I. EXPERIMENTAL TECHNIQUES ZONE ELECTROPHORESIS IN FILTER PAPER AND OTHER MEDIA BY ARNE TISELIUS Biokemiska Institutionen, Uppsala Universitet, Uppsala, Sweden Received 8th May, 1952 Zone electrophoresis in immobilized media (filter paper strips or powder, glass beads, starch, gels, etc.) has several advantages over boundary electrophoresis as a separation method. Completely separated zones may be obtained, and it is easy to work on a micro- scale and at low concentrations. The method is suitable also for the separation of many low molecular weight substances, like amino acids, peptides, nucleotides, etc. On the other hand electro-osmosis and adsorption phenomena interfere. There are certain ways of minimizing the influence of these and it is shown that, under suitable conditions, zone electrophoresis measurements may be used for determination of mobilities and isoelectric points.Electrophoresis of proteins and other biochemically important substances of large molecular weight is usually performed in free solution in a suitable U-tube apparatus. Observation of electrophoretic migration and separation (electro- phoretic analysis) in this modification of the method is essentially an observation of the migration of boundaries by optical or analytical methods. The separation obtained is a boundary separation of overlapping zones, not a complete separation into zones of different migration. Such a separation generally cannot be achieved in free solution, as convection due to gravity would upset the zones ; this requires some sort of an immobilized medium like glass powder, sand, various gels or filter paper, which were sometimes used in early work on ionic migration (Lodge, Hittorf).Several workers in this field have recently attempted to use such arrangements especially for preparative work, where accurate definition of mobilities is not so essential. Already in 1907 Field and Teague 1 observed the migration of diphtheria toxin and antitoxin in agar jelly. Ionic migration experiments in gels were performed by Kendall et al.2 for the separation of rare earths. Coolidge in 1939 separated serum proteins into zones in an electrophoresis tube med with ground glass ~001.3 Consden, Gordon and Martin in 1946 separated amino acids and peptides by migration through a slab of silica gel (ionophoresis).4 They also introduced an ingenious method of localizing the zones by taking a print of the surface of the gel on a sheet of filter paper which was subsequently stained with ninhydrin.Other modifications have recently been used by Butler and Stephen 5 by Gordon, Keil and Sebesta,6 by Haglund and Tiselius 7 and many others. Separations of isotopic ions have been achieved by Brewer, Madorsky and Westhower in layers of sand or glass powder.* One should also mention attempts to arrange zone electrophor- esis in free solutions by using a large number of U-tubes or other elements, coupled in series. Unless the units are very small and their number quite large, separations in such apparatus will tend to become less distinct, as convection currents are set up within each unit.They have, however, been used with advantage in some in- vestigations.9 Recently filter paper strips have been widely used with notable 2930 ZONE ELECTROPHORESIS success as an immobilizing medium for electrophoresis of proteins and substances of lower molecular weight. This has the advantage of providing a micro-method and requires only a very simple apparatus. Contributions to the development of this modification have been made by, for example, Wieland and Fischer,lo Turba and Enenke1,ll Durrum,l2 Cremer and Tiselius,l3 Kunkel and Tiselius 14 and many others. Obviously the great success of filter paper chromatography has been a source of inspiration in this work. We have proposed the term zone electrophoresis for the type of experiment in which a separation into zones is achieved, whereas the common electrophoresis in U-tubes with free solutions may be called boundary electrophoresis.* It is perhaps not generally realized that the advantages of zone electrophoresis are not only the possibility of complete separation, fairly easy isolation of the fractions and microscale operation ; in addition, much lower concentration can be used than in free boundary electrophoresis, where a certain minimum concen- tration is required for the stability of the boundaries against convections.Con- sequently the so-called " boundary anomalies " can be largely eliminated. These effects, which in boundary electrophoresis usually mark their presence by a pro- nounced difference in the sharpness and rate of migration of the boundaries in the two limbs of the U-tube, may cause grave errors and even make electrophoretical determinations impossible.They depend essentially on that contribution to the total conductivity of the solution which is due to the migrating substances (pro- teins, etc.) studied. With proteins-for a given concentration-the effect is usually small ; with amino acids, for example, it may be so large that it virtually makes the experiment meaningless. Amino acids, peptides and many other low molecular weight substances can, however, easily be studied electrophoretically in zone electrophoresis. If the con- centration is low, the zones are not spread out too much and a complete separation can be achieved. This was first made clearly evident in the work of Consden, Gordon and Martin.3 The advantages of the zone method over the boundary method have, however, to be gained at the cost of less well-defined conditions of migration in other respects, especially with regard to the accurate definition of the mobilities.There are three main factors to be taken into account, all of which depend upon the properties of the immobilizing medium used. First there is the purely geometrical effect of the structure of the medium, which makes the charged particles and the electric current travel in a manner which deviates considerably from the conditions in a homo- genous medium. Secondly, there is the possible interaction (by adsorption or other effects) between the medium and the migrating substances. And thirdly, we always have in media of this kind an electro-osmotic transport of the whole solution, superimposed upon the migration.As the migration is observed, relative to the solid medium (for example, filter paper) and we want to know the mobility relative to the solution, a correction term must be introduced. One also should pay attention to the reaction changes occurring at the interface between the ends of the filIing-medium column (or sheet of filter paper) and the buffer solution, which changes are analogous to membrane polarization. Thus the choice of filling medium is of great importance. Small particle size is favour- able from point of view of stabilizing the zones but will, on the other hand, tend * It may be qiiestionable to extend the use of the term electrophoresis, originally applied to colloids, proteins and other substances of large molecular weight, to the migration of substances of low molecular weight.However, as the experimental arrangements are essentially the same in both cases if a separation is the main purpose, it seems impractical to distinguish between electrophoresis, ionophoresis or electromigration and it would be preferable to use a common term. In any case these terms do not seem suitable for distinguishing between zone and boundary methods. The author is perhaps somewhat conservative in using the term electrophoresis. We notice the same difficulty in chromato- graphy, which nowadays is not limited to coloured substances by far.FIG. I a.-Apparatus for filter paper electrophoresis (Kunkel and Tiselius). (a) coiled platinum electrodes ; (b) glass wool at the juncture between electrode chambers ; (c) highly porous paper carrying liquid from electrode vessels to the filter paper ; (d) glass plates surrounding the filter paper.FIG. Ib.-Separation of a mixture of (left to right) crystalline lysozyme, purified myclorna protein, crystalline lacto to globulin and crystalline human serum albumin by paper electrophoresis ; staining with bromphenol blue. The line across the strip indicates the position of the original spot (Kunkel and Tiselius). [To face page 31ARNE TISELIUS 31 to increase adsorption effects. With gels, which would represent an extreme case, the resistance against the migration of large molecules due to adsorption or mechan- ical resistance may be appreciable. Thus silica gel is not suitable for proteins, whereas dilute agar jelly offers surprisingly small resistance to the migration of proteins of moderately high molecular weight.6 It is, however, difficult to avoid contamination by the agar when one attempts to elute the protein.Most investigators using these methods have been chiefly interested in separa- tion for preparative or analytical purposes or, in some cases, even for clinical diagnostics (blood plasma), It is important to take note of the factors mentioned above in these applications, but it is still more important when one attempts a quantitative interpretation of mobility. This problem is of considerable interest if one wanls to make use of zone electrophoresis for identification of components and determination of their electrophoretic properties in the same way as in ordinary boundary electrophoresis in free solution.Some results, obtained in filter paper electrophoresis by Kunkel and the author,l4 may be quoted as typical. The simple apparatus used, made of Perspex, is shown in fig. 1 . The strip of filter paper (Munktell no. 20 was found particularly useful) is enclosed between two thick plane glass plates, which are firmly clamped together. The surfaces received an anti-wetting treatment (Dow Coming Silicone 200, or a touch of silicon grease) and the edges are sealed with silicon grease. The ends of the strips are pressed against pads of filter paper which are partially immersed in the electrode vessels containing buffer solution. The technique for staining the strips with brom- phenol blue after the conclusion of an experiment is essentially that of Durrum.12 The " hanging strip " used by this author appears less suitable for mobility studies as it is difficult to obtain constant potential gradient along the strip, the quantity of buffer solution taken up by the filter paper being smaller at the top than at the bottom.Analysis of the protein contents of the different zones can be made by cutting the strip into segments and applying a micro-colorimetric method for protein,l4 or by evaluating the colour intensity of the bromphenol blue (or other suitable staining reagents) photometrically. Special photometers have been con- structed for this purpose. As the affinity between the dye and different proteins may vary considerably, great care has to be exercised in such procedures, however, even though they are very tempting to use because of convenience.Table 1 and fig. 2 shows the results from an experiment on normal human serum. Many similar experiments have been published during the last few years. TABLE 1 Comparison between the percentage composition of the various components of a normal serum as calculated from the areas under the curves obtained by paper electrophoresis and free electrophoresis. No attempt was made to correct the dye values for differences in affinity with the different components. (Experiments by Kunkel and Tiselius, J . Gen. PhysioE., 1951, 35, 89.) composition % method - alb. a1 a2 B Y mod. Folin method 52.9 6.3 12.2 10.1 18.5 paper electrophoresis- dye elution method 65.9 3.9 8-0 9.0 13.1 free electrophoresis- ascending 53.33 6.21 10.87 13-79 15-62 descending 52-06 5-28 11-26 14.66 16.74 For mobility determinations it is essential first to correct for the electro-osmotic flow.This can easily be done by introducing into the mixture to be studied a small amount of a sub.-tance of known mobility, preferably one with zero mobility.32 ZONE ELECTROPHORESIS As such an “ index substance ” we have used dextran which has the advantage of being uncharged and is also easily detected with the bromphenol blue reagent. If distances of migration are measured from the index spot, the effect of electro- osmosis is eliminated. There may, however, be disturbances due to this pheno- menon, especially at low buffer concentrations, if electro-osmosis sets in at the glass plate surfaces, and it is necessary to press the glass plates firmly against the strip to avoid trouble of this kind.The geometry of the migration can also be dealt with fairly simply.14 The migrating substances and the current follow the same somewhat crooked path through a porous medium, the length of which I’ is greater than the length of the strip 1. Thus an observed migration of d cm in the strip corresponds to an actual migration of Vd/l cm at a potential gradient of ljl’ . v/l and the apparent mobility U,, as observed on the strip after correction for electro-osmosis, is related to the real mobility U in the pores by the expression u = (l/Z’)2Ua. 0 ? FIG. 2.-Mobility values plotted against pH for human serum albumin on filter paper (O), and in free solution ( x ), The individual points (0) represent dextran mobilities (Kunkel and Tiselius).The factor l/Z’ can be determined by measuring the resistance of the strip when it has taken up a known weight of a standard conductivity solution. If both the electro-osmotic and geometrical corrections are thus applied, mobilities and iso- electric points agree remarkably well with those obtained in free solution (fig. 2). Adsorption phenomena which cause tailing, and sometimes even complele immobilization of the zones, are probably the greatest difficulties in this sort of work. Various qualities of filter paper differ in this respect (it is recommended that electro-dialyzed paper be used). Some proteins show rather low adsorption (e.g.serum proteins) but others, especially those which carry a strongly positive or negative charge, may be tenaciously held. Porath and Flodin in the author’s laboratory have had some success in using for such substances filter paper which by suitable pretreatment has been given a marked positive or negative charge (unpublished experiments). Columns of small, spherical glass beads or of ionic exchange resins (Dowex) have also been found useful in such cases. In the separation of larger quantities, it is convenient to substitute the paper strip in the apparatus fig. 1 with a rectangular trough containing, for example, filter paper powder. Such an arrangement is used in the author’s laboratory forARNE TISELIUS 33 analysis of mixtures of the basic amino acids.For continuous separation of still larger quantities, arrangements based upon a combination of vertical flow and horizontal electrophoresis in a glass powder layer or in a filter paper sheet have been worked 0~t.15~16 If gels are used as immobilizing media it appears possible to make use of the increasing resistance to migration of substances of increasing molecular weight for separation purposes.6917 Electro-osmosis will tend to transport all substances, charged or uncharged, through the gel and a kind of " zone ultrafiltration " will result, which appears to offer interesting applications. Zone electrophoresis in immobilized media is rapidly gaining importance, and has some obvious advantages, and in particular, the possibility of working with small quantities and of obtaining complete separation.Only very simple apparatus is required and yet, under suitable conditions, it is still possible to make quanti- tative measurements. The method is rapidly coming into general use in biochemical and clinical laboratories for the study of proteins, nucleic acids, nucleotides, polypeptides, peptides, amino acids and many other types of substances. Also in enzyme chemistry it offers great promise. As compared to chromatographic procedures its most valuable field of application is for substances of comparatively large molec- ular weight, but there are of course many cases where both methods may be used and where each supplements the other. 1 Field and Teague, J. Expt. Med., 1907, 9, 86, 225. 2 Kendall, Jette and West, J. Amer. Chem. SOC., 1926, 48, 31 14. 3 Coolidge, J. Biol. Chem., 1939, 127, 551. 4 Consden, Gordon and Martin, Biochem. J., 1946, 40, 33 ; 1947,41, 590. 5 Butler and Stephen, Nature, 1947, 160,468. 6 Gordon, Keil and Sebesta, Nature, 1949, 164,498. 7 Haglund and Tiselius, Acta Chem. Scund., 1950, 4, 957. 8 Brewer, Madorsky and Westhover, Science, 1946, 104, 156. 9 Reference is made to a review by Svensson, Advances in Protein Chemistry, voI. 4, 10 Wieland and Fischer, Naturwiss., 1948, 35, 29. 11 Turba and Enenkel, Naturwiss., 1950, 37, 93. IZDurrum, J. Amer. Chem. SOC., 1950, 72, 2943. 13 Cremer and Tiselius, Biochem. Z., 1950, 320, 273. 14 Kunkel and Tiselius, J. Gen. Physiol., 1951, 35, 89. 15 Svensson and Brattsten, Arkiv Kemi, 1949, 1, 401. 16 Grassmann and Hannig, 2. angew. Chem., 1950, 62, 170. 17 Synge and Tiselius, Biochem. J., 1950, 46,41. pp. 252-277 (Academic Press, Inc., New York, 1948).