Detection and identification of monoclonal gammopathies by capillary electrophoresis

^a Address correspondence to this author at: Klinisch Chemisch Laboratorium, Diagnostisch Centrum Centrum SSDZ, Postbus 5011, 2600 GA Delft, The Netherlands. Fax 31-15-2568103; e-mail henskens{at}compuserve.com.

	Abstract

Top Abstract Introduction Materials and Methods Results Discussion References

 
Capillary electrophoresis (CE) and immunosubtraction capillary electrophoresis (IS-CE) were compared with the conventional methods agarose gel electrophoresis (AGE) and immunofixation electrophoresis (IFE) for detection and identification of paraproteins. In total, 74 paraproteins out of 468 serum samples were detected by both methods. Seventy-three monoclonal bands with concentrations ranging from 0.6 to 50.9 g/L were detected by the routine method. With CE, 70 paraproteins were detected and quantified on the electropherogram. Four paraproteins were not detected by CE; three of these were IgG (0.6, 1.1, and 2.2 g/L, respectively), and one was a IgM paraprotein (20.3 g/L) that could be visualized by minor changes in the running conditions. In comparison with IFE, 69 paraproteins were typed identically using IS-CE; only one paraprotein (IgM, 14.9 g/L) could not be identified. On the other hand, a monoclonal IgA band that had not been detected by AGE was identified by CE and IS-CE. We conclude that, in general, CE could be a useful method for detection of paraproteins and that IS-CE is a good alternative to IFE. Additional studies are required to investigate the ionic strength and pH of the running buffer, because these prove to be the most crucial factors for routine CE separation of paraproteins.

	Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

 
Capillary electrophoresis (CE)¹ is an analytical tool for separating molecules on the basis of molecular size, electric charge, and hydrophobicity. Clinical applications of CE have been described for the detection and separation of hemoglobin variants (1)(2), PCR products (3)(4), and serum proteins (5, 6). CE has been suggested as an alternative for the conventional agarose gel electrophoresis (AGE) in separating human serum proteins because it allows fast protein separation with good resolution, using only small amounts of sample (7)(8)(9)(10)(11). AGE is widely used for the screening and monitoring of several pathological processes because it can separate human serum proteins into the albumin, -1, -2, ß-1, ß-2, and regions. Monoclonal immunoglobulin molecules (paraproteins) produced by single clones of plasma cells can be detected by AGE because they produce a narrow band in or near the region. The heavy and light chains of the paraprotein can be identified further by immunofixation electrophoresis (IFE), using specific antibodies. The capillary electropherogram, like the electropherogram produced with AGE, shows the conventional separation into five regions. CE can be used to screen for the presence of paraproteins, because these proteins produce sharp spikes (5)(11). Additional identification of paraproteins can also be performed on CE by the technique of immunosubtraction (IS-CE). Immunosubtraction as an alternative method for IFE was first described by Aguzzi and Poggi in 1977 (12). Classes or types of immunoglobulins are removed from a serum sample, using a binder specific for the immunoglobulin coupled to a solid support. The serum sample is exposed to five different binders: antibodies against the heavy chains of IgG, IgA, and IgM and the and light chains. Before and after exposure to each of the solid supports, the treated sera and an untreated control serum are separated by CE. The treated samples will show subtraction only if the specific protein is removed by the coupled antibody. Otherwise, the electrophoresis pattern will remain unchanged. The aim of the present study was to compare the established methods for investigating paraproteins (AGE and IFE) to CE and to IS-CE. Detection of paraproteins on a CE electropherogram was investigated. Subsequently these suspected paraproteins were identified by IS-CE. We present here a total of 74 paraproteins selected by the AGE and/or CE technique out of 468 serum samples that were presented to our laboratory for the routine investigation of the serum protein pattern.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

 
materials
CE experiments were performed on a P/ACE System 5000 software, a CAP Capillary Cartridge (100 x 200 µm aperture), and the P/ACE System Capillary Performance test kit (Beckman Instruments). The capillary tubings were purchased from Beckman Instruments. The reagents for immunosubtraction, the Paragon CZE 2000 IFE/s 50 test, were from Beckman Instruments. Densitometric scans of agarose gels were made on a EDC densitometer (Baxter). Immunofixation was performed using the immunofixation kit of Dako. All other reagents were from Merck or J. T. Baker.

patient samples
For the CE procedure, serum samples were diluted 20-fold in buffer containing 37.5 mmol/L sodium chloride, 1.6 mmol/L sodium phosphate, and 0.4 mmol/L potassium phosphate, pH 7.4 (fourfold-diluted phosphate-buffered saline).

For the immunosubtraction procedure by CE, the serum samples were diluted seven times in fourfold-diluted phosphate-buffered saline. Serum samples with small monoclonal bands (<0.5 g/L) were diluted two times.

ce procedure
Untreated fused-silica capillaries of 27 cm length and 50 µm i.d. were used. The capillary was thermostated at 20 °C. All samples were loaded on the 34-space inlet tray of the CE instrument. Diluted samples were introduced by pressure injection (2 s), followed by protein separation at a constant voltage of 12 kV. The assay conditions for CE used in this study were slightly different from those used in the study of Chen et al. (5). The running buffer used in this study was 150 mmol/L boric acid, pH 9.9, and the total run time was 4.4 min. The minor changes were made to improve the peak resolution in the region. Serum proteins were detected on-line at 200 nm. Between runs, the capillary was washed and reconditioned for 0.2 min with 1 mmol/L NaOH, 0.2 min with demineralized water, and 1 min with running buffer. Each new analysis was started with a rinsing period of 1 min with running buffer.

is-ce
Twenty microliters of a diluted serum sample were added to six wells of a microtiter plate. The solid supports in the Paragon CZE 2000 IFE/s segment were resuspended by shaking the segment on a vortex-type mixer at high speed for 10 s. Subsequently, the foil over the seven wells (one well remained empty as a control) of the segment was punctured with clean pipet tips without touching the solid support slurry. The six different solid support slurries, one containing only buffer, were added to the six wells of the microtiter plate containing the serum samples; the samples were then subjected to 10 cycles of aspiration and dispensing, using a multichannel pipet set on 75 µL. The mixing of serum samples and the solid support slurry was repeated four times in 1-min intervals, after which the solid supports were allowed to settle for 10 min. Twenty microliters of each supernatant, including the control without solid support slurry, were transferred to microvials and loaded on the inlet tray of the CE instrument. Then the CE procedure was started as described.

age and immunofixation
AGE was performed using gels made according to the procedure described by Johansson in 1972 (13). On every gel, a control sample (human serum pool) was used to control the electrophoresis process and the staining procedures. Immunological identification and characterization of monoclonal components was performed according to the standard immunofixation procedures described by Dako.

paraprotein quantification
The agarose gels used in the routine method were stained with Coomassie^® Brilliant Blue (2 g/L), and the bands were scanned at 610 nm. Relative peak areas were determined using a stand-alone integrator. The paraproteins on the CE electropherogram were calculated using the total protein value as concentration determined on a Paramax automatic analyzer (Baxter).

statistical analyses
Statistical analyses were performed using SPSS/Labostat 6.0 software.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

 
ce and is-ce methods
The electropherogram of a nondiseased serum sample obtained by CE is shown in Fig. 1 . Fig. 2 shows the difference between a monoclonal band (Fig. 2A , IgM) in the region and a polyclonal raised IgA in the region (Fig. 2B ). The IS-CE electropherograms of a serum sample with a monoclonal IgG band after incubating the serum sample with five different immobilized antibodies are shown in Fig. 3 . Only Fig. 3 , B and E, show the disappearance of the the large peak in the region, proving that the paraprotein is IgG and , respectively.

View larger version (18K): [in this window] [in a new window]   Figure 1. An electropherogram of a nondiseased serum sample on capillary electrophoresis.

View larger version (18K): [in this window] [in a new window]   Figure 2. Electropherograms on capillary electrophoresis of paraprotein IgM (A) and polyclonal raised -region IgA (B).

View larger version (29K): [in this window] [in a new window]   Figure 3. Electropherograms of an IgG band, using CE, before (A) and after immunosubtraction using solid binders with antibodies against IgG (B), IgM (C), IgA (D), (E), and (F).

comparison of paraprotein detection: age vs ce
An overview of the results of the prospective comparison between the two methods is presented in Fig. 4 and Table 1 . In total, 74 paraproteins were detected by both methods. The routine method (AGE) detected 73 paraproteins out of 468 serum samples. When densitometric scanning was used, 70 paraproteins, having protein concentrations ranging from 0.6 to 50.9 g/L, could be quantified from the agarose gel electropherogram. Of the three paraprotein bands that could not be quantified by routine densitometric scanning, two were too small and one was too high in concentration (Table 1 , numbers 6–8). When CE was used, 70 paraproteins were detected (Fig. 4 ). Four paraproteins were not detected by CE, three of which were IgG (0.6, 1.1, and 2.2 g/L, respectively). One was an IgM paraprotein (20.3 g/L), according to the routine method (Table 1 , numbers 2–5). Although the running conditions were optimized for paraprotein detection, the 20.3 g/L IgM band could not be separated by CE (Fig. 5 ). Changing the dilution of this sample did not reveal a peak in the region. On the other hand, changing the ionic strength (from 150 to 130 mmol/L) or the pH (from 9.9 to 10.0) of the running buffer did produce a peak in the region.

View larger version (22K): [in this window] [in a new window]   Figure 4. Detection, quantification, and identification of paraproteins by AGE and IFE or CE and IS-CE.

View larger version (15K): [in this window] [in a new window]   Figure 5. Electropherogram of the 20.3 g/L IgM paraprotein that could not be separated by CE using 150 mmol/L borate, pH 9.9, running buffer. The arrow in the upper lefthand corner of the AGE pattern (inset) indicates the IgM paraprotein.

Eleven of the seventy CE-detected paraproteins (Table 1 , numbers 1, and 7–16) could not be quantified on the electropherogram, using the integrating program of the Beckman software. The protein concentration could not be determined because the relative peak areas of the optically detectable bands in the region could not be calculated by this computer software. On the other hand, the large band (IgG, 53.9 g/L), which could not be quantified by routine densitometric scanning because the coloring of the band was to intense, was determined by CE (Table 1 , number 6). Furthermore, one small monoclonal band, not detected by AGE, was detected by CE and subsequently identified by by IS-CE (IgA; Table 1 , number 1). This paraprotein could be confirmed by routine IFE.

comparison of paraprotein quantification: age vs ce
Comparisons of monoclonal protein concentrations determined by AGE and CE are shown in Fig. 6 . The correlation between the two methods was 0.95. When the Bland and Altman (14) analysis was used to compare the values obtained for all the paraprotein types, the CE quantifications yielded higher mean protein values compared with AGE. These differences were statistically significant (Table 2 ).

View larger version (19K): [in this window] [in a new window]   Figure 6. Comparisons of monoclonal protein concentrations determined by AGE and CE. (A) Comparison of the protein concentrations of paraproteins determined on by densitometric scanning of bands separated by routine AGE and by CE. Regression equation: y = 1.48x + 0.13, r = 0.95, n = 58. (B) Bland and Altman analysis (14) shows the differences in protein quantification between AGE and CE plotted against the mean result (overall mean difference ± SD, -6.0 ± 5.6).

comparison of identification: ife vs is-ce
The routine method for paraprotein identification, IFE using Dako antibodies, typed all paraproteins that were detected on the agarose gel (n = 73). IS-CE and IFE identically typed 69 of the 70 CE-detected paraproteins, and only one paraprotein (IgM, 14.9 g/L) could not be identified (Table 1 , number 17). This monoclonal band was clearly present on the electropherogram but could not be immunosubtracted by any of the immobilized antibodies. On the other hand, repeating the routine IFE of this sample using Beckman antibodies also produced a monoclonal band. Furthermore, it was difficult to type small paraproteins other than IgG by IS-CE because the polyclonal IgG antibodies in the region were always immunosubtracted by the IgG antibodies.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

 
The use of CE for paraprotein analysis in a routine clinical laboratory has certain advantages. One is the replacement of the manual method for serum protein separation with automated instrumental analysis and on-line detection of paraproteins. Subsequently, identification of paraproteins can also be done using the same separation technique, namely, IS-CE. Most studies concerning the use of CE in the clinical laboratory have focused on the detection of monoclonal bands on the electropherogram (10)(11)(15). The main purpose of the present study was to investigate both detection and identification of paraproteins by the technique of CE. CE and IS-CE were compared with our established methods, AGE and IFE, using patient serum samples that were presented to our laboratory for routine protein electrophoresis. Our results show that the composition of the running buffer appears to be of major importance. The ionic strength and pH of the running buffer have a large influence on the CE pattern. Just how crucial the buffer pH is was demonstrated by a 20.3 g/L IgM paraprotein, not detected at pH 9.9, that reappeared in the pre- region at pH 10.0. This phenomenon was described by Jenkins and Guerin (16). They reported eight paraproteins that did not separate correctly on CE when the usual buffer conditions were used. These paraproteins could only be detected on the electropherogram by minor changes in buffer molarity and pH. The results of our study suggest that one standard buffer condition to detect all possible paraproteins might not be practicable. On the other hand, we want to contend that the traditional AGE technique is also not flawless, which means that the use of CE still proves to be advantageous. We demonstrate that a major advantage of CE, a result of the better resolution, is the detection of a gammopathy of the IgA type. When AGE is used, the IgA monoclonal band appears as a slightly increased ß-fraction that can be easily misinterpreted as an increased transferrin. In the CE electropherogram, even a small IgA peak is distinctly separated from the transferrin band. Except for one paraprotein, all CE-detected paraproteins could be correctly identified by the technique of IS-CE.

One IgM paraprotein could not be subtracted by the solid support-bound antibody in the Paragon IFE kit. This discrepancy was probably caused by failure of the antibody to bind this specific IgM paraprotein and had nothing to do with the IS-CE technique. In practice, the interpretation of the IS-CE pattern of small IgG bands on a polyclonal background was more difficult compared with the routine IFE pattern because the polyclonal IgG in the region was also subtracted by the anti-IgG antibody used in IS-CE. In general, quantification of protein concentrations was higher using CE compared with AGE and densitometric scanning. Quantification of large bands was more accurate on CE compared with AGE. In practice, protein staining of large bands on the agarose gel was often incomplete, and different bands displayed different dye uptake. On the other hand, eight small bands were more difficult to quantitate on CE because the computer software failed to recognize small bands on the CE electropherogram as separate peaks. Although these bands were visible on the CE electropherograms, the relative peak areas could not be calculated. We conclude that, in general, CE could be a useful method for detection of paraproteins and that IS-CE is a good alternative for IFE.

On the other hand, optimal running conditions for detection of monoclonal bands in the region appeared to miss one IgM paraprotein. Additional studies are required to optimize the ionic strength and pH of the running buffer because these were proven to be the most crucial factors for routine CE separation of paraproteins.

	Acknowledgments

 
We acknowledge the technical assistance of Wil van Bergen Henegouwen and Marion Manzoli-van Rijn.

	Footnotes

 
Klinisch Chemisch Laboratorium, Diagnostisch Centrum SSDZ, 2600 GA Delft, The Netherlands.

¹ Nonstandard abbreviations: CE, capillary electrophoresis; AGE, agarose gel electrophoresis; IFE, immunofixation electrophoresis; and IS-CE, immunosubtraction capillary electrophoresis.

	References

Top Abstract Introduction Materials and Methods Results Discussion References

 

1. Sahin A, Laleli YR, Ortancil R. Haemoglobin analysis by capillary zone electrophoresis. J Chromatogr 1995;709:121-125. [Medline] [Order article via Infotrieve]
2. Hempe JM, Craver RD. Quantification of hemoglobin variants by capillary isoelectric focusing. Clin Chem 1994;40:2288-2295. [Abstract/Free Full Text]
3. Kuypers AWHM, Willems PMW, Van der Schans MJ, Linssen PCM, Wessels HMC, De Bruijn CHMM, et al. Detection of point mutations in DNA using capillary electrophoresis in a polymer network. J Chromatogr 1993;621:149-156. [ISI][Medline] [Order article via Infotrieve]
4. Van de Locht LTF, Kuypers AWHM, Verbruggen BW, Linssen PCM, Novakova IRO, Mensink EJBM. Semi-automated detection of the factor V mutation by allele specific amplification and capillary electrophoresis. Thromb Haemostasis 1995;74:1276-1279. [ISI][Medline] [Order article via Infotrieve]
5. Chen F-TA, Liu C-M, Hsieh Y-Z, Sternberg JC. Capillary electrophoresis–a new clinical tool. Clin Chem 1991;37:14-19. [Abstract/Free Full Text]
7. Wang H-P, Liu C-M. Separation and identification of human serum proteins with capillary electrophoresis. Clin Chem 1992;38:963-964.
8. Dolnik V. Capillary zone electrophoresis of serum proteins: study of separation variables. J Chromatogr 1995;709:99-110.
9. Schwer C. Capillary electrophoresis focusing: a routine method for protein analysis. Electrophoresis 1995;16:2121-2126. [ISI][Medline] [Order article via Infotrieve]
10. Jenkins MA, Kullinskaya E, Martin HD, Guerin MD. Evaluation of serum protein separation by capillary electrophoresis: a prospective analysis of 1000 specimens. J Chromatogr 1995;672:241-251.
11. Jenkins MA, Guerin MD. Quantification of serum proteins using capillary electrophoresis. Ann Clin Biochem 1995;32:493-497.
12. Aguzzi F, Poggi N. Immunosubtraction electrophoresis: a simple method for identifying specific proteins producing the cellulose acetate electropherogram. Bull Inst Sierater Milan 1977;56:212-220.
13. Johansson BG. Agarose gel electrophoresis. Scand J Clin Lab Invest 1972;29:7-19.
14. Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1986;i:307-310.
15. Clark R, Katzmann JA, Wiegert E, Namyst-Goldberg C, Sanders L, Oda RP, et al. Rapid capillary electrophoretic analysis of human serum proteins: qualitative comparison with high-throughput agarose gel electrophoresis. J Chromatogr A 1996;744:205-213. [ISI][Medline] [Order article via Infotrieve]
16. Jenkins MA. Guerin MD. Optimization of serum protein separation by capillary electrophoresis [Letter]. Clin Chem 1996;42:1886.[Free Full Text]