63

Clinica Chimica Actu, 88 (1978) 63-70 @ Elsevier/North-Holland Biomedical Press

CCA 9499

FURTHER OBSERVATIONS ON ISOELECTRIC FOCUSSING OF SERUM PROTEINS USING MODIFIED CELLULOSE ACETATE GEL MEMBRANES, AND DIRECT ISOENZYME STAINING

JEFFREY Clinical

AMBLER

Chemistry,

(Received

February

General

Hospital,

Nottingham

NGl

6HA

(U.K.)

17th, 1978)

Summary 1. Treatment of cellulose acetate gel strips by boron trifluoride in methanol has been shown previously to produce a material suitable for isoelectric focussing. Further observations are reported here on the changes in the strips induced by the reagent. It has shown by conventional protein electrophoresis, for example, that not only is the degree of electroendosmosis reduced, but also that other properties are altered. 2. Minor modifications are introduced for the general isoelectric focussing method including improved electrode solutions. 3. Isoenzyme detection following isoelectric focussing of serum proteins is discussed in the light of experience gained using alkaline phosphatase and lactate dehydrogenase as examples.

Introduction Separation of charged species by analytical isoelectric focussing is usually carried out in polyacrylamide gel rods [l] or slabs [ 21. In a recent publication [3] a method was described for the treatment of cellulose acetate gel membranes (Cellogel@) by boron trifluoride in methanol to make it suitable for low voltage isoelectric focussing. The resulting inert material was ideal because of its technical ease of handling, and the low cost of apparatus and ampholytes. The modified material was also suited to the direct specific detection of proteins in a complex mixture and a method of immunofocussing was described where proteins were fixed and identified by specific antisera. Direct specific protein staining is now further extended to include isoenzymes. This type of identification has three general difficulties: an increased sample size must be applied to increase the sensitivity, the proteins must be fixed to the strip while the pH gradient is replaced by a uniform buffer at the optimum pH, and the

64

enzyme may need reactivating. These points are fully discussed allow a wider application than the limited examples given here.

in order

to

Materials and methods The electrophoresis tank, cooling platen and pre-treatment of the cellulose acetate gel have been described [3]. In summary, 25 Cellogel strips (5.4 X 14 cm) were equilibrated with absolute methanol before being placed in 150 ml 4.5% boron trifluoride in methanol at 45°C for 50 min. Excess reagent was washed out with methanol, and the strips were stored in this solvent. The strips were cut to 9 cm and equilibrated with ampholyte solution for at least 2 h before use. This was composed of the required pH ampholytes (Ampholines LKB) at 8% concentration in 5% glycerol. A lo-ml volume in a small equilibration box was sufficient for about 6 strips (5.4 X 9 cm). The strip was laid wet on the cooling platen and excess moisture was removed from the surface before the wicks (cellulose acetate gel strips dried by blotting) were overlaid from each electrode solution. The strips were then blotted quite heavily (Whatman No. 1 chromatography paper) before application of the samples. Some modifications have been made to the electrode solution recipes; all solutions now contain 5% glycerol and are composed as follows: pH 3.5-10, cathode 0.2 M ethanolamine, anode 0.2 M citric acid; narrow range gradients (e.g. pH 4-6, pH 5-S), cathode 0.2 M lysine (free base, Sigma Co.), anode 0.2 M acetic acid or 0.2 M citric acid. Samples (1 ~1) were applied using either a Cellogel@ micro or semi micro applicator, the anticondensation lid was placed over the strip and isoelectric focussing was carried out with cooling at 500 V for about 2 h. Protein staining Ampholytes were removed by washing in 100 ml 70% methanol in water for 20 min, and the strips were then shaken with 100 ml 0.15% Coomassie blue R (Sigma) in methanol/water/acetic acid (5 : 5 : 1, v/v) and excess stain was removed using the same solvent. Isoenzyme staining (i) Alkaline phosphatase. Serum samples were concentrated 4-fold using Lyphogel (Gelman Hawksley) and 1~1 applied on the same line at about the mid-point of a strip equilibrated with a pH 4-6 ampholyte gradient, containing 2.5 ml pH 4-6 and 0.5 ml pH 5-7 Ampholines (LKB) per 10 ml solution. Following isoelectric focussing, proteins were fixed by a 30-min wash in 100 ml fixing solution containing 50 ml saturated ammonium sulphate, 10 ml 1 M 2-amino-2-methylpropan-l-01 and 40 ml water. The staining solution comprised 14 mg sodium ol-naphthyl phosphate and 3 mg MgCl* in 100 ml of the fixing solution brought to pH 10.0 using 20% sodium hydroxide. The strip was stained in this solution for 2 h at 37°C before adding the colouring solution (10 mg 4-aminodiphenylamine diazosulphate (Sigma) in 100 ml 10 mM 2-amino-2-methylpropan-1-ol/HCl, pH 10.0). Excess stain was removed by washing with water. (ii) Lactate dehydrogenase. Serum samples were concentrated 4-fold using Lyphogel before application to the mid-point of a strip equilibrated with the

65

wide pH 3.5-10 gradient. Proteins were fixed by a 30-min wash in 100 ml 50% saturated ammonium sulphate in 0.1 M disodium hydrogen phosphate. The strip was then washed briefly (90 s) in 50 ml 0.1 M disodium hydrogen phosphate, pH 9.2 before being laid, absorbant face up in a small plastic tray. Excess moisture was removed and 10 ml staining solution in agar was poured on the strip and allowed to gel. The staining solution contained 5 ml aqueous 1% agar 1D (Oxoid) and 5 ml 0.2 M sodium lactate in 0.02 M disodium hydrogen phosphate containing 1 mg nitroblue tetrazolium (Sigma), 5 mg NAD (Sigma) and a small crystal of phenazonium methosulphate (Sigma). The enzyme reaction was incubated in the dark at 37°C for 1 h, and the strips were finally washed in 5% aqueous formaldehyde. Clearing the gel strips This may be accomplished for all the above methods using 30% diacetonyl alcohol in water and drying on a glass slide in an oven at about 70°C. Results and discussion 1. Investigation of the properties of the new isoelectric focusing material Methylation of cellulose acetate is necessary to produce a stable pH gradient during isoelectric focussing. Conventional electrophoresis has been carried out in barbitone buffer pH 8.6, I = 0.05, to investigate properties of Cellogel@ altered by the treatment. The same serum was applied on a treated and non-

Fig. 1. Zone electrophoresis of the same serum applied in three different positions (1, 2 and 3) on a methylated strip (left) and a non-treated strip. Separation was carried out in the same tank at the same time at 200 V for 1 h.

66

TABLE

I

RELATIVE STRIPS

ELECTROPHORETIC EITHER

TREATED

OR

MOBILITIES NOT

TREATED Position

strip

OF BY

PROTEINS BORON Length

ON

CELLULOSE

TRIFLUORIDE of

ACETATE

GEL

IN METHANOL

Distance

(cm)

from

origin

separation (cm) Treated

Distance

between zones

1

5.7

5.1

0.6

5.4

3.9

1.5

3

5.0

2.9

2.1

1

5.4

4.4

1.0

2

5.1

3.1

2.0

3

4.7

1.9

2.8

when

1

0.55

up (cm)

2

0.85

3

1.10

origins

are lined

y-Globulin

2

Non-treated

protein

Albumin

tail

treated strip in three different positions and was subjected to electrophoresis under identical conditions. The results are shown in Fig. 1 and Table I. Electroendosmosis is much reduced. There is decreased overall tailing of the y region on the treated strip. Comparison of P-globulins at position 3 on the two strips shows movement from the application point in different directions. Albumin has a faster mobility on the treated strip. Methylation may also affect properties other than electroendosmosis, the constant flow of buffer from anode to cathode. When the protein zones were lined up for corresponding positions on the two strips and the difference between points of application then measured, the difference in the anode region (1.10) was greater than in the cathode region (0.55) as shown in Table I. If only electroendosmosis were altered by methylation, a constant relative electrophoretic mobility value should be obtained, independent of the position of application. An explanation for the results might be that evaporation is reduced from the centre of the strip by the treatment. Unlike electroendosmosis, the direction of flow of buffer induced by evaporation is dependent upon position, being from the electrodes to the centre of the strip. Reduction of evaporation would lead to a reduced buffer flow from anode to the centre, and therefore a faster electrophoretic mobility would be obtained on the anode part of the strip. This would have significance in isoelectric focussing where more heat is generated than in electrophoresis. The mechanism by which evaporation is contained is not clear; however, it would appear not to be by a drastic reduction in the pore size of the medium since the electrophoretic mobility of large protein molecules, e.g. (Y2-macroglobulin, is not reduced compared to albumin. The new medium might be of value where low backward flow is desirable as in the Laurel1 rocket technique of electroimmunodiffusion. 2. General methodology (i) Methylation. The temperature and time that the strips are in contact with the reagent is important. At temperatures below 40°C and in less than 3% boron trifluoride in methanol, the strips were under-methylated and electroendosmosis occurred. More rigorous conditions than those specified resulted in

strips which were not suitable because of deformation and the production of a peculiar water-proofing effect which altered sample up-take. For the reaction, the strips should be contained in plastic boxes which are resistant to the reagent e.g. Cellogel@ staining trays, The strips should be kept flat at all stages of processing but especially when in the presence of hot methylating reagent; folds introduced at this stage cannot be removed. (ii) Ampholyte gradients and electrode solutions. New electrode solutions are recommended in this paper. The presence of 5% glycerol in the electrode solutions gave better overall results than without it. Lysine (free base) is a better cathode solution than ethanolamine for pH gradients up to pH 10.0, and especially below pH 6 where its use lessened the overheating effect caused by a large discontinuous pH drop at the junction of the wick and strip. Many narrow pY gradients have been tried and all have been successful. The most difficult gradient was pH 4-6; however, focussing could be accomplished by decreasing the effective size of the strip to about 8 cm by increasing wick overlap and focussing for 2.5 h. Although the ampholytes used in most of this work have been Ampholines (LKB), Servalyts (Serva) have also been used with equal success.

Fig. 2. Isoelectric focusing of five biological samples pH 3.5-10 (0.5 ~1 applied). 1 and 2, different IgG myeloma sera: 3, focusing marker mixture (bovine serum albumin PI = 4.93, P-lactoglobulin PI = 5.35 and 5.45, carbonic anhydrase principal pZ 6.16, sperm whale myoglobin PI = 8.27); 4, al-antitrypsin deficient serum; 5, urine containing Bence Jones protein (same patient as position 2).

68

Fig. 3. Densitometric scan of serum concentrated 4-f&d (Lyphogel) pH 3.5 and 10. The range PH 3.5-7 was chosen for scanning.

and isoelectric focussed between

(iii) Sample size, concentration and scanning. For qualitative purposes, one or two applications with the Cellogel@ micro applicator (0.5 ~1) were sufficient to produce clear protein zones. If the proteins of interest are in high concentration, e.g. myeloma proteins in serum, such small applications are mandatory, as shown in Fig. 2. However, in certain cases, it was necessary to increase the protein concentration to increase the sensitivity, for example in isoenzyme analysis, al-antitrypsin deficiency phenotyping, and in scanning cleared strips by densitometry. Concentration of proteins by Lyphogel was found to be ideal, since a calculated concentration could be obtained, and all proteins over 20 000 daltons were concentrated without a concomitant increase in salt concentration. Fig. 3 is a densitometric scan of a serum separation between pH 3.5 and 7, where the serum has been concentrated 4-fold using Lyphogel. 3. lsoenzyme methodology Staining isoelectric focussing separations for isoenzymes opens the way for studying the identification of specific proteins in specimens, for example, in forensic science. Although there are several problems to overcome in adapting isoenzyme methodology to separations on modified Cellogel, this medium is preferred to polyacrylamide gel [4] as it is a thin solid membrane, allowing easy diffusion of buffers and substrates. A general methodology for isoenzymes is given here, with some examples. fi) Concentration. Most isoenzymes require concentration to reach a satisfactory level of detection. The use of Lyphogel concentrates proteins without a concomitant increased concentration of salt. In Fig. 4, serum proteins have been concentrated 4-fold and isoelectric focussed in a pH 3.5-10 gradient. The strip has been cut down the middle and stained for lactate dehydrogenase on the left and proteins on the right. The protein staining shows that even at 4-fold concentration the protein zones remain relatively undistorted. (ii) ~emouaf of amp~o~~tes. The high concentration of ampholytes in the form of a pH gradient following focussing must ideally be removed and replaced by a buffer at the optimum pH of the enzyme. The isoenzymes must

69

obviously first be fixed without loss of activity and the most successful way of doing this, for the example given here, is to use a buffered 50% saturated ammonium sulphate solution. (iii) Removal of ammonium sulphate. This may not be necessary depending upon the enzyme. For example, alkaline phosphatase in the salted out form reacts with the detection reagents. If it is desirable to remove the large excess of ammonium sulphate this is accomplished by a brief (maximum 2 min) wash in the buffer at the correct pH. The proteins remain salted out during this wash and remain on the strip. If the enzyme retains its activity in the presence of ammonium sulphate, the final wash must be at the correct pH because ammonium sulphate effectively acts as one of a buffer pair. In the case of alkaline phosphatase the pH must be titrated to 10.0 using 20% sodium hydroxide solution. (iv) The enzyme reaction. This may be carried out in free solution where ammonium sulphate is present as is the case with alkaline phosphatase. If

1

2

3

Fig. 4. Sera concentrated 4-fold and isoelectric focussed pH 3.5-10. Left half stained for LDH activity, right half stained for proteins. 1, serum from patient with myocardial infarct: 2. serum from patient with myeloma; 3. serum (2) stained for proteins; 4, marker mixture (see Fig. 2). Fig. 5. Serum proteins concentrated C-fold, isoelectric focused phatase activity. 1. raised liver fraction: 2, raised bone fraction;

at pH 4-6 and stained for alkaline phos3, marginally raised liver fraction.

70

ammonium sulphate has been removed, it is better to apply the detection reagents in a 0.5% agar gel to stop diffusion of the isoenzymes from the strip and act as a sponge to soak up remaining ammonium sulphate, for example in the case of lactate dehydrogenase which is inactive when salted out. The alkaline phosphatase method is given as an example of reaction in free solution; however, the isoenzymes can be equally well detected after the removal of ammonium sulphate followed by a single combined substrate-dye reaction in 0.5% agar gel. Figs. 4 and 5 show the results obtained using the detection methods given here. Lactate dehydrogenase activity has been stained in Fig. 4 and this figure also demonstrates the added advantage of being able to stain lactate dehydrogenase on one half of the strip and for proteins on the other. An isoelectric focussing marker in position 4 allows an estimate to be made of the isoelectric points of the isoenzymes. The serum in position 1 was from a patient with a myocardial infarct and has raised LDH 1 and LDH 2 fractions. The serum separated in position 2 was from a patient with myeloma and shows an increase in LDH 5; in position 3 this serum has been stained for proteins. In Fig. 5, the separation has been stained for alkaline phosphatase activity. Good resolution is obtained between liver (positions 1 and 3) and bone, (position 2) isoenzymes. In sera examined here, only the slow liver and bone were detected, but in other cases a fast liver and intestinal isoenzymes have been detected. It is possible that a further aid to separation might be to include a surface active agent in the equilibration mixture as has been done with polyacrylamide gels [ 51. However, it is usually the presence of zones corresponding to the slow liver or bone which decides diagnosis for marginally raised alkaline phosphatase levels, and the use of Triton X does not appear to help greatly in separating these isoenzymes. Using the methodology outlined here and making modifications which other enzymes might demand, for example, different final ammonium sulphate concentrations and buffer pH values, it should be possible to detect many other isoenzymes of scientific interest. In a case where the isoelectric point of the isoenzyme is not too far removed from its optimum pH it may be possible to dispense with ammonium sulphate precipitation and carry out the enzyme reaction directly with buffered gelled reagents.

Ac,lowledgements I should like to thank Dr. G. Walker, head of department for helpful suggestions concerning this work. I should also like to thank Mr. Dallas Simpson and the Photography Department of the General Hospital for preparation of the photographs. References 1

Wrigley,

2

Vesterberg.

C.W.

3

Ambler,

4

WadstrGm,

5

Angellis,

(1968)

0.

J. (1978) T. and D.,

Science

(1972)

Inglis,

Tools

Biochim.

Clin. Smyth. N.R.

Chim.

Acta

C.J. and

15.17

Biophys.

Acta

257.

11

85.183-191

(1973)

Fishman,

Science W.H.

Tools (1976)

20,17 Am.

J. Clin.

Pathol.

66,

929