ANALYTICAL
Purification Two-Phase Anders
Persson
Biochemistry.
Received
204,
BIOCHEMISTRY
(199%)
of Plasma Membranes Affinity Partitioning’ and Bengt
Chemical
January
131-136
Center,
Jergil University
of Lund,
P.0. Box 124, S-221 00 Lund,
Sweden
23, 1992
A rapid method for purifying rat liver plasma membranes of high purity and yield is described. Squashed liver was homogenized in an aqueous polyethylene glycol-dextran two-phase system. After phase separation and reextraction of the bottom phase with fresh top phase, the combined polyethylene glycol-rich top phases were affinity partitioned in the presence of borate buffer with new bottom phase containing dextranlinked wheat-germ agglutinin. Under these conditions the lectin selectively pulled plasma membranes into the dextran-rich bottom phase, while other membranes preferentially distributed in the top phase. The lectincontaining bottom phase was reextracted with fresh top phase before collecting the purified plasma membranes by centrifugation. This protocol resulted in a preparation that was 30- to 40-fold enriched compared to the homogenate in plasma membrane markers for both the apical and basolateral domains and had yields of 5570%. The contamination by other membranes was low. The entire procedure was completed within 90 min. The method should be useful for purifying plasma membranes also from other sources. It Is92 Academic PT~SS, IIIC.
Mammalian liver plasma membranes are commonly separated from other membranes by utilizing differences in fragment size and density by a combination of differential and density-gradient centrifugation techniques (l-4). Other critical factors are the homogenization procedure and the medium used. Most methods are time-consuming, and yields of highly purified plasma membranes, as measured by marker enzyme activities, are often comparatively low. As an alternative, or a complement, methods based on partitioning in aqueous polymer two-phase systems
‘This work search Council.
by Aqueous
was
supported
by the
0003-x97/92 $5.00 Copyright 18 1992 by Academic Press, All rights of reproduction in any form
Swedish
Natural
Science
Re-
have been introduced (5-g), in which membranes are separated according to differences in surface properties (10). Two-phase partitioning has not turned out to be selective enough, however, for a facile purification of animal plasma membranes, although the technique has been used widely for preparation of plasma membranes from plants [for reviews see (11) and (12)]. In order to increase the selectivity of the two-phase partitioning technique, we are exploring the possibility of using biospecific ligands covalently coupled to phase polymers for membrane purification. Thus, the lectin wheat-germ agglutinin attached to dextran has turned out to selectively pull rat liver plasma membranes into the dextran-rich bottom phase of a polyethylene glycoldextran two-phase system (13). We have now developed a rapid and efficient method for the preparation of rat liver plasma membranes of high purity and yield. It includes homogenization and crude fractionation in a conventional aqueous polymer two-phase system followed by selective extraction of plasma membranes by biospecific affinity partitioning using wheat-germ agglutinin as a ligand.
MATERIALS
AND
METHODS
Chemicals Stock solutions in water of 20% (w/w) Dextran T500 (Pharmacia Fine Chemicals AB) and 40% (w/w) polyethylene glycol 3350 (Carbowax 3350; Union Carbide) were prepared as described by Lopez-Perez et al. (14). UDP-[14C]galactose and [3H]AMP were from Amersham International plc, [‘251]iodine was from New England Nuclear, 2,2,2-trifluoroethanesulfonyl chloride (tresyl chloride) was from Synthelec (Lund, Sweden) and wheat-germ agglutinin was from BoehringerMannheim. All other reagents used were of analytical grade. 131
Inc. reserved.
132
PERSSON
Activation of Dextran Dextran T500 was activated with tresyl chloride essentially as described by Persson and Olde (15). The dextran was dissolved in water and freeze-dried before use. Five grams of this dextran was dissolved in 25 ml of dimethyl sulfoxide, and then 1 ml of triethylamine was added slowly, followed by 5 ml of dichloromethane. After chilling on ice, 0.35 g of tresyl chloride was added dropwise while stirring. The mixture was stirred for 30 min on ice, for 60 min at 4°C and overnight at 20°C. The reaction was terminated by precipitating the dextran with 50 ml of dichloromethane. After washing and thorough kneading with several portions of dichloromethane, the precipitate was dissolved in water, dialyzed against water overnight, and freezedried. The activated dextran was stored at -20°C until use. The recovery of dextran was more than 95%. Freeze-drying was introduced to reduce the time required to dissolve dextran from several hours to a few minutes and to reduce the risk of tresyl chloride hydrolysis during activation. Coupling of Wheat-Germ Agglutinin to Tresyl Dextran The lectin wheat-germ agglutinin was coupled to tresyl dextran as described earlier (13). Two grams of tresyl dextran was dissolved in 10 ml of coupling buffer (0.1 M NaH,PO,/Na,HP0,/0.5 M NaCl, pH 7.5) before the addition of 10 mg of wheat-germ agglutinin dissolved in 1 ml of coupling buffer. The coupling proceeded overnight at 4°C with gentle agitation. The reaction was terminated by the addition of 0.1 M Tris/HCl, pH 7.5, to inactivate excess tresyl groups. Uncoupled wheat-germ agglutinin and salts were removed by ultrafiltration of the dextran-ligand mixture in a Filtron Omegacell (150ml size, cut-off 100 K) before freeze-drying. The recovery of dextran was more than 90%. When stored at -20°C for up to 6 months no loss of binding capacity for rat liver plasma membranes has been observed. The amount of wheat-germ agglutinin coupled to dextran was determined according to Bradford (16) using wheat-germ agglutinin as standard. Routinely 4 mg was bound per gram freeze-dried product. Purification Protocol for Plasma Membranes Adult male Sprague-Dawley rats (175-200 g) were killed by decapitation. Each liver was immediately chilled in ice-cold 0.25 M sucrose in 5 mM Tris/HCl, pH 8.0, squashed through a garlic press, and transferred to a Dounce homogenizer containing an aqueous twophase system. This system had a total weight of 15 g, including up to 2 g of liver, in 5.7% (w/w) each of Dextran T500 and polyethylene glycol 3350, diluted in 15 mM Tris/H,SO,, pH 7.8, from the stock solutions. After homogenization by 10 strokes with a loose-fitting pestle and 20 strokes with a tight-fitting one, the homogenization system was centrifuged at low speed (150g) for 5
AND
JERGIL
min in a swing-out rotor to facilitate phase separation. The top phase was collected and the bottom phase reextracted with an equal volume of fresh top phase obtained after preequilibration with bottom phase. The system was turned upside down 20 times, vortex-mixed, and turnedupside down another 20 times to ensure thorough mixing before phase separation as above. The combined top phases were added to a new bottom phase also containing dextran-bound wheat-germ agglutinin for affinity partitioning. This bottom phase had been prepared by preequilibration of a 10-g two-phase system containing 6.0% (w/w) of each polymer, including dextran-bound wheat-germ agglutinin as indicated in each case, and 20 mM boric acid adjusted to pH 7.8 with Tris base, and subsequent removal of the top phase. The volume of this bottom phase relative that of the combined top phases was not very critical and could be varied +50% compared to the amount given here. The combined top phases and the new bottom phase were thoroughly mixed as above, and the phases separated by gentle centrifugation (Wifug Doctor, 1250 rev/ min) for 5 min. The bottom phase obtained was reextracted with fresh, preequilibrated top phase twice in the same manner. After the final phase separation the bottom phase was diluted lo-fold in 0.25 M sucrose and 0.1 M N-acetylglucosamine in 5 mM Tris/HCl, pH 8.0, and membranes were pelleted by centrifugation for 60 min at 100,OOOg. The pellet was resuspended in 0.25 M sucrose and 5 mM Tris/HCl, pH 8. It should be noted that the performance of phase systems is sensitive to temperature. All phase separations were therefore done with solutions carefully adjusted to +4OC. Analyses The plasma membrane markers measured were 5’-nucleotidase [EC 3.1.3.5; (17)], alkaline phosphodiesterase I (EC 3.1.4.1; 18) and 1251-labeled asialoorosomucoid binding (4,8). Other markers measured were arylesterase [endoplasmic reticulum, EC 3.1.1.2; (19)], N-acetylglucosamine galactosyltransferase [Golgi membranes, EC 2.4.1.38; (20)], succinate-cytochrome c reductase [mitochondria, EC 1.3.99.1; (21)], N-acetyl-Pglucosaminidase [lysosomes, EC 3.2.1.30; (22)] and lactate dehydrogenase [cytosol, EC 1.1.1.27; (23)]. Protein was determined according to Bradford (16) using bovine serum albumin as standard. RESULTS
AND
DISCUSSION
We introduce a rapid method for purification of rat liver plasma membranes of high purity and yield. The procedure, as outlined in Fig. 1 and described in detail under Materials and Methods, includes homogenization of the tissue and crude fractionation in a conventional
AFFINITY
FIG. 1. Methods. agglutinin
PURIFICATION
Outline of the purification protocol for rat liver plasma Each addition of fresh phase, polyethylene glycol-containing (WGA-B), is indicated. Only phases designed for further
aqueous polymer two-phase system followed by selective purification of plasma membranes by lectin-affinity partitioning. After homogenization in the two-phase system and phase separation, the bottom phase is reextracted with fresh top phase to increase the yield. The two top phases are combined and partitioned in the presence of borate with the bottom phase containing dextran-linked wheat-germ agglutinin. Under these conditions membranes preferentially partition in the top phase, except plasma membranes which contain exposed N-acetylglucosamine residues and, therefore, are selectively pulled into the bottom phase by the lectin ligand. To improve purity, the wheat-germ agglutinincontaining bottom phase is reextracted twice with fresh top phase. The final bottom phase is diluted with Nacetylglucosamine to dissociate membranes from the ligand before sedimentation of the membranes.
OF
PLASMA
133
MEMBRANES
membranes. The procedure top phase (T) or bottom processing are labeled.
is described in detail under phase containing dextran-linked
Materials and wheat-germ
mains partitioned in the polyethylene glycol-rich top phase upon phase separation (Fig. 2), whereas enzyme markers for other membranes had a lower partition ratio, in line with our earlier observation (8). Reextraction of the bottom phase with fresh top phase increased the yield of plasma membranes to around 70% (Fig. 2). Although other membranes increased relatively more in the combined top phases by this second extraction, the final purity of plasma membranes was not affected, as contaminating membranes were efficiently removed during the subsequent affinity partitioning step (see below).
60
Composition
of the Homogenization
System
The partitioning of biological membranes in aqueous polymer two-phase systems is strongly influenced by polymer concentration and salt content (10). As previously reported (8) the partition ratios of rat liver membranes increase in the order: endoplasmic reticulum, mitochondria, lysosomes, Golgi, and plasma membranes. Close to the critical point [5.4% (w/w) of each polymer at 4”C] of a polyethylene glycol 3350-dextran T500 two-phase system different types of membranes partition almost equally between the phases. An increase in the concentration of the phase polymers selectively decreases the partition ratios of other membranes, particularly mitochondria and endoplasmic reticulum, relative to plasma membranes. The most selective system found consisted of 5.7% (w/w) of each polymer in 15 mM Tris/H,SO,, pH 7.8. After homogenization of squashed liver in this system, approximately 50% of the plasma membrane markers for both the apical (5’-nucleotidase and alkaline phosphodiesterase) and the basolateral (asialoorosomucoid binding) do-
1
2
3
4
MARKER
5
6
7
8
9
ENZYMES
FIG. 2. Partitioning of protein and marker enzymes after homogenization. One gram of squashed liver was homogenized in a 15-g twophase system followed by phase separation as described under Materials and Methods. The dextran-rich bottom phase was reextracted once with preequilibrated polyethylene glycol-rich top phase. The recoveries (percentage of the homogenate) of protein [l], asialoorosomucoid binding [2], alkaline phosphodiesterase [3], 5’nucleotidase [4], arylesterase [5], succinate-cytochrome c reductase [6], galactosyltransferase [7], N-acetylglucosaminidase [S], and lactate dehydrogenase [9] in each of the resulting top phases were determined. The first top phase, closed bars; the second, open bars. Results are averages of three independent preparations with essentially the same membrane distribution.
PERSSON
AND
JERGIL
combined top phases from the homogenate was attracted to the bottom phase by 80 pg of dextran-bound wheat-germ agglutinin, leaving 80%, or more, of markers for other membranes in the top phase (Fig. 3). Purity
U
100
200
WGA
300
hi9
FIG. 3. Affinity partitioning of rat liver membranes. The top phases obtained after homogenization and reextraction were combined. One gram of these, containing 4 mg of protein, was partitioned in the presence of increasing amounts of dextran-linked wheat-germ agglutinin. Each system had a total weight of 5 g and was treated as described under Materials and Methods. After two reextractions of the wheat-germ agglutinin containing bottom phase, the protein content (0) and the activities of alkaline phosphodiesterase (APDE, 0), 5’-nucleotidase (5-N, a), galactosyltransferase (GAL, Cl), arylesterase (AE, A), and N-acetylglucosaminidase (NACG, n ) were measured in the combined top phases and the final bottom phase. They are expressed as percentages in the combined top phases of the total activity in the phases. Recoveries were 80-105% of the starting material.
Affinity
Partitioning
In order to establish the amount of dextran-linked wheat-germ agglutinin required for pulling plasma membranes into the bottom phase, and to examine the selectivity of this step, the combined top phases of the homogenization system were partitioned with fresh bottom phase containing increasing amounts of ligand. Each of the resulting bottom phases was reextracted twice with fresh top phase before measuring marker enzyme activities in the combined top phases and the final bottom phase (Fig. 3). These separations were performed in the presence of borate, which nonselectively pushes membranes into the polyethylene glycol-rich top phase (13). As a consequence, the bulk of membranes was recovered in the combined top phases in the absence of ligand, whereas increasing amounts of ligand selectively pulled plasma membranes into the dextranrich bottom phase leaving other membranes in the top phase. Thus, approximately 90% of the plasma membrane markers contained in 4 mg protein present in the
and Yield of Plasma Membranes
As shown in Table 1 plasma membranes of both apical and basolateral domains were purified 30- to 40-fold over the homogenate by the procedure. The yield of the basolateral marker asialoorosomucoid binding was close to 55%, and that of the apical markers 5’-nucleotidase and alkaline phosphodiesterase 65-70%. The Golgi membrane and lysosomal markers were slightly enriched (6- and 3-fold, respectively) whereas marker activities for mitochondria, endoplasmic reticulum, and the cytosol were too low to be determined accurately. The conventional two-phase separation performed in the homogenization medium (including repartitioning of the first bottom phase with fresh top phase) removed 95% or more of mitochondria and endoplasmic reticulum (Table l), leaving significant amounts of Golgi membranes, lysosomes, and soluble proteins in the combined top phases. The subsequent affinity partitioning (including two reextractions of the wheat-germ agglutinin-containing bottom phase) removed most of the Golgi membranes and lysosomes but only a small amount of plasma membranes. The final centrifugation step pelleted the membranes present in the bottom phase, and removed remaining soluble proteins (as indicated by lactate dehydrogenase activity), thereby increasing the plasma membrane purification factor by approximately 2.5 times. The polymer plus the ligand were also removed in this step, as the final bottom phase was diluted with N-acetylglucosamine solution to dissociate ligand from membranes before the centrifugation. The degree of enrichment of plasma membrane markers by the two-phase technique was similar to that achieved by a method involving homogenization under well-defined conditions followed by centrifugations (4), but the yield was approximately four times higher. These two methods result in plasma membrane preparations consisting of both apical and basolateral domains. Other methods, in contrast, aim at separating these domains (2,3) by manipulating homogenization and centrifugation conditions. As various markers have different lateral distributions in the plasma membrane, higher enrichment factors than those reported here can be achieved by selective purification of plasma membrane domains (l-3,24) or, alternatively, by subfractionation of purified plasma membranes (8,25,26). Although the different domains have slightly different densities, we have only been able to separate these domains partly, so far, by density gradient centrifugation of plasma membranes prepared by the two-phase technique.
AFFINITY
PURIFICATION
OF
PLASMA
TABLE Distribution
1
of Membrane
Marker
Enzymes
Homogenate
Marker Protein Alkaline phosphodiesterase 5’-Nucleotidase Asialoorosomucoid binding Arylesterase Galactosyltransferase Succinate-cytochrome c reductase N-Acetylglucosaminidase Lactate dehydrogenase
Specific 114.6 42.3 13.6 3.99 717 0.44 36.4 2.41 3290
activity * f k f f f k k *
2.4 3.6 1.8 0.82 35 0.03 3.1 0.12 308
Affinity Combined top phases (%)
Combined top phases (%)
29.9 73.8 70.1 61.0 5.2 53.5 2.5 43.7 16.4
24.2 7.0 4.3 2.1 4.6 40.2
f +f f f +f -+ k
135
MEMBRANES
1.1 3.4 4.2 4.6 0.2 4.9 0.2 4.6 1.1
+ k 2 f k f
Purification Two-Phase Anders
Persson
Biochemistry.
Received
204,
BIOCHEMISTRY
(199%)
of Plasma Membranes Affinity Partitioning’ and Bengt
Chemical
January
131-136
Center,
Jergil University
of Lund,
P.0. Box 124, S-221 00 Lund,
Sweden
23, 1992
A rapid method for purifying rat liver plasma membranes of high purity and yield is described. Squashed liver was homogenized in an aqueous polyethylene glycol-dextran two-phase system. After phase separation and reextraction of the bottom phase with fresh top phase, the combined polyethylene glycol-rich top phases were affinity partitioned in the presence of borate buffer with new bottom phase containing dextranlinked wheat-germ agglutinin. Under these conditions the lectin selectively pulled plasma membranes into the dextran-rich bottom phase, while other membranes preferentially distributed in the top phase. The lectincontaining bottom phase was reextracted with fresh top phase before collecting the purified plasma membranes by centrifugation. This protocol resulted in a preparation that was 30- to 40-fold enriched compared to the homogenate in plasma membrane markers for both the apical and basolateral domains and had yields of 5570%. The contamination by other membranes was low. The entire procedure was completed within 90 min. The method should be useful for purifying plasma membranes also from other sources. It Is92 Academic PT~SS, IIIC.
Mammalian liver plasma membranes are commonly separated from other membranes by utilizing differences in fragment size and density by a combination of differential and density-gradient centrifugation techniques (l-4). Other critical factors are the homogenization procedure and the medium used. Most methods are time-consuming, and yields of highly purified plasma membranes, as measured by marker enzyme activities, are often comparatively low. As an alternative, or a complement, methods based on partitioning in aqueous polymer two-phase systems
‘This work search Council.
by Aqueous
was
supported
by the
0003-x97/92 $5.00 Copyright 18 1992 by Academic Press, All rights of reproduction in any form
Swedish
Natural
Science
Re-
have been introduced (5-g), in which membranes are separated according to differences in surface properties (10). Two-phase partitioning has not turned out to be selective enough, however, for a facile purification of animal plasma membranes, although the technique has been used widely for preparation of plasma membranes from plants [for reviews see (11) and (12)]. In order to increase the selectivity of the two-phase partitioning technique, we are exploring the possibility of using biospecific ligands covalently coupled to phase polymers for membrane purification. Thus, the lectin wheat-germ agglutinin attached to dextran has turned out to selectively pull rat liver plasma membranes into the dextran-rich bottom phase of a polyethylene glycoldextran two-phase system (13). We have now developed a rapid and efficient method for the preparation of rat liver plasma membranes of high purity and yield. It includes homogenization and crude fractionation in a conventional aqueous polymer two-phase system followed by selective extraction of plasma membranes by biospecific affinity partitioning using wheat-germ agglutinin as a ligand.
MATERIALS
AND
METHODS
Chemicals Stock solutions in water of 20% (w/w) Dextran T500 (Pharmacia Fine Chemicals AB) and 40% (w/w) polyethylene glycol 3350 (Carbowax 3350; Union Carbide) were prepared as described by Lopez-Perez et al. (14). UDP-[14C]galactose and [3H]AMP were from Amersham International plc, [‘251]iodine was from New England Nuclear, 2,2,2-trifluoroethanesulfonyl chloride (tresyl chloride) was from Synthelec (Lund, Sweden) and wheat-germ agglutinin was from BoehringerMannheim. All other reagents used were of analytical grade. 131
Inc. reserved.
132
PERSSON
Activation of Dextran Dextran T500 was activated with tresyl chloride essentially as described by Persson and Olde (15). The dextran was dissolved in water and freeze-dried before use. Five grams of this dextran was dissolved in 25 ml of dimethyl sulfoxide, and then 1 ml of triethylamine was added slowly, followed by 5 ml of dichloromethane. After chilling on ice, 0.35 g of tresyl chloride was added dropwise while stirring. The mixture was stirred for 30 min on ice, for 60 min at 4°C and overnight at 20°C. The reaction was terminated by precipitating the dextran with 50 ml of dichloromethane. After washing and thorough kneading with several portions of dichloromethane, the precipitate was dissolved in water, dialyzed against water overnight, and freezedried. The activated dextran was stored at -20°C until use. The recovery of dextran was more than 95%. Freeze-drying was introduced to reduce the time required to dissolve dextran from several hours to a few minutes and to reduce the risk of tresyl chloride hydrolysis during activation. Coupling of Wheat-Germ Agglutinin to Tresyl Dextran The lectin wheat-germ agglutinin was coupled to tresyl dextran as described earlier (13). Two grams of tresyl dextran was dissolved in 10 ml of coupling buffer (0.1 M NaH,PO,/Na,HP0,/0.5 M NaCl, pH 7.5) before the addition of 10 mg of wheat-germ agglutinin dissolved in 1 ml of coupling buffer. The coupling proceeded overnight at 4°C with gentle agitation. The reaction was terminated by the addition of 0.1 M Tris/HCl, pH 7.5, to inactivate excess tresyl groups. Uncoupled wheat-germ agglutinin and salts were removed by ultrafiltration of the dextran-ligand mixture in a Filtron Omegacell (150ml size, cut-off 100 K) before freeze-drying. The recovery of dextran was more than 90%. When stored at -20°C for up to 6 months no loss of binding capacity for rat liver plasma membranes has been observed. The amount of wheat-germ agglutinin coupled to dextran was determined according to Bradford (16) using wheat-germ agglutinin as standard. Routinely 4 mg was bound per gram freeze-dried product. Purification Protocol for Plasma Membranes Adult male Sprague-Dawley rats (175-200 g) were killed by decapitation. Each liver was immediately chilled in ice-cold 0.25 M sucrose in 5 mM Tris/HCl, pH 8.0, squashed through a garlic press, and transferred to a Dounce homogenizer containing an aqueous twophase system. This system had a total weight of 15 g, including up to 2 g of liver, in 5.7% (w/w) each of Dextran T500 and polyethylene glycol 3350, diluted in 15 mM Tris/H,SO,, pH 7.8, from the stock solutions. After homogenization by 10 strokes with a loose-fitting pestle and 20 strokes with a tight-fitting one, the homogenization system was centrifuged at low speed (150g) for 5
AND
JERGIL
min in a swing-out rotor to facilitate phase separation. The top phase was collected and the bottom phase reextracted with an equal volume of fresh top phase obtained after preequilibration with bottom phase. The system was turned upside down 20 times, vortex-mixed, and turnedupside down another 20 times to ensure thorough mixing before phase separation as above. The combined top phases were added to a new bottom phase also containing dextran-bound wheat-germ agglutinin for affinity partitioning. This bottom phase had been prepared by preequilibration of a 10-g two-phase system containing 6.0% (w/w) of each polymer, including dextran-bound wheat-germ agglutinin as indicated in each case, and 20 mM boric acid adjusted to pH 7.8 with Tris base, and subsequent removal of the top phase. The volume of this bottom phase relative that of the combined top phases was not very critical and could be varied +50% compared to the amount given here. The combined top phases and the new bottom phase were thoroughly mixed as above, and the phases separated by gentle centrifugation (Wifug Doctor, 1250 rev/ min) for 5 min. The bottom phase obtained was reextracted with fresh, preequilibrated top phase twice in the same manner. After the final phase separation the bottom phase was diluted lo-fold in 0.25 M sucrose and 0.1 M N-acetylglucosamine in 5 mM Tris/HCl, pH 8.0, and membranes were pelleted by centrifugation for 60 min at 100,OOOg. The pellet was resuspended in 0.25 M sucrose and 5 mM Tris/HCl, pH 8. It should be noted that the performance of phase systems is sensitive to temperature. All phase separations were therefore done with solutions carefully adjusted to +4OC. Analyses The plasma membrane markers measured were 5’-nucleotidase [EC 3.1.3.5; (17)], alkaline phosphodiesterase I (EC 3.1.4.1; 18) and 1251-labeled asialoorosomucoid binding (4,8). Other markers measured were arylesterase [endoplasmic reticulum, EC 3.1.1.2; (19)], N-acetylglucosamine galactosyltransferase [Golgi membranes, EC 2.4.1.38; (20)], succinate-cytochrome c reductase [mitochondria, EC 1.3.99.1; (21)], N-acetyl-Pglucosaminidase [lysosomes, EC 3.2.1.30; (22)] and lactate dehydrogenase [cytosol, EC 1.1.1.27; (23)]. Protein was determined according to Bradford (16) using bovine serum albumin as standard. RESULTS
AND
DISCUSSION
We introduce a rapid method for purification of rat liver plasma membranes of high purity and yield. The procedure, as outlined in Fig. 1 and described in detail under Materials and Methods, includes homogenization of the tissue and crude fractionation in a conventional
AFFINITY
FIG. 1. Methods. agglutinin
PURIFICATION
Outline of the purification protocol for rat liver plasma Each addition of fresh phase, polyethylene glycol-containing (WGA-B), is indicated. Only phases designed for further
aqueous polymer two-phase system followed by selective purification of plasma membranes by lectin-affinity partitioning. After homogenization in the two-phase system and phase separation, the bottom phase is reextracted with fresh top phase to increase the yield. The two top phases are combined and partitioned in the presence of borate with the bottom phase containing dextran-linked wheat-germ agglutinin. Under these conditions membranes preferentially partition in the top phase, except plasma membranes which contain exposed N-acetylglucosamine residues and, therefore, are selectively pulled into the bottom phase by the lectin ligand. To improve purity, the wheat-germ agglutinincontaining bottom phase is reextracted twice with fresh top phase. The final bottom phase is diluted with Nacetylglucosamine to dissociate membranes from the ligand before sedimentation of the membranes.
OF
PLASMA
133
MEMBRANES
membranes. The procedure top phase (T) or bottom processing are labeled.
is described in detail under phase containing dextran-linked
Materials and wheat-germ
mains partitioned in the polyethylene glycol-rich top phase upon phase separation (Fig. 2), whereas enzyme markers for other membranes had a lower partition ratio, in line with our earlier observation (8). Reextraction of the bottom phase with fresh top phase increased the yield of plasma membranes to around 70% (Fig. 2). Although other membranes increased relatively more in the combined top phases by this second extraction, the final purity of plasma membranes was not affected, as contaminating membranes were efficiently removed during the subsequent affinity partitioning step (see below).
60
Composition
of the Homogenization
System
The partitioning of biological membranes in aqueous polymer two-phase systems is strongly influenced by polymer concentration and salt content (10). As previously reported (8) the partition ratios of rat liver membranes increase in the order: endoplasmic reticulum, mitochondria, lysosomes, Golgi, and plasma membranes. Close to the critical point [5.4% (w/w) of each polymer at 4”C] of a polyethylene glycol 3350-dextran T500 two-phase system different types of membranes partition almost equally between the phases. An increase in the concentration of the phase polymers selectively decreases the partition ratios of other membranes, particularly mitochondria and endoplasmic reticulum, relative to plasma membranes. The most selective system found consisted of 5.7% (w/w) of each polymer in 15 mM Tris/H,SO,, pH 7.8. After homogenization of squashed liver in this system, approximately 50% of the plasma membrane markers for both the apical (5’-nucleotidase and alkaline phosphodiesterase) and the basolateral (asialoorosomucoid binding) do-
1
2
3
4
MARKER
5
6
7
8
9
ENZYMES
FIG. 2. Partitioning of protein and marker enzymes after homogenization. One gram of squashed liver was homogenized in a 15-g twophase system followed by phase separation as described under Materials and Methods. The dextran-rich bottom phase was reextracted once with preequilibrated polyethylene glycol-rich top phase. The recoveries (percentage of the homogenate) of protein [l], asialoorosomucoid binding [2], alkaline phosphodiesterase [3], 5’nucleotidase [4], arylesterase [5], succinate-cytochrome c reductase [6], galactosyltransferase [7], N-acetylglucosaminidase [S], and lactate dehydrogenase [9] in each of the resulting top phases were determined. The first top phase, closed bars; the second, open bars. Results are averages of three independent preparations with essentially the same membrane distribution.
PERSSON
AND
JERGIL
combined top phases from the homogenate was attracted to the bottom phase by 80 pg of dextran-bound wheat-germ agglutinin, leaving 80%, or more, of markers for other membranes in the top phase (Fig. 3). Purity
U
100
200
WGA
300
hi9
FIG. 3. Affinity partitioning of rat liver membranes. The top phases obtained after homogenization and reextraction were combined. One gram of these, containing 4 mg of protein, was partitioned in the presence of increasing amounts of dextran-linked wheat-germ agglutinin. Each system had a total weight of 5 g and was treated as described under Materials and Methods. After two reextractions of the wheat-germ agglutinin containing bottom phase, the protein content (0) and the activities of alkaline phosphodiesterase (APDE, 0), 5’-nucleotidase (5-N, a), galactosyltransferase (GAL, Cl), arylesterase (AE, A), and N-acetylglucosaminidase (NACG, n ) were measured in the combined top phases and the final bottom phase. They are expressed as percentages in the combined top phases of the total activity in the phases. Recoveries were 80-105% of the starting material.
Affinity
Partitioning
In order to establish the amount of dextran-linked wheat-germ agglutinin required for pulling plasma membranes into the bottom phase, and to examine the selectivity of this step, the combined top phases of the homogenization system were partitioned with fresh bottom phase containing increasing amounts of ligand. Each of the resulting bottom phases was reextracted twice with fresh top phase before measuring marker enzyme activities in the combined top phases and the final bottom phase (Fig. 3). These separations were performed in the presence of borate, which nonselectively pushes membranes into the polyethylene glycol-rich top phase (13). As a consequence, the bulk of membranes was recovered in the combined top phases in the absence of ligand, whereas increasing amounts of ligand selectively pulled plasma membranes into the dextranrich bottom phase leaving other membranes in the top phase. Thus, approximately 90% of the plasma membrane markers contained in 4 mg protein present in the
and Yield of Plasma Membranes
As shown in Table 1 plasma membranes of both apical and basolateral domains were purified 30- to 40-fold over the homogenate by the procedure. The yield of the basolateral marker asialoorosomucoid binding was close to 55%, and that of the apical markers 5’-nucleotidase and alkaline phosphodiesterase 65-70%. The Golgi membrane and lysosomal markers were slightly enriched (6- and 3-fold, respectively) whereas marker activities for mitochondria, endoplasmic reticulum, and the cytosol were too low to be determined accurately. The conventional two-phase separation performed in the homogenization medium (including repartitioning of the first bottom phase with fresh top phase) removed 95% or more of mitochondria and endoplasmic reticulum (Table l), leaving significant amounts of Golgi membranes, lysosomes, and soluble proteins in the combined top phases. The subsequent affinity partitioning (including two reextractions of the wheat-germ agglutinin-containing bottom phase) removed most of the Golgi membranes and lysosomes but only a small amount of plasma membranes. The final centrifugation step pelleted the membranes present in the bottom phase, and removed remaining soluble proteins (as indicated by lactate dehydrogenase activity), thereby increasing the plasma membrane purification factor by approximately 2.5 times. The polymer plus the ligand were also removed in this step, as the final bottom phase was diluted with N-acetylglucosamine solution to dissociate ligand from membranes before the centrifugation. The degree of enrichment of plasma membrane markers by the two-phase technique was similar to that achieved by a method involving homogenization under well-defined conditions followed by centrifugations (4), but the yield was approximately four times higher. These two methods result in plasma membrane preparations consisting of both apical and basolateral domains. Other methods, in contrast, aim at separating these domains (2,3) by manipulating homogenization and centrifugation conditions. As various markers have different lateral distributions in the plasma membrane, higher enrichment factors than those reported here can be achieved by selective purification of plasma membrane domains (l-3,24) or, alternatively, by subfractionation of purified plasma membranes (8,25,26). Although the different domains have slightly different densities, we have only been able to separate these domains partly, so far, by density gradient centrifugation of plasma membranes prepared by the two-phase technique.
AFFINITY
PURIFICATION
OF
PLASMA
TABLE Distribution
1
of Membrane
Marker
Enzymes
Homogenate
Marker Protein Alkaline phosphodiesterase 5’-Nucleotidase Asialoorosomucoid binding Arylesterase Galactosyltransferase Succinate-cytochrome c reductase N-Acetylglucosaminidase Lactate dehydrogenase
Specific 114.6 42.3 13.6 3.99 717 0.44 36.4 2.41 3290
activity * f k f f f k k *
2.4 3.6 1.8 0.82 35 0.03 3.1 0.12 308
Affinity Combined top phases (%)
Combined top phases (%)
29.9 73.8 70.1 61.0 5.2 53.5 2.5 43.7 16.4
24.2 7.0 4.3 2.1 4.6 40.2
f +f f f +f -+ k
135
MEMBRANES
1.1 3.4 4.2 4.6 0.2 4.9 0.2 4.6 1.1
+ k 2 f k f
