Serum protein electrophoresis and immunofixation by a semiautomated electrophoresis system

Clinical Chemistry 44:5 944 949 (1998) Enzymes and Protein Markers Serum protein electrophoresis and immunofixation by a semiautomated electrophoresis system Xavier Bossuyt, * Ann Bogaerts, Gilberte Schiettekatte, and Norbert Blanckaert Semiautomated agarose electrophoresis and immunofixation performed with Hydrasys-Hyrys (Sebia) were compared with conventional, manually performed methods, including cellulose acetate electrophoresis, immunoelectrophoresis, and immunofixation. Reference intervals for agarose electrophoresis with Hydrasys-Hyrys were determined. Within-run imprecision (CV) for fraction quantitation with the semiautomated system was between 1% (albumin) and 4.5% ( -globulin). Total imprecision (CV) was between 2.7% (albumin) and 7.3% ( -globulin). Semiautomated agarose electrophoresis showed linear correlation with cellulose acetate electrophoresis. Thirty-four specimens with monoclonal components were analyzed by manual immunoelectrophoresis and immunofixation and by Hydrasys. In one case, a light-chain disease was missed with Hydrasys when the sample was diluted 1:3 (the routine dilution) but not when the sample was assayed undiluted. In another case, the Hydrasys system revealed a small IgG monoclonal component in addition to the IgA monoclonal component detected by the manual methods. In the other cases, no differences between the manual methods and the semiautomated method were seen with respect to paraprotein identification. Serum protein electrophoresis is widely used in clinical laboratories, especially for the detection and identification of paraproteins. Traditional clinical electrophoretic procedures are manual methods that use agarose gels or cellulose acetate membranes as the separation bed. Quantitation of the five major serum fractions is done by densitometric scanning of the gel or the membrane. Clinical interpretation is based on the alteration of the content Central Clinical Laboratory, Department of Clinical Pathology, University Hospital of Leuven, Kapucijnenvoer 33, B-3000 Leuven, Belgium. *Address correspondence to this author at: Central Clinical Laboratory, Clinical Pathology Department, University Hospital Leuven, Kapucijnenvoer 33, B-3000 Leuven, Belgium. Fax 32 16 332896; e-mail xavier.bossuyt @uz.kuleuven.ac.be. Received July 15, 1997; revision accepted February 11, 1998. of one or more of the five fractions. Agarose as the supporting medium for protein electrophoresis is reported to give better resolution than cellulose acetate, with increased ability to detect paraproteins (1 2). Classification of paraproteins is accomplished by technically demanding immunoelectrophoresis or immunofixation. Immunoelectrophoresis is lengthy in terms of obtaining the results and is less sensitive than immunofixation (3 6). Several automated protein electrophoresis systems, such as REP and its variants manufactured by Helena, have been released during the last few decades. Recently, a new semiautomated system for agarose electrophoresis of human serum proteins has become commercially available (Hydrasys-Hyrys, Sebia). The system automatically carries out the different phases of electrophoresis: sample application, migration, incubation, staining, destaining, and drying. The system is also adapted for immunofixation. We compared the semiautomated Hydrasys-Hyrys system and conventional manual serum protein separations that use cellulose acetate membrane electrophoresis. In addition, we also compared immunofixation performed by Hydrasys-Hyrys with manually performed immunofixation and immunoelectrophoresis. Materials and Methods electrophoretic methods Semiautomated agarose electrophoresis and immunofixation was performed with the Hydrasys automate according to the manufacturer s instructions, using Hydragel 15 and 30 protein(e) and Hydragel 2 and 4 immunofixation gels (Sebia). For protein electrophoresis, 10 L of sample was applied manually to the sample template. The sample was allowed to diffuse for 5 min in the template. The application (30 s), electrophoresis (ph 8.6, 20 W, and 20 C for 7 min), and drying (65 C for 10 min) were then performed automatically in the migration compartment of the instrument. The gel was manually transferred to the staining compartment, in which staining (4 min with amido black), destaining (three times, for 3 min, 2 min, and 1 min, respectively), and drying (75 C for 8 min) 944

Clinical Chemistry 44, No. 5, 1998 945 were done automatically. The gels were scanned with the Hyrys densitometer. The immunofixation procedure on the Hydrasys automate consisted of three steps. In the first step, the sample was applied by use of a template to 6 different positions on an agarose gel (Hydragel 2 and 4 immunofixation, Sebia), and electrophoretic separation (20 W and 20 C for 9 min) was done automatically. Either fixative or monospecific antisera (IgG, IgA, IgM,, or ) was then applied to the electrophoresis lanes. This was followed by a 5 min incubation step to allow for fixation and immunoprecipitation. The final step consisted of staining (4 min with violet blue), destaining (3, 2, and 6 min, respectively), and drying (8 min at 75 C). This was done automatically in the staining compartment of the instrument. Detection of monoclonal bands was by visual inspection of stained gels. The system simultaneously processes four immunofixation samples. The manufacturer s recommendation for serum dilution before immunofixation was 1:6 for the IgG lane and 1:3 for the other lanes. The dilution was adapted depending on the concentration of the paraprotein. Generally, if the immunochemical measurement of the paraprotein exceeded 25 g/l, the dilution was 1:10 for the IgG lane and 1:5 for the other lanes. On the other hand, if the immunochemical measurement of the paraprotein was 10 g/l, the dilution was 1:4 for the IgG lane and 1:2 for the other lanes. Bence Jones detection and characterization with Hydrasys was performed with nonconcentrated urine and Hydragel 2 and 4 Bence Jones gels (Sebia). The antibodies used for immunofixation were from Sebia. Cellulose acetate electrophoresis was done manually with Sepharose cellulose polyacetate electrophoresis strips for the Microzone System (Gelman Sciences). The gels were equilibrated in diethylbarbital buffer, ph 8.6. We applied 1 L of each sample and carried out the electrophoresis at 220 V for 25 min in diethylbarbital buffer (ph 8.6). Proteins were stained by incubating the gels in Ponceau S (Analis) for 10 min. The individual fractions were quantified by densitometry (Beckman Appraise, Beckman Instruments). For conventional immunofixation, the Paragon kit (Beckman Instruments) was used according to the manufacturer s instructions. For small paraproteins ( 10 g/l by immunochemical measurement), a 1:5 dilution was used. Immunoelectrophoresis was performed as described by Grabar and Williams (7), using barbital buffer, ph 8.6, and antibodies from Sanofi Diagnostics Pasteur. The gels were from Kallestad Laboratories. Detection and identification of Bence Jones proteins with the manual techniques were done after concentration of the urine using Minicon (Amicon). evaluation details Reference intervals were established in accordance with NCCLS guideline C28-A (8). Cellulose acetate protein electrophoresis and nephelometric determination (IFCCstandardized) of IgG, IgA, IgM, haptoglobin, C3c, albumin, transferrin, and 1 -acid glycoprotein (see below) were performed on all candidate samples. We excluded samples that were hemolytic, icteric, lipemic, that displayed an M-component, or that had an abnormal value for one of the specific proteins determined. Within-assay and total imprecision were calculated as described in NCCLS EP5-T2 (9) from 50 determinations of a serum pool analyzed in duplicate over 13 days. Each day, two assays with two aliquots of the pool were analyzed. The assays were separated by a minimum of 2 h. The serum pool was prepared from patient specimens filtered through 8 m (pore size) filters (Elkay Products). The pool was aliquoted and stored at 20 C until assay. The pool concentration was 77 g/l. Method comparisons were carried out with linear regression analysis on 50 patient specimens analyzed over 5 days, as described in NCCLS guideline EP9-A (10). other analyses Quantification of serum proteins (IgG, IgA, IgM, kappa light chains, lambda light chains, haptoglobin, C3c, albumin, transferrin, and 1 -acid glycoprotein) was done by endpoint nephelometry using a Behring BN 100 instrument (Behringwerke). All reagents and calibrators were from Behring (Behringwerke). Total protein was determined using the Boehringer Mannheim reagent kit and application on a Hitachi 747 analyzer (Boehringer Mannheim). statistical analysis Statistical analysis was performed with the use of the Statistical Analysis System (SAS Institute). Results and Discussion protein electrophoresis and immunofixation by hydrasys Agarose electrophoresis performed by Hydrasys on serum samples from hospitalized patients is shown in Fig. 1. Batchwise, 30 serum protein samples could be electrophoresed simultaneously every 20 min with the Hydrasys system. The processing of the first series took 40 min. In comparison, Fig. 2 shows cellulose acetate electrophoresis of 16 of the same samples processed by the Hydrasys system. Samples 1 8 of the upper cellulose acetate gel correspond to samples 3 10 of Fig. 1. Samples 1 8 of the lower gel correspond to samples 18 25 of Fig. 1. The resolution is higher and the clarity of the electrophoresis bands is better with agarose (Hydragel) than with cellulose acetate (Microzone). For example, the oligoclonal pattern that is obvious in lane 8 on agarose gel electrophoresis (Fig. 1) is less clear on cellulose acetate electrophoresis (Fig. 2, upper gel, lane 6). Fig. 3 shows immunofixation performed by Hydrasys on four samples. The immunofixation revealed a IgG paraprotein (serum 1, lower lefthand panel), an IgG monoclonal protein (serum 2, lower righthand panel), absence of a monoclonal protein (serum 3, upper lefthand panel), and an IgG monoclonal protein (serum 4, upper righthand panel).

946 Bossuyt et al.: Serum protein electrophoresis Fig. 1. Agarose protein electrophoresis using Hydrasys was performed on 30 samples from hospitalized patients as described in Materials and Methods. Fig. 3. Immunofixation was performed on four samples, using Hydrasys as described in Materials and Methods. The immunofixation revealed a IgG paraprotein (serum 1, lower lefthand panel), a IgG paraprotein (serum 2, lower righthand panel), absence of paraproteins (serum 3, upper lefthand panel), and an IgG paraprotein (serum 4, upper righthand panel). Fig. 4 shows the manual immunofixation (Paragon) of the lower righthand sample of Fig. 3. The immunoelectrophoresis of the same sample is shown in Fig. 5. Both manual techniques reveal a IgG paraprotein. reference intervals Reference intervals for the five electrophoretic serum protein fractions were determined with Hydrasys-Hyrys from a population of 76 healthy men and 76 healthy women. Median values and the central 95% of the distribution, expressed as a fraction percentage and in absolute values (g/l), are presented in Table 1. All distributions were gaussian. No statistically significant (P 0.05) sexrelated differences were observed (Student s t-test). Fig. 2. Cellulose acetate electrophoresis was performed as described in Materials and Methods. Samples 1 8 (top to bottom) oftheupper gel correspond to samples 3 10 of Fig. 1. Samples 1 8 (top to bottom) ofthelower gel correspond to samples 18 25 of Fig. 1. imprecision In a preliminary precision test, 20 aliquots of a serum were assayed by Hydrasys-Hyrys in sequence, as proposed by NCCLS EP5-T2 (3). The CV values were 0.90%, 4.61%, 1.95%, 1.86%, and 2.4% for the albumin, 1 -globulin, 2 -globulin, -globulin, and -globulin fractions, respectively. The total imprecision and within-run imprecision for protein electrophoresis by Hydrasys-Hyrys are listed in Table 2. These CV values are clearly better than previously published CV values for manually performed cellulose acetate electrophoresis (11, 12) and CV values for cellulose acetate electrophoresis obtained in our laboratory. The latter were 1.6%, 8.8%, 6.3%, 7.9%, and 6.2% (n 49; over 25 days) for the albumin, 1 -globulin, 2 -globulin, -globulin, and -globulin fractions, respectively. The excellent CV values that can be obtained with Hydrasys-

Clinical Chemistry 44, No. 5, 1998 947 Table 1. Reference intervals for the five electrophoretic serum protein fractions. a Absolute values Median, g/l 95% range, g/l Albumin 48 41.9 53.5 1 -Globulin 1.5 1.0 2.1 2 -Globulin 6.7 5.1 8.1 -Globulin 6.9 4.7 9.2 -Globulin 7.8 4.7 12.2 Fraction percentage Median, % 95% range, % Albumin 67.4 60.8 72.2 1 -Globulin 2.2 1.4 3.1 2 -Globulin 9.5 7.1 11.8 -Globulin 9.6 6.9 12.5 -Globulin 11.2 7.1 15.9 a Reference intervals were estimated from a population of 152 healthy adults, ages 18 65 years (76 men and 76 women). Fig. 4. Immunofixation performed with the Paragon kit (see Materials and Methods) of the case presented in the lower righthand panel of Fig. 3. The sample was applied at six different positions on an agarose gel, and proteins were separated by electrophoresis. Monospecific antisera was then applied to five of the electrophoresis lanes (IgG, IgA, IgM,, and ). After fixation, the electrophoresis gel was washed, and the proteins were stained. Immunofixation revealed an IgG paraprotein. Hyrys probably are related to the automation of most steps of the procedure, including sample application, migration, and staining. method comparison studies To compare Hydrasys-Hyrys with cellulose acetate electrophoresis, 50 samples were analyzed with both systems. The five major serum fractions were quantified. The linear regression plots are shown in Fig. 6. For each of the five fractions, good linear correlation was found between both methods, as reflected by the good coefficients of correlation ( 0.98). monoclonal immunoglobulin and light chain disease Thirty-four specimens with a monoclonal component were analyzed with Hydrasys and with the two reference methods, immunoelectrophoresis and immunofixation. We investigated 5 cases of light chain disease (4 and 1 ), 2 of IgA (1 IgA and 1 IgA ), 23 of IgG (14 IgG and 9 IgG ), and 4 of IgM (2 IgM and 2 IgM ) monoclonal paraproteins. In addition, two samples with biclonal paraproteins were also evaluated. The results of the comparison study are summarized in Table 3. Generally, no clinically significant differences between the manual methods and Hydrasys were seen with respect to characterization of monoclonal paraproteins. In one case, the Hydrasys system revealed a small IgG monoclonal component in addition to the IgA monoclonal component detected by the manual comparison methods. In another case, the heavy chain but not the light chain could be identified by Hydrasys using the Fig. 5. Immunoelectrophoresis of the case presented in the lower righthand panel of Fig. 3 was performed as described in Materials and Methods. The troughs contained antibodies against human proteins, IgG, IgA, IgM,, and, respectively. Wells 1 7 contained control serum (odd numbers) or patient serum (even numbers). Table 2. Precision. a Mean, % SD, % CV, % Albumin Within-run 65.30 0.65 1.0 Total 65.30 1.78 2.7 1 -Globulin Within-run 1.96 0.07 3.6 Total 1.96 0.11 5.8 2 -Globulin Within-run 7.87 0.23 3.0 Total 7.87 0.36 4.6 -Globulin Within-run 10.80 0.49 4.5 Total 10.80 0.79 7.3 -Globulin Within-run 14.00 0.47 3.4 Total 14.00 0.92 6.6 a Within-run precision and total precision for protein electrophoresis with Hydrasys-Hyrys were estimated according to NCCLS EP5-T2 (3) on 52 determinations analyzed in duplicate over 13 days.

948 Bossuyt et al.: Serum protein electrophoresis Fig. 6. Linear correlation between cellulose acetate electrophoresis (x-axis) and Hydrasys-Hyrys (y-axis) for each of the five serum protein fractions. Linear regression analysis was performed on 50 patient specimens analyzed over 5 days. The parameter estimates of the regression analysis for, respectively, albumin, 1 -globulin, 2 -globulin, -globulin, and -globulin are as follows: slope (1.01 0.02, 0.81 0.05, 0.89 0.04, 1.00 0.08, and 0.92 0.03); intercept ( 0.01 0.08, 0.01 0.01, 0.16 0.03, 0.01 0.05, and 0.01 0.03); and correlation coefficient (0.99, 0.91, 0.95, 0.89, and 0.98). standard dilution. The light chain type was revealed by immunoelectrophoresis. The group of paraproteins that we studied also included five cases of light chain disease. Of these five cases, one case was missed with Hydrasys when the usual 1:3 serum dilution was used. However, when the serum sample was assayed undiluted, free light chains could be detected by this system. Free light chains in serum are more easily detected by immunoelectrophoresis than by immunofixation. This is because immunoelectrophoresis includes a diffusion step that guarantees antibody-antigen equivalence. With immunofixation, great care in the control of the antigen concentration to maximize sensitivity is necessary. It should be pointed out that the usual dilution recommended by the manufacturer is not appropriate for all samples. Each sample should be evaluated individually and the serum dilution adapted according to the amount of paraprotein. Therefore, high-resolution agarose electrophoresis should be available before immunofixation is performed. In addition to monoclonal paraproteins, we have also investigated two cases of biclonal (oligoclonal) disease. In

Clinical Chemistry 44, No. 5, 1998 949 IgG (n) Table 3. Paraprotein identification. IgA (n) IgM (n) Quantitation, g/l Hydrasys Hyrys Manual IEF IFE a Hydrasys Hyrys Manual IEF IFE Hydrasys Hyrys Manual IEF IFE 50 2 2 1 1 1 1 30 50 7 (4) b 7 (6) 10 30 14 c (1) 14 (1) 1 d 1 1 1 10 2 2 a IEF, immunoelectrophoresis; and IFE, immunofixation. b The number in parentheses indicates the number of samples in which free light chains were found. c In one case, the heavy chain but not the light chain could be identified with the use of the usual dilution recommended by the manufacturer. d In addition to the IgA monoclonal component, an IgG was found with Hydrasys Hyrys but not with manual immunoelectrophoresis and immunofixation. one case, Hydrasys revealed an IgG in combination with an IgM and free, whereas immunoelectrophoresis and manual immunofixation disclosed IgG and free. In another case, the manual techniques revealed an IgG in combination with IgG, whereas the same paraproteins with an additional IgG was found with Hydrasys. We also evaluated 20 randomly selected samples that did not reveal a monoclonal component by the manual techniques. None of them revealed a monoclonal protein by Hydrasys. urinalysis Identification of Bence Jones proteins in urine was evaluated on 16 samples. The samples included 10 and 6 Bence Jones proteins. An IgG monoclonal protein was present in 4 of the 16 cases. No discordances were found between Hydrasys and the two manual methods. In four additional cases, a polyclonal increase of the light chains was equally detected by all three techniques. The manual method used concentrated urine, whereas the Hydrasys system used nonconcentrated urine, which is a timesaving feature. Concentrated urine can also be used by Hydrasys with the Hydragel protein(e) gel and might be the more sensitive method. In conclusion, Hydrasys provides for reproducible and rapid serum protein electrophoresis. The system also provides for accurate identification of paraproteins, which ascertains the quality and appropriateness of the specific antibodies. We acknowledge H. Raveschot and M. Artoos for their expert technical assistance. We wish to thank G. Marien and E. Stevens for their contribution in generating the immunoelectrophoresis data. References 1. Jeppson JO, Laurell CB, Franzén B. Agarose gel electrophoresis. Clin Chem 1979;25:629 38. 2. Aguzzi F, Jayakar SD, Merlini G, Petrini C. Electrophoresis: cellulose acetate vs agarose, visual inspection vs densitometry. Clin Chem 1981;27:1944 5. 3. Gerard S, Chen K, Khayam-Bashi H. Immunofixation compared with immunoelectrophoresis for the routine characterization of paraprotein disorders. Am J Clin Pathol 1987;88:198 203. 4. Duc J, Morel B, Peitrequin R, Frei PC. Identification of monoclonal gammopathies: a comparison of immunofixation, immunoelectrophoresis and measurements of kappa- and lambda-immunoglobulin levels. J Clin Lab Immunol 1988;26:141 6. 5. Guinan JEC, Kenny DF, Gatenby P. Detection and typing of paraproteins: comparison of different methods in a routine diagnostic laboratory. Pathology 1989;21:35 41. 6. Chu SY, MacLeod JE, Bocci L, Monteith M. Characterization of small monoclonal protein bands with Beckman s Paragon immunofixation system. Clin Chem 1987;617. 7. Grabar P, Williams CA. Méthode permettant l étude conjugée des propriétés électrophoreétique et immuno-chimiques d un mélange de protéines. Application en serum sanguin. Biochim Biophys Acta 1953;10:193 4. 8. National Committee for Clinical Laboratory Standards. How to define and determine reference intervals in the clinical laboratory: approved guideline. NCCLS Document C28-A. Villanova, PA: NCCLS, 1992;15. 9. National Committee for Clinical Laboratory Standards. Evaluation of precision performance of clinical chemistry devices, 2nd ed. Tentative guideline. NCCLS Document EP5 T2. Villanova, PA: NCCLS 1992;12. 10. National Committee for Clinical Laboratory Standards. Method comparison and bias estimation using patient samples: approved guideline. NCCLS Document EP9-A. Villanova, PA: NCCLS, 1995; 15. 11. Baars JD, Lombarts AJ. Imprecision of protein electrophoresis. Clin Chem 1986;32:1425 6. 12. Kahn SN, Strony LP. Imprecision of quantification of serum protein fractions by electrophoresis on cellulose acetate. Clin Chem 1986;32:356 7.