOO mol. wt) in sedimentation ~~ILiilibriui~ as compared to norm;tJ I.chains (Fig. 6F). Similarly. Av r
1 (Al

2 ~

M

mol.

Neutral sugar content was estimated CJII L-chains from D.A Ln and Lh after reduction and H-~L separation through Biogel PhO. Percentages of neutral sugar were low: O.i3”,, Tot Lh (one residue per molecule) and 0.X”,, for Ln (less than one residue). Amino sugars wcw not detectable in DA Ln nor in normal L-chains or in two Lm chains. Various percent-

Increased

mol. wt in Myeionw

L-Chains

‘to3

Fig. 5. SDS--PAGE of elutcd fraction tubes from rsduc~d Dj 1gG. separated on GfOO Sephadex. in urea--acetic acld medium: wells I 3 = first peak (H-chaini: \vells S-10 = second peak (Ln chain); Lh chain mas eluted in tubes S-7: wells I I. II = non-treated Dj IgG. Lh and Ln chains were eluted in different volumes.

0.

0. 0.

a!

0.t

0.1

w

0.7

0.6

0.6

MW

404

JUAN-PIERRE

BOUVET.

JACOLJES

PILLOT

and PANAYOTIS

LIACOPOULOS

r

LOI 1

(0

Rf

I

Cn

0.2-

0.3.

0.4.

0.5,

0-6,

0.7.

9.6,

Fig. 6. Mol. wt estimation of Lh chains using different techniques. A, SDS- PAGE. DA Ln = 24,000 mol. wt. DA Lh = 30,000 mol. wt (+25:,); B, gel filtration. DA tn = 24.000. DA Lh = 29,500 mol. wt (+20.&i:,); C, sedimentation equilibrium, DA Ln < DA Lb ( + 15”,;.27.600 mol. wt); D, SDS--PAGE. BE Lb = 28.500 mol. wt, BE Lm = 26,000 mol. wt. BE Ln = 2i.500 mol. \+I: E. SDS--PAGE. Av Lh = ~O,O~ mol. +vt; F. s~dimeutatiou equilibrium. Nor~~~~~lLn < BE Lb (+ 15.6”,,1 and Normal Ln < Av Lh (+ 15.4”,;,).

ages of amino sugars were found in DA Lh, BO Ln, Cl Lh, and in Dj Ln with 0.2 1‘:,A,0.29X, 0.58% and 0.47:( respectively. It was therefore considered that these molecules contained one or less residues of amino sugar. The influence

of sialic acid content on mol. wt estimation was established directly by comparison of neuraminidase treated and non-treated chains when analysed by SDSPAGE (Fig. 7). The apparent mol. wt of Ln and Lm did not seem to he

Increased

I abab

mol. WI in Myelomo

2M3 ab

L-Chains

4 5 abab

Fig. 7. Influence of neuraminidase treatment on mol. wt estimation by (Lh + Lm). (2) IgM;. (Lm), (3) BJj. (Lm). (4) BJK(Ln). (5) BJi (Ln). (6) DA IgCh and b are non-treated and treated samples respectively.

modified by neuraminidase treatment. The mol. wt of the Lh chain was only slightly modified, indicating the presence of one or two sialic residues per molecule. Lack of‘ large ll~~ro~~~~~?~c p~~~~~~~ irz Lh chains

Thirteen L-chain preparations were digested with trypsin in order to examine the possibility of hydrophobic intra-chain peptides which may be responsible for the unusual migration after SDS interaction. An insoluble pellet was found in three out of the four Lh chains, one out of the three Lm chains, and three out of the six Ln. It therefore became obvious that no relationship existed between the apparent mol. wt and the presence of hydrophobic peptides.

From the examination of myeloma proteins from 105 patients, it appeared that 34 had L-chains with an increased apparent mol. wt. Such L-chains with abnormal mol. wt were not found, even in trace amounts, in normal sera. Although H-chain fragments of 30,000 mol. wt could be detected in stored samples from norma1 or myeloma sera, the L-chain nature of Lh

~6 iab

SDS-PAGE. (I) I@fk M. Mol. wt markers. a

was proved as they only reacted with specific anti-L-chain sera, when tested as pure preparations (BJ proteins) or in combined form with H-chain when normal sized L-chains were absent. The increased mol. wt did not seem to be due to a migration anomaly in SDS-PAGE as it is the case with an allelic form of p2 microglobulin when unreduced (Graf et al., 1982) or with some glycoproteins (Poduslo, 1981). Indeed all samples were examined after reduction and further examination of three Lh chains by gel filtration and sedimentation equilibrium (Fig. 6) constantly showed that Lh were heavier than normal L-chains, although excess mol. wt. as indicated by sedimentation equilibrium, seemed to be less than that estimated by the two other methods. The molecular nature of the increased mol. wt has not been established, but variations in the mol. wt of Lh and Lm might be due to variations in length of the glucidic chain. Preud’homme et al. (1980) reported a case of visceral L-chain deposition in which bone marrow cells synthetized large (28,000 mol. wt) K chains. Although the carbohydrate moiety of these chains was not determined. the authors

406

.IC.+IN-I’IERRI IIOIIL’ET. .14(‘Q~:~.S PILL07 and PAYAYOTIS ~.I~~(‘oP()[‘L()s

suggested ;I hypothesis of glycosylation because of the strong acid-Schiff-positivit) of the tissue deposits. The study of Savvidou CJ~(I/. (19Xlrr. I9Xlh) favours this hvpothesis since lhey described I x chain present in both glycosylatcd and i~~~~~-~~~~osvlate~iforms in the same serum. The dilfercncc ;n mol. wt es~irn~t~i~~ll of the two forms by SDS PAGE were significant (~0.000 and 3.001) mol. wt 1. although no dif%xences were found in the amino-acid sequences. The only dilltrence between the two types of molecule was the presence of a glitcidic tnoiet! in the larger. In contrast

to

this

finding,

estimation

of

sugar composition of DA LII chain did not provide any evidcncc for a large glucidic chain, since it showed only 3 or 4 sugar residues per molecule. and this is iJ~slt~~ci~]lt to explain the mol. wt increase. Presence of amino sugars has been sought in several Lh. Lm, Ln and normal L-chains. Such sugars were not found in most of them. When detected. their proportions were similar in Lli and Ln. and not higher than 0.5x0,,, i.e. one or less amino sugar residue pet L-chrtin molecule. Although such dcterminations could underestimate amino sugar content owing to some partial dcstructioti during h+rolysis, there was 110 correlation between L-chain size and the presence of amino sugrtrs. On the other hand. since amino sugars arc linked to asp~~t-a&iile of immunoglob~tlir~ chains through their glucidic moiety (Sou & Hood. 1970) it would be expected that Lh chains have additional asparagine residues. Howcvcr. total amino-acid determination of DA Lh and DA Ln did not show any dilference in asparagine content which could favour glycosylation of DA Lh (Bouvct ct trl.. 1983). Sialic acid content did not seem either to be present in proportions higher than one or Iwo residues in Lh chains as estimated by the slight si~iftitl~ of Lh bands in SDS-PAGE (but not of Lm hands) after lleitran~iiii~~as~ treatment (Fig. 7). It is therefore unlikely that all Lh chains have branched glucidic moieties with enough sialic residues to explain the increased mol. wt (I S-Y,,). An alternate explanation would be the presence of an additional peptide in the protein chain. in Daudi cells. membrane IgM comprise ;t [urge /
change in mobilit! in SDS -PA<;E occurrod after treatment with neuraminidase and glj~cosidases (Kennel. 1973). 2Y-terminal hydraphobic peptide is present in L-chain prccnrsor~ (Milstein r’f LII.. 1973: Jilka & Pcska. 1077) atxi one mouse L-chain (MPC I1 1 with 1i t”xir:~ residues at its !~‘-~~rnlitttts. has hccn rcportcti (Smith, 1973). in the case of D.4 Lh. the 25 ,Y-tertninal amino-acids wcrc 4itnilur to Ihi~ of the VtiIV xtb-group. although they had diiferences at positions Lc. 5. 15. IS and 21. iii comparison with those of DA Lii chain (Houvet c’[ tr/., 1983). The C-tertnin:tl region :IIQ) seemed fret of an extra tail piccc. since it \\ ;i:,
found in normal IgG. though ii contaiiicd 30”,, of i L-chains. Simiiarly. no clear difti-renccs iti asparagine

content.

favoring

pref~rel~t~~t~ glyx-

ylation. were found to cvist betwecti I; :rtxi c chains (Kabat P! l/i., 1979). in the extra-pcptidc

hypothesis, some unknown tiifiereticcs in li and i gene organization. anri,,ot- c~prcsaiott. could explain the more frequent occitrrencc of abrroi mal expression of j. type L-chains. Prcfcroiitial association of L-chain mol. wt hctcrog2ncit) and H-chain tnultiple handing in SDS PA(;E was also striking. On the other hand. only h cases of L,-chain heterogeneity were found among 9il cases with one single H-chain band or a HJ pro&in. In cnn?rast. 5 cxcs with L-ol1ain

Iloterogclleil~

occur-t-cd

111 I 1 p;“‘:‘p““-

Increased

mol. wt in Myeloma

banding multiple H-chain with teins (P < 0.05). On the other hand, out of 19 cases with more than one L-chain bands, 7 also presented multiple H-chain banding, whereas among 86 cases with one single L-chain, only 4 had more than one H-chain bands (P < 0.01) (Table 2). This preferential association between L-chain heterogeneity and H-chain band multiplicity is very suggestive of a common mechanism allowing expression and secretion of several M-components, with size abnormalities, in the same individual. Whether this anomaly is linked with the malignant process or represents some transient intermediate step of normal Ig chain formation, immortalized by neoplastic transformation. it could be a matter of speculation at the present time. The authors wish to thank R. Pi& for technical assistance. Dr. A. Fritsctt for performing the ullracentrifuge experiments and Mme J. L&isse for secretarz~l assistance. We also thank Dr. R. Wilhamson who rc~icwcd the manuscript and Drs. A. GLIIlabert and L. GI-angcot-KCros who kindly provided us a part of the studied ‘iera. .~~hr,t,~t,lc,lil[,~17(,f~f.s-

i3is

cxccllent

REFERENCES

Bouvet J. P.. Liacopoulos P.. Pillot J.. Banda R., Tung E. & Wang A. C. (19X0) Three M-components IgAi + IgGr;n i IgGh-h in one patient (DA): lack of shared idiotypic determill~~nts between IgA and IgG and the presence of an unusual h-h chain of 30,000 mol. wt. .I. ~/~~~l~~//i. t25, 2 13.

Bouvet J. P.. Liacopoulos P.. Pillot J.. Bande R.. Tung E. YL Wang A. C. (19X3) Three M-components IgAi + l&pin + IgGh-h 1n one patient (DA) II: Structurally d~l~crent K,, and K,, chains associated with Ho,) chains sharing idiotypic determinants. J. Imfnuf~. (In press). Dorrington K. J. & Rochey J. H. (1970) Difercnces in the molrcular size of the heavy chams from ;‘A,- and yAZ-globulms. Hiwhirn. hioph~s. ilc,t~~ 200, 5X4. Filittl-Wurmser S.. TempewGaillourdet M. Kr Hartmann L. ( l97Oi Various molecular weights for the heavy chains of human monoclonal IgM ~mmunogiobuii~is. 1n2inio70~/i~,~~7;srF~ 7, 34i. Fine J. M. & Steinbucb M. (1Y701 A simple tcchmque for the isolation of n~onoclonal IgG and IgA. Rer. Ertrop. Eru&,\ (‘/ill. ?T Nili. 15, 1I 15. Fieischman J. B.. Pain R. & Porter R. R. (1961) Reduction of ;-globulins. 4r( 1)s Bioclfr0t. Bioplfys, Suppl. 1, 174. Graf L.. Peterson P. A.. Sege K. & Robinson P. J. (1982) isolation and charactcrtzatlon of mouse /I, mxroglohuhn allotvpes. !~lo/uc~. f~n011~n.19, X77. Isobe T. &’ Osscrman E. F. (lY74) Plasma ccl1 dyscrasia assoc~a~cd with the nroduction of incomnlete I’.)deleted~ I

L-Chains

407

IgGi molecules, gamma heavy chains. and free lambda chains containlng carbohydrate: description of the first case. Blood 43, 505. Jilka R. L. & Pestka S. (lY77) Amino acid sequence of the precursor region of MOPC-315 mouse immunoglohulin heavy chain. Proc. JIQ~II.Acutf. Sci. C.S.A. 74, 5692. Kabat E. A., Wu T. & Bilofsky H. (1979) Sequences of immunoglobulin chains. NIH Publication No. X&200X. Kennel J. (1974) The h- chains of irnrn~~noglo~~~~li~~from a continuous culture of human lymphocytes (Daudl) have an unusual molecular size. J. esp. Mrti. 139, 1031. Laemmli U. K. (1970) Cleavage of structural protems during the assembl) of the head of bactel-iophage T,. .vrrtrl~c~.Lond 227, 6X0. Lev\y G. A. & McAllan .A (1959) The .“r’-acetylatton and estimation of henosamincs. Biochrtn. .J. 73. 177. Lewis A. F., Bergsngel 11). E., Bruce-Robertson A.. Schachter R. K. 61 Connell G. E. (1968) An atypical ~mmunoglobulin. E/o& 32, 189. Milstcin C.. Brownlee G. G., Harrison T. M. & Mathews M. B. (1972) 4 possible precursor of immunoglobulin light chains. Vcr~ut’c, :V’r\v Bin&y. 239, 117. Pod&o J. F. (I 98 1) Giycoprotein molccu~~lr-sleight estini~~tion using sodium dodecyi sulfate-pore gradient clcctrophoresis: comparison of Tris-glycinc and Trisborate-EDTA but&z systems. 4rrctiyr. Btoch~vn. 114. I3 I. Preud’homme J. L.. Morel-Maroger L.. Brouet J. C., Ceri M.. Mignon F.. Guglielmi P. & Sehgmann M. (19x0) S\nthesls of ahnormal immunoglohulins. ,IUI .J. ,%lt&. 69, 703. Reiasig J. L.. Strommper J. L. & Lelou L. F-. (l9jj) A modified calorimetric method for the estimation of h’-acetyl amino-sugars. .I. hiol. Che~n. 217. 959. Savvidou G.. Horne C.. Hofmann T.. Carver J.. Klelll M. & Dorrington K. (19XIa) A human IpG, monoclonal ore_

Savvidou G.. Klein ht.. Howe C.. HoEmann T. & Dorrington K. J. (tY8lhl 4 ~i~~~oclo~~al i~~rn~~noglobull~~ GI m xvhich some molecules possess glycosylatcd hght chains: I stc of gtycosylation. Jl&t immtri7.18, 79.3. Scott T. A.. Jr & Melvin E. H. (1953) Determination of Dextrnn with anthrone. Actrxrl~r.Clwn. 11. 1656. Small P. A.. Jr & Lamm M. E. (1966) Polypcptide chain structure of rabbit irnmunoglobuhn 1: ~G-immunoglohulin. B/oc~irc,/~~~r~,~~~ 5. 159. Smith G. P. (1073) Mouse immunoglohulin kappa chain MPC I I : extra amino-terminal rcsiduea. Sc,~e,ic,c,. \‘. y 181, 941. Spira G.. Silvian I., Tatarsky 1. & Hazmi A. (19x0) A calcium bindIng IgG my&ma protem. .%trnc/. J. H~rcwl~tol. 24. 193. Sox H. C. Jr. & Hood L. (1970) Attachment of carhohydrate to the variable regions of myeloma i~lrn~inoglobulin light chains. I+cw. win. Awri. Sci. Li.S.II. 66, 975 Vweila G. & Coelho I, M. (1074) Unexpected mobilit\ of human lambda chains m sodium dodecyl sulfate-poliacrylamide gel electrophoresis. Ilt~r,tu,lo~itc,rIli\rr,. I I, 157. Webcr K. & Oshorn M. (1969) The reliability oi molecular weight determination bc dodecyl sulfate-polyacrylamide gel electrophoresls. .I. hiol. Ciww 244, 4406. .,. .Yphantls 1). A. (1964) Equihhrium ultracentrlftlgatlon of dilute solutions. Bioc~/wwi.\rr~~ 3, 197.