928
Oxidative Interactions of Cholesterol in the Milk Fat Globule Membrane 1 S.K. Kim and W.W. Nawar* Department of Food Science, University of Massachusetts, Amherst, Massachusetts 01003
The effects of oxidative interactions between cholesterol and milk fat globule membrane (MFGM) components,/.e., nonlipid fraction, total lipid, nonpolar lipid and polar lipid, on cholesterol oxidation were studied in the presence and absence of water. In the dry state, cholesterol natively present in MFGM appeared to be protected at 135~ The nonpolar lipid and nonlipid fraction contributed to the protective effect of MFGM. Added cholesterol accelerated the oxidation of membrane lipid fractions. At 75~ pure cholesterol and membrane lipid fractions did not show significant interaction. However, cholesterol and other lipids in M F G M were less stable than when these were heated separately. When cholesterol and membrane lipids were mixed in an aqueous medium at 75~ each accelerated the oxidation of the other. The MFGM exhibited a high protective effect on cholesterol oxidation in an aqueous environment. The nonlipid fraction p r o tected cholesterol against oxidation, whereas the lipid fraction was destructive. In the absence of water, the net balance between these two opposing factors was destructive. The presence of water reversed the balance in favor of protection. Lipids 27, 928-932 (1992). Bovine milk contains approximately 3-5% fat in the form of globules and these are surrounded by a thin membrane (1). This milk fat globule membrane (MFGM) consists of proteins, glycoproteins, triacylglycerols, cholesterol, enzymes and other minor components. The unsaturated lipids in the membrane were reported to play a role in initiating the oxidation of milk fat (2). Chemical and microscopical data (3-7) indicate t h a t the M F G M originates from the plasma membrane of the secretory cells. Therefore, M F G M can be defined as a modified plasma memb r a n e Oxidation of membrane lipids would naturally be expected to cause alterations in membrane function which eventually lead to disturbance in cell physiology. The objective of this work was to investigate the influence of the interaction among the membrane components on their oxidative stability using M F G M as a model membrane system.
EXPERIMENTAL PROCEDURES
Materials. Cholesterol and silicic acid were purchased from Sigma Chemical Co. (St. Louis, MO). Cholesterol used in this study contained less than 0.9% of contaminants, including 7-ketocholesterol, 7/3-hydroxycholesterol, 5,6~epoxides and cholest-3,5-dien-7-one. Cholesterol oxide standards and silylating agents were obtained from Steraloid Inc (Wilton, NH) and Pierce Chemical Co. (Rockford, IL), 1Based on a paper presented at the Symposium on Milk Lipids held at the AOCS Annual Meeting, Baltimore, MD, April 1990. *To whom correspondence should be addressed. Abbreviations: BSTFA, N,O-bis-(trimethylsilyl)trifiuoroacetamide; HPLC, high-performance liquid chromatography; MFGM, milk fat globule membrane; NL, nonlipid fraction; NPL, nonpolar Iipids; PL, polar lipids; TL, total lipids; TLC, thin-layer chromatography; TMCS, trimethylchlorosilane. LIPIDS, Vol. 27, no. 11 (1992)
respectively. Solvents were high-performance liquid chromatography (HPLC) grade from various sources.
Preparation, fractionation and characterization of MFGM. Raw milk was obtained from the University of Massachusetts farm and pasteurized at 63~ for 30 min. Cream was separated by centrifugation, then washed twice with 3 vol of Tris-HC1 buffer (10 mM, pH 7.5 containing 0.25 M of sucrose and 1.0 mM of MgC12) and once with 1 vol of 50 mM Tris-HC1 buffer (pH 7.5) (4). The washed cream was resuspended in 50 mM Tris-HC1 buffer, and the cream structure was destroyed by freezing and thawing to liberate M F G M from the cream structure (8). Butter was separated from the MFGM suspension by centrifugation at 2,500 X g for 20 min. The b u t t e r was washed with 1 vol of distilled water to recover the residual MFGM, and the water part was combined with the M F G M suspension (9). The membrane suspension was centrifuged at 100,000 • g for 1 h to obtain the membrane (10). The pellets were dispersed in distilled water and quickly recentrifuged to remove the buffer salts from the membrane fraction. The membrane was then. freeze-dried and stored at - 4 0 ~ The freeze-dried M F G M was washed 3 times with 5 vol of CHC13/MeOH (2:1, vol/vol) and c e n t r i f u g e d at 5,000 • g for 30 min to obtain the nonlipid fraction (NL) of MFGM, which was dried and stored at - 4 0 ~ Solvent in the extract was removed on a rotary evaporator to obtain crude M F G M lipid. The crude lipid was purified (11) and fractionated into polar and nonpolar fractions by silicic acid column chromatography. Chloroform and methanol were used to elute the nonpolar lipids (NPL) and the polar lipids {PL), respectively (12). PL and N P L fractions were separated into their components by thin-layer chromatography (TLC) (silica gel G, 0.25 nun thickness). The solvent system used for NPL was petroleum hydrocarbon/diethyl ether/acetic acid (90:10:1, vol/vol]vol) (12), and that used for PL was CHCI3/MeOH/ water/18.4% aqueous NH3 (130:70:8:0.5, vol/vol/vol/vol). TLC fractions were visualized by heating the plates at 120~ for 25 min after spraying with a potassium dichromate/sulfuric acid solution. The lipid classes were quantitated with a TLC densitometer. Sample treatment. Solutions of cholesterol and membrane lipid in chloroform (1 mg each) were placed in vials, and the solvent was evaporated on a rotary evaporator to make a thin film on the b o t t o m of each vial. For the dry samples, the vials were placed in an oven at 135~ for 40 h or at 75~ for 40 d. For the aqueous samples, 1 mL of 10 mM Tris buffer was added to each vial and incubated at 75~ for 6 d. The lipids in the reaction tubes were extracted with 3.5 mL of chloroform/methanol (2:1, vol/vol) and divided into two parts, one for cholesterol analysis and the other for f a t t y acid analysis. Cholesterol analysis. Cholesterol and its oxides were separated from the heated mixtures of cholesterol and membrane lipids by silicic acid column chromatography. The NPL, mainly triacylglycerols, were eluted with hexane/diethyl ether (95:5, vol]vol), followed by cholesterol and its oxidation products with 100% diethyl ether. Phospholipids remained on the column. The ether fraction was
929
OXIDATIVE INTERACTIONS IN MEMBRANE LIPIDS collected and silylated with N,O-bis-(trimethylsilyl)trifluoroacetamide (BSTFA) and 1% trimethylchlorosilane (TMCS) at 80~ for 1 h. The trimethylsflyl ethers were separated on a Varian 3700 GC (Varian Associates, Paio Alt~ CA) equipped with an Ultra-1 capillary column (50 m, 0.2 mm i.d. and 0.33 ~m film thickness, HewlettPackard, Avondale, PA) with temperature programming from 100~ to 300~ at the rate of 10~ 5aAndrostan-3/3-ol-17-one acetate was used as an internal standard for the analysis of cholesterol and its oxides. Fatty acid analysis. Oxidation of triacylglycerols and free fatty acids was traced according to the loss of their fatty acyl chains. Fatty acid analysis was done by methylating the whole lipid portion directly without using silicic acid column chromatography. The lipid samples containing 1-2 mg of lipid were saponified with 0.5 mL of 2N methanolic KOH solution at 65~ for 25 min, then methylated with 0.5 mL of 14% BF3/methanol solution at 65~ After 25 min reaction, residual methylating agent was deactivated by adding 0.5 mL of water. The fatty acid methyl esters were extracted with hexane and analyzed by gas chromatography on a Supelcowax-10 capillary column (30 m, 0.32 mm i.d. and 0.25 ~m film thickness, Supelco Ina, Bellefonte, PA). Triheptadecanoin was used as an internal standard. The graphs in this report represent the averages of triplicate analyses.
100 ,-, o~
80-
m
O
60-
i
.c O
40-
.~.~m
E
0-
rr
CHOL only
9
"Jl:
0
0
I0
2'0
3o
40
Time (h) FIG. 1. Effect of membrane components on cholesterol oxidation in the dry state at 135~ MFGM, milk fat globule membrane; NL, nonlipids; TL, total lipids; CHOL, cholesterol.
RESULTS
Composition of milk fat globule membrane. The water content of the dried MFGM was 7.6% (w/w). The MFGM solids, 92.4%, consisted of 59.7% total lipids (TL) and 32.7% NL. Spectrophotometric analysis of the protein in MFGM using bovine serum albumin as a standard showed a total protein content of 32.0% (w/w),indicating that 98% of the nonlipid fraction was protein. TL were fractionated into NPL and PL on a silicic acid column. The NPL were eluted first with chloroform followed by PL with methanol. The NPL and PL in the dried MFGM accounted for 46.6% and 13.1%, respectively. The NPL contained mainly triacylglycerols (42.5%), diacylglycerols (32.0%)and free fatty acids (6.3%). The PL contained 17.5% phospholipids (29.7% sphingomyelin, 24.8% phosphatidylcholine, 17.4% phosphatidylethanolamine, 13.3% phosphatidylinositol and 14.8% others) and 82.5% of unidentified compounds, probably glycerides. The lipid fractions (TL, NPL and PL)did not show significant differences in their fatty acid profiles (38.6-41.4% palmitate, 10.3-17.8% stearate, 12.6-15.5% oleate, eta). Effect of M F G M and its components on cholesterol oxidation in the dry state. At 135~ cholesterol in the MFGM appears to be protected against oxidation when compared to pure cholesterol (Fig. 1). The TL fraction inhibited cholesterol destruction. The NL fraction of MFGM was more protective than the TL fraction. When the whole MFGM (containing 32.7% NL and 59.7% TL) was added to pure cholesterol, no changes in cholesterol amount were observed until 5 h heating at 135~ when it started to decrease Finally, the remaining cholesterol reached a level in between those produced by NL and TL (Fig. 1). The added MFGM showed a net protective effect on cholesterol oxidation. When the PL fraction of MFGM was added to cholesterol at 135~ it did not influence the stability of
100_~
"~ Ja_
O ~:] :~ ~
I ~ . e L
I J
",,,= =
~
CHOLonly
- ~
=
~
_
I
01 0
10
20
"30
40
Time (h) FIG. 2. Effect of membrane lipid fractions on cholesterol oxidation in the dry state at 135~ TL, total lipids; NPL, nonpolar lipids; PL, polar lipids; CHOL, cholesterol.
cholesterol (Fig. 2). The NPL protected cholesterol against oxidation. The effect of TL (containing 21.9% PL and 78.1% NPL) on cholesterol oxidation was in between the effects of its two components, Le., polar and nonpolar. At 75~ cholesterol in MFGM was much less stable than pure cholesterol (Fig. 3). Solid cholesterol is very LIPIDS, Vol. 27, no. 11 (1992)
930 S.K. KIM AND WTW.NAWAR
* MF'GM
10t
I/I
~_~
~
~
.~
Q,r
eto, § PL
4
~
.
ET. O O
Ol
o
lb
2'0
3o
C .'e'-
40
o
Time (Day) FIG. 3. Effect of membrane components on cholesterol oxidation in the dry state at 75~ MFGM, milk fat globule membrane; NL, nonlipids; TL, total lipids; CHOL, cholesterol,
stable at low t e m p e r a t u r ~ Molecular orientation or the surrounding components in M F G M m a y have accelerated cholesterol oxidation. Addition of M F G M , T L or N L did not show a significant effect on cholesterol oxidation at 75~ Only a slight protective effect of added N L and M F G M was observed after 40 d of heating (Fig. 3). The lack of an effect of added M F G M and its c o m p o n e n t s on cholesterol oxidation at 75~ m a y be due to the lack of interaction between cholesterol and these components. When the solvent was evaporated during sample preparation, cholesterol was crystMlized and separated from the
other components added. Therefore, at 75~ the opportunity for interaction was minimized, whereas at 135~ the cholesterol crystals were melted and well mixed. The difference in cholesterol stability between the mixtures and dry M F G M at 75~ appears to reflect the effectiveness of the natural mixing in M F G M . In the presence of M F G M and its components, cholesterol produced very low levels of oxides, although substrate cholesterol was decreased during heating. Only when the P L fraction was added did cholesterol produce relatively large amounts of 7-ketocholesterol and cholesta-3,5-dien-7-one {Fig. 4). Since the P L fraction of M F G M contains phosphatidylcholine and sphingomyelin as the main components, the oxide profile produced was similar to that of phosphatidylcholine and sphingomyelin, i.a,formation of 7-ketocholesterol at the beginning of heating followed by its conversion to cholesta-3,5-dien-7-one. Effect of M F G M and its components on cholesterol oxidation in aqueous system. As d e m o n s t r a t e d in Figure 5, M F G M cholesterol was very stable during six days of incubation in buffer at 75~ Sonication of the M F G M to destroy or reorganize its native structure did not affect the stability of cholesterol in the m e m b r a n e When M F G M was added to pure cholesterol, it exhibited a
LIPIDS, Vol. 27, no. 11 (1992)
20
3o
40
T i m e (h) FIG. 4. Effect of membrane polar lipids (PL) on the production of 7-ketocholesterol {7-Keto)and cholesta-3,5-dien-7-one (3,5-Dien-7mne) in the dry state at 135~
00
.~ 80 84 l~
0
* MFGM
6O
ro .c_
.~_ o
; cHoLow t 40-
E n-
20-
f
O.
o
j 6
Time ( D a y ) FIG. 5. Effect of membrane components on cholesterol oxidation in dispersion at 75~ pH 7.4. MFGM, milk fat globule membrane; NL, nonHpids; TL, total Hpids; CHOL, cholesterol.
protective effect. In contrast to the d r y state, in aqueous dispersion the lipid portion of M F G M (TL, N P L and PL) accelerated cholesterol oxidation at the beginning of incubation, i.e., the first 12 h of heating. The stability of cholesterol in the presence of N P L or P L was similar to t h a t in the presence of T L (Fig. 5). The N L fraction
931 OXIDATIVE INTERACTIONS IN MEMBRANE LIPIDS appears to be the most protective among the M F G M components tested. In the presence of whole membrane materials or N L fraction of MFGM, cholesterol did not produce appreciable amounts of cholesterol oxides. However, in the presence of the lipid portion (TL, N P L or PL) of MFGM, approximately the same level of total oxides was produced as t h a t observed in the pure cholesterol control. Higher amounts of epoxides and 7-hydroxycholesterols also were observed when the M F G M lipid was added. The lipid fraction favored 7-hydroxylation but did not have a significant effect on the ~- to a-form ratios of 7-hydroxycholesterols and epoxides. Effect of added cholesterol on the oxidation of membrane lipid. The saturated f a t t y acids exhibited higher stability than the unsaturated ones in dry M F G M heated at 135~ and in MFGM, TL, PL and NPL heated at 75~ in the presence of water. However, saturated short chains in the lipid fractions (TL, PL and NPL) of M F G M showed low stability when heated at 135~ in the dry s t a t s Figure 6 shows t h a t the short chain f a t t y acids were less stable than the longer chains, both saturated and unsaturated. An explanation for the more rapid disappearance of the short chains cannot be provided at this time. When cholesterol was added to M F G M or its lipid fractions at 135~ it accelerated the destruction of the C18 chains, i.e., stearate, oleate and linolenate, of membrane lipid, but the short chain f a t t y acids became more stable. The destructive effect of cholesterol on the oxidation of C18 chains was much greater when it was added to membrane P L {Fig. 7). In contrast to the behavior at 135~ addition of cholesterol to M F G M did not affect the stability of membrane lipids at 75~ As discussed earlier, at 75~ in the dry state, cholesterol and added components did not appear to interact.
100
18:0
0
18:1
~]~
16:0
0
18:3
t
100
80-
9
60.
0 = 40 "E
20 0
=m
0
2o
3b
Time (h) FIG. 7. Effect of cholesterol on the oxidation of oleate in membrane NPL, nonpolar lipids; PL, polar lipids; CHOL, cholesterol.
lipids at 135~
100
A
9-.. 80
o
t
70 r
'~ 60'
14:0 c
40
o
. m
12:0
c
. m
E
2
/.
Time (Day) 2 0 84
FIG. 8. Effect of cholesterol on the oxidation of oleate in membrane lipids in dispersion at 75~ pH 7.4. MFGM, milk fat globule membrane; TL, total lipids; CHOL, cholesterol.
10
20 Time (h)
30
4O
FIG. 6. Stability of fatty acyl chains in membrane total lipid fraction during heating in the dry state at 135~
The lipids in M F G M were very stable in aqueous disper ~ sions when compared to each of its lipid fractions incubated alone {Fig. 8). Stability of P L and N P L in the absence or presence of cholesterol also showed similar patterns to those of TL. When cholesterol and whole M F G M
LIPID& Vol. 27, no. 11 (1992)
932
S.K. KIM AND WNr NAWAR were mixed in aqueous dispersion, they protected other from oxidation. However, when cholesterol was ed with the M F G M lipid portion, they accelerated other's oxidation at the beginning of incubation and reached a plateau (Figs. 5 and 8).
each mixeach then
TABLE 1 Interactions Between Cholesterol and Other Milk Fat Globule Membrane Components a
Dry C
L
C
L
Aqueous 75~ C L
--
+ +
= +
= +
+ --
135~
DISCUSSION The M F G M consists of various components, such as proteins, triacylglycerols, phospholipids and other compounds, including monoacylglycerols, diacylglycerols and free f a t t y acids. In the dry state at 135~ the effects of added membrane components on cholesterol oxidation and the oxidation of cholesterol natively present in M F G M showed consistent results, Le., M F G M and its components protected cholesterol from oxidation, and cholesterol accelerated the oxidation of m e m b r a n e lipids. However, because of poor interaction between cholesterol and added M F G M components at 75~ only cholesterol in M F G M showed a significant difference from the cholesterol control. Cholesterol natively present in M F G M is well surrounded by other m e m b r a n e components. The interactions between cholesterol and neighboring molecules, and perhaps their physical arrangement, appears to have accelerated cholesterol oxidation. When the heating t e m p e r a t u r e was increased from 75 to 135~ the stability of pure cholesterol changed drastically whereas in M F G M , the interaction between cholesterol and neighboring molecules counteracted the temperature effect on the stability of cholesterol Consequently, the neighboring c o m p o n e n t s in M F G M which are destructive at 75~ showed a p p a r e n t protection at 135~ When incubated alone, cholesterol oxidized faster in aqueous dispersion t h a n in the d r y state at the same t e m p e r a t u r e (Figs. 3 and 5). However, cholesterol natively present in M F G M exhibited high stability in aqueous dispersion. The lipid portion of M F G M (TL, NPL, PL) accelerated cholesterol oxidation at the beginning of incubation, then did not show a significant effect on further cholesterol oxidation in the aqueous system. Added cholesterol accelerated the oxidation of M F G M lipid fractions (TL, NPL, PL) in b o t h d r y and aqueous systems, and it also promoted oxidation of the lipids in whole M F G M in the d r y state. However, in the aqueous system, added cholesterol protected the lipids in M F G M against oxidation. Cholesterol tended to accelerate the oxidation of the other lipids in the m e m b r a n e However, in the presence of water, the N L fraction of M F G M appeared to provide a condition where b o t h cholesterol and the other lipid c o m p o n e n t s protected each other (Table 1). The M F G M exhibited a high protective effect on cholesterol oxidation in an aqueous environment. I t s N L
LIPIDS, Vol.'27, no. 11 (1992)
Individual rni~ing In membrane
75~
+ --
ac, cholesterol; L, membrane lipid; --, oxidation decreased relative to control; +, oxidation increased; =, no difference.
fraction protects cholesterol from oxidation, while the lipid fraction was destructive in the early stages of cholesterol oxidation. In the absence of water, the net balance between these two opposing factors was destructive. The presence of water tipped the balance in favor of protection. The results of this s t u d y suggest t h a t both biochemical and physic~chemical factors are i m p o r t a n t in protecting the m e m b r a n e against destruction.
ACKNOWLEDGMENTS This research was supported in part by Massachusetts Agricultural Experiment Station Hatch Project No. 654 and a grant from the Dairy Bureau of Canad& REFERENCES 1. Dowben, R.M., Brunner, J.R., and Philpott, D.E. (1967) Biochim. Biophys. Acta 135, 1-10. 2. Bouzas, J., Kamarei, A.R., and Karel, M. (1985)J. Food Process. Presera 9, 11-24. 3. Keenan, TW., and Huang, C.M. (1972)I Dairy Sci. 55, 1586-1596. 4. Mather, I.M., and Keenan, T.W.(1975)J. Memb. BIOL 21, 65-85. 5. Patten, S., and Keenan, T.W.(1975)Biochim. Biophys. Acta 415, 273-309. 6. Eigel, W.N., Butler, J.E., Ernstron, C.A., Farrell, Jr., H.M., Ha~ walkar, V.R., Jenness, R., and Whitney, R.M.L. (1984) J. Dairy Sci. 67, 1599-1631. 7. Kanno, C., Hattori, H., and Yamauchi, K. (1987) Agric. Biol. Chem. 51, 1325-1332. 8. McPherson, A.V., and Kitchen, B.J. (1983) J. Dairy Res. 50, 107-133. 9. Kitchen, B.J. (1974) Biochim. Biophys. Acta 356, 257-269. 10. Kitchen, B.J. (1977)J. Dairy Res. 44, 469-482. 11. Folch, J., Lees, M., and Stanley, G.H.S. (1957)I Biol. Chem. 226, 497-509. 12. Price, RB., and Parsons, J.G. (1974) Lipids 9, 560-566. [Received January 15, 1991, and in revised form May 21, 1991; Revision accepted September 7, 1992]
Oxidative Interactions of Cholesterol in the Milk Fat Globule Membrane 1 S.K. Kim and W.W. Nawar* Department of Food Science, University of Massachusetts, Amherst, Massachusetts 01003
The effects of oxidative interactions between cholesterol and milk fat globule membrane (MFGM) components,/.e., nonlipid fraction, total lipid, nonpolar lipid and polar lipid, on cholesterol oxidation were studied in the presence and absence of water. In the dry state, cholesterol natively present in MFGM appeared to be protected at 135~ The nonpolar lipid and nonlipid fraction contributed to the protective effect of MFGM. Added cholesterol accelerated the oxidation of membrane lipid fractions. At 75~ pure cholesterol and membrane lipid fractions did not show significant interaction. However, cholesterol and other lipids in M F G M were less stable than when these were heated separately. When cholesterol and membrane lipids were mixed in an aqueous medium at 75~ each accelerated the oxidation of the other. The MFGM exhibited a high protective effect on cholesterol oxidation in an aqueous environment. The nonlipid fraction p r o tected cholesterol against oxidation, whereas the lipid fraction was destructive. In the absence of water, the net balance between these two opposing factors was destructive. The presence of water reversed the balance in favor of protection. Lipids 27, 928-932 (1992). Bovine milk contains approximately 3-5% fat in the form of globules and these are surrounded by a thin membrane (1). This milk fat globule membrane (MFGM) consists of proteins, glycoproteins, triacylglycerols, cholesterol, enzymes and other minor components. The unsaturated lipids in the membrane were reported to play a role in initiating the oxidation of milk fat (2). Chemical and microscopical data (3-7) indicate t h a t the M F G M originates from the plasma membrane of the secretory cells. Therefore, M F G M can be defined as a modified plasma memb r a n e Oxidation of membrane lipids would naturally be expected to cause alterations in membrane function which eventually lead to disturbance in cell physiology. The objective of this work was to investigate the influence of the interaction among the membrane components on their oxidative stability using M F G M as a model membrane system.
EXPERIMENTAL PROCEDURES
Materials. Cholesterol and silicic acid were purchased from Sigma Chemical Co. (St. Louis, MO). Cholesterol used in this study contained less than 0.9% of contaminants, including 7-ketocholesterol, 7/3-hydroxycholesterol, 5,6~epoxides and cholest-3,5-dien-7-one. Cholesterol oxide standards and silylating agents were obtained from Steraloid Inc (Wilton, NH) and Pierce Chemical Co. (Rockford, IL), 1Based on a paper presented at the Symposium on Milk Lipids held at the AOCS Annual Meeting, Baltimore, MD, April 1990. *To whom correspondence should be addressed. Abbreviations: BSTFA, N,O-bis-(trimethylsilyl)trifiuoroacetamide; HPLC, high-performance liquid chromatography; MFGM, milk fat globule membrane; NL, nonlipid fraction; NPL, nonpolar Iipids; PL, polar lipids; TL, total lipids; TLC, thin-layer chromatography; TMCS, trimethylchlorosilane. LIPIDS, Vol. 27, no. 11 (1992)
respectively. Solvents were high-performance liquid chromatography (HPLC) grade from various sources.
Preparation, fractionation and characterization of MFGM. Raw milk was obtained from the University of Massachusetts farm and pasteurized at 63~ for 30 min. Cream was separated by centrifugation, then washed twice with 3 vol of Tris-HC1 buffer (10 mM, pH 7.5 containing 0.25 M of sucrose and 1.0 mM of MgC12) and once with 1 vol of 50 mM Tris-HC1 buffer (pH 7.5) (4). The washed cream was resuspended in 50 mM Tris-HC1 buffer, and the cream structure was destroyed by freezing and thawing to liberate M F G M from the cream structure (8). Butter was separated from the MFGM suspension by centrifugation at 2,500 X g for 20 min. The b u t t e r was washed with 1 vol of distilled water to recover the residual MFGM, and the water part was combined with the M F G M suspension (9). The membrane suspension was centrifuged at 100,000 • g for 1 h to obtain the membrane (10). The pellets were dispersed in distilled water and quickly recentrifuged to remove the buffer salts from the membrane fraction. The membrane was then. freeze-dried and stored at - 4 0 ~ The freeze-dried M F G M was washed 3 times with 5 vol of CHC13/MeOH (2:1, vol/vol) and c e n t r i f u g e d at 5,000 • g for 30 min to obtain the nonlipid fraction (NL) of MFGM, which was dried and stored at - 4 0 ~ Solvent in the extract was removed on a rotary evaporator to obtain crude M F G M lipid. The crude lipid was purified (11) and fractionated into polar and nonpolar fractions by silicic acid column chromatography. Chloroform and methanol were used to elute the nonpolar lipids (NPL) and the polar lipids {PL), respectively (12). PL and N P L fractions were separated into their components by thin-layer chromatography (TLC) (silica gel G, 0.25 nun thickness). The solvent system used for NPL was petroleum hydrocarbon/diethyl ether/acetic acid (90:10:1, vol/vol]vol) (12), and that used for PL was CHCI3/MeOH/ water/18.4% aqueous NH3 (130:70:8:0.5, vol/vol/vol/vol). TLC fractions were visualized by heating the plates at 120~ for 25 min after spraying with a potassium dichromate/sulfuric acid solution. The lipid classes were quantitated with a TLC densitometer. Sample treatment. Solutions of cholesterol and membrane lipid in chloroform (1 mg each) were placed in vials, and the solvent was evaporated on a rotary evaporator to make a thin film on the b o t t o m of each vial. For the dry samples, the vials were placed in an oven at 135~ for 40 h or at 75~ for 40 d. For the aqueous samples, 1 mL of 10 mM Tris buffer was added to each vial and incubated at 75~ for 6 d. The lipids in the reaction tubes were extracted with 3.5 mL of chloroform/methanol (2:1, vol/vol) and divided into two parts, one for cholesterol analysis and the other for f a t t y acid analysis. Cholesterol analysis. Cholesterol and its oxides were separated from the heated mixtures of cholesterol and membrane lipids by silicic acid column chromatography. The NPL, mainly triacylglycerols, were eluted with hexane/diethyl ether (95:5, vol]vol), followed by cholesterol and its oxidation products with 100% diethyl ether. Phospholipids remained on the column. The ether fraction was
929
OXIDATIVE INTERACTIONS IN MEMBRANE LIPIDS collected and silylated with N,O-bis-(trimethylsilyl)trifluoroacetamide (BSTFA) and 1% trimethylchlorosilane (TMCS) at 80~ for 1 h. The trimethylsflyl ethers were separated on a Varian 3700 GC (Varian Associates, Paio Alt~ CA) equipped with an Ultra-1 capillary column (50 m, 0.2 mm i.d. and 0.33 ~m film thickness, HewlettPackard, Avondale, PA) with temperature programming from 100~ to 300~ at the rate of 10~ 5aAndrostan-3/3-ol-17-one acetate was used as an internal standard for the analysis of cholesterol and its oxides. Fatty acid analysis. Oxidation of triacylglycerols and free fatty acids was traced according to the loss of their fatty acyl chains. Fatty acid analysis was done by methylating the whole lipid portion directly without using silicic acid column chromatography. The lipid samples containing 1-2 mg of lipid were saponified with 0.5 mL of 2N methanolic KOH solution at 65~ for 25 min, then methylated with 0.5 mL of 14% BF3/methanol solution at 65~ After 25 min reaction, residual methylating agent was deactivated by adding 0.5 mL of water. The fatty acid methyl esters were extracted with hexane and analyzed by gas chromatography on a Supelcowax-10 capillary column (30 m, 0.32 mm i.d. and 0.25 ~m film thickness, Supelco Ina, Bellefonte, PA). Triheptadecanoin was used as an internal standard. The graphs in this report represent the averages of triplicate analyses.
100 ,-, o~
80-
m
O
60-
i
.c O
40-
.~.~m
E
0-
rr
CHOL only
9
"Jl:
0
0
I0
2'0
3o
40
Time (h) FIG. 1. Effect of membrane components on cholesterol oxidation in the dry state at 135~ MFGM, milk fat globule membrane; NL, nonlipids; TL, total lipids; CHOL, cholesterol.
RESULTS
Composition of milk fat globule membrane. The water content of the dried MFGM was 7.6% (w/w). The MFGM solids, 92.4%, consisted of 59.7% total lipids (TL) and 32.7% NL. Spectrophotometric analysis of the protein in MFGM using bovine serum albumin as a standard showed a total protein content of 32.0% (w/w),indicating that 98% of the nonlipid fraction was protein. TL were fractionated into NPL and PL on a silicic acid column. The NPL were eluted first with chloroform followed by PL with methanol. The NPL and PL in the dried MFGM accounted for 46.6% and 13.1%, respectively. The NPL contained mainly triacylglycerols (42.5%), diacylglycerols (32.0%)and free fatty acids (6.3%). The PL contained 17.5% phospholipids (29.7% sphingomyelin, 24.8% phosphatidylcholine, 17.4% phosphatidylethanolamine, 13.3% phosphatidylinositol and 14.8% others) and 82.5% of unidentified compounds, probably glycerides. The lipid fractions (TL, NPL and PL)did not show significant differences in their fatty acid profiles (38.6-41.4% palmitate, 10.3-17.8% stearate, 12.6-15.5% oleate, eta). Effect of M F G M and its components on cholesterol oxidation in the dry state. At 135~ cholesterol in the MFGM appears to be protected against oxidation when compared to pure cholesterol (Fig. 1). The TL fraction inhibited cholesterol destruction. The NL fraction of MFGM was more protective than the TL fraction. When the whole MFGM (containing 32.7% NL and 59.7% TL) was added to pure cholesterol, no changes in cholesterol amount were observed until 5 h heating at 135~ when it started to decrease Finally, the remaining cholesterol reached a level in between those produced by NL and TL (Fig. 1). The added MFGM showed a net protective effect on cholesterol oxidation. When the PL fraction of MFGM was added to cholesterol at 135~ it did not influence the stability of
100_~
"~ Ja_
O ~:] :~ ~
I ~ . e L
I J
",,,= =
~
CHOLonly
- ~
=
~
_
I
01 0
10
20
"30
40
Time (h) FIG. 2. Effect of membrane lipid fractions on cholesterol oxidation in the dry state at 135~ TL, total lipids; NPL, nonpolar lipids; PL, polar lipids; CHOL, cholesterol.
cholesterol (Fig. 2). The NPL protected cholesterol against oxidation. The effect of TL (containing 21.9% PL and 78.1% NPL) on cholesterol oxidation was in between the effects of its two components, Le., polar and nonpolar. At 75~ cholesterol in MFGM was much less stable than pure cholesterol (Fig. 3). Solid cholesterol is very LIPIDS, Vol. 27, no. 11 (1992)
930 S.K. KIM AND WTW.NAWAR
* MF'GM
10t
I/I
~_~
~
~
.~
Q,r
eto, § PL
4
~
.
ET. O O
Ol
o
lb
2'0
3o
C .'e'-
40
o
Time (Day) FIG. 3. Effect of membrane components on cholesterol oxidation in the dry state at 75~ MFGM, milk fat globule membrane; NL, nonlipids; TL, total lipids; CHOL, cholesterol,
stable at low t e m p e r a t u r ~ Molecular orientation or the surrounding components in M F G M m a y have accelerated cholesterol oxidation. Addition of M F G M , T L or N L did not show a significant effect on cholesterol oxidation at 75~ Only a slight protective effect of added N L and M F G M was observed after 40 d of heating (Fig. 3). The lack of an effect of added M F G M and its c o m p o n e n t s on cholesterol oxidation at 75~ m a y be due to the lack of interaction between cholesterol and these components. When the solvent was evaporated during sample preparation, cholesterol was crystMlized and separated from the
other components added. Therefore, at 75~ the opportunity for interaction was minimized, whereas at 135~ the cholesterol crystals were melted and well mixed. The difference in cholesterol stability between the mixtures and dry M F G M at 75~ appears to reflect the effectiveness of the natural mixing in M F G M . In the presence of M F G M and its components, cholesterol produced very low levels of oxides, although substrate cholesterol was decreased during heating. Only when the P L fraction was added did cholesterol produce relatively large amounts of 7-ketocholesterol and cholesta-3,5-dien-7-one {Fig. 4). Since the P L fraction of M F G M contains phosphatidylcholine and sphingomyelin as the main components, the oxide profile produced was similar to that of phosphatidylcholine and sphingomyelin, i.a,formation of 7-ketocholesterol at the beginning of heating followed by its conversion to cholesta-3,5-dien-7-one. Effect of M F G M and its components on cholesterol oxidation in aqueous system. As d e m o n s t r a t e d in Figure 5, M F G M cholesterol was very stable during six days of incubation in buffer at 75~ Sonication of the M F G M to destroy or reorganize its native structure did not affect the stability of cholesterol in the m e m b r a n e When M F G M was added to pure cholesterol, it exhibited a
LIPIDS, Vol. 27, no. 11 (1992)
20
3o
40
T i m e (h) FIG. 4. Effect of membrane polar lipids (PL) on the production of 7-ketocholesterol {7-Keto)and cholesta-3,5-dien-7-one (3,5-Dien-7mne) in the dry state at 135~
00
.~ 80 84 l~
0
* MFGM
6O
ro .c_
.~_ o
; cHoLow t 40-
E n-
20-
f
O.
o
j 6
Time ( D a y ) FIG. 5. Effect of membrane components on cholesterol oxidation in dispersion at 75~ pH 7.4. MFGM, milk fat globule membrane; NL, nonHpids; TL, total Hpids; CHOL, cholesterol.
protective effect. In contrast to the d r y state, in aqueous dispersion the lipid portion of M F G M (TL, N P L and PL) accelerated cholesterol oxidation at the beginning of incubation, i.e., the first 12 h of heating. The stability of cholesterol in the presence of N P L or P L was similar to t h a t in the presence of T L (Fig. 5). The N L fraction
931 OXIDATIVE INTERACTIONS IN MEMBRANE LIPIDS appears to be the most protective among the M F G M components tested. In the presence of whole membrane materials or N L fraction of MFGM, cholesterol did not produce appreciable amounts of cholesterol oxides. However, in the presence of the lipid portion (TL, N P L or PL) of MFGM, approximately the same level of total oxides was produced as t h a t observed in the pure cholesterol control. Higher amounts of epoxides and 7-hydroxycholesterols also were observed when the M F G M lipid was added. The lipid fraction favored 7-hydroxylation but did not have a significant effect on the ~- to a-form ratios of 7-hydroxycholesterols and epoxides. Effect of added cholesterol on the oxidation of membrane lipid. The saturated f a t t y acids exhibited higher stability than the unsaturated ones in dry M F G M heated at 135~ and in MFGM, TL, PL and NPL heated at 75~ in the presence of water. However, saturated short chains in the lipid fractions (TL, PL and NPL) of M F G M showed low stability when heated at 135~ in the dry s t a t s Figure 6 shows t h a t the short chain f a t t y acids were less stable than the longer chains, both saturated and unsaturated. An explanation for the more rapid disappearance of the short chains cannot be provided at this time. When cholesterol was added to M F G M or its lipid fractions at 135~ it accelerated the destruction of the C18 chains, i.e., stearate, oleate and linolenate, of membrane lipid, but the short chain f a t t y acids became more stable. The destructive effect of cholesterol on the oxidation of C18 chains was much greater when it was added to membrane P L {Fig. 7). In contrast to the behavior at 135~ addition of cholesterol to M F G M did not affect the stability of membrane lipids at 75~ As discussed earlier, at 75~ in the dry state, cholesterol and added components did not appear to interact.
100
18:0
0
18:1
~]~
16:0
0
18:3
t
100
80-
9
60.
0 = 40 "E
20 0
=m
0
2o
3b
Time (h) FIG. 7. Effect of cholesterol on the oxidation of oleate in membrane NPL, nonpolar lipids; PL, polar lipids; CHOL, cholesterol.
lipids at 135~
100
A
9-.. 80
o
t
70 r
'~ 60'
14:0 c
40
o
. m
12:0
c
. m
E
2
/.
Time (Day) 2 0 84
FIG. 8. Effect of cholesterol on the oxidation of oleate in membrane lipids in dispersion at 75~ pH 7.4. MFGM, milk fat globule membrane; TL, total lipids; CHOL, cholesterol.
10
20 Time (h)
30
4O
FIG. 6. Stability of fatty acyl chains in membrane total lipid fraction during heating in the dry state at 135~
The lipids in M F G M were very stable in aqueous disper ~ sions when compared to each of its lipid fractions incubated alone {Fig. 8). Stability of P L and N P L in the absence or presence of cholesterol also showed similar patterns to those of TL. When cholesterol and whole M F G M
LIPID& Vol. 27, no. 11 (1992)
932
S.K. KIM AND WNr NAWAR were mixed in aqueous dispersion, they protected other from oxidation. However, when cholesterol was ed with the M F G M lipid portion, they accelerated other's oxidation at the beginning of incubation and reached a plateau (Figs. 5 and 8).
each mixeach then
TABLE 1 Interactions Between Cholesterol and Other Milk Fat Globule Membrane Components a
Dry C
L
C
L
Aqueous 75~ C L
--
+ +
= +
= +
+ --
135~
DISCUSSION The M F G M consists of various components, such as proteins, triacylglycerols, phospholipids and other compounds, including monoacylglycerols, diacylglycerols and free f a t t y acids. In the dry state at 135~ the effects of added membrane components on cholesterol oxidation and the oxidation of cholesterol natively present in M F G M showed consistent results, Le., M F G M and its components protected cholesterol from oxidation, and cholesterol accelerated the oxidation of m e m b r a n e lipids. However, because of poor interaction between cholesterol and added M F G M components at 75~ only cholesterol in M F G M showed a significant difference from the cholesterol control. Cholesterol natively present in M F G M is well surrounded by other m e m b r a n e components. The interactions between cholesterol and neighboring molecules, and perhaps their physical arrangement, appears to have accelerated cholesterol oxidation. When the heating t e m p e r a t u r e was increased from 75 to 135~ the stability of pure cholesterol changed drastically whereas in M F G M , the interaction between cholesterol and neighboring molecules counteracted the temperature effect on the stability of cholesterol Consequently, the neighboring c o m p o n e n t s in M F G M which are destructive at 75~ showed a p p a r e n t protection at 135~ When incubated alone, cholesterol oxidized faster in aqueous dispersion t h a n in the d r y state at the same t e m p e r a t u r e (Figs. 3 and 5). However, cholesterol natively present in M F G M exhibited high stability in aqueous dispersion. The lipid portion of M F G M (TL, NPL, PL) accelerated cholesterol oxidation at the beginning of incubation, then did not show a significant effect on further cholesterol oxidation in the aqueous system. Added cholesterol accelerated the oxidation of M F G M lipid fractions (TL, NPL, PL) in b o t h d r y and aqueous systems, and it also promoted oxidation of the lipids in whole M F G M in the d r y state. However, in the aqueous system, added cholesterol protected the lipids in M F G M against oxidation. Cholesterol tended to accelerate the oxidation of the other lipids in the m e m b r a n e However, in the presence of water, the N L fraction of M F G M appeared to provide a condition where b o t h cholesterol and the other lipid c o m p o n e n t s protected each other (Table 1). The M F G M exhibited a high protective effect on cholesterol oxidation in an aqueous environment. I t s N L
LIPIDS, Vol.'27, no. 11 (1992)
Individual rni~ing In membrane
75~
+ --
ac, cholesterol; L, membrane lipid; --, oxidation decreased relative to control; +, oxidation increased; =, no difference.
fraction protects cholesterol from oxidation, while the lipid fraction was destructive in the early stages of cholesterol oxidation. In the absence of water, the net balance between these two opposing factors was destructive. The presence of water tipped the balance in favor of protection. The results of this s t u d y suggest t h a t both biochemical and physic~chemical factors are i m p o r t a n t in protecting the m e m b r a n e against destruction.
ACKNOWLEDGMENTS This research was supported in part by Massachusetts Agricultural Experiment Station Hatch Project No. 654 and a grant from the Dairy Bureau of Canad& REFERENCES 1. Dowben, R.M., Brunner, J.R., and Philpott, D.E. (1967) Biochim. Biophys. Acta 135, 1-10. 2. Bouzas, J., Kamarei, A.R., and Karel, M. (1985)J. Food Process. Presera 9, 11-24. 3. Keenan, TW., and Huang, C.M. (1972)I Dairy Sci. 55, 1586-1596. 4. Mather, I.M., and Keenan, T.W.(1975)J. Memb. BIOL 21, 65-85. 5. Patten, S., and Keenan, T.W.(1975)Biochim. Biophys. Acta 415, 273-309. 6. Eigel, W.N., Butler, J.E., Ernstron, C.A., Farrell, Jr., H.M., Ha~ walkar, V.R., Jenness, R., and Whitney, R.M.L. (1984) J. Dairy Sci. 67, 1599-1631. 7. Kanno, C., Hattori, H., and Yamauchi, K. (1987) Agric. Biol. Chem. 51, 1325-1332. 8. McPherson, A.V., and Kitchen, B.J. (1983) J. Dairy Res. 50, 107-133. 9. Kitchen, B.J. (1974) Biochim. Biophys. Acta 356, 257-269. 10. Kitchen, B.J. (1977)J. Dairy Res. 44, 469-482. 11. Folch, J., Lees, M., and Stanley, G.H.S. (1957)I Biol. Chem. 226, 497-509. 12. Price, RB., and Parsons, J.G. (1974) Lipids 9, 560-566. [Received January 15, 1991, and in revised form May 21, 1991; Revision accepted September 7, 1992]
