Subclass Typing of IgG Paraproteins by Immunofixation Electrophoresis

Transcription

1 CLIN. CHEM. 41/10, (1995) #{149} Enzymes and Protein Markers Subclass Typing of IgG Paraproteins by Immunofixation Electrophoresis Mariam Klouche, Arthur R. Bradwell,2 Dorothea Wilhelm, and Holger Kirchner We present a simple method for subclass typing of IgG paraproteins, with which we have demonstrated a large number of paraproteins that were undetected by conventional immunofixation techniques. The types and distribution of lgg subclass paraproteins were analyzed in 92 human sera in which lgg paraproteins had been demonstrated. The lgg subclass paraproteins were separated by agarose gel electrophoresis rapidly and simply and then typed with the use of sheep anti-human monospecific lggl-lgg4 antibodies. In 24 of the sera analyzed, IgG subclass typing revealed 25 additional monoclonal bands that were not detected by conventional immunofixation electrophoresis with anti-igg antisera. Most of these belonged to a different subclass type. The overall subclass frequencies were 68% IgGi, 13% IgG2, 16% IgG3, and 3% lgg4. The distribution of paraprotein subclasses, however, was different in monoclonal gammopathies of undetermined significance in which more lgg3 was shown, whereas in non-hodgkin lymphomas the number of lgg2 paraproteins was greater than expected; this finding may have diagnostic and prognostic significance. Indexing Terms: IgG subclass/paraprotein/electrophoresis, agarose/monoclonal gammopathy/non-hodgkin Iymphoma/multiple myeloma Several B-cell-proliferativedisorders are associated with the production of IgG paraproteins. These disorders include multiple myeloma, the benign monoclonal gammopathies of undetermined significance (MGUS), and the non-hodgkin lymphomas (NHL). Over recent decades an increasing incidence of multiple myelomas has been noted; it is now the second most common lymphoid malignancy (1). Some of the clinical manifestations, including bone destruction, anemia, hyperviscosity, chronic infection, and renal failure, can directly be ascribed to the excess production of plasma cells or monoclonal paraproteins (2). Other manifestations may be related to abnormal cytokine production by the malignant plasma cells or by accessory cells (3, 4). Conversely, cytokines may also direct the paraprotein isotype or subtype and thereby influence the clinical picture (5). On the basis of structural differences in the heavy chains, IgG can be divided into four subclasses with distinct physicochemical, biological,and immunologi- Institute of Immunology and Transfusion Medicine, University of L#{252}beckMedical School, L#{252}beck, Germany. 2 Department of Immunology, University of Birmingham Medical School, B15 2TJ, UK. Author for correspondence. Fax 0451/ Received February 21, 1995; accepted June 21, cal properties (6, 7). Accordingly, the pathogenetic potential of each IgG subclass differs, as is apparent from the distribution of subclasses in distinct disorders. Hyperviscosity syndrome as a complication of multiple myeloma is particularly associated with IgG3 proteins, which may autoaggregate (8), whereas patients with IgG2 paraproteins are more likely to have depressed residual immunoglobulmns (9). In addition, other studies have reported a different frequency of paraprotein subclasses in MGUS from that expected on the basis of the normal serum IgG subclass distribution (2). A distinguishable pathology with different IgG subclasses is also evident in other disorders that are characterized by pathological antibody production. As exemplified in lupus glomerulonephritis, proliferative lesions predominantly contain IgGi and IgG3, which are believed to initiate local cell proliferation by their pronounced complement fixation (10). In contrast, membranous lesions were found in association with the poor complement-fixing IgG2 or IgG4 (10). Similarly, the subclass composition of IgG paraproteins may be involved in the distinct manifestations of different B-cell neoplasias (11). No simple routine clinical laboratory techniques are currently available for subtyping IgG paraproteins. Isoelectric focusing immunoblot analysis (12) and immunoblotting after electrophoresis (13) or after highresolution electrophoresis (14) result in highly sensitive detection of IgG subclasses of paraproteins, but these methods require experience to perform and interpret and are not practical in routine laboratories. We therefore developed a simple method in which immunofixation electrophoresis is used; this method is applicable in most protein laboratories. We then used the method to compare the subclass distribution of paraproteins in three different B-cell lymphoproliferative disorders-multiple myeloma, MGUS, and NHLfor which little information is available. Materials and Methods Patients and Samples The sera used in this study were obtained from 36 patients with multiple myeloma, 30 patients with MGUS, 26 patients with NHL (23 B-cell chronic lymphocytic leukemia, 3 diffuse small lymphocytic lymphomas), and 10 healthy blood donors of the University of Lubeck Medical School. The procedures followed were in accordance with the Helsinki Declaration of 1975, updated in The median age of the patients was 72 years; 52 men and 40 women entered the study. The presence of an IgG paraprotein in the 92 sera had been demonstrated previously by conventional immunofix- CLINICAL CHEMISTRY, Vol. 41, No. 10,

2 ation electrophoresis with the use of anti-igg, IgA, 1gM, K, and A antisera (Beckman, Brea, CA). The presence of free light chains was confirmed with the use of different IgG antisera (Dakopatts, Glostrup, Denmark; Behring, Marburg, Germany). Diagnosis of multiple myeloma, MGUS, or NHL was based on clinical history or bone marrow or lymph node pathology and other standard diagnostic criteria. The patients were newly diagnosed and had not received prior treatment. The sera had been collected over the past 3 years and stored at -80 #{176}C until use. Human lgg Subclass-Specific Polyclonal Antisera Sheep were immunized with highly purified IgG subclass proteins prepared from the sera of patients with multiple myeloma by chromatographic techniques. The resulting antisera were adsorbed to monospecificity with the use of pure subclass paraproteins on Sepharose immunoabsorption columns. The antisera were then affinity purified on their respective immunoabsorption columns with the use of the same purified paraproteins as stated above. Particular attention was paid to ensuring a balanced antiserum against different subclass allotypes and K and A paraproteins of the same monospecificity. The specificities of the resulting affinity-purified antisera were demonstrated by immunoelectrophoresis and radioimmunodiffusion against a panel of purified IgG subclass paraproteins comprising eight IgGi, eight IgG2, eight IgG3, and five IgG4 of both light chain types (The Binding Site, Birmingham, UK). Antisera showing reactions with other subclass paraproteins were further purified to nonspecificity.contaminating antibodies against other IgG subclasses were estimated to be <1% and to be of low affinity. The antisera were purchased from The Binding Site. Sample Preparation Serum samples were diluted to -1.5 g/l IgG per subclass with the use of B-2 barbital buffer (10 mmoijl 5,5-diethylbarbituric acid, 50 mmol/l 5,5-diethylbarbituric acid sodium salt, ph 8.6) (ParagonTM Gel System; Beckman). For optimal precipitation, sera were supplemented with 20 g/l polyethylene glycol 6000 (Serva, Heidelberg, Germany). The concomitant presence of immune complexes was excluded by using commercially available tests for the detection of immune complex IgG, -IgA and -1gM (Behring) as well as C3d and Clq (Medac, Hamburg, Germany). lgg Subclass-Specific Immunofixation Electrophoresis A modified immunofixation electrophoresis procedure was followed for subclass typing of IgG paraproteins. Prediluted sera (5 L, corresponding to 6-10 g of the respective IgG subclass) were electrophoretically separated in agarose gels (The Binding Site or Beckman) for 30 mm at 100 V with use of the Paragon system (Beckman). After electrophoresis, the proteins were precipitated with 70 j.ll of the sheep anti-human monospecific IgG1-IgG4 antibodies. To allow specific immunoprecipitation in situ, we incubated the gels at 45 #{176}C for 30 mm. After repeatedly washing the gels in saline (8.5 g/l NaC1) solution, we dried and stained them with 5 g/l Paragon blue stain followed by 50 ml/l acetic acid solution (Beckman). The serum from each patient was tested for all IgG subclasses, and on each agarose plate a control serum was run simultaneously. Specificity of the IgG subclass antisera was documented with the use of IgGl-IgG4 myeloma proteins, the subclass of which had been determined by monoclonal antibodies (The Binding Site). In addition, well-characterized monoclonal antihuman IgG subclass antisera were used in immunofixation to confirm monospecificity of the polyclonal antibodies IgGi (clone MH161-1), IgG2 (CDC clone HP6014), IgG3 (clone MH163-1), and IgG4 (clone MH164-4). Because the monoclonal antibodies did not readily precipitate their corresponding antigens, the IgG subclass concentrations were also quantitatively determined by enzyme immunoassay with the use of monoclonal anti-human IgG subclass antisera (CLB, Amsterdam, The Netherlands). The results obtained correlated with the paraprotein subclasses determined by immunofixation. Statistical Analysis Fisher s exact test and test were used to determine differences in the subclass distribution of paraproteins in multiple myeloma, MGUS, and NHL. Differences were considered significant at P <0.05. Analysis was carried out with the use of Stat-X-Act (Cytel, Cambridge, MA). Results Detection of Subclass Types of lgg Paraproteins In the presence of 20 g/l polyethylene glycol the anti-igg subclass antibodies readily precipitated their corresponding antigens without demonstrating any nonspecific binding against the other three subclasses. By applying -5,.g of each IgG subclass to the agarose gels, we were able to classify IgG paraproteins according to their subclass type in all 92 sera examined. The frequency distribution of the IgG subclass paraprotein bands was 62 IgG1, 12 IgG2, 15 IgG3, and 3 IgG4. Identification of Additional Monoclonal Bands In 24 of these sera, TgG subclass typing revealed 25 additional paraproteins, which had not been detected after immunofixation electrophoresis with anti-igg antisera (Table 1).Twenty-one of these additional bands belonged to a different IgG subclass from that already detected, including nine IgGl-IgG2, nine IgG1-IgG3, one IgG1-IgG2-IgG4, and one IgG2-IgG4 paraprotein combination. In addition, two IgGi-IgGi and two IgG3-IgG3 paraprotein combinations with distinct electrophoretic mobility were detected. Only subclass typing permitted the identification of the two IgG1A paraproteins, which had been misidentified as free A light chains with three different conventional antisera CLINICALCHEMISTRY,Vol. 41, No. 10, 1995

3 Table 1. Identification of additional paraproteins by IgG subclass-specific immunofixation electrophoresis (IFE) that had not been detected by conventional immunofixation with anti-lgg antisera. Patient Conventional lgg subclass-specific no. IFE (anti-igg) IFE (anti-iggi,2,3,4) 1 IgGK, freea lgglk, IgG1A 2 IgGK IgG1K, IgG2k 3 IgGK IgGlK, lgg2k 4 IgGK IgG1K, lgg2k 5 IgGK IgG1K, IgG2K 6 IgGK lgglk, IgG2K, IgG4K 7 IgGK IgGlK, IgG2K 8 IgGK IgGlK, lgg2k 9 IgGA I9G1A, I9G2A 10 IgGK IgGlK,IgG2K 11 IgGA IgG1A, IgG3A 12 lgga IgG1A, IgG3A 13 I9GA IgG1A, IgG3A 14 IgGK IgG1K, lgg3 15 IgGK IgG1K, IgG3K 16 lggk lgglk, IgG3K 17 IgGK IgG1K, lgg3 18 IgGK IgG1K,IgG3K 19 IgGK IgGlx, IgG3K 20 IgGK lgg2k, IgG4K 21 lggk, free A lgglk, IgG1A 22 IgGK lgglk, IgGi 23 IgGA IgG3A, lgg3 24 IgGA IgG3A, lgg3a Additional monoclonal bands from immune complexes were excluded by the absence of immune complexbound IgG, IgA, 1gM, C3c, and C3d. Four of the additional IgG subclass paraproteins were present at concentrations too low for typing the light chains. Heterogeneous Electrophoretic Mobility of Paraproteins The majority of the additional paraproteins identified displayed electrophoretic mobilities identical to that of the bands already detected (Fig. 1, Lanes 1 and 2). In contrast, additional monoclonal bands with clearly different mobility occurred in about one-third of the 24 sera with more than one paraprotein (Fig. 1, lane 3). IgG2 and IgG4 paraproteins were mainly or exclusively positioned towards the anode, but IgG3 paraproteins demonstrated pronounced shifts towards the cathode. However, the electrophoretic movement in our system showed considerable heterogeneity, particularly for IgGi paraproteins, with extreme cathodal movement in 29, moderate anodal movement in 9, and the absence of mobility in 26. Distinctive Subclass Distribution in Different B-Cell Tumors We detected a distinctive and statistically different = 14.2; dl = 6; P = by Fisher s exact test) subclass distribution of IgG paraprotemns in different B-cell lymphoproliferative disorders (Table 2). The 36 multiple myeloma patients showed subclass distribution similar to concentrations found in normal serum, whereas IgG3 paraproteins predominated in the sera of the 30 MGUS patients. In contrast, an equally high frequency of IgGi and IgG2 paraproteins was observed in sera of the 26 patients with NHL. Discussion This simple clinical laboratory technique for subclass typing of IgG paraproteins in human sera is easy to perform and to interpret. Agarose gel electrophoresis of human sera followed by immunofixation with monospecific polyclonal anti-human IgG subclass antisera resulted in the specific detection of the subclasses of A B 25!flflQ!UIIIUI Fig. 1. Comparison of conventional immunofixation (A) with IgG subclass-specific rnmunofixation electrophoresis (B): (1) detection of monoclonal lgglk and I9G2K paraproteins; (2) identification of monoclonal IgGl K and IgG3K paraproteins; (3) detection of IgGl K and an additional lgg3 with different electrophoretic mobility. SPE IgG IgA 1gM K SPE IgGI lgg2 lgg3 lgg4 K SPE, serum protein electrophoresis.. i#{224}i1i: CLINICAL CHEMISTRY, Vol. 41, No. 10,

4 Table 2. Distribution of lgg subclasses of paraproteins in various B-cell lymphoproliferative disorders. Paraprotein subclass, number (%) MM MGUS NHL IgGi lgg2 lgg3 lgg4 Total 31 (66) 9(19) 5(11) 2(4) 47 (100) 21 (53) 4(10) 14(35) 1(2) 40 (100) 11 (37) 10 (33) 7 (23) 2 (7) 30 (100) MM, multiple myeloma. the IgG paraproteins in all 92 sera tested. Specific binding of the anti-human IgG subclass antisera was checked against purified IgGl-IgG4 myeloma proteins. When using polyclonal antibodies, nonspecific reactions cannot be excluded; however, we observed specific interactions of the anti-human IgG subclass antibodies with the IgG paraproteins in all sera analyzed. This benefited from the use of 20 g/l polyethylene glycol. Because immune precipitation is enhanced in the presence of polyethylene glycol, and because multiple IgG subclasses have also been detected in type II cryoglobulins (15), we excluded the presence of immune complexes in the 92 sera analyzed. Therefore, the additional monoclonal bands detected in 26% of our sera underline the presence of multiple IgG subclasses within one single band, a finding that supports the results reported by Withold and Rick (13). Furthermore, the use of monospecific anti-igg subclass antisera results in an increase in diagnostic sensitivity by reducing the background of polyclonal IgG (13). On the other hand, multiple paraproteins of different subclasses possibly open a broader view on B-cell neoplasias, allowing for the production of more than one molecular species of immunoglobulin by one cell or the involvement of more than one neoplastic B-cell clone. The failure to define the light-chain type of four additional paraproteins, which is also a problem with immunoblotting, may be explained by their low serum concentration (16). Therefore, the umbrella effect of the light chains of the polyclonal immunoglobulmns might obscure light-chain typing of the additional paraproteins. Recently, highly sensitive detection of subclasses of IgG paraproteins was performed by isoelectricfocusing immunoblot analysis (12) and immunoblotting after high-resolution (14) or agarose gel electrophoresis (13). These refined methods are not generally practicable in routine laboratories, are more time consuming, and often result in multiband patterns that may be difficult to interpret. In addition, multiple paraprotein heavy or light chain bands are detected more often with immunoblotting than after immunofixation (17) and may be caused by patient-specific microheterogeneity patterns as revealed by high-resolution two-dimensional electrophoresis (18). Whereas our method allows the analysis of all four IgG subclasses, including the light-chain type, on one gel, immunoblotting requires four separate blots for each subclass. Simple quantitative determinations of IgG subclasses in patients with IgG paraproteins (19) is possible only if the concentration of the specific paraprotein subclass exceeds the corresponding upper reference limit; this is particularly difficult for IgG3 and IgG4, because of their very low concentrations in unaffected subjects. The subclass frequency distribution of the IgG paraproteins in multiple myeloma paralleled the normal serum distribution, which is in accordance with results published by others (9, 12, 13). In contrast, we detected much greater involvement of IgG3 paraproteins in MGUS patients than did Kyle and Gleich (2). Currently, we have no explanation for this except that the MGUS patients tested in our study were previously untreated and did not demonstrate the same pronounced extent of residual immunoglobulin reduction; they may therefore represent a different stage of the disease. At present, no general prognostic marker predicts the clinical manifestations and the outcome of multiple myeloma or MGUS. Both multiple myeloma and MGUS are characterized by the presence of monoclonal paraproteins, of which --60% are IgG class. Because of the considerable differences of physicochemical, biological, and immunological properties (6, 7), the IgG subclass composition of paraproteins may be crucial to the clinical manifestations of these and possibly other B-cell lymphoproliferative diseases (11). To our knowledge, this is the first report on the IgG subclass distribution of paraproteins in NHL, showing a higher-than-expected occurrence of IgG2 paraproteins. We demonstrated a clear difference in the subclass composition of paraproteins in multiple myeloma, MGUS, and NHL. However, prospective and longitudinal studies to confirm this difference and possibly to establish a relation between the subclass of an IgG paraprotein and manifestations of disease are warranted. We conclude that the method described was successful in subclass typing of the 92 sera analyzed, is suitable for routine diagnostic laboratories, and may therefore serve to further clarify the possible diagnostic and prognostic value of the subclass determination of IgG paraproteins. The differential involvement of IgG paraprotein subclasses in three separate B-cell lymphoproliferative disorders may indicate an additional pathophysiological role of the subclass type of paraproteins in these diseases. We thank Viola Nissen for helpful technical assistance and T. Kohlmann of the Institute of Social Medicine for statistical advice. References 1. Osterborg A, Mellstedt H. Monoclonal and biclonal globulin-producing disorders. Eur J Haematol 1989;43(Suppl 51): Kyle RA, Gleich GJ. IgG subclasses in monoclonal gammopathy of undetermined significance. J Lab Clin Med 1982;100: Kawano M, Iwato K, Asaoku H, Tanabe 0, Tanaka H, Ishikawa H, Kuramoto A. Altered cytokine activities are related to suppression of synthesis of normal immunoglobulin in multiple myeloma. Am J Hematol 1989;30: CLINICALCHEMISTRY,Vol. 41, No. 10, 1995

5 4. Sonneveld P, Schoester M, De Leeuw K. In vitro Ig-synthesis and proliferative activity in multiple myeloma stimulated by different growth factors. Br J Haematol 1991;79: Snapper CM, Mond JJ. Towards a comprehensive view of immunoglobulin class switch. Immunol Today 1993;14: Hamilton RG. Human IgG subclass measurements in the clinical laboratory. Clin Chem 1987;33: Schur PH. IgG subclasses: a review. Ann Allergy 1987;58: Capra JD, Kunkel HG. Aggregation of yg3 proteins: relevance to the hyperviscosity syndrome. J Clin Invest 1970;49: Papadea C, Reimer CB, Check IJ.IgG subclass distribution in patients with multiple myeloma or with monoclonal gammopathy of undetermined significance. Ann Clin Lab Sci 1989;19: Roberts JL, Wyatt R,J, Schwartz MM, Lewis EJ. Differential characteristics of immune-bound antibodies in diffuse proliferative and membranous forms of lupus glomerulonephritis. Clin Immunol Immunopathol 1983;29: Schur PH, Kyle RA, Bloch KJ, Hammack WJ, Rivers SL, Sargent A, et al. IgG subclasses: relationship to clinical aspects of multiple myeloma and frequency distribution among M-components. Scand J Haematol 1974;12: Fasullo FJ, Fritsche HA, Liu FJ, Hamilton RG. IgG heavychain subclass typing of myeloma paraproteins by isoelectric focusing immunoblot analysis. Clin Chem 1989;35: Withold W, Rick W. An immunoblotting procedure following agarose gel electrophoresis for subclass typing of IgG paraproteins in human sera. Eur J Clin Chem Clin Biochem 1993;31: RadI J, Wels J, Hoogeveen CM. Immunoblotting with (sub) class-specific antibodies reveals a high frequency of monoclonal gammopathies in persons thought to be immunodeficient. Clin Chem 1988;34: Musset L, Duarte F, Gaillard 0, Thi Huong Du L, Bilala J, Galli J, et al. Immunochemical characterization of monoclonal IgG containing mixed cryoglobulins. Clin Immunol Immunopathol 1994;70: Kahn SN, Bina M. Sensitivity of immunofixation electrophoresis for detecting 1gM paraproteins in serum. Clin Chem 1988;34: Norden AG, Fuicher LM, Heys AD. Rapid typing of serum paraproteins by immunoblotting without antigen-excess artifacts. Clin Chem 1987;33: Harrison HH. Patient-specific microheterogeneity patterns of monoclonal immunoglobulin light chains as revealed by highresolution, two-dimensional electrophoresis. Clin Biochem 1992; 25: Magnusson CGM, Delacroix DL, Vaerman JP, Masson Typing of subclasses and light chains of human monoclonal immunoglobulins by particle counting immunoassay (PACIA). J Immunol Methods 1984;69: CLINICAL CHEMISTRY, Vol. 41, No. 10,