IgM Pyroglobulinemia Hyperviscosity II. Biochemical
Syndrome
B.Ch:
M.D.
MARTIN M. OKEN, M.D. AKHOURI A. SINHA, Ph.D. Minneapolis,
Presenting as
Properties and Mechanisms of Pyrogel Formation
SHAUN R. MCCANN, MB., HORACE H. ZINNEMAN,
with Erythrocytosis
Minnesota
An immunogiobuiin M with kappa light chains (IgMK) pyroglobulin from a patient with hyperviscosity syndrome, erythrocytosis and coagulation defects has been studied for its immunochemical properties. At physiologic temperatures the purified macropyroglobulin showed a striking tendency to aggregate in the pentamer as well as in the monomer form. This property was also observed in its H chains. Aggregate formation of the pentamers may have contributed to the blood viscosity and coagulation defects. Formation of pyrogel at 56% was observed with pentamers as well as monomers, but not with separated H or L chains. Amino acid analysis showed quantitative abnormalities of aspartic acid, glycine, cystine and leucine within the H chains. Solubility of the pyrogel in sodium dodecyl sulfate, the pyroglobulin’s tendency to aggregate and the cystine deficit of H chains implicate conformational changes leading to hydrophobic bonding at 56OC in the formation of pyrogel. The study of a patient with macropyroglobulinemia and erythrocytosis, presenting as hyperviscosity syndrome [ 11, has been extended to additional
biochemical
and immunochemical
inquiries.
Previous
re-
in respect to thermoprecipitability of monomers (IgM,) and the mechanisms of pyrogel formation. Previously Zinneman and Seal [2] presented evidence that hydrophobic bonding was essential in the formation of pyrogel in an IgG pyroglobulin. This study investigates the biochemical properties of IgMK(Bi) [I] and provides further evidence for hydrophobic bonding ports of IgM pyroglobulins
in the formation MATERIALS
have differed
of pyrogels.
AND METHODS
The first steps toward isolation and purification of the pyroglobulin consisted of the separation of immunoglobulins from other serum components by starch From the Section of Infectious Diseases, Hematology, and Department of Laboratory Medicine and Pathology, Veterans Administration Hospital and University of Minnesota School of Medicine, Minneapolis, Minnesota. Requests for reprints should be addressed to Dr. Horace H. Zinneman, Veterans Administration Hospital, 54th Street and 48th Avenue South, Minneapolis, Minnesota 554 17. Manuscript accepted January 12. 1978. Present address: Department of Clinical Haematology. Meath Hospital, Dublin 8, Ireland. l
block electrophoresis. IgM could then be separated from immunoglobulins of lesser molecular weight by passage through a 2.5 by 100 cm column of Sephadex@ G200 (Pharmacia), using 0.1 M phosphate buffer of pH 7.8. The purity of the frontally eluted IgM was confirmed by immunoelectrophoresis against specific heavy and light chain antiserums. Ultracentrifugation was carried out at 60,000 rpm in a Spinco (Model E) analytical centrifuge. Reduction of IgMK(Bi) to monomers (IgM,K) was obtained by reduction with 0.03M 2-mercaptoethanol and alkylation by iodoacetamide. Ultracentrifugation of this product showed one 8s peak, confirming the reduction to monomers. Heavy and light chains were prepared
September1976
The American Journal ol Medicine
Volume 61
321
PYROGLOBULIN AND PYROGEL FORMATION-MCCANN
ET AL.
Figure 1. Gel filtration of immunoglobulins on a column of Sepklex G200. The frontal &ate of IgM has a shoulder of polymerization.
by reduction of 30 mg IgMK(Bi) in 0.312 ml of l.OM 2-mercaptoethanol and alkylation with 920 mg of iodoacetamide. Separation of heavy from light chains was obtained by passage through a column of Sephadex G75 (Pharmacia) in a medium of 0.5 N propionic acid. The purity of these preparations was verified by immunoelectrophoresis against specific antiserums as well as by ultracentrifugation. IgMK(Bi) as well as its heavy and light chains were subjected to acid hydrolysis and measured for amino acid content by. means of ion exchange chromatography on columns of amberlite resin, in a Spinco, Model 120 automatic amino acid analyzer. IgMK(Bi) was assayed for sialic acid content by the method of Warren [a]. Fucose, hexose and hexosamine were quantified according to Winzler [4]. In order to study the pyrogel’s ultrastructure, IgMK(Bi) was heated to 56’C for 30 minutes, and the resulting gel centrifuged to remove some of the trapped physiologic sodium chloride solution. The resulting pyroglobulin pellet was then fixed for 2 hours in cold 3 per cent glutaraldehyde in O.lM phosphate buffer and post-fixed in buffered 1 per cent OSmium tetroxide for 1 hour. The pellet was then dehydrated in graded ethyl alcohol and embedded in Epona 812 [5]. Thin sections were stained with a combination of uranyl acetate and lead citrate, and examined with an RCA-EMU-4 electron microscope. Unstained sections were also examined. RESULTS
After separation of immunoglobulins from other serum components by starch block electrophoresis, IgMK(Bi) could be obtained in pure form by gel filtration on columns of Sephadex G-200. On elution, the frontal IgM peak had a “shoulder” of heavier weight molecules, indicating a tendency to aggregation (Figure 1). The
322
September
1976
The American
Journal of Medlclne
Figure 2. Comparison of /gMK( Bo with five IgM’s of cirrhotic subjects. Anti-h rabbit serum in center well. /gMK( Bl) in well Number 1, and wells Number 2-6 contain purified IgM from five cirrhotic subjects. /gMK( Bi) forms lines of identity with normal IgM.
purity of the frontal peak as well as of its “shoulder” (as IgM) was confirmed by immunoelectrophoresis, which also established the monoclonal character of this IgM macropyroglobulin, which had neither euglobulin nor antiglobulin activity. IgMK(Bi) was found by double immunodiffusion to be antigenically identical with IgM from five patients with Laennec’s cirrhosis (Figure 2). As described in part I of this study [ 11, ultracentrifugation showed peaks of 19s and 22S, confirming the presence of aggregates of IgMK(Bi). Monomers of this macropyioglobulin, with sedimentation constants of 8S retained their thermoprecipitability at 56OC, but heavy and light chains did not. However, ultracentrifugation of H chains did not show the usual symmetrical peak, but one which extended from 3S to 6S, indicating the formation of H chain aggregates (Figure 3). Amino acid analysis of IgMK(Bi) showed an abnormal number of aspartic acid, glycine, cystine and leucine residues, compared to studies of normal IgM in this laboratory (Table I). The nonpolar residues of glycine and leucine were increased. The aspartic acid content was increased, whereas the cystine content was decreased. These abnormalities were noted in the heavy chains only. The carbohydrate components of IgMK(Bi) were within the normal range (Table II). Treatment of the pyrogel with 6 M urea broke it up into strands, but failed to bring it into solution. Concentrated sodium hydroxide seemed to solubilize the pyrogel, but electrophoresis showed that the pyrogel had been hydrolyzed. Heating of the patient’s IgA and IgG failed to produce precipitates or gels, confirming that IgMK(Bi) alone was the pyroglobulin. Addition of
Volume
61
PYROGLO6ULlN AND PYROGEL FORMATION-MCCANN
TABLE II
ET AL.
Carbohydrate Components of the Pyroglobulin Calculated as Mol/Mol of IgM, IgM,K(Bi)
Sialic acid Fucose Hexose Hexosamine
2.0 4.0 28.0 20.0
Normal Range 2.0-3.0 2.0-5.0 10.0-42.0 14.0-26.0
Polymerization of H chains from 35 to 6s obFjgure 3. served at ultracentrifugation at 60,000 rpm after 32 minutes.
TABLE I ____-.
Amino Acid Composition of IgM,K(Bi) and Its Heavy and Light Chains Normal K Normal Range* Chain* Range*
Amino Acid
IgM,K(Bi)”
Normal Range*
Chlin”
Lysine Histidine Arginine Aspartic acid+ Threonine Serine Glutamicacid Proline Glycrne+ Alanrne Cystine/2t Valine Methionine lsoleucine Leucine+ Tyrosine
66 20 46 160 136 190 132 122 220 158 26 148 10 46 120 34
64-86 18-24 34-46 108-118 124-148 188-208 126-138 83-130 198-212 126-l 60 30-34 138-I 56 12-14 36-50 110-112 3442
24 8 19 54 48 68 44 47 75 55 8 54 4 16 39 10
21-30 8-10 16-19 39-44 45-56 65-69 42-48 44-51 68-72 46-56 12-15 46-62 3-6 15-20 35-38 lo-12
9 2 4 22 19 25 21 14 32 23 4 16 0 6 20 6
14
14-18
6
Phenylalanine ___
40
38-48
9-10 2-3 4-6 15-20 16-20 25-30 19-22 12-14 29-32 21-23 4-6 14-17 0 5-7 16-19 5-7 6-9
*Calculated as moles of amino acid per mole of protein. + Denotes amino acids with quantitative p chain abnormalities.
Figure 4. Electrophoresis on a supporting medium of cellulose acetate, comparing the mobility of normal serum (a and f) with IgMK( Bfl (b), the SDS-IgMK-complex (c), Hand L chains of /gMK( So (d) and /gMK( Bf) in 6 M urea (e). When complexed with SDS, /gMK(SJ has acquired additional negative charges, which impart anodal mobility exceeding that of albumin.
September 1976
The American Journal of Medicine Volume 61
323
PYROGLOBULIN AND PYROGEL FORMATION-MCCANN
ET AL.
Figure 5. An electron micrograph, original magnification X 103,300, reduced by 14 per cent; shows a meshwork of IgM molecules which appear translucent and are outlined by the darker counter stain (uranyl acetate and lead citrate). Arrows indicate some of the pentamer formsof IgIUK(Sq which lie within the plane of the cut.
these immunoglobulins to IgMK(Bi) did not result in coprecipitation upon heating to 56OC. Alteration of pH by appropriate buffers within a pH range from 3 to 9 failed to inhibit pyroprecipitation, although the gel appeared to be maximal at pH 7.3. Variations of ionic strength using sodium chloride in concentrations between 3 and 10 per cent failed to inhibit pyrogel formation. The pyrogel could .be brought completely into solution by addition of sodium dodecyl sulfate (SDS) at a ratio of 30 mol of SDS to 1 mol of IgMK(Bi). Electrophoresis of the SDS-IgMK(Bi) complex showed that it had acquired a considerably greater electrophoretic mobility than the original IgtvlK(Bi) (Figure 4). Electron microscopy was helpful in providing additional information concerning the structure of the pyrogel. Figure 5 shows a section of the gel which appears to consist of a mesh of IgM pentamers. COMMENTS Previous studies designed to elucidate the nature of pyrogel formation have yielded conflicting results. Patterson and co-workers [6] described a patient with IgM pyroglobulinemia and lymphoma. This pyroglobulin lost thermoprecipitability when broken into 7s monomers. Stefanini et al. [7] described six patients with Waldenstrtim’s disease and IgM pyroglobulinemia.
324
September1976 TheAmericanJournalof
Medlclne
Treatment with 2-mercaptoethanol and iodoacetamide partially inhibited thermoprecipitability. However, the degree of reduction by P-mercaptoethanol was not documented by sedimentation coefficients. Our findings concerning the thermoprecipitability of the 7s monomers are in agreement with those of lnvernizzi et al [8]. The decreased number of cystine residues and the increase in the glycine and leucine residues in the H chains of IgMK(Bi) may be of significance, since they may represent a relative excess of nonpolar residues. Whereas carbohydrate components and their abnormalities have been shown to influence the solubility of cryoglobulins [g-12], they have been found to be normal in one IgG pyroglobulin studied by Zinneman and Seal [2] as well as in IgMK(Bi) of the present study. IgMK(Bi) was solubilized with sodium dodecyl sulfate (SDS), as was the IgG pyroglobulin previously reported by Zinneman and Seal. Solubility with SDS leading to SDS-IgM complexes implicates hydrophobic bonding in the pyrogel formation. The tendency of IgMK(Bi) and its H chains to aggregate is compatible with cystine and methionine deficit and loss of polarity under physiological conditions. The cited abnormalities of amino acid composition, particularly the low number of cystine residues, may be responsible for abnormal secondary and tertiary conformation, resulting in loss of surface polarity of
Volume61
PYROGLOBULIN
AND PYROGEL FORMATION-MCCANN
ET AL.
IgMK(Bi) to a point where solubility can be maintained
bonds and solubility changes on cooling of the former.
only by means of aggregation.
Pyroglobulins, however, have a temperature-dependent instability, which leads to irreversible hydrophobic bonding at 56OC. The occasional combination of cryoprecipitation and pyrogel formation [ 13,141 would presume the presence of both defects in the same molecule.
Further conformational
changes, due to heating, expose additional nonpolar residues of the molecule’s hydrophobic interior. Whereas the abnormal H chains alone do not form pyrogels, they do so together with L chains which normally show a change of solubility at 56OC, and strong hydrophobic bonds, involving both H and L chains, result in pyrogel. A basic difference between cryoglobulins and pyroglobulins is the appearance of reversible noncovalent
ACKNOWLEDGMENT We thank Mrs. Marilyn G. Goodsell for valuable technical assistance.
REFERENCES
2.
3. 4.
5. 6.
7. 8.
McCann SR, Zinneman HH, Oken MM, Leary MC, Swaim WR, Moore M: IgM pyroglobulinemia with erythrocytosis presenting as hyperviscosity syndrome. I. Clinical features and viscometric studies. Am J Med 61: 316, 1976. Zinneman HH, Seal US: The role of hydrophobic bonding in the thermoprecipitation of a pyroglobulin. Rev Europ Etud Clin Biol 6: 668, 1971. Warren L: The thiobarbituric acid assay of sialic acids. J Biol Chem 234: 1971, 1959. Winzler RJ: Methods of biochemical analysis, chap 2. Determination of Glycoproteins, vol2 (Glick D, ed), New York, Interscience. 1955. Luft JH: Improvements in epoxy resin embedding methods. J Biophys Biochem Cytol 9: 409, 1961. Patterson R, Roberts M, Rambach W, Falleroni A: An IgM pyroglobulin associated with lymphosarcoma. Am J Med 48: 503, 1970. Stefanini M. McConnell EE, Andracki EG. Swanstro WJ, Durr P: Macropyroglobulinemia. Am J Clin Pathol 54: 94, 1970. lnvernizzi F, Cattaneo R, Rossodi SAN, et al.: Pyroglobu-
9. 10.
11.
12.
13.
14.
September 1976
linemia. Acta Haematol 50: 65, 1973. Zinneman HH, Levi D, Seal US: On the nature of cryoglobulins. J lmmunol 100: 594. 1968. McIntosh RM, Kaufman DE%,Kulvinskas C, Grossman BJ: Cryoglobulins. I. Studies on the nature, incidence and clinical significance of serum cryoproteins in glomerulonephritis. J Lab Clin Med 75: 566, 1970. McIntosh RM, Kulvinskas C, Kaufman DB: Cryoglobulins. II. The biological and chemical properties of cryoproteins to acute post streptococcal glomerulonephritis. Int Arch Allergy Appl lmmunol 41: 700, 1971. Zlotnick A, Shim S, Eliakim M: Mixed cryoglobulinemia with a monoclonal IgM component associated with chronic liver disease. Israel J Med Sci 8: 1068, 1972. Meltzer M, Franklin EC, Elias K, et al.: Cryoglobulinemia-a clinical and laboratory study. II. Cryoglobulins with rheumatoid factor activity. Am J Med 40: 837, 1966. Watanabe A, Kitamura M, Shimizu M: lmmunogiobulin A (IgA) with properties of both cryoglobulin and pyroglobulin. Clin Chim Acta 52: 231, 1974.
The American Journel of Medicine
Volume 61
325