Journal of Dairy Research (1975), 42, 267-275
267
The precipitation of proteins by carboxymethyl cellulose BY J. G. ZADOW AND R. D. HILL Division of Food Research, Dairy Research Laboratory, C.S.I.R.O, Melbourne, Victoria, Australia {Received 10 October 1974) SUMMARY. Carboxymethyl cellulose (CMC) formed insoluble complexes with /?lactoglobulin, bovine serum albumin and Na caseinate. Maximum precipitation of the /?-lactoglobulin-CMC complex occurred at pH 3-2, whereas maximum precipitation of the bovine serum albumin-CMC complex and the Na caseinate-CMC complex occurred at pH 2-8. The ratio of CMC to protein for maximum precipitation depended on the protein, being greatest for Na caseinate and least for bovine serum albumin. The percentage of protein precipitated by CMC decreased with increasing ionic strength of the solution, the rate of decrease being least for bovine serum albumin. At a given ionic strength, more protein was precipitated by CMC of high degree of substitution than by CMC of low degree of substitution. The change in pH (ApH) occurring on mixing CMC and unbuffered protein solutions, each initially at the same pH, was measured. ApH was negative for /?-lactoglobulin-CMC mixtures over the pH range 7-2 (minimum at pH 5-5). For bovine serum albumin-CMC and Na caseinate-CMC mixtures, ApH was positive between pH 7 and 3-2 (maximum at pH 4-5), zero at pH 3-2 and negative between pH 3-2 and 2-0 (minimum at pH 2-8).
The formation of insoluble complexes between carboxymethyl cellulose (CMC) and the proteins in whey has been suggested by Hansen, Hidalgo & Gould (1971) as a means of reclaiming whey proteins. The extent of precipitation of the protein depends on the degree of substitution of the CMC, and on the dilution (i.e. ionic strength) of the whey (Hill & Zadow, 1974). It has been suggested that these insoluble CMCprotein complexes are formed as a result of charge neutralization between the positively charged protein and the negatively charged CMC (Hidalgo & Hansen, 1969). In an attempt to obtain more information about the mechanism of complex formation, the conditions for precipitation by CMC of /ff-lactoglobulin, casein, and bovine serum albumin have been studied. As interactions between protein and CMC can be expected to alter the ionization of the CMC (Hill & Zadow, 1974), the changes in pH that occurred on mixing solutions of CMC and protein, each initially at the same pH, were also studied. „'
268
J. G. ZADOW AND R. D. HILL EXPERIMENTAL
Proteins Bovine serum albumin. Bovine serum albumin was supplied by the Commonwealth Serum Laboratories. Sodium caseinate. Casein was precipitated from skim-milk by adjustment of the pH to 4-5 using HC1. The casein was separated and redissolved using NaOH, the pH of the resulting solution being maintained at approximately 7-0. The Na caseinate solution was then freeze-dried. fi-Lactoglobulin. /?-Lactoglobulin was prepared by selective precipitation of this protein from whey in the form of a CMC complex (Hidalgo & Hansen, 1971). After concentration of the complex by centrifugation at 200 g for 20 min, a small volume of water was added to the sediment and the pH of the dispersion adjusted to 7-0 with 1 M-NaOH. After stirring overnight at 2 °C, the mixture was centrifuged at 2000 g for 20 min and the sediment discarded. The supernatant was made 4 M with respect to (NH4)2SO4 and the resulting precipitate separated by centrifugation at 2000 g for 20 min. This precipitate was redispersed in 4 M-(NH 4 ) 2 SO 4 and recentrifuged. The sediment was dissolved in water and dialysed exhaustively against water. The protein solution was checked periodically for ammonia. When ammonia free, the solution was freeze-dried. The CMC content of the dried mixture was 2-4%. The purity of the protein as /Mactoglobulin was confirmed by starch-gel electrophoresis. /?-Lactoglobulin was also prepared by method I l a as described by McKenzie (1967). Carboxymethyl cellulose CMC with degrees of substitution (DOS) from 0-64 to 1-47 were prepared by the method of Klug & Tinsley (1950) modified as described by Hill & Zadow (1974). Determination of the optimum pH of complex formation, the optimum ratio of carboxymethyl cellulose to protein, the efficiency of precipitation of protein by carboxymethyl cellulose and the influence of ionic strength on complex formation These factors were determined by the methods of Hill & Zadow (1974). pH shift The change in pH on mixing samples of CMC and protein, each initially at the same pH and ionic strength, were determined as follows. CMC of DOS 0-80 was dissolved in 0-1 M-KC1. Hydrochloric acid (0-1 M) was added to 20-ml quantities of the CMC solution, the volume added ranging from zero to that required for complete titration of the carboxyls, in 0-1 ml increments. The total volume of each sample was made up to 25 ml by addition of 0-1 M-KC1. The samples were shaken vigorously for 20 min, and stored overnight under N 2 at 2 °C. The pH of the samples was then determined at 25 °C. The proteins used in the trials were yff-lactoglobulin, bovine serum albumin and Na caseinate, dissolved in a solution of 0-1M-KC1. Twenty-ml quantities of the protein solution were then adjusted to the pH of the CMC solutions by addition of 0-1 M-HC1. The volume of each protein solution was then adjusted to 25 ml by addition of 0-1 M-
Precipitation of proteins by CMC
269
KC1 solution. The pH of the protein solution was rechecked, and if necessary, readjusted to that of the corresponding CMC solution. Only a small volume (< 0-1 ml) of 0-1 M-HCI was required for this final adjustment. The temperatures of the CMC solution and the corresponding protein solution were adjusted to 25 °C, the solutions mixed thoroughly and the pH of the mixture determined. Protein concentrations were all 1-6 g/1 in the CMC-protein mixture. CMC concentrations in the mixture were 0-6 g/1 for bovine serum albumin, 0-8 g/1 for /Mactoglobulin and 1-2 g/1 for Na caseinate. Analytical The DOS of the carboxymethyl groups in CMC was determined by the non-aqueous method of the American Society for Testing and Materials (1965). pH was determined using a Model 26 pH meter (Radiometer Copenhagen N.V., Denmark) fitted with Radiometer combined electrode 6K2301C. Carboxymethyl cellulose content of /Mactoglobulin preparations was determined by the method of Hansen & Chang (1968). RESULTS AND DISCUSSION
The efficiency of precipitation of a protein by CMC was defined as the percentage of protein originally in solution which formed an insoluble complex with CMC (Hill & Zadow, 1974). The influence of pH on the efficiency of precipitation of /Mactoglobulin, bovine serum albumin and Na caseinate by CMC of DOS 0-8 is shown in Fig. 1. Protein concentration was 2 g/1 and ionic strength 0-1 in each case. CMC concentrations were 0-8 g/1 for bovine serum albumin trials, 1 g/1 for /Mactoglobulin trials and 1-5 g/1 for Na caseinate trials. It was possible to precipitate more than 90 % of each of the proteins as an insoluble CMC complex at the correct pH. Other trials (Hill & Zadow, unpublished data) have shown that the pH of maximum precipitation of a proteinCMC complex does not depend on the DOS of the CMC. The pH of maximum precipitation was 2-8 for bovine serum albumin-CMC complexes and Na caseinate-CMC complexes, and 3-2 for yff-lactoglobulin-CMC complexes. yff-Lactoglobulin isolated by both methods behaved similarly in these trials. An optimum pH of 4-0 has been reported by Hidalgo & Hansen (1969) for formation of complexes between /Mactoglobulin and CMC. The concentration of /Mactoglobulin solution used by these authors was 6 g/1. The ratio of CMC to protein for maximum precipitation depends on both the CMC DOS and on the particular protein, as is shown in Fig. 2. Similar results for whey-CMC mixtures have been included in this Figure, the ordinate being expressed as the ratio of CMC to total protein in the whey. The ratio of CMC to yff-lactoglobulin, Na caseinate and bovine serum albumin for maximum precipitation remained fairly constant over the CMC DOS range 0-6-1-10. Above this value the ratio for maximum precipitation increased. The results shown in Fig. 2 contrast with those reported by Hidalgo & Hansen (1969), who found that the ratio of CMC to /Mactoglobulin for maximum precipitation decreased with increasing CMC DOS. The selection of the optimum ratio of CMC to protein can be difficult, as high efficiencies of precipitation are observed over a wide range of ratios at low ionic
270
J. G. ZADOW AND R. D. H I L L
ioor 90
80
tate
70
60
'u
a u
a.
50
tei
c
o t_ a.
40 30
20
10
22
24
2-6
2-8
30
32
34
3-6
38
Pig. 1. The effect of pH on the precipitation of yJ-lactoglobulin, sodium caseinate and bovine serum albumin by CMC of degree of substitution 0-8. Ionic strength 0-1. Protein concentration 2 eft- •> /J-lactoglobulin-CMC, CMC concentration, 1 g/1; A. Na caseinate-CMC, CMC concentration, 1-5 g/1; O> bovine serum albumin-CMC, CMC concentration, 0-8 g/1.
strength. This effect is shown in Fig 3, which shows the percentage of bovine serum albumin precipitated by CMC at pH 2-8 as a function of the ratio of CMC to protein. The results are reported at ionic strengths of 0-05, 0-10 and 0-20. The CMC had a DOS of 1-10, and the protein concentration was 2-5 g/1. At an ionic strength of 0-05, more than 90 % of the protein was precipitated if the ratio of CMC to protein lay between 0'23 and 0'54. At an ionic strength of 0-1, more than 90 % of the protein was precipitated if the ratio of CMC to protein lay between 0-32 and 0-48. At an ionic strength of 0-2, a ratio close to 0-4 was necessary. The optimum ratios of CMC to protein reported in Fig. 2 were determined at an ionic strength of 0-1. This value was chosen for 2 reasons. First, the curve at ionic strength 0-1 (Fig. 3) shows a fairly sharp rate of change of efficiency of precipitation with the ratio of CMC to protein. Secondly, at ionic strengths sensibly above 0-1, it is not possible to achieve high efficiencies of precipitation with CMC of low DOS (Hill & Zadow, 1974). In such cases, the optimum ratio was often difficult to select. The effect of ionic strength on the percentage of bovine serum albumin precipitated by CMC of varying DOS is shown in Fig. 4. Similar data for Na caseinate and
Precipitation of proteins by CMC
271
1-5
2
a. 10 U V o
1 E 0-5 3
06
07
08
09
10
1-1 DOS
1-2
1-3
1-4
15
Fig. 2. The effect of degree of substitution (DOS) of CMC on the ratio of CMC to protein required for maximum precipitation, determined at an ionic strength of 0-1. O» Bovine serum albuminCMC; • , CMC-/Mactoglobulin; A, CMC-Na caseinate; A, CMC-whey. 100
r
90
S
80
70
60 50 0-10
0-20
0-30 0-40 0-50 Ratio CMC:bovine serum albumin
060
Fig. 3. The percentage of bovine serum albumin precipitated at different ratios of CMC to protein and at different ionic strengths. Protein concentration 2-5 g/1, CMC degree of substitution 1-10. O. Ionic strength 0-05; # , ionic strength 0-10; A. ionic strength 0-20.
yff-lactoglobulin are shown in Figs 5 and 6 respectively. Protein concentration in all cases was 2 g/1. The ratio of CMC to protein was the optimum value for the particular CMC DOS and protein, determined from Fig. 2. In each case, the amount of protein precipitated decreased with increasing ionic strength. At a given ionic strength, more protein was precipitated by CMC of higher DOS than by CMC of lower DOS. This effect was greatest at high ionic strengths. Under given conditions of ionic strength and of CMC DOS, approximately the same percentages of /?-lactoglobulin and Na caseinate were precipitated. However, more bovine serum albumin was precipitated under the same conditions.
272
J. G. ZADOW AND R. D. H I L L 100 -
IUU
DOS 1-47 1 10
tate
90
80
80 -
70
70 \ \ \ \ \
40
0
I 0 64
30
10 20
0
1 60 g
50
a. c
tei
0 80
60
'a. 'u
90 -
a. c 2 o
50 40
Q_
30 20 10
l 0-10 Ionic strength
1
020
Fig. 4. The influence of ionic strength on the percentage of bovine serum albumin precipitated by CMC of different degrees of substitution (DOS). Protein concentration 2 g/1. O, CMC DOS 1-47; « , CMC DOS 1-10; A, CMC DOS 0-80; A, CMC DOS 0-64.
0
0-1
0-2
Ionic strength
Fig. 5. The influence of ionic strength on the percentage of Na caseinate precipitated by CMC of different degrees of substitution (DOS). Protein concentration 2 g/1. O, CMC DOS 1-47; « , CMC DOS 1-10; A, CMC DOS 0-80; A, CMC DOS 0-64.
The change in pH on mixing solutions of protein and CMC, each initially at the same pH, is shown in Fig. 7. The ratio of CMC to protein was selected to be close to the optimum value for the particular protein as shown in Fig. 2. The value of ApH was defined as the difference in pH of the CMC-protein complex and the pH of the component solutions. The results in Fig. 7 show that the effects that occur on mixing CMC and the different proteins are complex. Negative pH shifts are the consistent effect at low pH, when the protein has its greatest net positive charge. At such pH values (»3) the CMC is normally only partly ionized (a«0-2) (Zadow, unpublished data), and the interaction with positively charged protein modifies the electric field around the CMC, inducing further ionization of the CMC carboxyl groups. Calculations based on the formation of an electrically neutral CMC-/?-lactoglobulin complex have shown that for CMC of DOS 0*8 an increase in the degree of dissociation of 042 was required for the formation of an electrically neutral, precipitatedc omplex with /?lactoglobulin at pH 3-2. This theoretical increase accords well with a figure of 0403 calculated from the pH shifts observed in this study (Table 1). These calculations are detailed in the Appendix. However, negative pH shifts on mixing /Mactoglobulin and CMC also occur at pH values above 4-7 (Fig. 7) when the protein is negatively charged.
Precipitation of proteins by CMC 100
273
r
90 80
^
70
1 60 50 S o
40 30 20 0-64
10
01
0-2
Ionic strength
Fig. 6. The influence of ionic strength on the percentage of /?-lactoglobulin precipitated by CMC of differing degrees of substitution (DOS). Protein concentration 2 g/1. O, CMC DOS 1-47; « , CMC DOS 1-10; A, CMC DOS 0-80; A, CMC DOS 0-64.
-0-5 20
30
40 50 pH of individual components
60
Fig. 7. The change in pH (ApH) on mixing solution of CMC and protein each initially at thesame pH. — — —, CMC-/J-lactoglobulin; , CMC-bovine serum albumin; ,. CMC-Na casemate.
274
J. G. ZADOW AND R. D. H I L L
Table 1. The values of AH+ for fi-lactoglobulin-CMC mixtures pH of individual components 600 5-50 5-00 4-50 4-00 3-50 300
pH on mixing 5-53 5-00 4-52
AH+
ApH
-0-47 -0-50 -0-48 -0-40 -0-31 -0-21
410
3-69 3-30 2-89
-on
(mole/1 xlO6) 2 7
20 50 100 180 290
This effect is much smaller in terms of proton release than that at lower pH values (Table 1), and may be due to the interaction of the CMC with regions of the protein that have a concentration of positively charged groups. In /Mactoglobulin, for example, there are 6 basic residues between residues 126 and 150, and 6 basic residues between residues 76 and 102 (Braunitzer & Chen, 1972; Braunitzer et al. 1972). The positive values of ApH observed at pH values above 3-2 for bovine serum albumin and casein (Fig. 7) represent a net uptake of proteins from solution and must be the result of a different process, such as the protonation of a previously uncharged amino group or the discharging of a previously charged carboxyl group. For example, an ion pair (protonated amine and ionized carboxyl) may continue to exist in a hydrophobic region of a protein at a pH well below the normal pK of the carboxyl partner. If such a pair were broken during the interaction of the protein with CMC, either as a direct result of a local interaction or because of a conformational change, the amine would remain protonated, but the carboxyl group would be discharged by taking up a proton from solution. The actual pH shift that occurs when the protein and CMC are mixed must therefore be the net result of several factors. For the 3 proteins studied, however, there is a negative pH shift under the conditions where the CMC-protein complex is precipitated most efficiently. CMC is highly hydrophilic, but it becomes less so as pH is reduced when its ionization, and hence hydration, are also reduced. The interaction will therefore be strongest at the lowest pH at which an electrically neutral complex can be formed, when the complex will have its least hydrophilic character. The technical assistance of Mr J. F. Hardham is gratefully acknowledged. This work was supported in part by grants from the Dairy Industry Research Fund administered by the Australian Dairy Produce Board.
APPENDIX
Calculation of change in degree of dissociation of carboxymethyl cellulose required for formation of a neutral carboxymethyl cellulose-fi-lactoglobulin complex A. Based on acid strength of carboxymethyl cellulose The degree of substitution (DOS) of carboxymethyl cellulose (CMC) may be defined as the average number of carboxymethyl groups introduced per anhydroglucose unit
Precipitation of proteins by CMC
275
in the cellulose. The average mol. wt per CMC anhydroglucose unit may be calculated from the formula: M = 162 + 80Z), where M is the average mol. wt per CMC anhydroglucose unit, the mol. wt of a cellulose anhydroglucose unit being 162 and the net increase in mol. wt of an anhydroglucose unit for each Na carboxymethyl group added being 80. The Na salt of CMC of DOS 0-80 therefore has an average anhydroglucose unit mol. wt of 226. Thus, 1 g of this CMC contains 0-80/226 (i.e. 3-54 x lO"3) moles of carboxyl groups. /?-Lactoglobulin of mol. wt 36000 contains 0-0278 x 10~3 moles/g. In solution at pH 3-2 and ionic strength 0-15, /?-lactoglobulin has a net positive charge of approx. 26 groups/molecule (Basch & Timasheff, 1967). Under such conditions, y#-lactoglobulin will contain 0-72 x 10~3 moles of net positive charge/g. The CMC-/ff-lactoglobulin complex precipitated at pH 3-2 and ionic strength 0-1 contains approx 65 % protein (Hill & Zadow, unpublished data). Thus, 1 g of complex contains 0-65 g of yff-lactoglobulin, which in turn should carry 0-65 x 0-72 x 10~3 (i.e. 0-47 x 10~3) moles of net positive charge. This charge must be neutralized by the anionic charge of the CMC carboxyl groups. Thus, 0-35 g of CMC must contain 0-47 x 10~3 moles of ionized carboxyl groups, i.e. 1-34 x 10~3 moles of ionized carboxyl groups/g. This requires a degree of dissociation of (1-34 x 10~3)/(3-54 x 10~3) i.e. 0-38. The degree of dissociation of CMC of DOS = 0-80 at ionic strength of 0-1 and at pH 3-2 has been estimated to be 0-26 (Zadow, unpublished data). Thus, an increase in the degree of dissociation of the CMC of approx. 0-12 would be necessary to form a neutral complex between CMC and /Mactoglobulin under these conditions. B. Based on the pH shift reported in Table 1 The CMC concentration used in these trials was 0-8 g/1 and thus 11 of such solution would contain 0-8 x 3-54 x 10~3 (i.e. 2-83 x 10~3) moles of carboxyl groups. An increase in H + content of 290xl0~ 6 moles/1 was observed on mixing CMC and /Mactoglobulin at pH 3-2 and ionic strength 0-1. This increase in H + content would be equivalent to a change in the degree of dissociation of the CMC of (290 x 10~6)/ (2-83 xlO- 3 ), i.e. 0-103. REFERENCES AMERICAN SOCIETY FOB TESTING AND MATERIALS (1965). ASTM Standard D-1439-65.
BASCH, J. J. & TIMASHEFF, S. M. (1967). Archives of Biochemistry and Biophysics 118, 37. BRAUKITZEB, G. & CHEN, R. (1972). Hoppe-Seyler's Zeitschrift fur Physiologische Chemie 353, 674. BRAUNITZER, G., CHEN, R., SCHRANK, B. & STASTGL, A. (1972). Hoppe-Seyler's Zeitschift fur Physiologische
Chemie 353, 832. HANSEN, P. M. T. & CHANG, J. C. (1968). Journal of Agricultural and Food Chemistry 16, 77. HANSEN, P. M. T., HIDALGO, J. & GOULD, I. A. (1971). Journal of Dairy Science 54, 830.
HIDALGO, J. & HANSEN, P. M. T. (1969). Journal of Agricultural and Food Chemistry 17, 1089. HIDALGO, J. & HANSEN, P. M. T. (1971). Journal of Dairy Science 54, 1270. HILL, R. D. & ZADOW, J. G. (1974). Journal of Dairy Research 41, 373. KLUG, E. G. & TINSLEY, J. G. (1950). U.S. Patent 2,517,577. MCKENZIE, H. A. (1967). Advances in Protein Chemistry 22, 55. Printed in Great Britain
267
The precipitation of proteins by carboxymethyl cellulose BY J. G. ZADOW AND R. D. HILL Division of Food Research, Dairy Research Laboratory, C.S.I.R.O, Melbourne, Victoria, Australia {Received 10 October 1974) SUMMARY. Carboxymethyl cellulose (CMC) formed insoluble complexes with /?lactoglobulin, bovine serum albumin and Na caseinate. Maximum precipitation of the /?-lactoglobulin-CMC complex occurred at pH 3-2, whereas maximum precipitation of the bovine serum albumin-CMC complex and the Na caseinate-CMC complex occurred at pH 2-8. The ratio of CMC to protein for maximum precipitation depended on the protein, being greatest for Na caseinate and least for bovine serum albumin. The percentage of protein precipitated by CMC decreased with increasing ionic strength of the solution, the rate of decrease being least for bovine serum albumin. At a given ionic strength, more protein was precipitated by CMC of high degree of substitution than by CMC of low degree of substitution. The change in pH (ApH) occurring on mixing CMC and unbuffered protein solutions, each initially at the same pH, was measured. ApH was negative for /?-lactoglobulin-CMC mixtures over the pH range 7-2 (minimum at pH 5-5). For bovine serum albumin-CMC and Na caseinate-CMC mixtures, ApH was positive between pH 7 and 3-2 (maximum at pH 4-5), zero at pH 3-2 and negative between pH 3-2 and 2-0 (minimum at pH 2-8).
The formation of insoluble complexes between carboxymethyl cellulose (CMC) and the proteins in whey has been suggested by Hansen, Hidalgo & Gould (1971) as a means of reclaiming whey proteins. The extent of precipitation of the protein depends on the degree of substitution of the CMC, and on the dilution (i.e. ionic strength) of the whey (Hill & Zadow, 1974). It has been suggested that these insoluble CMCprotein complexes are formed as a result of charge neutralization between the positively charged protein and the negatively charged CMC (Hidalgo & Hansen, 1969). In an attempt to obtain more information about the mechanism of complex formation, the conditions for precipitation by CMC of /ff-lactoglobulin, casein, and bovine serum albumin have been studied. As interactions between protein and CMC can be expected to alter the ionization of the CMC (Hill & Zadow, 1974), the changes in pH that occurred on mixing solutions of CMC and protein, each initially at the same pH, were also studied. „'
268
J. G. ZADOW AND R. D. HILL EXPERIMENTAL
Proteins Bovine serum albumin. Bovine serum albumin was supplied by the Commonwealth Serum Laboratories. Sodium caseinate. Casein was precipitated from skim-milk by adjustment of the pH to 4-5 using HC1. The casein was separated and redissolved using NaOH, the pH of the resulting solution being maintained at approximately 7-0. The Na caseinate solution was then freeze-dried. fi-Lactoglobulin. /?-Lactoglobulin was prepared by selective precipitation of this protein from whey in the form of a CMC complex (Hidalgo & Hansen, 1971). After concentration of the complex by centrifugation at 200 g for 20 min, a small volume of water was added to the sediment and the pH of the dispersion adjusted to 7-0 with 1 M-NaOH. After stirring overnight at 2 °C, the mixture was centrifuged at 2000 g for 20 min and the sediment discarded. The supernatant was made 4 M with respect to (NH4)2SO4 and the resulting precipitate separated by centrifugation at 2000 g for 20 min. This precipitate was redispersed in 4 M-(NH 4 ) 2 SO 4 and recentrifuged. The sediment was dissolved in water and dialysed exhaustively against water. The protein solution was checked periodically for ammonia. When ammonia free, the solution was freeze-dried. The CMC content of the dried mixture was 2-4%. The purity of the protein as /Mactoglobulin was confirmed by starch-gel electrophoresis. /?-Lactoglobulin was also prepared by method I l a as described by McKenzie (1967). Carboxymethyl cellulose CMC with degrees of substitution (DOS) from 0-64 to 1-47 were prepared by the method of Klug & Tinsley (1950) modified as described by Hill & Zadow (1974). Determination of the optimum pH of complex formation, the optimum ratio of carboxymethyl cellulose to protein, the efficiency of precipitation of protein by carboxymethyl cellulose and the influence of ionic strength on complex formation These factors were determined by the methods of Hill & Zadow (1974). pH shift The change in pH on mixing samples of CMC and protein, each initially at the same pH and ionic strength, were determined as follows. CMC of DOS 0-80 was dissolved in 0-1 M-KC1. Hydrochloric acid (0-1 M) was added to 20-ml quantities of the CMC solution, the volume added ranging from zero to that required for complete titration of the carboxyls, in 0-1 ml increments. The total volume of each sample was made up to 25 ml by addition of 0-1 M-KC1. The samples were shaken vigorously for 20 min, and stored overnight under N 2 at 2 °C. The pH of the samples was then determined at 25 °C. The proteins used in the trials were yff-lactoglobulin, bovine serum albumin and Na caseinate, dissolved in a solution of 0-1M-KC1. Twenty-ml quantities of the protein solution were then adjusted to the pH of the CMC solutions by addition of 0-1 M-HC1. The volume of each protein solution was then adjusted to 25 ml by addition of 0-1 M-
Precipitation of proteins by CMC
269
KC1 solution. The pH of the protein solution was rechecked, and if necessary, readjusted to that of the corresponding CMC solution. Only a small volume (< 0-1 ml) of 0-1 M-HCI was required for this final adjustment. The temperatures of the CMC solution and the corresponding protein solution were adjusted to 25 °C, the solutions mixed thoroughly and the pH of the mixture determined. Protein concentrations were all 1-6 g/1 in the CMC-protein mixture. CMC concentrations in the mixture were 0-6 g/1 for bovine serum albumin, 0-8 g/1 for /Mactoglobulin and 1-2 g/1 for Na caseinate. Analytical The DOS of the carboxymethyl groups in CMC was determined by the non-aqueous method of the American Society for Testing and Materials (1965). pH was determined using a Model 26 pH meter (Radiometer Copenhagen N.V., Denmark) fitted with Radiometer combined electrode 6K2301C. Carboxymethyl cellulose content of /Mactoglobulin preparations was determined by the method of Hansen & Chang (1968). RESULTS AND DISCUSSION
The efficiency of precipitation of a protein by CMC was defined as the percentage of protein originally in solution which formed an insoluble complex with CMC (Hill & Zadow, 1974). The influence of pH on the efficiency of precipitation of /Mactoglobulin, bovine serum albumin and Na caseinate by CMC of DOS 0-8 is shown in Fig. 1. Protein concentration was 2 g/1 and ionic strength 0-1 in each case. CMC concentrations were 0-8 g/1 for bovine serum albumin trials, 1 g/1 for /Mactoglobulin trials and 1-5 g/1 for Na caseinate trials. It was possible to precipitate more than 90 % of each of the proteins as an insoluble CMC complex at the correct pH. Other trials (Hill & Zadow, unpublished data) have shown that the pH of maximum precipitation of a proteinCMC complex does not depend on the DOS of the CMC. The pH of maximum precipitation was 2-8 for bovine serum albumin-CMC complexes and Na caseinate-CMC complexes, and 3-2 for yff-lactoglobulin-CMC complexes. yff-Lactoglobulin isolated by both methods behaved similarly in these trials. An optimum pH of 4-0 has been reported by Hidalgo & Hansen (1969) for formation of complexes between /Mactoglobulin and CMC. The concentration of /Mactoglobulin solution used by these authors was 6 g/1. The ratio of CMC to protein for maximum precipitation depends on both the CMC DOS and on the particular protein, as is shown in Fig. 2. Similar results for whey-CMC mixtures have been included in this Figure, the ordinate being expressed as the ratio of CMC to total protein in the whey. The ratio of CMC to yff-lactoglobulin, Na caseinate and bovine serum albumin for maximum precipitation remained fairly constant over the CMC DOS range 0-6-1-10. Above this value the ratio for maximum precipitation increased. The results shown in Fig. 2 contrast with those reported by Hidalgo & Hansen (1969), who found that the ratio of CMC to /Mactoglobulin for maximum precipitation decreased with increasing CMC DOS. The selection of the optimum ratio of CMC to protein can be difficult, as high efficiencies of precipitation are observed over a wide range of ratios at low ionic
270
J. G. ZADOW AND R. D. H I L L
ioor 90
80
tate
70
60
'u
a u
a.
50
tei
c
o t_ a.
40 30
20
10
22
24
2-6
2-8
30
32
34
3-6
38
Pig. 1. The effect of pH on the precipitation of yJ-lactoglobulin, sodium caseinate and bovine serum albumin by CMC of degree of substitution 0-8. Ionic strength 0-1. Protein concentration 2 eft- •> /J-lactoglobulin-CMC, CMC concentration, 1 g/1; A. Na caseinate-CMC, CMC concentration, 1-5 g/1; O> bovine serum albumin-CMC, CMC concentration, 0-8 g/1.
strength. This effect is shown in Fig 3, which shows the percentage of bovine serum albumin precipitated by CMC at pH 2-8 as a function of the ratio of CMC to protein. The results are reported at ionic strengths of 0-05, 0-10 and 0-20. The CMC had a DOS of 1-10, and the protein concentration was 2-5 g/1. At an ionic strength of 0-05, more than 90 % of the protein was precipitated if the ratio of CMC to protein lay between 0'23 and 0'54. At an ionic strength of 0-1, more than 90 % of the protein was precipitated if the ratio of CMC to protein lay between 0-32 and 0-48. At an ionic strength of 0-2, a ratio close to 0-4 was necessary. The optimum ratios of CMC to protein reported in Fig. 2 were determined at an ionic strength of 0-1. This value was chosen for 2 reasons. First, the curve at ionic strength 0-1 (Fig. 3) shows a fairly sharp rate of change of efficiency of precipitation with the ratio of CMC to protein. Secondly, at ionic strengths sensibly above 0-1, it is not possible to achieve high efficiencies of precipitation with CMC of low DOS (Hill & Zadow, 1974). In such cases, the optimum ratio was often difficult to select. The effect of ionic strength on the percentage of bovine serum albumin precipitated by CMC of varying DOS is shown in Fig. 4. Similar data for Na caseinate and
Precipitation of proteins by CMC
271
1-5
2
a. 10 U V o
1 E 0-5 3
06
07
08
09
10
1-1 DOS
1-2
1-3
1-4
15
Fig. 2. The effect of degree of substitution (DOS) of CMC on the ratio of CMC to protein required for maximum precipitation, determined at an ionic strength of 0-1. O» Bovine serum albuminCMC; • , CMC-/Mactoglobulin; A, CMC-Na caseinate; A, CMC-whey. 100
r
90
S
80
70
60 50 0-10
0-20
0-30 0-40 0-50 Ratio CMC:bovine serum albumin
060
Fig. 3. The percentage of bovine serum albumin precipitated at different ratios of CMC to protein and at different ionic strengths. Protein concentration 2-5 g/1, CMC degree of substitution 1-10. O. Ionic strength 0-05; # , ionic strength 0-10; A. ionic strength 0-20.
yff-lactoglobulin are shown in Figs 5 and 6 respectively. Protein concentration in all cases was 2 g/1. The ratio of CMC to protein was the optimum value for the particular CMC DOS and protein, determined from Fig. 2. In each case, the amount of protein precipitated decreased with increasing ionic strength. At a given ionic strength, more protein was precipitated by CMC of higher DOS than by CMC of lower DOS. This effect was greatest at high ionic strengths. Under given conditions of ionic strength and of CMC DOS, approximately the same percentages of /?-lactoglobulin and Na caseinate were precipitated. However, more bovine serum albumin was precipitated under the same conditions.
272
J. G. ZADOW AND R. D. H I L L 100 -
IUU
DOS 1-47 1 10
tate
90
80
80 -
70
70 \ \ \ \ \
40
0
I 0 64
30
10 20
0
1 60 g
50
a. c
tei
0 80
60
'a. 'u
90 -
a. c 2 o
50 40
Q_
30 20 10
l 0-10 Ionic strength
1
020
Fig. 4. The influence of ionic strength on the percentage of bovine serum albumin precipitated by CMC of different degrees of substitution (DOS). Protein concentration 2 g/1. O, CMC DOS 1-47; « , CMC DOS 1-10; A, CMC DOS 0-80; A, CMC DOS 0-64.
0
0-1
0-2
Ionic strength
Fig. 5. The influence of ionic strength on the percentage of Na caseinate precipitated by CMC of different degrees of substitution (DOS). Protein concentration 2 g/1. O, CMC DOS 1-47; « , CMC DOS 1-10; A, CMC DOS 0-80; A, CMC DOS 0-64.
The change in pH on mixing solutions of protein and CMC, each initially at the same pH, is shown in Fig. 7. The ratio of CMC to protein was selected to be close to the optimum value for the particular protein as shown in Fig. 2. The value of ApH was defined as the difference in pH of the CMC-protein complex and the pH of the component solutions. The results in Fig. 7 show that the effects that occur on mixing CMC and the different proteins are complex. Negative pH shifts are the consistent effect at low pH, when the protein has its greatest net positive charge. At such pH values (»3) the CMC is normally only partly ionized (a«0-2) (Zadow, unpublished data), and the interaction with positively charged protein modifies the electric field around the CMC, inducing further ionization of the CMC carboxyl groups. Calculations based on the formation of an electrically neutral CMC-/?-lactoglobulin complex have shown that for CMC of DOS 0*8 an increase in the degree of dissociation of 042 was required for the formation of an electrically neutral, precipitatedc omplex with /?lactoglobulin at pH 3-2. This theoretical increase accords well with a figure of 0403 calculated from the pH shifts observed in this study (Table 1). These calculations are detailed in the Appendix. However, negative pH shifts on mixing /Mactoglobulin and CMC also occur at pH values above 4-7 (Fig. 7) when the protein is negatively charged.
Precipitation of proteins by CMC 100
273
r
90 80
^
70
1 60 50 S o
40 30 20 0-64
10
01
0-2
Ionic strength
Fig. 6. The influence of ionic strength on the percentage of /?-lactoglobulin precipitated by CMC of differing degrees of substitution (DOS). Protein concentration 2 g/1. O, CMC DOS 1-47; « , CMC DOS 1-10; A, CMC DOS 0-80; A, CMC DOS 0-64.
-0-5 20
30
40 50 pH of individual components
60
Fig. 7. The change in pH (ApH) on mixing solution of CMC and protein each initially at thesame pH. — — —, CMC-/J-lactoglobulin; , CMC-bovine serum albumin; ,. CMC-Na casemate.
274
J. G. ZADOW AND R. D. H I L L
Table 1. The values of AH+ for fi-lactoglobulin-CMC mixtures pH of individual components 600 5-50 5-00 4-50 4-00 3-50 300
pH on mixing 5-53 5-00 4-52
AH+
ApH
-0-47 -0-50 -0-48 -0-40 -0-31 -0-21
410
3-69 3-30 2-89
-on
(mole/1 xlO6) 2 7
20 50 100 180 290
This effect is much smaller in terms of proton release than that at lower pH values (Table 1), and may be due to the interaction of the CMC with regions of the protein that have a concentration of positively charged groups. In /Mactoglobulin, for example, there are 6 basic residues between residues 126 and 150, and 6 basic residues between residues 76 and 102 (Braunitzer & Chen, 1972; Braunitzer et al. 1972). The positive values of ApH observed at pH values above 3-2 for bovine serum albumin and casein (Fig. 7) represent a net uptake of proteins from solution and must be the result of a different process, such as the protonation of a previously uncharged amino group or the discharging of a previously charged carboxyl group. For example, an ion pair (protonated amine and ionized carboxyl) may continue to exist in a hydrophobic region of a protein at a pH well below the normal pK of the carboxyl partner. If such a pair were broken during the interaction of the protein with CMC, either as a direct result of a local interaction or because of a conformational change, the amine would remain protonated, but the carboxyl group would be discharged by taking up a proton from solution. The actual pH shift that occurs when the protein and CMC are mixed must therefore be the net result of several factors. For the 3 proteins studied, however, there is a negative pH shift under the conditions where the CMC-protein complex is precipitated most efficiently. CMC is highly hydrophilic, but it becomes less so as pH is reduced when its ionization, and hence hydration, are also reduced. The interaction will therefore be strongest at the lowest pH at which an electrically neutral complex can be formed, when the complex will have its least hydrophilic character. The technical assistance of Mr J. F. Hardham is gratefully acknowledged. This work was supported in part by grants from the Dairy Industry Research Fund administered by the Australian Dairy Produce Board.
APPENDIX
Calculation of change in degree of dissociation of carboxymethyl cellulose required for formation of a neutral carboxymethyl cellulose-fi-lactoglobulin complex A. Based on acid strength of carboxymethyl cellulose The degree of substitution (DOS) of carboxymethyl cellulose (CMC) may be defined as the average number of carboxymethyl groups introduced per anhydroglucose unit
Precipitation of proteins by CMC
275
in the cellulose. The average mol. wt per CMC anhydroglucose unit may be calculated from the formula: M = 162 + 80Z), where M is the average mol. wt per CMC anhydroglucose unit, the mol. wt of a cellulose anhydroglucose unit being 162 and the net increase in mol. wt of an anhydroglucose unit for each Na carboxymethyl group added being 80. The Na salt of CMC of DOS 0-80 therefore has an average anhydroglucose unit mol. wt of 226. Thus, 1 g of this CMC contains 0-80/226 (i.e. 3-54 x lO"3) moles of carboxyl groups. /?-Lactoglobulin of mol. wt 36000 contains 0-0278 x 10~3 moles/g. In solution at pH 3-2 and ionic strength 0-15, /?-lactoglobulin has a net positive charge of approx. 26 groups/molecule (Basch & Timasheff, 1967). Under such conditions, y#-lactoglobulin will contain 0-72 x 10~3 moles of net positive charge/g. The CMC-/ff-lactoglobulin complex precipitated at pH 3-2 and ionic strength 0-1 contains approx 65 % protein (Hill & Zadow, unpublished data). Thus, 1 g of complex contains 0-65 g of yff-lactoglobulin, which in turn should carry 0-65 x 0-72 x 10~3 (i.e. 0-47 x 10~3) moles of net positive charge. This charge must be neutralized by the anionic charge of the CMC carboxyl groups. Thus, 0-35 g of CMC must contain 0-47 x 10~3 moles of ionized carboxyl groups, i.e. 1-34 x 10~3 moles of ionized carboxyl groups/g. This requires a degree of dissociation of (1-34 x 10~3)/(3-54 x 10~3) i.e. 0-38. The degree of dissociation of CMC of DOS = 0-80 at ionic strength of 0-1 and at pH 3-2 has been estimated to be 0-26 (Zadow, unpublished data). Thus, an increase in the degree of dissociation of the CMC of approx. 0-12 would be necessary to form a neutral complex between CMC and /Mactoglobulin under these conditions. B. Based on the pH shift reported in Table 1 The CMC concentration used in these trials was 0-8 g/1 and thus 11 of such solution would contain 0-8 x 3-54 x 10~3 (i.e. 2-83 x 10~3) moles of carboxyl groups. An increase in H + content of 290xl0~ 6 moles/1 was observed on mixing CMC and /Mactoglobulin at pH 3-2 and ionic strength 0-1. This increase in H + content would be equivalent to a change in the degree of dissociation of the CMC of (290 x 10~6)/ (2-83 xlO- 3 ), i.e. 0-103. REFERENCES AMERICAN SOCIETY FOB TESTING AND MATERIALS (1965). ASTM Standard D-1439-65.
BASCH, J. J. & TIMASHEFF, S. M. (1967). Archives of Biochemistry and Biophysics 118, 37. BRAUKITZEB, G. & CHEN, R. (1972). Hoppe-Seyler's Zeitschrift fur Physiologische Chemie 353, 674. BRAUNITZER, G., CHEN, R., SCHRANK, B. & STASTGL, A. (1972). Hoppe-Seyler's Zeitschift fur Physiologische
Chemie 353, 832. HANSEN, P. M. T. & CHANG, J. C. (1968). Journal of Agricultural and Food Chemistry 16, 77. HANSEN, P. M. T., HIDALGO, J. & GOULD, I. A. (1971). Journal of Dairy Science 54, 830.
HIDALGO, J. & HANSEN, P. M. T. (1969). Journal of Agricultural and Food Chemistry 17, 1089. HIDALGO, J. & HANSEN, P. M. T. (1971). Journal of Dairy Science 54, 1270. HILL, R. D. & ZADOW, J. G. (1974). Journal of Dairy Research 41, 373. KLUG, E. G. & TINSLEY, J. G. (1950). U.S. Patent 2,517,577. MCKENZIE, H. A. (1967). Advances in Protein Chemistry 22, 55. Printed in Great Britain
