Microheterogeneity of paraproteins. I. Diagnostic value of isoelectric focusing followed by immunoblotting

271

Clinica Chin&a Acta, 171 (1988) 271-278 Elsevier

CCA 04055

Microheterogeneity of paraproteins. I. Diagnostic value of isoelectric focusing followed by immunoblotting Hayo Ihnken Schipper a, Gernot Bertram a and Ulrich Kaboth b Departments of 0 Neurology and b Medicine, University of Goettingen (FRG) (Received

7 April 1987; revision received 20 September

Key words: Monoclonal

immunoglobulin;

1987; accepted

Paraprotein Immunoblotting

after revision

microheterogeneity;

29 September

Isoelectric

1987)

focusing;

Isoelectric focusing (IEF) in thin-layer polyacrylamide gels followed by immunoblotting on nitrocellulose membranes is presently the most sensitive method in the routine detection of IgG paraproteins. With this technique, immunoglobulin class and light chain composition can be as reliably identified as in immunoelectrophoresis. The problem of firm adherence between IEF polyacrylamide gels and nitrocellulose membranes can be overcome by brief incubation in sodium dodecyl sulfate. After isoelectric focusing, IgG paraproteins display a characteristic pattern of limited electrophoretic heterogeneity. This pattern is easily recognized even in the few cases with a constant tendency to aggregate under IEF conditions and in the surprisingly high percentage of paraproteins with very alkaline isoelectric points in which it is altered due to a cathodal collection effect. It is independent of the total amount of IgG in serum and remains stable intraindividually over extended observation periods. On the other hand, there is a very high degree of interindividual variability while common paraprotein characteristics still remain recognizable.

Introduction In clinical routine, paraproteins are usually at first detected by serum fractionation with zone electrophoretic techniques. In a second step, the identity of immuno-

Correspondence to: Priv.-Doz. Dr. med. Koch-Str. 40, D-3400 Goettingen, FRG.

0009-8981/88/$03.50

Hayo

Schipper,

0 1988 Elsevier Science Publishers

Neurologische

B.V. (Biomedical

Universitaetsklinik,

Division)

Robert-

212

globulin class and light chain type is established by diffusion against specific antisera, most frequently by immunoelectrophoresis. Isoelectric focusing (IEF), due to its high resolving power, is even more suitable for the detection of paraproteins. With this technique paraproteins of all classes previously regarded as highly homogeneous - display a characteristic pattern of limited electrophoretic heterogeneity [l-3]. This monoclonal pattern consists of an individual number of broad, relatively uniform and closely spaced bands spread over a limited range within the spectrum of normal electrophoretic mobility of immunoglobulins. It is apparently mainly due to rapid postsynthetic loss of charged amino groups from glutamine and asparagine residues [4,5]. The monoclonal pattern can easily be differentiated from the varying (and in general much lesser) thickness and irregular spacing of oligoclonal immunoglobulins and from the diffuse polyclonal immunoglobulin background. In an earlier study on cerebrospinal fluid IgG paraproteins [6] we were able to demonstrate that IEF in thin-layer polyacrylamide gels is clearly more sensitive in the detection of low-intensity paraproteins than other techniques: 25% more IgG paraproteins could be found compared to cellulose acetate electrophoresis and immunoelectrophoresis. Sinclair et al [7], using IEF in agarose gels, reported similar results. However, as stated above, the diagnosis of monoclonality also requires further analysis of immunoglobulin class and light chain type after IEF by fractionation in a second dimension, preferably by immunoblot (immunofixation). This procedure has successfully been performed following IEF in agarose gels by transfer to cellulose acetate membranes [8]. Yet, there are several minor drawbacks which have prevented the widespread use of this technique in the routine screening for paraproteins, e.g. lack of skill in casting the gels [9]. This can easily be overcome by using IEF in commercially available thin-layer polyacrylamide gels. In addition, polyacrylamide gels provide higher resolution compared to agarose gels, thus permitting a more detailed analysis of monoclonal immunoglobulins. A second possible improvement consists of the use of nitrocellulose in immunoblot techniques. Nitrocellulose membranes have several advantages over cellulose acetate since they do not shrink during staining procedures and provide firm and reliable protein binding regardless of concentration. Consequently, they have replaced other membranes in immunoblot techniques following IEF in agarose gels. However, when IEF in polyacrylamide gels and blotting on nitrocellulose membranes are combined, serious problems are encountered since nitrocellulose membranes tend to stick firmly to the surface of the polyacrylamide gel. If removed by force, the gel is irreversibly damaged. Having solved this problem, we describe here a two-dimensional procedure involving IEF in polyacrylamide gels followed by immunoblot transfer to nitrocellulose membranes. With this technique, the IEF pattern characteristics are examined in a large number of known IgG paraproteins - the most abundant - and compared with conventional electrophoretic techniques in order to provide a firm basis for routine clinical evaluation.

273

Material and methods

Sera from 125 patients were investigated in which the presence of an IgG paraprotein had previously been proved using established immunoelectrophoretic criteria. In 6 of these patients, sequential examinations were possible over a period of up to 6 years. A total of 121 sera in which a clear paraprotein pattern was evident on IEF were included in the immunoblot study. IEF was performed on commercially available, thin-layer 5% polyacrylamide gels, pH 3.5-9.5 (PAG-plates; LKB, Bromma) as described earlier [lo]. For extension of pH gradients into the alkaline range up to pH 11, several gels were incubated for 12 h with a mixture containing 1% ampholytes of pH 9-11, 1.5% of pH 3-10, and 0.5% of pH 7-9 (Ampholines; LKB, Bromma). Serum specimens were diluted with distilled water to obtain comparable paraprotein concentrations: for routine protein staining with Coomassie blue, 40 pg total IgG per track in 5 ~1 sample volume were applied. For the i~unoblot technique with IgG-, kappa and lambda light chain-specific antisera, much smaller amounts were sufficient: 1, 2 and 3 pg IgG, respectively, Prior to immunoblotting after isoelectric focusing, nitrocellulose membrane strips (BA 85; Schleicher and Schiill, G&tingen) were soaked for 12 h with rabbit antisera specific for human IgG, kappa and lambda light chains (Dako, Copenhagen, Denmark) which were previously diluted with 0.05 mol/l Tris buffer, pH 7.4 (1: 1000, 1: 500, 1: 500, respectively). After washing (3 X 5 min) with the same buffer for removal of nonbound proteins, the strips were incubated for 10 min in a buffer solution containing 0.1% sodium dodecyl sulfate (SDS). The same SDS incubation procedure was applied to the gels after IEF. Thereafter, gels were carefully cleaned with a soft paper towel. The nitrocellulose strips were placed on the cathodal part of the IEF gel (where the IgG was expected) for 60 min under pressure. After washing with Tris buffer (3 X 5 min)‘and blocking with 1% bovine serum albumin (in order to prevent nonspecific staining), the strips were incubated with peroxidase-conjugated rabbit antisera specific for human IgG, kappa and lambda light chains (Dako) and stained essentially as described by Diirries and ter Me&en [ll] (with the 5% albumin incubation step omitted). Results

In 121 out of the 125 sera the characteristic IEF paraprotein pattern could be found: between 3 and 12 typical (as described above) bands, closely arranged over a distance from 0.2 to 1.6 pH units (Fig. 1). Underneath this typical pattern there is a great interindividual variability in number and thickness of bands as well as arrangement among each other and pattern position within the range of polyclonal IgG. In our group we did not find any two paraprotein patterns which were completely identical. On the other hand, the intruindividual pattern is very constant: in the 6 patients where lon~tudinal observations over up to 6 years were possible, no qualitative pattern changes could be detected (Fig. 2).

274

%

n404

40 20 Ill’

%

11&&

n = 83

t

co.45 Fig. 1. Number of bands paraproteins are excluded effect).

0.65

0.85

a0

pH Units

and size (in pH units) of IgG paraprotein banding patterns (very alkaline in which exact determinations were not possible due to cathodal collection

Four of the 125 sera exhibited only a conspicuous zone of presumably aggregated protein particles without a discernible banding pattern, spread over an individual pH range. This individual tendency to aggregate under IEF running conditions is also constant, as could be demonstrated in 2 patients where sequential exa~nations were made over 2 yr (from the four patients’ case histories, no particular properties, e.g. presence of cold agglutinins, were evident). Although the majority of banding patterns is assembled between pH 6.5 and 8.5 (the most frequent range of oligoclonal and polyclonal IgG isoelectric points), a relatively high percentage (36%) of paraproteins is found at the cathode, indicating an isoelectric point of more than approximately pH 8.7. In IEF using carrier ampholytes, proteins with isoelectric points beyond the pH range of the gel do not enter the electrode strips but are collected at the margin of the gel. Due to this collection effect, the typical distances between the single bands disappear more and

275

1983 I

1984 1985 1982 1983 1984 1985 1986

Fig. 2. Intraindividual stability in two different paraprotein patterns. Cathode to the left,

more (depending on the actual pl) until finally only one single broad band appears in the very alkaline paraproteins. However, after extension of the pH range of the IEF gel up to pH 11, the typical pattern readily emerges also in these paraproteins (Fig. 3).

Fig. 3. Cathodal collection effect of very alkaline paraproteins. Tracks 1 and 3: pH range 3.5-9.5; tracks 2 and 4: same paraproteins, appearance of a typical pattern after extension of IEF pH gradients up to pH 11.

276

anti IgG anti

Kappa

anti lambda

I

anti IgG anti Kappa

J

Fig. 4. IEF paraprotein patterns (tracks 2-4 and 6-8, peroxidase staining).

1 and 5, Coomassie

anti lambda blue staining)

after immunoblotting

(tracks

The intensity of banding patterns is apparently not correlated with the amount of paraprotein: if total serum IgG content and number of paraprotein bands or size of paraprotein pattern (in pH units) are plotted against each other, no significant correlation is evident (regression equations/coefficients of correlation: y = 23.5 1.8x/r = 0.001 and y = 22 - 16.2x/r = 0.15, respectively). Also, if paraproteins with an intensive pattern (numerous bands spreading over a broad pH range) are serially diluted (n = 3) the bands of an individual paraprotein tend to disappear at similar dilutions. On the other hand, if paraproteins with few bands spreading over a narrow pH range are concentrated (n = 3), no additional bands appear. After immunoblot transfer the paraprotein staining pattern with antisera specific for IgG or kappa/lambda light chains is identical with the pattern after Coomassie blue staining. In the 121 sera we found 72 kappa and 46 lambda light chain type paraproteins as well as 3 biclonal cases (kappa only; lambda only; kappa plus lambda). Additional free light chains were observed in 3 cases. There were no qualitative or quantitative differences in the paraprotein pattern between kappa and lambda light chain type paraproteins (Fig. 4). No serious discrepancies with the immunoelectrophoretic results were evident. Only in one of the biclonal gammopathies had the second paraprotein not been detected by immunoelectrophoresis.

211

Discussion Although isoelectric focusing is presently the most sensitive technique in the detection of paraproteins, this report cannot focus on the question of sensitivity since only (IgG) paraproteins are included which had been found at first by a less sensitive method (immunoelectrophoresis). Hence, it was not surprising that all paraproteins in our study could easily be identified by IEF. (A comparative study concerning the matter of sensitivity is currently under way involving paraproteins which were first discovered by IEF and later analyzed by immunoelectrophoresis.) The firm adherence of polyacrylamide IEF gels and nitrocellulose membranes can easily be reversed by a short SDS incubation step of IEF gels and nitrocellulose strips prior to blotting. This does not alter the electrophoretic paraprotein pattern nor the quality of the protein transfer: after immunoblotting and staining with the respective antiserum, the complete paraprotein pattern could be recognized in all cases. Therefore, even in the routine screening of large patient groups a misdiagnosis is hardly possible. We were able to demonstrate that the monoclonal pattern has a very high degree of interindividual variability among different patients, while the above mentioned common typical paraprotein characteristics still remain easily recognizable. In a recent study with IEF in immobilized pH gradients, we found that this variability is even more complex than previously assumed from conventional IEF with carrier ampholytes: in the majority of IgG paraproteins the main bands revealed an extended subfractionation [13]. On the other hand, there is a marked intruindividual stability over extended time periods. Using an experimental tumor cell line, Awdeh et al [12] demonstrated that paraproteins are synthetized in a single banded pattern but that microheterogeneity starts to emerge already within the plasma cell. This process continues for a relatively short period after secretion until a stable individual pattern is reached. Even in our most malignant myeloma cases where one would expect a certain degree of change with progression of the disease, this pattern stability persisted. The pattern stability also extends to the four cases with aggregating IgG. It should be mentioned that, unlike smaller immunoglobulins (as IgG, IgE, IgD, free heavy and light chains), IEF in polyacrylamide gels is not as well suited for the fractionation of very large or aggregating paraproteins (as IgM and IgA). This problem can be overcome by pretreatment of sera with 2-mercaptoethanol [8] which was not necessary in our study since it was restricted to IgG paraproteins. Paraproteins whose IEF pattern is altered due to a cathodal collection effect should roughly correspond to the section of IgG which moves cathodally from the application well in immunoelectrophoresis. Yet, when taking former findings into account in which only a minor percentage of IgG has very high isoelectric points [14], the large number in our study is somewhat surprising. As stated above, number of bands and size of a paraprotein pattern are individual properties and not dependent on the total amount of monoclonal IgG. However, a significant correlation of the pattern configuration with the degree of malignancy of the underlying monoclonal gammopathy was evident in an earlier

278

report: multiple myeloma cases had more bands extending over a wider pH range than patients with benign monoclonal gammopathy [4]. A more detailed study concerning this correlation will be published soon. References 1 Fahey JL. heterogeneity of myeloma proteins. J Clin Invest 1963;42:111-123. 2 Williamson AR, Salaman MR, Kreth HW. Microheterogeneity and allomorphism of proteins. Ann NY Acad Sci 1973;209:210-224. 3 Latner AL, Marshall T, Gambie M. Microheterogeneity of serum myeloma immunoglobulins revealed by a technique of high resolution two-dimensional electrophoresis. Electrophoresis 1980;1:82-89. 4 Brendel S, Mulder J, Verhaar MAT. Heterogeneity of monoclonal immunoglobulin-G proteins studied by isoelectric focusing. Clin Chim Acta 1974;54:243-248. 5 Wil~amson AR. Isoelectric focusing of j~unoglobulins. In: Weir DM, ed. Handbook of experimental immunology, 3rd ed. London: Blackwell, 1978;9.1-9.30. 6 Schipper HI, Prange HW. Cerebrospinal fluid paraproteins in neurological disorders. J Neural Sci 1984;64:305-314. 7 Sinclair D, Kumaratne DS, Stott DI. Detection and identification of monoclonal immunoglobulin by immunoisoelectric focusing. Limits of sensitivity and use during relapse of multiple myeloma. J Clin Path01 1984;37:255-262. 8 Sinclair D, Kumaratne DS, Forrester JB, Lamont A, Stott DI. The application of isoelectric focusing to routine screening of paraproteinaemia. J Immunol Methods 1983;64:147-156. 9 Normansell DE. The author’s reply - more for research. Am J Clin Path01 1986;85:532. 10 Schipper HI, &use H, Reiber H. Silver staining of oligoclonal IgG subfractions in cerebrospinal fluid after isoelectric focusing in thin layer polyacrylamide gels. Science Tools 1984;31:5-6. 11 Dorries R, Ter Meulen V. Detection and identification of virus-specific, oligoclonal IgG in unconcentrated cerebrospinal fluid by imm~obiot technique. J Neuroi~unoi 1984;7:77-89. 12 Awdeh ZL, Williamson AR, Askonas BA. One cell - one immuno~obulin: origin of limited heterogeneity of myeloma proteins. Biochem J 1970:116:241-248. 13 Schipper HI, Weser J, Bertram G, Goerg A. IgG paraprotein microheterogeneity after isoelectric focusing: lmmobilized pH gradients vs. carrier ampholytes. In: Peeters H, ed. Protides of the biological fluids, Vol. 34. Oxford: Pergamon Press, 1986;827-830. 14 Fateh-Moghadam A. Paraproteinaemische Haemoblastosen. In: Schwiegk H, ed. Handbuch der Inneren Medizin, Vol. H/5. Berlin-Heidelberg-New York: Springer, 1974;245-452.