J Clin Pathol: first published as 10.1136/jcp.2.3.161 on 1 August 1949. Downloaded from J. clin. Path. (1949), 2, 161 SERUM PROTEINS: A Review BY J. R. MARRACK AND H. HOCH From the Department of Chemical Pathology, London Hospital Medical College (RECEIVED FOR PUBLICATION, JuNE 3, 1949) CONTENTS Introduction ... ... ... ... ... 161 IV. Composition, Physical, and Physiological Properties ... ... ... ... 168 Composition L. Electrophoresis: Method and Interpretation 162 Molecular Weights Immunology H. Fractionation by Precipitation Methods ... 163 Physiological Properties Salting Out V. Serum Proteins in Disease ... 173 Low Salt Concentrations Response to Injury and Infection Quantitative Methods Effect of Deficiency of Protein Low-Temperature-Low-Salt-Low-Dielec- Lipaemia Liver Disease copyright. tric-Constant Fractionation Myelomatosis Miscellaneous Diseases III. Normal Concentrations of Serum Proteins 167 Other Abnormal Proteins Total Protein VI. Flocculation Reactions .... ... 185 Electrophoretic Fractions Erythrocyte Sedimentation Rate Serum Proteins in Pregnancy and Infancy Discussion ... ... ... ... ... 187 http://jcp.bmj.com/ For many years it has been realized that ser detected. Cohn and his colleagues at Harvard globulin is not a single homogeneous protein, have studied the factors that affect the solubility comprises a variety of proteins which have 4 of proteins. On the basis of these studies they ferent physical, chemical, and physiological p have developed a method of separation of perties. The salt fractionation methods that h proteins by independent variation of these been 'used for many years do not yield fracti factors. This method has revealed the presence which are sharply distinguished one from anot in serum of a great variety of distinct proteins. on September 27, 2021 by guest. Protected in any properties. A striking example is that It has now been used extensively and on a large the fractionation of diphtheria antitoxic ser scale for the preparation of active fractions rela- with ammonium sulphate; antitoxin is found tively free from other protein. Methods have a series of fractions ranging from that precipita been devised by which the molecular weights of between 30 and 35% saturation to that E protein molecules can be calculated and tentative cipitated between 52 and 56% saturation (E estimates can be made of their shapes and sizes. and Glennie, 1931). The separation of fractions by electrophoresis During recent years new methods have b is based on one property only, that is, mobility used for the separation and study of the E in an electric field. The molecules contained in teins of serum and plasma. In particular a fraction which, on the 'basis of mobility, is moving boundary method of electrophoretic an: homogeneous, may have different shapes, weights, sis developed by Tiselius (1937) raised great hol and sizes, different chemical compositions and dif- for by this method at least five fractions of ( ferent physiological properties. The globulin frac- ferent mobility can be recognized in human seru tions separated by Cohn's methods and, still more, the mobilities of these fractions are distinct E those separated by salt fractionation, contain pro- little material of intermediate mobility can teins of different mobilities. The grouping of 0* J Clin Pathol: first published as 10.1136/jcp.2.3.161 on 1 August 1949. Downloaded from 162 J. R. MARRACK AND- H. HOCH serum globulins, therefore, depends on the choice of property on the basis of which they are classified. K The mobilities of the components of the frac- tions prepared by Cohn's methods have been measured. A very large number of abnormal sera have been studied by electrophoresis; from K+L :1 these results it is possible to draw conclusions as ., to the changes of electrophoretic pattern that may 1+ be expected in various types of disease, and the significance of these changes. Few abnormal sera have been studied by the other new methods. We will, therefore, use the electrophoretic patterns as I~~ ~ ~ ~ ~K+L+M the basis of discussion in this article. Comprehensive reviews of investigations into the I. 33 concentration, distribution, and significance of pro- teins in serum have been published by Janeway I X ~I M I (1943), Stern and Reiner (1946), Luetscher (1947), gL+M, and Gutman (1948). 1. Electrophoresis: Method and Interpretation The object of electrophoresis is to demonstrate the a b presence of constituents which have different mobilities FIG. 1 Distribution of three components, K, L, and and to measure the relative concentrations of these M, of different mobilities: (a) Before electro- constituents. The conditions of pH and salt concen- phoresis; (b) after electrophoresis. tration influence the results and they have to be copyright. specified. The range of pH commonly used for during the passage of the current, the U-tube is plasma or serum is between 7.7 and 8.6; in this pH immersed in a water-b'ath of a temperature near the range all serum proteins carry negative charges and density maximum of the solution (0-4' C.). therefore move towards the anode. The salt concen- The methods of observing the number and positions tration usually chosen is equivalent to an ionic of the boundaries and of measuring the relative strength of 0.1 or 0.2.* The experiments are made concentrations of the fractions are based on the in the special apparatus designed by Tiselius which change of refractive index produced by the changehttp://jcp.bmj.com/ consists of a U-tube with a rectangular cross section, in the concentration of protein at the boundary. On the upper enas of which are connected to two elec- account of this change of refractive index a horizontal trodes. The solution to be analysed is run into the ray of light passing through the tube is deflected bottom part of the U-tube, buffer solution is placed downwards. The extent of the deflection depends on on it and in the remaining space, including the the rate of change of the refractive index. In the electrode vessels. space immediately above and below the boundary the At the beginning there is one boundary between the concentration does not change and a ray of light pass- protein solution and the supernatant buffer in each ing through the tube is not deflected. In Thovert's on September 27, 2021 by guest. Protected limb of the U-tube (Fig. la). On passing an electric cylindrical lens (Thovert, 1914; Philpot, 1938; current these two boundaries move away from the Svensson, 1939) and in Longsworth's Schlieren scan- original positions at velocities equal to the velocities ning (1939) methods the vertical deflection is trans- of the protein ions below these moving boundaries. lated by optical and mechanical devices into a lateral If several protein constituents of different mobilities deflection which is recorded on a screen. The are presetit the original boundaries will split into boundary appears on the screen as a peak. On several boundaries moving with different speeds. In migration the sharpness of the peak changes. Vari- addition to these moving boundaries there are two ation in conductivity anrd in pH have each the effect almost stationary boundaries, 8 and e, which are due of sharpening the peak in one limb and of broaden- to changes in conductivity at the positions of the ing it in the other. These effects may be in the oppo- original boundaries (Fig. 1). Before it is put in the site or in the same direction. Spreading by. diffusion U-tube the protein solution is dialysed against the is superimposed in both limbs. If a boundary peak buffer. In order to reduce to a minimum convection remains single during migration over a long distance currents arising from differences in temperature within it is usually inferred that the protein is electro- the U-tube between the inner and the marginal parts phoretically homogeneous. But this need not neces- sarily be the case, since a mixture of similar proteins, *The ionic strensth is defined by j (%+c . the mobilities of which are distributed in a smooth where c1,c,,. are the ion concentrations and z2,z,.. frequency curve, can also show a single peak after are the valenI J Clin Pathol: first published as 10.1136/jcp.2.3.161 on 1 August 1949. Downloaded from SERUM PROTEINS 163 migration over a long distance. Criteria to define the -y,-globulin and lies between the /3- and y-globulin. the degree of non-homogeneity have been proposed. The concentrations of these components are propor- (For literature see Alberty, 1948.) tional to the areas (measured with a planimeter) under Fig. 2 shows the concentration distribution along these peaks; the relative concentrations are calculated the vertical direction in a boundary. The difference from the relative proportions of these areas, and, if between the concentrations at A and that at A' is the total concentration of protein in serum has been proportional to the difference between the refractive determined, the absolute concentrations of these frac- indices at A and that at A'. This can be shown to be tions can be calculated.* This simple statement needs proportional to the area under the peak ADA'. When some qualification. In the first place it assumes that the protein solution used is a solution of serum pro- the rate of change of refractive index with change of C concentration (specific refractive increment) is the same for each protein and that no gradients of refrac- tive index are produced across the boundaries by concentration gradients of other substances. In the second place, the molecules which move in groups and cause the changes of refractive index are not simple proteins, in the sense of compounds built of amino-acids, but are aggregates of protein, carbo- hydrate, and lipids more or less firmly united. The assumption that the specific refractive incre- ment is the same for each protein is approxi- mately correct if the concentration of " protein" used as a basis of calculation is the weight of the total protein-lipid-carbohydrate aggregate that moves A' A X as a unit.