Journal of Immunological Methods, 98 (1987) 119-122
119
Elsevier JIM 04276
Comparison of polyacrylamide and agarose support media for detection of oligoclonal IgG bands in CSF J.P. N o r t h and H.M. Chapel Department of lmmunology, John Radcliffe Hospital, Oxford, U.K.
(Received 8 August1986, revisedreceived19 November1986, accepted4 December1986)
Parallel samples of cerebrospinal fluid (CSF) from 120 patients clinically suspected of having multiple sclerosis were electrophoresed using agarose or 7% polyacrylamide. 57 samples showed oligoclonal IgG patterns in polyacrylamide but only 34 showed bands after concentration and electrophoresis in agarose. The concentrations of IgG and albumin were also measured in each sample and IgG/albumin ratios calculated. The sensitivities of the three methods for the detection of intrathecally synthesised IgG were compared. Electrophoresis in 7% polyacrylamide gave the highest proportion of significant results. In an attempt to increase the sensitivity of the agarose method, silver staining was performed on unconcentrated samples run in agarose. This did not alter the pattern of electrophoresis in samples containing oligoclonal IgG, but the sensitivity of the technique was lower than that of polyacrylamide. Key words: Agarose; Multiplesclerosis; IgG, oligoclonal;Polyacrylamide
Introduction
Materials and methods
Oligoclonal IgG bands in the CSF are usually indicative of intrathecal synthesis of immunoglobulin (Thompson and Johnson, 1982). Since they are present in a high proportion of patients with MS, their detection is a useful adjunct to the diagnosis (Thompson, 1977). Electrophoresis of unconcentrated CSF in 7% polyacrylamide with differential staining is often used, although this method is time consuming for a routine laboratory. We have compared results obtained using polyacrylamide or agarose media and identical samples of CSF from patients with suspected MS; both traditional and silver staining methods have been used.
Samples
Correspondence to: H.M. Chapel,Departmentof Immunology,John RadcliffeHospital,Oxford,U.K. Abbreviations: MS, multiple sclerosis; CSF, cerebro-spinal fluid.
Samples containing at least 3 ml of CSF were obtained by lumbar puncture from 120 patients with clinically suspected MS. Samples were collected in plain, sterile glass or plastic universals and stored at 4°C until use. Concentrations of albumin and IgG in each sample were measured using a Beckman ICS II rate nephelometer; the percentage ratio of IgG to albumin was then calculated. The ratios were interpreted according to the data obtained by Roberts-Thompson et al. (1976). Values up to 22% are normal, those over 28% are raised and those between are doubtful. The remaining CSF from each sample was then split and electrophoresed in either agarose or 7% polyacrylamide. Polyacrylamide
The polyacrylamide gels were prepared as de-
0022-1759/87/$03.50 © 1987 ElsevierSciencePubfishersB.V.(BiomedicalDivision)
120
scribed by Johnson and Thompson (1982). Protein concentrations were estimated and 100 mg and 200 mg of protein respectively were applied to each of two 7% polyacrylamide gel columns. These were run simultaneously in a Pharmacia GE-2/4 LS electrophoresis tank with a Trisma/glycine buffer at 50 mA for 4-5 h. The gel containing 100 mg of protein was subsequently stained with Coomassie brilliant blue G250 for 2 h in a 56 o C water bath. The other gel was stained with naphthalene black for 1 h at room temperature. The Coomassie-stained gels were destained overnight with 7% acetic acid and the naphthalene black-stained gels were destained either over a week, or latterly with a Pharmacia GD-4 II gel destainer for 1 h. Cathodal (3'4, 3'5) bands or diffuse cathodal (3'4, 3'5) staining (stronger in the Coomassie gel) were regarded as indicative of MS. Anodal (3'1, ,/2, 3'3) banding, stronger in the Coomassie gel, was interpreted as intrathecal synthesis of unknown significance (Thompson et al., 1979). The gel stained with naphthalene black permitted differentiation between basic IgG and more acidic proteins (e.g., haptoglobin) which also run in the gamma region.
subsequently applied to the gel which was electrophoresed for 40 min in pH 8.6 barbitone buffer at 100 mV in a Beckman Paragon electrophoresis system. The gels were then fixed in methanol/ acetic acid solution for 10 min, dried and then stained with the supplied Paragon Violet (C.I. Acid Violet 17). As haptoglobin and haemoglobin run in the a region, 3' region bands on the agarose gels were presumed to be IgG and so indicative of MS.
Silver stain A silver precipitation technique (Lubahn and Silverman, 1984) was used to stain duplicate agarose gels in which 5/~1 of unconcentrated CSF was run. The method requires fixing, drying and treatment of the agarose with Triton X-100 and dithiothreitol before staining with a silver nitrate suspension. Densitometry 28 of the agarose gels were examined on an Appraise densitometer (Beckman Instruments) and the number of gamma region peaks counted in accordance with manufacturer's instructions.
Agarose Superfine SPE II agarose gels were kindly supplied by Beckman Instruments and run according to manufacturer's instructions. Minicon CS15 concentrators were used to concentrate 2.5 ml of CSF 20-fold in one step. 5 /~1 of concentrate were
Results
43 polyacrylamide gels showed T4, T5 staining consistent with MS; a further 14 showed patterns consistent with intrathecal synthesis of unknown
TABLE I C O M P A R I S O N OF R E S U L T S OF E L E C T R O P H O R E S I S ON P O L Y A C R Y L A M I D E A N D A G A R O S E A N D IgG : A L B U M I N RATIOS F O R M A T C H E D CSF SAMPLES F R O M 120 PATIENTS C L I N I C A L L Y S U S P E C T E D O F MS Findings on polyacrylamide gel
Findings on matched agarose gel
no. of samples with IgG : Alb with > 28% 23%-28%
Total
Cathodal Ig bands Diffuse cathodal Ig
3' bands present
10 5
1 1
9 2
20 8
Cathodal Ig bands Diffuse cathodal Ig
3' bands absent
1 1
1 3
5 4
7 8
Anodal Ig bands
3' bands present 3' bands absent
1 0
1 0
4 8
6 8
N o t consistent with intrathecal synthesis
3, bands present T bands absent
1 0
2 3
3 54
6 57 120
< 23%
121 significance. A total of 57 of 120 CSF samples run in polyacrylamide stained for IgG bands in the 3' region (Table I). In contrast, IgG bands were only demonstrated in 40 of the agarose gels from the same CSF samples, y region bands were found in the agarose gels without being present in the 3' region of the corresponding polyacrylamide gels in six cases. Almost half of the samples (42%) showing cathodal IgG on polyacrylamide also had raised IgG: albumin ratios. Only one sample out of 63 with negative staining in polyacrylamide and only one of 14 with anodal IgG staining had a raised IgG : albumin ratio. The silver precipitation technique was used to stain 24 unconcentrated samples run on agarose. Seven of these samples were previously positive when concentrated and electrophoresed on agarose and stained with acid violet. Only four of these seven previously positive samples showed T bands when stained by the silver method; the poor sensitivity of the silver method was due to high background staining. The banding patterns seen with the four positive samples was identical to that seen when run previously after concentration. Silver staining did not reveal any band not seen previously on conventionally stained gels.
Discussion 15 samples run on agarose failed to show ~, bands whereas results of the same samples run on 7% polyacrylamide were consistent with MS. Although the agarose method is easy and relatively quick to use, it is not sensitive enough for general use, it appears not to have the resolution of polyacrylamide technique. In our hands, silver staining did not result in more gamma bands being detected when used in conjunction with the agarose method. The relatively poor rate of detection of oligoclonal IgG bands in agarose may partly be due to loss of IgG during the concentration step. The identical banding patterns obtained by the concentrated/acid violet-stained method and by the unconcentrated/silver-stained method shows that the electrophoretic property of the immunoglobulin bands is not changed by concentration; the loss of sensitivity cannot be due to bands appearing in different regions in the gel as sug-
gested by Glasner (1978). This was a technical comparison of two methods, but the presence of bands on six agarose gels which were not apparent on polyacrylamide prompted us to examine the clinical history of the patients concerned. Three of these six patients had symptoms clinically suggestive of MS but the IgG: albumin ratios were in the doubtful or normal range. Of the other three one had symptoms of peripheral sensory loss and a raised I g G : a l bumin ratio (33%) and the remaining two had symptoms of dementia or myelopathy and normal CSF ratios. Oligoclonal IgG bands have been demonstrated by agarose electrophoresis in CSF from patients with all three conditions (Gerson et al., 1981). Why these bands appeared on the agarose gels and not on polyacrylamide is not clear; they may be non-IgG bands but unfortunately insufficient sample remained for immuno-fixation.
Acknowledgements We would like to thank Mrs. J. Wilkinson for the preparation of the polyacry!amide gels; Mrs. J. Goble for typing the manuscript and Miss Elizabeth Murphy of Beckman Instruments Inc. for kindly supplying equipment and materials.
References Gerson, B., Cohen, S.R., Gerson, I.M. and Guest, G.H. (1981) Myelin basic protein, oligoclonal bands, and IgG in cerebrospinal fluid as indicators of multiple sclerosis. Clin. Chem. 27, 1974. Glasner, H. (1978) Immunologicalmethods in multiple sclerosis. J. Neurol. 218, 73. Johnson, M.H. and Thompson, E.J. (1982) Measurement of body fluid proteins by polacrylamidegel electrophoresis.J. Clin. Pathol. 35, 1328. Link, H. and Muller, R. (1971) Immunoglobulins in multiple sclerosis and infections of the nervous system. Arch. Neurol. 25, 326. Lubahn, D.B. and Silverman, L.M. (1984) A rapid silver-stain procedure for use with routine electrophoresisof cerebrospinal fluid on agarose gels. Clin. Chem. 30, 1689. Roberts-Thompson, P.J., Esiri, M.M., Young, A.C. and Maclennan, I.C.M. (1976) Cerebrospinal fluid immunoglobulin quotients, kappa/lamdba ratios, and viral antibody titres in neurologicaldisease. J. Clin. Pathol. 29, 1105.
122 Thompson, E.J. (1977) Laboratory diagnosis of multiple sclerosis: Immunological and Biochemical aspects. Br. Med. Bull. 33, 28. Thompson, E.J. and Johnson, M.M. (1982) Electrophoresis of CSF protein, Br. J. Hosp. Med. 28, 600.
Thompson, E.J., Kaufman, P., Shortman, R.C., Rudge, P. and Mcdonald, W.I. (1979) Ohgoclonal immunoglobulius and plasma cells in spinal fluid of patients with multiple sclerosis. Br. Med. J. 1, 16.