- Email: [email protected]
Studies on the binding of haemoglobin by haptoglobin using electrofocusing and gradient electrophoresis
Biochimica et Biophysica Acta, 420 (1976) 57-68
© Elsevier Scientific Publishing Company, Amsterdam- Printed in The Netherlands BBA 37227 STUDIES ON T H E B I N D I N G OF H A E M O G L O B I N BY H A P T O G L O B I N U S I N G E L E C T R O F O C U S I N G A N D G R A D I E N T ELECTROPHORESIS
ALAN V. EMES, ALBERT L. LATNER, JOHN A. MARTIN* and FIONA A. MULLIGAN University Department of Clinical Biochemistry, Royal Victoria Infirmary, Newcastle upon Tyne, NEI 4LP (U.K )
(Received July 14th, 1975)
SUMMARY 1. Gel electrofocusing followed by gel gradient electrophoresis separated the haptoglobins and their complexes with haemoglobin into characteristic two-dimensional patterns of protein bands. 2. Molecular weights of 107 000, 139 000 and 168 000 were obtained for the three bands seen after a purified preparation of haptoglobin type 1 was partially saturated with haemoglobin. This indicated that free haptoglobin, the intermediate haptoglobin-haemoglobin complex containing one half-haemoglobin and the saturated complex with two half-haemoglobins were present. 3. The three proteins showed considerable microheterogeneity and gave a number of isoelectric points in the pH ranges 4.58-4.77, 5.20-5.40 and 5.74-5.93, free haptoglobin type 1 being the lowest group. These ranges were all 0.15-0.30 pH units lower if other values were taken for the isoelectric points of markers used to calibrate the pH gradient. 4. All three proteins were present over a wide range of haemoglobin concentrations, from 0.5 ~ to 92 ~o of that required for saturation. This would be expected if both binding sites have similar affinities for haemoglobin.
INTRODUCTION The haptoglobins have been extensively investigated because of their ability to form stable complexes with haemoglobin [1]. There have been several reports concerning the number of binding sites available [2] and the reactive unit of haemoglobin which takes part in complex formation with haptoglobin [3, 4]. Both free haptoglobin and the saturated complex show considerable microheterogeneity [5]. We have found that a combination of gel electrofocusing and polyacrylamide gel gradient electrophoresis separates the haptoglobins and their complexes into well-defined patterns of protein bands. We now report the use of these two separation techniques, both independently and combined as a two-dimensional procedure. They have provided further information on the binding process and the complexes, mainly * Present address: Department of Biochemistry,Bolton Royal Infirmary, Bolton, Lancs., U.K.
58 with purified haptoglobin type 1. The notation for the haptoglobin types is that used by Sutton [1]. MATERIALS AND METHODS
Electrofocusing and electrophoresis Gel electrofocusing followed by electrophoresis through a polyacrylamide gel gradient was performed as described elsewhere [6]. The electrophoresis stage was also carried out alone to give a one-dimensional separation. The gels were stained for protein with Coomassie Brilliant Blue G [7] and for haemoglobin with benzidine [8]. For the determination of molecular weights, polyacrylamide gradient gels were calibrated using haemopexin, transferrin, yeast alcohol dehydrogenase (EC 1.1.1.1), glutamate dehydrogenase (EC 1.4.1.3) and az-macroglobulin as marker proteins of known molecular weight [9, 10]. These were identified by comparison with published data [11] and after location of the enzymes using a reagent mixture based on those given by Shaw and Koen [12]. This contained 0.5 M KH2PO4, Na2HPO4 buffer, pH 7.0 (20 ml); 95 % (v/v) ethanol (5 ml); 1 M sodium glutamate in buffer, pH 7.0 (20 ml); NAD (10 mg); MTT tetrazolium [3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide] (5 mg); phenazine methosulphate (2 mg); water (20 ml). A linear relationship between log molecular weight and log migration distance [13] was obtained after electrophoresis for 48 h. Electrofocusing was also carried out in thin-layer polyacrylamide gels containing carrier ampholyte mixtures designed to produce more or less linear pH gradients over the pH ranges 3.5-9.5 or 2.5-6 [14]. For the pH 3.5-9.5 range, it was possible to reduce the recommended carrier ampholyte concentration by a half without noticeable effect. The Multiphor apparatus and Type 3371C power supply (LKBProdukter AB, S-161 25 Bromma 1, Sweden) were used. Gels were stained for protein with Coomassie Brilliant Blue R [14]. Ferritin, bovine serum albumin, fl-lactogloglobulin, carbonic anhydrase, equine and whale myoglobins and cytochrome c were used as pH markers. Linear pH gradients were obtained between about pH 4.5 and pH 9.5 using the isoelectric points given by Bours [15] or Radola [16], although there was a difference of 0.15-0.30 pH units between the two sets of values. Using the narrower range mixture of carrier ampholytes, the pH gradient was S-shaped rather than linear and became somewhat steeper around pH 5. It probably did not extend beyond pH 5.5.
Estimation of proteins Haptoglobin was estimated as its haemoglobin binding capacity [17]. Haemoglobin was obtained from normal human red blood cells by lysis with chloroform and measured spectrophotometrically [18]. Total protein was assayed using the Lowry method [19] except that reagent solutions were stored separately and mixed immediately before use; human serum albumin (Blood Products Laboratory, Lister Institute, Elstree, Herts., U.K.) was used as a standard.
Preparation of haptoglobin type l Human serum (10 ml) containing haptoglobin type 1 was mixed with 0.87 ml of acid citrate dextrose solution [20] and 10 ml of the mixture fractionated using Cohn's
59 method 10 [21]. We found that most of the haptoglobin appeared with albumin in Cohn fraction V. This fraction (78 ml) was concentrated to 24.5 ml by ultrafiltration [22] and dialysed against 0.08 M potassium acetate buffer, p H 5.7. The dialysis residue was subjected to chromatography on DEAE-Sephadex A-50, a m m o n i u m sulphate precipitation and gel filtration on Sephadex G-150, as described by Betlach and McMillan [23]. RESULTS
Haptoglobin type 1 preparation The purified preparation of haptoglobin type 1 contained 2.8 mg of protein/ml and 1.1 mg of haemoglobin binding capacity/ml. This represented 1.7 mg of haptoglobin/ml, assuming molecular weights for haptoglobin and haemoglobin of 100 000 [9] and 64 500 [24], respectively. The protein separation obtained from the preparation after electrofocusing and gradient electrophoresis is shown in Fig. 3A. There was some contamination with albumin and small amounts of other proteins can be seen, including the intermediate haptoglobin-haemoglobin complex.
Resolution of the haptoglobins and their haemoglobin complexes Gel electrofocusing of human serum followed by electrophoresis through a polyacrylamide gel gradient produced a characteristic two-dimensional pattern of separated proteins. In normal unhaemolysed serum, haptoglobin appeared as a single protein spot or as a single column of spots, depending on the haptoglobin type present A
B
Fig. 1. Separated proteins from 5/~1 of serum containing haptoglobin type 2, after electrofocusing followed by gradient electrophoresis. The gel was cut to give two identical slices; one was stained with Coomassie Brilliant Blue G (A) and the other with benzidine reagent (B). The electrofocusing gel, at the top, was positioned so that the pH gradient increased from left to right. Electrophoresis proceeded in a downward direction towards the anode. The diagonal rows of spots comprise 1, four proteins; 2, five proteins; 3, six proteins. All but the most acidic protein in each row contain haemoglobin.
60 [6]. In some sera, demonstrated to contain haemoglobin, large numbers of protein spots were resolved and these were arranged in diagonal rows rising towards the more basic end of the electrofocusing gel, that is, increasing in isoelectric point and molecular weight. Fig. 1 shows such a separation, in this case of a serum from a patient with an IgG myeloma and containing haptoglobin type 2. The gel was sliced to provide two identical slices; one was stained in Coomassie Brilliant Blue G (Fig. IA) and the other in benzidine reagent (Fig. 1B). The array of protein spots can be clearly seen, together with the large number shown to contain haemoglobin. The lowest diagonal row comprised four protein spots (barely visible), the next row consisted of five protein spots, the next of six spots and so on. All these proteins contained haemoglobin except for the spot with the most acidic isoelectric point in each row. The pattern of spots demonstrated in Fig. 1 most probably arose from partial saturation of the haptoglobin due to slight haemolysis of the specimen before separation of the serum [2]. We have therefore used our separation technique with haptoglobin type 1 where the picture is not confused by the presence of large numbers of polymers.
Haemoglob#z complexformation by haptoglobin type 1 Fig. 2A shows the protein pattern obtained from pooled serum containing haptoglobin type 1 while Fig. 2B shows the same serum after the addition of haemoglobin in excess of the amount required to saturate the haptoglobin. The haptoglobinhaemoglobin complex occupied a position consistent with a rise in isoelectric point and molecular weight compared with the original haptoglobin. Also shown are two proteins migrating close to the position of free haptoglobin type 1 and which probably appeared as contaminants in the purified haptoglobin preparation. This pooled serum was used in preparing the haptoglobin. A
B
I
i
t3
Fig. 2. Separated proteins from 5 ~1 of pooled serum containing haptoglobin type 1, after electrofocusing followed by gradient electrophoresis. The gels are positioned as in Fig. I. (A) Serum alone. (B) Serum after the addition of 14 ,ug of haemoglobin, l, free haptoglobin type 1; 2, saturated haptoglobin-haemoglobin complex; 3, free haemoglobin; 4 and 5, proteins probably contaminating the purified haptoglobin preparation.
61 t~
B
o4 ~4
Fig. 3. Electrofocusingfollowedby gradientelectrophoresisof purifiedhaptoglobintype 1. The gels are positionedas in Fig. 1. (A) 15/~1of the preparation alone. (B) 5/~1(containing5.5/~g of haemoglobin bindingcapacity)after mixingwith 3.5/~gof haemoglobin. 1, freehaptoglobintype 1; 2, intermediate haptoglobin-haemoglobincomplex; 3, saturated complex; 4, albumin. The pattern of separated proteins from a 15/~1 sample of the purified haptoglobin type 1 preparation is shown in Fig. 3A. A smaller sample (5 #1), containing about 5.5/~g of haemoglobin binding capacity, was mixed with 3.5/~g of haemoglobin and, after electrofocusing and electrophoresis, gave the protein pattern shown in Fig. 3B. Three well-defined spots were resolved, arranged in the expected diagonal line, and representing from left to right free haptoglobin, the intermediate haptoglobinhaemoglobin complex and the saturated complex [2]. In order to investigate the appearance of the two complexes samples of the haptoglobin, each possessing 5.5 #g of haemoglobin binding capacity were mixed with from 0.05 #g to 14.0 #g of haemoglobin and subjected to electrophoresis through a polyacrylamide gradient gel. The separated proteins are shown in Fig. 4 and the haemoglobin concentrations at which the various proteins appeared and disappeared are summarised in Table I. While the intermediate complex appeared first, the saturated complex became visible after the addition of a further small amount of haemoglobin. All three components were present as the amount of added haemoglobin was increased, free haptoglobin finally disappearing just before saturation was achieved. Free haemoglobin appeared after the addition of 1.38 #g of haemoglobin//~g of haemoglobin binding capacity. By comparison with the staining intensity of haemoglobin bands of known concentration, the excess haemoglobin in this mixture amounted to about 0.1 #g of haemoglobin/#g of haemoglobin binding capacity. Therefore, by this method saturation was estimated to have occurred after the addition of about i.28 #g of haemoglobin//~g of haemoglobin binding capacity. This obviously differs from the estimation of haemoglobin binding capacity by the method of Ratcliffe and Hardwicke [17] and probably reflects the difficulty in accurately assessing protein concentrations using staining intensity on gels. On the other hand, gradient electrophoresis of haemoglobin solutions of known concentration showed that it was possible to detect 0.35-040/~g of haemoglobin as a stained band on the gel.
62 ~,~r~II~'~
~
~ ~''l-v~'`
~ .....
e
d C
p b
i
1
2
3
4
5
6
7
8
Fig. 4. Samples of purified haptoglobin type 1 (each containing 5.5yg of haemoglobin binding capacity) mixed with haemoglobin before separation by gradient electrophoresis. This proceeded in a downward direction towards the anode. Haemoglobin was added as follows: 1, none; 2, 0.05 yg; 3, 1.7yg; 4, 3.5 fig; 5, 6.1 yg; 6, 7.0 f~g; 7, 7.?/~g; 8, 8.4fig. a, albumin; b, free haemoglobin; c, free haptoglobin type 1; d, intermediate haptoglobin-haemog]obin complex; e, saturated complex.
TABLE I THE E F F E C T O F A D D I N G I N C R E A S I N G A M O U N T S O F H A E M O G L O B I N TO HAPTOGLOBIN TYPE 1 Samples of a haptoglobin type 1 preparation were mixed with various amounts of haemoglobin and the proteins in the mixture separated by gradient electrophoresis. The haemoglobin concentration is expressed as a percentage of that required to just saturate the haptoglobin. Haemoglobin concentration Intermediate complex appears Saturated complex appears Free haptoglobin disappears Excess haemoglobin appears
0.5 1.8 92 108
63
Proteins of known molecular weight and a sample of haptoglobin type 1 preparation partially saturated with haemoglobin were subjected to electrophoresis for 48 h through a polyacrylamide gradient gel. From the linear calibration plot [13] values of 107 000, 139 000 and 168 000 were obtained for the molecular weights of
4-
/ w
// N
1
2
3
4
Fig. 5. Thin-layer gel electrofocusing, over the pH range 2.5-6, of marker proteins and the purified haptoglobin type 1 preparation. The marker solutions contained 10 mg of ferritin or 1 mg of fl-lactoglobulin in 1 ml of 2% (w/v) carrier ampholytes, pH range 2.5-4, and 10#1 were applied to each piece of filter paper. 1, purified haptoglobin type 1 (22/~g of haemoglobin binding capacity); 2, haptoglobin type 1 (14.7/~g of haemoglobin binding capacity) mixed with 9.3/~g of haemoglobin; 3, 4, marker proteins, a, ferritin; b, albumin contaminating the haptoglobin preparation; c, fl-lactoglobulin; d, haptoglobulin type 1.
64 free haptoglobin type 1, the intermediate haptoglobin-haemoglobin complex and the saturated complex, respectively.
Electrofocusing of haptoglobin type 1 and its haemoglobin complexes After thin-layer gel electrofocusing over the p H range 2.5-6, the purified haptoglobin preparation showed a number of closely-spaced protein bands. When the haptoglobin preparation, partially saturated with haemoglobin (0.64 #g of haemoglobin/ k~g of haemoglobin binding capacity), was subjected to electrofocusing over the same p H range most of the protein bands were reduced in intensity and almost disappeared. The remaining prominent band probably represented contaminating albumin. We were unable to show the haptoglobin-haemoglobin complexes by electrofocusing over this narrower p H range since the saturated complex would have migrated beyond the limit of the p H gradient, while the intermediate complex was probably obscured by the filter paper used for applying the sample. Application at the anode side of the gel did not give a satisfactory separation. The results are shown in Fig. 5 and the range of isoelectric points given by haptoglobin type 1 is given in Table II. TABLE II ISOELECTRIC POINTS OF HAPTOGLOBIN TYPE 1 AND ITS COMPLEXES WITH HAEMOGLOBIN Haptoglobin type 1 was mixed with haemoglobin (0.64 and 2.55 #g of haemoglobin//~g of haemoglobin binding capacity) to give mixtures containing free haptoglobin, the intermediate complex, the saturated complex and free haemoglobin. The mixtures were subjected to thin-layer gel electrofocusing over the pH range 3.5-10. Electrofocusing of haptoglobin type 1 alone was also carried out over the pH range 2.5-6. The isoelectric points for marker proteins were obtained from Bouts [15] or Radola [16]. Bands obviously due to the albumin contaminating the haptoglobin preparation have been ignored. There were other minor, faintly-staining bands with isoelectric points outside the ranges given. Isoelectric points pH range used Reference for isoelectric points of markers Free haptoglobin type 1 Intermediate haptoglobin-haemoglobin complex Saturated haptoglobin-haemoglobin complex Free haemoglobin
2.5-6 16 4.38-4.69* -
3.5-10 15 4.58-4.77 5.20-5.40 5.74--5.93 7.38-7.64
3.5-10 16 4.43-4.61 5.04-5.17 5.52-5.70 7.07-7.31
* There were two prominent bands with isoelectric points at pH 4.55 and pH 4.61. The haptoglobin preparation was mixed with sufficient haemoglobin partially or fully to saturate the haptoglobin (0.64 and 2.55/zg of haemoglobin//~g of haemoglobin binding capacity respectively). The mixtures and the haptoglobin preparation alone were subjected to thin layer gel electrofocusing over the p H range 3.5-10. Three distinct groups of bands were resolved, corresponding to free haptoglobin, the intermediate complex and the saturated complex (Fig. 6). In the sample fully saturated with haemoglobin, the excess haemoglobin focused as a group of red-coloured bands. The isoelectric points are given in Table II. In addition, a number of faintly staining bands were seen especially in the samples containing haemoglobin. These bands focused at a wide range of isoelectric points from about p H 4,5 to 6.4.
65
~ c
3--k-E ~ d
~ f
g
1
.
.
.
.
.
4
.5
(3
7
Fig. 6. Thin-layer gel electrofocusing, over the pH range 3.5-10, of marker proteins, purified haptoglobin type 1 and the haptoglobin-haemoglobin complexes. Marker solutions contained 1 mg of each marker (except ferritin, 10 mg) in 1 ml of 2 ~ (w/v) carrier ampholytes, pH range 3.5-10, and 10 pl was applied to each piece of filter paper. 1-4, marker proteins; 5, purified haptoglobin type 1 (22tLg of haemoglobin binding capacity); 6, haptoglobin type 1 (14.7#g of haemoglobin binding capacity) mixed with 9.3/zg of haemoglobin; 7, haptoglobin type 1 (7.3/~g of haemoglobin binding capacity) mixed with 18.7 pg of haemoglobin, a, ferritin; b, albumin; e, ~-lactoglobulin; d, carbonic anhydrase; e, equine myoglobin; f, whale myoglobin; g, cytochrome c; h, haptoglobin type 1; j, intermediate haptoglobin-haemoglobin complex; k, saturated complex; 1, haemoglobin. DISCUSSION Our method for the preparation of haptoglobin type 1 followed the later stages of the procedure given by Betlach and McMillan [23] but we found that the haptoglobin appeared with the albumin during the zinc-ethanol precipitation [21]. This has been demonstrated by others [25] who found that only haptoglobin types 2 and 2-1 could be separated from albumin by the precipitation procedure. In our preparation most of the albumin was removed later by ion-exchange chromatography. The haptoglobin preparation was contaminated with two proteins which appeared very close to haptoglobin type 1 after electrofocusing and gel gradient electrophoresis. These may be clearly seen in the serum used for the preparation as well as in the haptoglobin preparation itself, remaining behind in the position occupied by haptoglobin before the addition of excess haemoglobin (Figs. 2 and 4), The hapto-
66 globin preparation may have been contaminated with a trace of the intermediate haemoglobin complex (Fig. 3A) due to slight haemolysis of the serum before the purification was started. The patterns of separated proteins obtained after electrofocusing and electrophoresis, both of serum and the purified haptoglobin preparation showed that haptoglobin-haemoglobin complexes were present in numbers which confirmed the findings of Ogawa and Kawamura [2]. Haptoglobin type 1 was shown to possess two binding sites for the reactive unit of haemoglobin (Fig. 3). The haptoglobin type 2 component with the lowest molecular weight had three binding sites and there was an increase of one binding site for each step up the polymer series (Fig. 1). The fully saturated complexes showed a higher peroxidase activity than the other proteins. Our molecular weight values of 107 000, 139 000 and 168 000 for haptoglobin type 1 and its two complexes are in close agreement with values obtained elsewhere [26, 27] although much higher than the values of Zwaan and Maki [4], who also used polyacrylamide gel electrophoresis. The increases in molecular weight during the formation of the complexes (32 000 and 29 000) agree well with similar increases found by others [4, 27] and correspond almost exactly to half the molecular weight of haemoglobin [24]. This confirms that the reactive unit of haemoglobin which binds to haptoglobin during complex formation is half a haemoglobin molecule [3]. The microheterogeneity of haptoglobin [5] has been demonstrated by the resolution of our haptoglobin preparation into numerous bands after electrofocusing. Most of these bands were reduced in intensity when haemoglobin was added to the haptoglobin preparation (Figs. 5 and 6), the bands remaining being due to impurities, mainly albumin, which had been shown to be present in the preparation (Fig. 3A). We were unable to count accurately the number of bands due to haptoglobin type 1, although at least seven bands could be seen, with isoelectric points in the range given in Table II. Two bands were more prominent than the others. It has been shown elsewhere [5] that each of the multiple components of haptoglobin type 1 contained a different amount of sialic acid and was pre-existing in the blood, rather than an artefact of the isolation procedure. We have found no evidence to suggest that the various glycoproteins have different binding properties or affinities for haemoglobin. In fact the components of the free haptoglobin type 1 and its two complexes all showed the same spread of isoelectric points (viz. about 0.2 pH units) indicating that the binding of a half-haemoglobin probably altered the isoelectric point of each component by an equal amount. The isoelectric points for haptoglobin type 1 and its complexes given in Table 1I are considerably higher than those obtained by Yang and Przybylska [5]. These authors give values of pH 3.9-4.2 and pH 4.9-5.3 for the isoelectric points of haptoglobin type 1 and its saturated complex respectively. If our pH gradient was calibrated using the reference values for marker proteins given by Radola [16] our isoelectric points for the components of free human haemoglobin are in close agreement with the data of Drysdale and co-workers [28]. However, our isoelectric points varied by 0.150.30 pH units depending on the source of the information used to calibrate the pH gradient [15, 16]. Yang and Przybylska [5] found gel electrofocusing unsuccessful for the study of the haptoglobin-haemoglobin complexes because of extensive dissociation in the presence of 2 ~ (w/v) carrier ampholytes. Instead they employed electrofocusing in a
67 sucrose density gradient containing 1 ~ (w/v) carrier ampholytes. We were able to use the thin-layer gel electrofocusing method over the pH range 3.5-10 with gels containing about 1.1 ~ (w/v) carrier ampholytes. Therefore, the level of dissociation seems to depend upon the concentration of carrier ampholytes present during electrofocusing. In our experiments, some dissociation of the haemoglobin complexes probably did take place and, together with competitive binding of carrier ampholytes to the protein, gave rise to protein bands of low concentration showing isoelectric points in a wide range from about pH 4.5 up to about pH 6.4. Nevertheless, the haptoglobin-haemoglobin complexes were still clearly visible (Fig. 6). We were not successful in showing the haemoglobin complexes using electrofocusing over the pH range 2.5-6 and gels containing 1.6 ~ (w/v) carrier ampholytes (Fig. 5). Since the focused bands due to haptoglobin were reduced in intensity after the addition of haemoglobin, it is unlikely that the haemoglobin complexes were extensively dissociated during electrofocusing to reform free haptoglobin. Some binding with carrier ampholytes could have taken place but it seems more probable that the complexes either migrated into the cathode electrode strip or were obscured by the filter paper square visible in Fig. 5. Attempts to produce a satisfactory separation of focused proteins after application on filter paper at the anode side of the gel were consistently unsuccessful since our preparations were probably unstable under the acid conditions produced near the anode. In agreement with Vesterberg [29] best results were obtained with most proteins after application near the cathode. As shown in Table I, the saturated complex was seen even after the addition of very small amounts of haemoglobin and three bands (free haptoglobin type 1 and the two complexes) were present over a very wide range of haemoglobin concentrations. This amplifies the results obtained elsewhere [2] with a small number of haemoglobin concentrations from 5 3 ~ to 120~ of that required for saturation. Judging from staining intensity, all three bands were present in about equal concentrations at half the haemoglobin concentration required for saturation (Fig. 3B). This would be expected if both binding sites bound haemoglobin with about the same affinity. Further information should become available after studies on haemoglobin binding by polymers of haptoglobin types 2 and 2-1 where there can be larger numbers of binding sites on each molecule. ACKNOWLEDGMENTS The Medical Research Council assisted J. A. M. with a Further Education Award. Miss B. Mackay provided skilled technical assistance. REFERENCES 1 Sutton, H. E. (1970) Progr. Meal. Genet. 7, 163-216 20gawa, A. and Kawamura, K. (1966) Proc. Jap. Acad. 42, 413-417 3 Kawamura, K., Kagiyama, S., Ogawa, A. and Yanase, T. (1972) Biochim. Biophys. Acta 285, 15-21 4 Zwaan, J. and Maki, T. N. (1968) Nature 218, 476-478 5 Yang, H. J. and Przybylska, M. (1973) Can. J. Biochem. 51, 597-605 6 Emes, A. V., Latner, A. L. and Martin, J. A. (1975) Clin. Chim, Acta 64, 69-78
68 7 Diezel, W., Kopperschl~iger, G. and Hofmann, E. (1972) Anal. Biochem. 48, 617-620 8 Clarke, J. T. (1964) Ann. N.Y. Acad. Sci. 121,428-436 9 Schultze, H. E. and Heremans, J. F. (1966) Molecular Biology of Human Proteins, Vol. 1, pp. 176-235, Elsevier, Amsterdam 10 Barman, T. E. (1969) Enzyme Handbook, Vol. 1, pp. 23 and 170, Springer-Verlag, Berlin 11 Margolis, J. and Kenrick, K. G. (1968) Anal. Biochem. 25, 347-362 12 Shaw, C. R. and Koen, A. L. (1968) in Chromatographic and Electrophoretic Techniques (Smith, I., ed), 2nd edn, Vol. 2, pp. 325-364 13 Slater, G. G. (1969) Anal. Chem. 41, 1039-1041 14 Davies, H. (1975) in Isoelectric Focusing (Arbuthnott, J. P. and Beeley, J. A., eds), pp. 97-113, Butterworths, London 15 Bours, J. (1973) Science Tools 20, 29-34 16 Radola, B. J. (1973) Biochim. Biophys. Acta 295, 412-428 17 Ratcliffe, A. P. and Hardwicke, J. (1964) J. Clin. Path. 17, 676-679 18 Dacie, J. V. and Lewis, S. M. (1968) Practical Haematology, 4th edn, pp. 41-43, Churchill, London 19 Lowry, O. H., Rosebrough, N. J., Farr, A. L. and Randall, R. J. (1951) J. Biol. Chem. 193, 265275 20 Lever, W. F., Gurd, F. R. N., Uroma, E., Brown, R. K., Barnes, B. A., Schmid, K. and Schultz, E. L. (1951) J. Clin. Invest. 30, 99-111 21 Cohn, E. J., Gurd, F. R. N . Surgenor, D. M., Barnes, B. A., Brown, R. K., Derouaux. G., Gillespie, J. M., Kahnt, F. W., Lever, W. F., Liu, C. H., Mittelman, D., Mouton, R. F., Schmid, K. and Uroma, E. (1950) J. Am. Chem. Soc. 72, 465-474 22 Gregory, M. E. (1954) Br. J. Nutr. 8, 340-347 23 Betlach, C. J. and McMillan, D. E. (1972) Anal. Biochem. 49, 103-108 24 Perutz, M. F. (1969) Proc. R. Soc. Ser. B 173, 113-140 25 Guinand, S., Tonnelat, J., Boussier, G. and Jayle, M. F. (1956) Bull. Soc. Chim. Biol. 38,329-341 26 Jayle, M. F. and Moretti, L (1962) Progr. Haematol. 3, 342-359 27 Ogawa, A., Kagiyama, S. and Kawamura, K. (1968) Proc. Jap. Acad. 44, 1054--1059 28 Drysdale, J. W., Righetti, P. and Bunn, H. F. (1971) Biochim. Biophys. Acta 229, 42-50 29 Vesterberg, O. (1975) in Isoelectric Focusing (Arbuthnott, J. P. and Beeley, J. A., eds), pp. 7896, Butterworths, London
© Elsevier Scientific Publishing Company, Amsterdam- Printed in The Netherlands BBA 37227 STUDIES ON T H E B I N D I N G OF H A E M O G L O B I N BY H A P T O G L O B I N U S I N G E L E C T R O F O C U S I N G A N D G R A D I E N T ELECTROPHORESIS
ALAN V. EMES, ALBERT L. LATNER, JOHN A. MARTIN* and FIONA A. MULLIGAN University Department of Clinical Biochemistry, Royal Victoria Infirmary, Newcastle upon Tyne, NEI 4LP (U.K )
(Received July 14th, 1975)
SUMMARY 1. Gel electrofocusing followed by gel gradient electrophoresis separated the haptoglobins and their complexes with haemoglobin into characteristic two-dimensional patterns of protein bands. 2. Molecular weights of 107 000, 139 000 and 168 000 were obtained for the three bands seen after a purified preparation of haptoglobin type 1 was partially saturated with haemoglobin. This indicated that free haptoglobin, the intermediate haptoglobin-haemoglobin complex containing one half-haemoglobin and the saturated complex with two half-haemoglobins were present. 3. The three proteins showed considerable microheterogeneity and gave a number of isoelectric points in the pH ranges 4.58-4.77, 5.20-5.40 and 5.74-5.93, free haptoglobin type 1 being the lowest group. These ranges were all 0.15-0.30 pH units lower if other values were taken for the isoelectric points of markers used to calibrate the pH gradient. 4. All three proteins were present over a wide range of haemoglobin concentrations, from 0.5 ~ to 92 ~o of that required for saturation. This would be expected if both binding sites have similar affinities for haemoglobin.
INTRODUCTION The haptoglobins have been extensively investigated because of their ability to form stable complexes with haemoglobin [1]. There have been several reports concerning the number of binding sites available [2] and the reactive unit of haemoglobin which takes part in complex formation with haptoglobin [3, 4]. Both free haptoglobin and the saturated complex show considerable microheterogeneity [5]. We have found that a combination of gel electrofocusing and polyacrylamide gel gradient electrophoresis separates the haptoglobins and their complexes into well-defined patterns of protein bands. We now report the use of these two separation techniques, both independently and combined as a two-dimensional procedure. They have provided further information on the binding process and the complexes, mainly * Present address: Department of Biochemistry,Bolton Royal Infirmary, Bolton, Lancs., U.K.
58 with purified haptoglobin type 1. The notation for the haptoglobin types is that used by Sutton [1]. MATERIALS AND METHODS
Electrofocusing and electrophoresis Gel electrofocusing followed by electrophoresis through a polyacrylamide gel gradient was performed as described elsewhere [6]. The electrophoresis stage was also carried out alone to give a one-dimensional separation. The gels were stained for protein with Coomassie Brilliant Blue G [7] and for haemoglobin with benzidine [8]. For the determination of molecular weights, polyacrylamide gradient gels were calibrated using haemopexin, transferrin, yeast alcohol dehydrogenase (EC 1.1.1.1), glutamate dehydrogenase (EC 1.4.1.3) and az-macroglobulin as marker proteins of known molecular weight [9, 10]. These were identified by comparison with published data [11] and after location of the enzymes using a reagent mixture based on those given by Shaw and Koen [12]. This contained 0.5 M KH2PO4, Na2HPO4 buffer, pH 7.0 (20 ml); 95 % (v/v) ethanol (5 ml); 1 M sodium glutamate in buffer, pH 7.0 (20 ml); NAD (10 mg); MTT tetrazolium [3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide] (5 mg); phenazine methosulphate (2 mg); water (20 ml). A linear relationship between log molecular weight and log migration distance [13] was obtained after electrophoresis for 48 h. Electrofocusing was also carried out in thin-layer polyacrylamide gels containing carrier ampholyte mixtures designed to produce more or less linear pH gradients over the pH ranges 3.5-9.5 or 2.5-6 [14]. For the pH 3.5-9.5 range, it was possible to reduce the recommended carrier ampholyte concentration by a half without noticeable effect. The Multiphor apparatus and Type 3371C power supply (LKBProdukter AB, S-161 25 Bromma 1, Sweden) were used. Gels were stained for protein with Coomassie Brilliant Blue R [14]. Ferritin, bovine serum albumin, fl-lactogloglobulin, carbonic anhydrase, equine and whale myoglobins and cytochrome c were used as pH markers. Linear pH gradients were obtained between about pH 4.5 and pH 9.5 using the isoelectric points given by Bours [15] or Radola [16], although there was a difference of 0.15-0.30 pH units between the two sets of values. Using the narrower range mixture of carrier ampholytes, the pH gradient was S-shaped rather than linear and became somewhat steeper around pH 5. It probably did not extend beyond pH 5.5.
Estimation of proteins Haptoglobin was estimated as its haemoglobin binding capacity [17]. Haemoglobin was obtained from normal human red blood cells by lysis with chloroform and measured spectrophotometrically [18]. Total protein was assayed using the Lowry method [19] except that reagent solutions were stored separately and mixed immediately before use; human serum albumin (Blood Products Laboratory, Lister Institute, Elstree, Herts., U.K.) was used as a standard.
Preparation of haptoglobin type l Human serum (10 ml) containing haptoglobin type 1 was mixed with 0.87 ml of acid citrate dextrose solution [20] and 10 ml of the mixture fractionated using Cohn's
59 method 10 [21]. We found that most of the haptoglobin appeared with albumin in Cohn fraction V. This fraction (78 ml) was concentrated to 24.5 ml by ultrafiltration [22] and dialysed against 0.08 M potassium acetate buffer, p H 5.7. The dialysis residue was subjected to chromatography on DEAE-Sephadex A-50, a m m o n i u m sulphate precipitation and gel filtration on Sephadex G-150, as described by Betlach and McMillan [23]. RESULTS
Haptoglobin type 1 preparation The purified preparation of haptoglobin type 1 contained 2.8 mg of protein/ml and 1.1 mg of haemoglobin binding capacity/ml. This represented 1.7 mg of haptoglobin/ml, assuming molecular weights for haptoglobin and haemoglobin of 100 000 [9] and 64 500 [24], respectively. The protein separation obtained from the preparation after electrofocusing and gradient electrophoresis is shown in Fig. 3A. There was some contamination with albumin and small amounts of other proteins can be seen, including the intermediate haptoglobin-haemoglobin complex.
Resolution of the haptoglobins and their haemoglobin complexes Gel electrofocusing of human serum followed by electrophoresis through a polyacrylamide gel gradient produced a characteristic two-dimensional pattern of separated proteins. In normal unhaemolysed serum, haptoglobin appeared as a single protein spot or as a single column of spots, depending on the haptoglobin type present A
B
Fig. 1. Separated proteins from 5/~1 of serum containing haptoglobin type 2, after electrofocusing followed by gradient electrophoresis. The gel was cut to give two identical slices; one was stained with Coomassie Brilliant Blue G (A) and the other with benzidine reagent (B). The electrofocusing gel, at the top, was positioned so that the pH gradient increased from left to right. Electrophoresis proceeded in a downward direction towards the anode. The diagonal rows of spots comprise 1, four proteins; 2, five proteins; 3, six proteins. All but the most acidic protein in each row contain haemoglobin.
60 [6]. In some sera, demonstrated to contain haemoglobin, large numbers of protein spots were resolved and these were arranged in diagonal rows rising towards the more basic end of the electrofocusing gel, that is, increasing in isoelectric point and molecular weight. Fig. 1 shows such a separation, in this case of a serum from a patient with an IgG myeloma and containing haptoglobin type 2. The gel was sliced to provide two identical slices; one was stained in Coomassie Brilliant Blue G (Fig. IA) and the other in benzidine reagent (Fig. 1B). The array of protein spots can be clearly seen, together with the large number shown to contain haemoglobin. The lowest diagonal row comprised four protein spots (barely visible), the next row consisted of five protein spots, the next of six spots and so on. All these proteins contained haemoglobin except for the spot with the most acidic isoelectric point in each row. The pattern of spots demonstrated in Fig. 1 most probably arose from partial saturation of the haptoglobin due to slight haemolysis of the specimen before separation of the serum [2]. We have therefore used our separation technique with haptoglobin type 1 where the picture is not confused by the presence of large numbers of polymers.
Haemoglob#z complexformation by haptoglobin type 1 Fig. 2A shows the protein pattern obtained from pooled serum containing haptoglobin type 1 while Fig. 2B shows the same serum after the addition of haemoglobin in excess of the amount required to saturate the haptoglobin. The haptoglobinhaemoglobin complex occupied a position consistent with a rise in isoelectric point and molecular weight compared with the original haptoglobin. Also shown are two proteins migrating close to the position of free haptoglobin type 1 and which probably appeared as contaminants in the purified haptoglobin preparation. This pooled serum was used in preparing the haptoglobin. A
B
I
i
t3
Fig. 2. Separated proteins from 5 ~1 of pooled serum containing haptoglobin type 1, after electrofocusing followed by gradient electrophoresis. The gels are positioned as in Fig. I. (A) Serum alone. (B) Serum after the addition of 14 ,ug of haemoglobin, l, free haptoglobin type 1; 2, saturated haptoglobin-haemoglobin complex; 3, free haemoglobin; 4 and 5, proteins probably contaminating the purified haptoglobin preparation.
61 t~
B
o4 ~4
Fig. 3. Electrofocusingfollowedby gradientelectrophoresisof purifiedhaptoglobintype 1. The gels are positionedas in Fig. 1. (A) 15/~1of the preparation alone. (B) 5/~1(containing5.5/~g of haemoglobin bindingcapacity)after mixingwith 3.5/~gof haemoglobin. 1, freehaptoglobintype 1; 2, intermediate haptoglobin-haemoglobincomplex; 3, saturated complex; 4, albumin. The pattern of separated proteins from a 15/~1 sample of the purified haptoglobin type 1 preparation is shown in Fig. 3A. A smaller sample (5 #1), containing about 5.5/~g of haemoglobin binding capacity, was mixed with 3.5/~g of haemoglobin and, after electrofocusing and electrophoresis, gave the protein pattern shown in Fig. 3B. Three well-defined spots were resolved, arranged in the expected diagonal line, and representing from left to right free haptoglobin, the intermediate haptoglobinhaemoglobin complex and the saturated complex [2]. In order to investigate the appearance of the two complexes samples of the haptoglobin, each possessing 5.5 #g of haemoglobin binding capacity were mixed with from 0.05 #g to 14.0 #g of haemoglobin and subjected to electrophoresis through a polyacrylamide gradient gel. The separated proteins are shown in Fig. 4 and the haemoglobin concentrations at which the various proteins appeared and disappeared are summarised in Table I. While the intermediate complex appeared first, the saturated complex became visible after the addition of a further small amount of haemoglobin. All three components were present as the amount of added haemoglobin was increased, free haptoglobin finally disappearing just before saturation was achieved. Free haemoglobin appeared after the addition of 1.38 #g of haemoglobin//~g of haemoglobin binding capacity. By comparison with the staining intensity of haemoglobin bands of known concentration, the excess haemoglobin in this mixture amounted to about 0.1 #g of haemoglobin/#g of haemoglobin binding capacity. Therefore, by this method saturation was estimated to have occurred after the addition of about i.28 #g of haemoglobin//~g of haemoglobin binding capacity. This obviously differs from the estimation of haemoglobin binding capacity by the method of Ratcliffe and Hardwicke [17] and probably reflects the difficulty in accurately assessing protein concentrations using staining intensity on gels. On the other hand, gradient electrophoresis of haemoglobin solutions of known concentration showed that it was possible to detect 0.35-040/~g of haemoglobin as a stained band on the gel.
62 ~,~r~II~'~
~
~ ~''l-v~'`
~ .....
e
d C
p b
i
1
2
3
4
5
6
7
8
Fig. 4. Samples of purified haptoglobin type 1 (each containing 5.5yg of haemoglobin binding capacity) mixed with haemoglobin before separation by gradient electrophoresis. This proceeded in a downward direction towards the anode. Haemoglobin was added as follows: 1, none; 2, 0.05 yg; 3, 1.7yg; 4, 3.5 fig; 5, 6.1 yg; 6, 7.0 f~g; 7, 7.?/~g; 8, 8.4fig. a, albumin; b, free haemoglobin; c, free haptoglobin type 1; d, intermediate haptoglobin-haemog]obin complex; e, saturated complex.
TABLE I THE E F F E C T O F A D D I N G I N C R E A S I N G A M O U N T S O F H A E M O G L O B I N TO HAPTOGLOBIN TYPE 1 Samples of a haptoglobin type 1 preparation were mixed with various amounts of haemoglobin and the proteins in the mixture separated by gradient electrophoresis. The haemoglobin concentration is expressed as a percentage of that required to just saturate the haptoglobin. Haemoglobin concentration Intermediate complex appears Saturated complex appears Free haptoglobin disappears Excess haemoglobin appears
0.5 1.8 92 108
63
Proteins of known molecular weight and a sample of haptoglobin type 1 preparation partially saturated with haemoglobin were subjected to electrophoresis for 48 h through a polyacrylamide gradient gel. From the linear calibration plot [13] values of 107 000, 139 000 and 168 000 were obtained for the molecular weights of
4-
/ w
// N
1
2
3
4
Fig. 5. Thin-layer gel electrofocusing, over the pH range 2.5-6, of marker proteins and the purified haptoglobin type 1 preparation. The marker solutions contained 10 mg of ferritin or 1 mg of fl-lactoglobulin in 1 ml of 2% (w/v) carrier ampholytes, pH range 2.5-4, and 10#1 were applied to each piece of filter paper. 1, purified haptoglobin type 1 (22/~g of haemoglobin binding capacity); 2, haptoglobin type 1 (14.7/~g of haemoglobin binding capacity) mixed with 9.3/~g of haemoglobin; 3, 4, marker proteins, a, ferritin; b, albumin contaminating the haptoglobin preparation; c, fl-lactoglobulin; d, haptoglobulin type 1.
64 free haptoglobin type 1, the intermediate haptoglobin-haemoglobin complex and the saturated complex, respectively.
Electrofocusing of haptoglobin type 1 and its haemoglobin complexes After thin-layer gel electrofocusing over the p H range 2.5-6, the purified haptoglobin preparation showed a number of closely-spaced protein bands. When the haptoglobin preparation, partially saturated with haemoglobin (0.64 #g of haemoglobin/ k~g of haemoglobin binding capacity), was subjected to electrofocusing over the same p H range most of the protein bands were reduced in intensity and almost disappeared. The remaining prominent band probably represented contaminating albumin. We were unable to show the haptoglobin-haemoglobin complexes by electrofocusing over this narrower p H range since the saturated complex would have migrated beyond the limit of the p H gradient, while the intermediate complex was probably obscured by the filter paper used for applying the sample. Application at the anode side of the gel did not give a satisfactory separation. The results are shown in Fig. 5 and the range of isoelectric points given by haptoglobin type 1 is given in Table II. TABLE II ISOELECTRIC POINTS OF HAPTOGLOBIN TYPE 1 AND ITS COMPLEXES WITH HAEMOGLOBIN Haptoglobin type 1 was mixed with haemoglobin (0.64 and 2.55 #g of haemoglobin//~g of haemoglobin binding capacity) to give mixtures containing free haptoglobin, the intermediate complex, the saturated complex and free haemoglobin. The mixtures were subjected to thin-layer gel electrofocusing over the pH range 3.5-10. Electrofocusing of haptoglobin type 1 alone was also carried out over the pH range 2.5-6. The isoelectric points for marker proteins were obtained from Bouts [15] or Radola [16]. Bands obviously due to the albumin contaminating the haptoglobin preparation have been ignored. There were other minor, faintly-staining bands with isoelectric points outside the ranges given. Isoelectric points pH range used Reference for isoelectric points of markers Free haptoglobin type 1 Intermediate haptoglobin-haemoglobin complex Saturated haptoglobin-haemoglobin complex Free haemoglobin
2.5-6 16 4.38-4.69* -
3.5-10 15 4.58-4.77 5.20-5.40 5.74--5.93 7.38-7.64
3.5-10 16 4.43-4.61 5.04-5.17 5.52-5.70 7.07-7.31
* There were two prominent bands with isoelectric points at pH 4.55 and pH 4.61. The haptoglobin preparation was mixed with sufficient haemoglobin partially or fully to saturate the haptoglobin (0.64 and 2.55/zg of haemoglobin//~g of haemoglobin binding capacity respectively). The mixtures and the haptoglobin preparation alone were subjected to thin layer gel electrofocusing over the p H range 3.5-10. Three distinct groups of bands were resolved, corresponding to free haptoglobin, the intermediate complex and the saturated complex (Fig. 6). In the sample fully saturated with haemoglobin, the excess haemoglobin focused as a group of red-coloured bands. The isoelectric points are given in Table II. In addition, a number of faintly staining bands were seen especially in the samples containing haemoglobin. These bands focused at a wide range of isoelectric points from about p H 4,5 to 6.4.
65
~ c
3--k-E ~ d
~ f
g
1
.
.
.
.
.
4
.5
(3
7
Fig. 6. Thin-layer gel electrofocusing, over the pH range 3.5-10, of marker proteins, purified haptoglobin type 1 and the haptoglobin-haemoglobin complexes. Marker solutions contained 1 mg of each marker (except ferritin, 10 mg) in 1 ml of 2 ~ (w/v) carrier ampholytes, pH range 3.5-10, and 10 pl was applied to each piece of filter paper. 1-4, marker proteins; 5, purified haptoglobin type 1 (22tLg of haemoglobin binding capacity); 6, haptoglobin type 1 (14.7#g of haemoglobin binding capacity) mixed with 9.3/zg of haemoglobin; 7, haptoglobin type 1 (7.3/~g of haemoglobin binding capacity) mixed with 18.7 pg of haemoglobin, a, ferritin; b, albumin; e, ~-lactoglobulin; d, carbonic anhydrase; e, equine myoglobin; f, whale myoglobin; g, cytochrome c; h, haptoglobin type 1; j, intermediate haptoglobin-haemoglobin complex; k, saturated complex; 1, haemoglobin. DISCUSSION Our method for the preparation of haptoglobin type 1 followed the later stages of the procedure given by Betlach and McMillan [23] but we found that the haptoglobin appeared with the albumin during the zinc-ethanol precipitation [21]. This has been demonstrated by others [25] who found that only haptoglobin types 2 and 2-1 could be separated from albumin by the precipitation procedure. In our preparation most of the albumin was removed later by ion-exchange chromatography. The haptoglobin preparation was contaminated with two proteins which appeared very close to haptoglobin type 1 after electrofocusing and gel gradient electrophoresis. These may be clearly seen in the serum used for the preparation as well as in the haptoglobin preparation itself, remaining behind in the position occupied by haptoglobin before the addition of excess haemoglobin (Figs. 2 and 4), The hapto-
66 globin preparation may have been contaminated with a trace of the intermediate haemoglobin complex (Fig. 3A) due to slight haemolysis of the serum before the purification was started. The patterns of separated proteins obtained after electrofocusing and electrophoresis, both of serum and the purified haptoglobin preparation showed that haptoglobin-haemoglobin complexes were present in numbers which confirmed the findings of Ogawa and Kawamura [2]. Haptoglobin type 1 was shown to possess two binding sites for the reactive unit of haemoglobin (Fig. 3). The haptoglobin type 2 component with the lowest molecular weight had three binding sites and there was an increase of one binding site for each step up the polymer series (Fig. 1). The fully saturated complexes showed a higher peroxidase activity than the other proteins. Our molecular weight values of 107 000, 139 000 and 168 000 for haptoglobin type 1 and its two complexes are in close agreement with values obtained elsewhere [26, 27] although much higher than the values of Zwaan and Maki [4], who also used polyacrylamide gel electrophoresis. The increases in molecular weight during the formation of the complexes (32 000 and 29 000) agree well with similar increases found by others [4, 27] and correspond almost exactly to half the molecular weight of haemoglobin [24]. This confirms that the reactive unit of haemoglobin which binds to haptoglobin during complex formation is half a haemoglobin molecule [3]. The microheterogeneity of haptoglobin [5] has been demonstrated by the resolution of our haptoglobin preparation into numerous bands after electrofocusing. Most of these bands were reduced in intensity when haemoglobin was added to the haptoglobin preparation (Figs. 5 and 6), the bands remaining being due to impurities, mainly albumin, which had been shown to be present in the preparation (Fig. 3A). We were unable to count accurately the number of bands due to haptoglobin type 1, although at least seven bands could be seen, with isoelectric points in the range given in Table II. Two bands were more prominent than the others. It has been shown elsewhere [5] that each of the multiple components of haptoglobin type 1 contained a different amount of sialic acid and was pre-existing in the blood, rather than an artefact of the isolation procedure. We have found no evidence to suggest that the various glycoproteins have different binding properties or affinities for haemoglobin. In fact the components of the free haptoglobin type 1 and its two complexes all showed the same spread of isoelectric points (viz. about 0.2 pH units) indicating that the binding of a half-haemoglobin probably altered the isoelectric point of each component by an equal amount. The isoelectric points for haptoglobin type 1 and its complexes given in Table 1I are considerably higher than those obtained by Yang and Przybylska [5]. These authors give values of pH 3.9-4.2 and pH 4.9-5.3 for the isoelectric points of haptoglobin type 1 and its saturated complex respectively. If our pH gradient was calibrated using the reference values for marker proteins given by Radola [16] our isoelectric points for the components of free human haemoglobin are in close agreement with the data of Drysdale and co-workers [28]. However, our isoelectric points varied by 0.150.30 pH units depending on the source of the information used to calibrate the pH gradient [15, 16]. Yang and Przybylska [5] found gel electrofocusing unsuccessful for the study of the haptoglobin-haemoglobin complexes because of extensive dissociation in the presence of 2 ~ (w/v) carrier ampholytes. Instead they employed electrofocusing in a
67 sucrose density gradient containing 1 ~ (w/v) carrier ampholytes. We were able to use the thin-layer gel electrofocusing method over the pH range 3.5-10 with gels containing about 1.1 ~ (w/v) carrier ampholytes. Therefore, the level of dissociation seems to depend upon the concentration of carrier ampholytes present during electrofocusing. In our experiments, some dissociation of the haemoglobin complexes probably did take place and, together with competitive binding of carrier ampholytes to the protein, gave rise to protein bands of low concentration showing isoelectric points in a wide range from about pH 4.5 up to about pH 6.4. Nevertheless, the haptoglobin-haemoglobin complexes were still clearly visible (Fig. 6). We were not successful in showing the haemoglobin complexes using electrofocusing over the pH range 2.5-6 and gels containing 1.6 ~ (w/v) carrier ampholytes (Fig. 5). Since the focused bands due to haptoglobin were reduced in intensity after the addition of haemoglobin, it is unlikely that the haemoglobin complexes were extensively dissociated during electrofocusing to reform free haptoglobin. Some binding with carrier ampholytes could have taken place but it seems more probable that the complexes either migrated into the cathode electrode strip or were obscured by the filter paper square visible in Fig. 5. Attempts to produce a satisfactory separation of focused proteins after application on filter paper at the anode side of the gel were consistently unsuccessful since our preparations were probably unstable under the acid conditions produced near the anode. In agreement with Vesterberg [29] best results were obtained with most proteins after application near the cathode. As shown in Table I, the saturated complex was seen even after the addition of very small amounts of haemoglobin and three bands (free haptoglobin type 1 and the two complexes) were present over a very wide range of haemoglobin concentrations. This amplifies the results obtained elsewhere [2] with a small number of haemoglobin concentrations from 5 3 ~ to 120~ of that required for saturation. Judging from staining intensity, all three bands were present in about equal concentrations at half the haemoglobin concentration required for saturation (Fig. 3B). This would be expected if both binding sites bound haemoglobin with about the same affinity. Further information should become available after studies on haemoglobin binding by polymers of haptoglobin types 2 and 2-1 where there can be larger numbers of binding sites on each molecule. ACKNOWLEDGMENTS The Medical Research Council assisted J. A. M. with a Further Education Award. Miss B. Mackay provided skilled technical assistance. REFERENCES 1 Sutton, H. E. (1970) Progr. Meal. Genet. 7, 163-216 20gawa, A. and Kawamura, K. (1966) Proc. Jap. Acad. 42, 413-417 3 Kawamura, K., Kagiyama, S., Ogawa, A. and Yanase, T. (1972) Biochim. Biophys. Acta 285, 15-21 4 Zwaan, J. and Maki, T. N. (1968) Nature 218, 476-478 5 Yang, H. J. and Przybylska, M. (1973) Can. J. Biochem. 51, 597-605 6 Emes, A. V., Latner, A. L. and Martin, J. A. (1975) Clin. Chim, Acta 64, 69-78
68 7 Diezel, W., Kopperschl~iger, G. and Hofmann, E. (1972) Anal. Biochem. 48, 617-620 8 Clarke, J. T. (1964) Ann. N.Y. Acad. Sci. 121,428-436 9 Schultze, H. E. and Heremans, J. F. (1966) Molecular Biology of Human Proteins, Vol. 1, pp. 176-235, Elsevier, Amsterdam 10 Barman, T. E. (1969) Enzyme Handbook, Vol. 1, pp. 23 and 170, Springer-Verlag, Berlin 11 Margolis, J. and Kenrick, K. G. (1968) Anal. Biochem. 25, 347-362 12 Shaw, C. R. and Koen, A. L. (1968) in Chromatographic and Electrophoretic Techniques (Smith, I., ed), 2nd edn, Vol. 2, pp. 325-364 13 Slater, G. G. (1969) Anal. Chem. 41, 1039-1041 14 Davies, H. (1975) in Isoelectric Focusing (Arbuthnott, J. P. and Beeley, J. A., eds), pp. 97-113, Butterworths, London 15 Bours, J. (1973) Science Tools 20, 29-34 16 Radola, B. J. (1973) Biochim. Biophys. Acta 295, 412-428 17 Ratcliffe, A. P. and Hardwicke, J. (1964) J. Clin. Path. 17, 676-679 18 Dacie, J. V. and Lewis, S. M. (1968) Practical Haematology, 4th edn, pp. 41-43, Churchill, London 19 Lowry, O. H., Rosebrough, N. J., Farr, A. L. and Randall, R. J. (1951) J. Biol. Chem. 193, 265275 20 Lever, W. F., Gurd, F. R. N., Uroma, E., Brown, R. K., Barnes, B. A., Schmid, K. and Schultz, E. L. (1951) J. Clin. Invest. 30, 99-111 21 Cohn, E. J., Gurd, F. R. N . Surgenor, D. M., Barnes, B. A., Brown, R. K., Derouaux. G., Gillespie, J. M., Kahnt, F. W., Lever, W. F., Liu, C. H., Mittelman, D., Mouton, R. F., Schmid, K. and Uroma, E. (1950) J. Am. Chem. Soc. 72, 465-474 22 Gregory, M. E. (1954) Br. J. Nutr. 8, 340-347 23 Betlach, C. J. and McMillan, D. E. (1972) Anal. Biochem. 49, 103-108 24 Perutz, M. F. (1969) Proc. R. Soc. Ser. B 173, 113-140 25 Guinand, S., Tonnelat, J., Boussier, G. and Jayle, M. F. (1956) Bull. Soc. Chim. Biol. 38,329-341 26 Jayle, M. F. and Moretti, L (1962) Progr. Haematol. 3, 342-359 27 Ogawa, A., Kagiyama, S. and Kawamura, K. (1968) Proc. Jap. Acad. 44, 1054--1059 28 Drysdale, J. W., Righetti, P. and Bunn, H. F. (1971) Biochim. Biophys. Acta 229, 42-50 29 Vesterberg, O. (1975) in Isoelectric Focusing (Arbuthnott, J. P. and Beeley, J. A., eds), pp. 7896, Butterworths, London