Journal of Neurochemistry Raven Press, Ltd., New York 0 199 I International Society for Neurochemistry
Isolation of Hippocampal Synaptosomes on Percoll Gradients: Cholinergic Markers and Ligand Binding Sites B. Thorne, S. Wonnacott, and *P. R. Dunkley Department of Biochemistry, University of Bath, Bath, England; and *The Neuroscience Group, Faculty of Medicine, University of Newcastle, New South Wales, Australia
Abstract: The S 1 Percoll procedure, devised empirically for cortical tissue, provides highly purified, functionally viable synaptosomes on a four-step Percoll gradient. Here, for the first time, the procedure has been applied to rat hippocampus, and the gradient fractions have been analysed with respect to cholinergic markers and the synaptosomal index, lactate dehydrogenase. The presynaptic cholinergic markers choline acetyltransferaseand [3H]cholineuptake were most enriched in fraction 4. In contrast, acetylcholinesterase activity was broadly distributed across the gradient, consistent with the separation of synaptic plasma membranes (in fractions 1 and 2) from synaptosomes(in fractions 3 and 4). This is supported by the recovery of muscarinic binding sites labelled with [3H]quinuclidinylbenzilate in fractions 1 and 2. (-H3H]-
Nicotine binding sites, however, were most enriched in fraction 4, consistent with their predominantly presynaptic localisation in the CNS. These results demonstrate the applicability of the S1 Percoll method to discrete brain regions for the recovery of homogeneous and viable synaptosome fractions. The separation of presynaptic terminals from postsynaptic membranes is a further advantage of this technique. Key Words: Choline acetyltransferase-Choline uptakeQuinuclidinylbenzilate binding sites-Nicotine binding sites-Hippocampus-Presynaptic receptors. Thorne B. et al. Isolation of hippocampal synaptosomes on Percoll gradients: Cholinergic markers and ligand binding sites. J. Neurochem. 56,479-484 (1991).
The release of acetylcholine (ACh) from neurons in the CNS is subject to positive and negative feedback regulation via presynaptic nicotinic and muscarinic receptors, respectively (Raiteri et al., 1984; Wonnacott et al., 1989). To facilitate the study of these and other mechanisms influencing ACh release, a preparation of highly purified, functionally viable synaptosomes would be an advantage. The Sl Percoll procedure (Dunkley et al., 1988; Harrison et al., 1988) has provided such a preparation from rat cortex. The technique involves only two medium-speed centrifugation steps: the first of 10 min results in an S 1 fraction, and the second of 5 min uses a discontinuous four-step Percoll gradient, from which five major fractions (interfacial fractions 1-4 and pellet fraction 5) are obtained under near isotonic conditions. In addition to its speed and convenience, the procedure offers several advantages over traditional subcellular fractionation techniques for preparing synaptosomes
(Dunkley et al., 1987, 1988). Because the Percoll procedure is a nonequilibrium method which separates particles on the basis of their size, as well as their density (in contrast to conventional isopycnic gradients), the highly purified synaptosome fractions, especially fraction 4, are devoid of myelin, free mitochondria, and synaptic plasma membranes. Moreover, the Percoll procedure separates synaptosomes of different size and mitochondria1 content (Dunkley et al., 1988). Small synaptosomes (with only 8% containing mitochondria) are found in fraction 1, whereas fraction 5 contains large synaptosomeswith 83%containing mitochondria. Those synaptosomes without mitochondria are not functionally viable, and hence this method is unique in separating viable from nonviable synaptosomes. Comparison of markers for various transmitter systems suggests that populations of synaptosomes containing different neurotransmitters may be separated on the Percoll gradient (Harrison et al., 1988).
Received February 27, 1990; revised manuscript received May 25, 1990; accepted July 23, 1990. Address correspondence and reprint requests to Dr. S. Wonnacott at Department of Biochemistry, University of Bath, Bath BA2 7AY,
siology, Albert Einstein College of Medicine, Bronx, NY 10467, USA. Abbreviations used: ACh, acetylcholine; AChE, acetylcholinesterase; ChAT, choline acetyltransferase;GABA, y-aminobutync acid; LDH, lactate dehydrogenase; QNB, quinuclidinylbenzilate.
~~~
U.K.
The present address of Dr. B. Thorne is Department of Anesthe-
4 79
~
~
480
B. THORNE ET AL
T h e Percoll procedure was established empirically for rat cortical tissue, but has been used successfully in the preparation of synaptosomes from rat striatum (Robinson and Lovenberg, 1986; Robinson et al., 1989). In preliminary studies, we have shown that hippocampal synaptosomes prepared by this method can be used for monitoring ACh release (Thorne et al., 1988). In t h e present study, the fractionation of cholinergic nerve terminals from rat hippocampus on Percoll density gradients has been characterised, with respect to acetylcholinesterase (AChE; EC 3.1.1.7), choline acetyltransferase (ChAT; E C 2.3.1.6), [3H]choline uptake, and cholinergic ligand binding sites.
MATERIALS AND METHODS Materials Percoll was obtained from Pharmacia (Uppsala, Sweden). [3H]Choline (80 Ci/mmol), 1-quinuclidinyl[ phen~d-4-~H]benzilate (QNB 42 Ci/mmol), (-)-[N-me~hy/-~H]nicotine(72 Ci/mmol), and [3H]acetyl-coenzyme A (3 Ci/mmol) were obtained from Amersham International (Amersham, Bucks, U.K.). All general reagents were supplied by either BDH (Poole, Donet, U.K.), Fisons (Loughborough, Leicestershire, U.K.), or Sigma Chemical Company (Poole, Dorset, U.K.).
Subcellular fractionation Male Wistar rats (200-250 g) were killed by cervical dislocation, and the brains were rapidly removed and dissected. Six hippocampi (0.6-0.65 g wet weight) were homogenised (lo%, wt/vol) in 0.32 M sucrose, pH 7.4, and synaptosomes were isolated as described by Dunkley et al. (1988) on four Percoll gradients. Each gradient fraction was collected and washed twice with 10 ml of Krebs bicarbonate buffer (composition, in m M NaCl, 1 18; KCI, 2.35; CaCI, 2H20. 2.40; KH2P04, I .20; MgS04 * 7H20, 1.20; NaHC03, 25; glucose, 10;gassed with 95% 0 2 / 5 % C 0 2to pH 7.4) by centrifugation at 15,000 g for 15 min. Each fraction (pooled from four gradients) was resuspended in Krebs buffer (fractions 1-4 in 1.5 ml; fraction 5 in 0.7 ml, unless otherwise stated). e
Enzyme assays Lactate dehydrogenase (LDH; EC I . 1 .I .27). Enzyme activity in samples (0.1 ml) of the gradient fractions and SI was assayed as described by Johnson ( 1960) and modified by Marchbanks (1967). The inclusion of Triton X- 100 to rupture the synaptosomes gives a measure of occluded LDH, indicative of synaptosomal integrity. AChE. AChE activity in the subcellular fractions (S I, 0.1 ml; gradient fractions, 0.5 ml) was determined by the method of Ellman et al. ( 196 1). C U T . The protocol used to determine ChAT activity was based on the liquid cation-exchange method described by Fonnum (1975). The washed gradient fractions were resuspended in phosphate buffer [0.05M , pH 7.4, containing 0.2 M NaC1, 1 mM EDTA, and 0.5% (vol/vol) Triton X-1001 to give a protein concentration of 1 mg/ml. Samples (10 p l ) were incubated for 10 rnin at 37°C with 20 p1 of assay buffer (see below) in the presence and absence of 20 p1 of formic acid. The assay buffer, freshly prepared, comprised 12.5 m M choline chloride, 0.1 mM neostigmine bromide, 0.2 mM [3H]acetyl-coenzyme A (sp. act. 10 mCi/mmol), and 5 mg of bovine serum albumin/ml in phosphate buffer. After 10 min, the reaction was stopped by adding 20 pI of formic acid. J. Neurochern.. Vol. 56, No. 2, 1991
[3H]AChwas extracted in 0.3 ml of heptan-2-one containing tetraphenylboron ( 1 5 mg/ml). A sample ( 150 p l ) of the upper, organic phase was counted for radioactivity in 5 ml of Optiphase-Safe (LKB) in a Packard scintillation spectrometer. The counting efficiency was 25%. Nonspecific acetyltransferase activity was determined in the presence of bromoacetylcholine (0.1 mM), a ChAT inhibitor (Roskoski et al., 1974). Bromoacetylcholine was synthesized as described by Chiou and Sastry (1968).
Uptake assays [3HJCholine tiptake. The washed gradient fractions (1 20pl samples) were preincubated at 37°C for 10 min before incubation with [3H]choline (final concentration 0.8 p M , sp. act. 3 I Ci/mmol) for 20 min at 37°C. Samples (100 pl) were then filtered on Whatman GF/C filters and washed with 5 ml of Krebs buffer. Nonspecific uptake was determined in the presence of I pM hemicholinium-3.
Ligand binding assays (-)-[ 'HINicotine binding. Gradient fractions were washed a further two times in 10 ml of Krebs buffer to remove any residual traces of Percoll, which itself displays specific [3H]nicotine binding. The fractions were then resuspended in 3 ml of Krebs buffer and assayed essentially as previously described (MacAllan et al., 1988). Samples (0.25 ml; neat, twofold and fourfold dilutions) were incubated with (-)[3H]nicotine (20 nM) for 30 min at 20°C, in the presence and absence of excess unlabelled nicotine ( M ) , to determine nonspecific binding. Mitscurinic binding. Washed gradient fractions were resuspended in 1 ml of phosphate-buffered saline and assayed for ['HIQNB binding by the method of Yamamura and Snyder ( 1974). Samples (10-50-fold dilutions; 1 ml) were incubated with [3H]QNB(2 nM) for 60 min at 30"C, in the presM ) to determine nonence and absence of atropine ( specific binding. The samples were rapidly filtered through a double thickness of Whatman GF/C filter paper, using a Brandell cell harvester. Soaking the filter paper in 0.1% polyethylenimine reduced blank values (from 300 cpm to 150 cpm) and improved the reproducibility of replicates.
Protein determination Protein was measured using the method of Lowry et al. (195 I), using bovine serum albumin as standard. For samples containing detergent (ChAT assay), protein determination was camed out using the modification of the Lowry method as described by Markwell et al. (1978).
Statistical analyses Comparison of the specific activities of markers in S1 with their specific activities in the gradient fractions was made using Student's two-tailed, nonpaired t test. Statistical significance at the 95% level of confidence was accepted.
RESULTS Subcellular fractionation on Percoll gradients: general markers Fractionation of the low-speed supernatant (S 1) from the hippocampal homogenate on four-step Percoll density gradients (Dunkley et al., 1988) yielded five major fractions: interfacial fractions 1-4 and a pellet, fraction 5. T h e distribution of protein across the gradient (Table 1) showed that fraction 2 had two to five
CHOLINERGIC S YNAPTOSOMES FROM HIPPOCAMPUS TABLE 1. Distribution of hippocampal protein on Percoll gradients Subcellular fraction
mg of protein/g of tissue
S1
24.3 2 2.2
Fraction 1 Fraction 2 Fraction 3 Fraction 4 Fraction 5 Sum of fractions 1-5
1.7 f 0.4 6.4 f 0.4 3.4 f 0.2 2.0 f 0.3 1.1 f 0.1 14.6 f 1.4
% recovery
100
7.0 26.3 14.0 8.2 4.5 60.0
synaptosomes in fractions 2-4. Fraction 5, the mitochondrial fraction, had a lower level of LDH activity, and only 38% of this was occluded. Cholinergic markers Gradient fractions from rat hippocampus were assayed for cholinergic markers. Of the AChE activity measured in the S1 fraction, 68% was recovered in the combined gradient fractions. This enzyme was found in all the gradient fractions (Fig. 2a), and its activity
la
2.5-
S1 fractions were prepared from rat hippocampus and applied to
Percoll gradients. After centrifugation, fractions 1-5 were collected and washed twice with Kreb's buffer before determination of protein; values are calculated per gram of original tissue. Protein was also measured in the S1 fraction after two washes. The recovery of protein is expressed as a percentage of this value. Values are the means f SEM (n = 6).
L
.2" .> h c 0
Isolation of Hippocampal Synaptosomes on Percoll Gradients: Cholinergic Markers and Ligand Binding Sites B. Thorne, S. Wonnacott, and *P. R. Dunkley Department of Biochemistry, University of Bath, Bath, England; and *The Neuroscience Group, Faculty of Medicine, University of Newcastle, New South Wales, Australia
Abstract: The S 1 Percoll procedure, devised empirically for cortical tissue, provides highly purified, functionally viable synaptosomes on a four-step Percoll gradient. Here, for the first time, the procedure has been applied to rat hippocampus, and the gradient fractions have been analysed with respect to cholinergic markers and the synaptosomal index, lactate dehydrogenase. The presynaptic cholinergic markers choline acetyltransferaseand [3H]cholineuptake were most enriched in fraction 4. In contrast, acetylcholinesterase activity was broadly distributed across the gradient, consistent with the separation of synaptic plasma membranes (in fractions 1 and 2) from synaptosomes(in fractions 3 and 4). This is supported by the recovery of muscarinic binding sites labelled with [3H]quinuclidinylbenzilate in fractions 1 and 2. (-H3H]-
Nicotine binding sites, however, were most enriched in fraction 4, consistent with their predominantly presynaptic localisation in the CNS. These results demonstrate the applicability of the S1 Percoll method to discrete brain regions for the recovery of homogeneous and viable synaptosome fractions. The separation of presynaptic terminals from postsynaptic membranes is a further advantage of this technique. Key Words: Choline acetyltransferase-Choline uptakeQuinuclidinylbenzilate binding sites-Nicotine binding sites-Hippocampus-Presynaptic receptors. Thorne B. et al. Isolation of hippocampal synaptosomes on Percoll gradients: Cholinergic markers and ligand binding sites. J. Neurochem. 56,479-484 (1991).
The release of acetylcholine (ACh) from neurons in the CNS is subject to positive and negative feedback regulation via presynaptic nicotinic and muscarinic receptors, respectively (Raiteri et al., 1984; Wonnacott et al., 1989). To facilitate the study of these and other mechanisms influencing ACh release, a preparation of highly purified, functionally viable synaptosomes would be an advantage. The Sl Percoll procedure (Dunkley et al., 1988; Harrison et al., 1988) has provided such a preparation from rat cortex. The technique involves only two medium-speed centrifugation steps: the first of 10 min results in an S 1 fraction, and the second of 5 min uses a discontinuous four-step Percoll gradient, from which five major fractions (interfacial fractions 1-4 and pellet fraction 5) are obtained under near isotonic conditions. In addition to its speed and convenience, the procedure offers several advantages over traditional subcellular fractionation techniques for preparing synaptosomes
(Dunkley et al., 1987, 1988). Because the Percoll procedure is a nonequilibrium method which separates particles on the basis of their size, as well as their density (in contrast to conventional isopycnic gradients), the highly purified synaptosome fractions, especially fraction 4, are devoid of myelin, free mitochondria, and synaptic plasma membranes. Moreover, the Percoll procedure separates synaptosomes of different size and mitochondria1 content (Dunkley et al., 1988). Small synaptosomes (with only 8% containing mitochondria) are found in fraction 1, whereas fraction 5 contains large synaptosomeswith 83%containing mitochondria. Those synaptosomes without mitochondria are not functionally viable, and hence this method is unique in separating viable from nonviable synaptosomes. Comparison of markers for various transmitter systems suggests that populations of synaptosomes containing different neurotransmitters may be separated on the Percoll gradient (Harrison et al., 1988).
Received February 27, 1990; revised manuscript received May 25, 1990; accepted July 23, 1990. Address correspondence and reprint requests to Dr. S. Wonnacott at Department of Biochemistry, University of Bath, Bath BA2 7AY,
siology, Albert Einstein College of Medicine, Bronx, NY 10467, USA. Abbreviations used: ACh, acetylcholine; AChE, acetylcholinesterase; ChAT, choline acetyltransferase;GABA, y-aminobutync acid; LDH, lactate dehydrogenase; QNB, quinuclidinylbenzilate.
~~~
U.K.
The present address of Dr. B. Thorne is Department of Anesthe-
4 79
~
~
480
B. THORNE ET AL
T h e Percoll procedure was established empirically for rat cortical tissue, but has been used successfully in the preparation of synaptosomes from rat striatum (Robinson and Lovenberg, 1986; Robinson et al., 1989). In preliminary studies, we have shown that hippocampal synaptosomes prepared by this method can be used for monitoring ACh release (Thorne et al., 1988). In t h e present study, the fractionation of cholinergic nerve terminals from rat hippocampus on Percoll density gradients has been characterised, with respect to acetylcholinesterase (AChE; EC 3.1.1.7), choline acetyltransferase (ChAT; E C 2.3.1.6), [3H]choline uptake, and cholinergic ligand binding sites.
MATERIALS AND METHODS Materials Percoll was obtained from Pharmacia (Uppsala, Sweden). [3H]Choline (80 Ci/mmol), 1-quinuclidinyl[ phen~d-4-~H]benzilate (QNB 42 Ci/mmol), (-)-[N-me~hy/-~H]nicotine(72 Ci/mmol), and [3H]acetyl-coenzyme A (3 Ci/mmol) were obtained from Amersham International (Amersham, Bucks, U.K.). All general reagents were supplied by either BDH (Poole, Donet, U.K.), Fisons (Loughborough, Leicestershire, U.K.), or Sigma Chemical Company (Poole, Dorset, U.K.).
Subcellular fractionation Male Wistar rats (200-250 g) were killed by cervical dislocation, and the brains were rapidly removed and dissected. Six hippocampi (0.6-0.65 g wet weight) were homogenised (lo%, wt/vol) in 0.32 M sucrose, pH 7.4, and synaptosomes were isolated as described by Dunkley et al. (1988) on four Percoll gradients. Each gradient fraction was collected and washed twice with 10 ml of Krebs bicarbonate buffer (composition, in m M NaCl, 1 18; KCI, 2.35; CaCI, 2H20. 2.40; KH2P04, I .20; MgS04 * 7H20, 1.20; NaHC03, 25; glucose, 10;gassed with 95% 0 2 / 5 % C 0 2to pH 7.4) by centrifugation at 15,000 g for 15 min. Each fraction (pooled from four gradients) was resuspended in Krebs buffer (fractions 1-4 in 1.5 ml; fraction 5 in 0.7 ml, unless otherwise stated). e
Enzyme assays Lactate dehydrogenase (LDH; EC I . 1 .I .27). Enzyme activity in samples (0.1 ml) of the gradient fractions and SI was assayed as described by Johnson ( 1960) and modified by Marchbanks (1967). The inclusion of Triton X- 100 to rupture the synaptosomes gives a measure of occluded LDH, indicative of synaptosomal integrity. AChE. AChE activity in the subcellular fractions (S I, 0.1 ml; gradient fractions, 0.5 ml) was determined by the method of Ellman et al. ( 196 1). C U T . The protocol used to determine ChAT activity was based on the liquid cation-exchange method described by Fonnum (1975). The washed gradient fractions were resuspended in phosphate buffer [0.05M , pH 7.4, containing 0.2 M NaC1, 1 mM EDTA, and 0.5% (vol/vol) Triton X-1001 to give a protein concentration of 1 mg/ml. Samples (10 p l ) were incubated for 10 rnin at 37°C with 20 p1 of assay buffer (see below) in the presence and absence of 20 p1 of formic acid. The assay buffer, freshly prepared, comprised 12.5 m M choline chloride, 0.1 mM neostigmine bromide, 0.2 mM [3H]acetyl-coenzyme A (sp. act. 10 mCi/mmol), and 5 mg of bovine serum albumin/ml in phosphate buffer. After 10 min, the reaction was stopped by adding 20 pI of formic acid. J. Neurochern.. Vol. 56, No. 2, 1991
[3H]AChwas extracted in 0.3 ml of heptan-2-one containing tetraphenylboron ( 1 5 mg/ml). A sample ( 150 p l ) of the upper, organic phase was counted for radioactivity in 5 ml of Optiphase-Safe (LKB) in a Packard scintillation spectrometer. The counting efficiency was 25%. Nonspecific acetyltransferase activity was determined in the presence of bromoacetylcholine (0.1 mM), a ChAT inhibitor (Roskoski et al., 1974). Bromoacetylcholine was synthesized as described by Chiou and Sastry (1968).
Uptake assays [3HJCholine tiptake. The washed gradient fractions (1 20pl samples) were preincubated at 37°C for 10 min before incubation with [3H]choline (final concentration 0.8 p M , sp. act. 3 I Ci/mmol) for 20 min at 37°C. Samples (100 pl) were then filtered on Whatman GF/C filters and washed with 5 ml of Krebs buffer. Nonspecific uptake was determined in the presence of I pM hemicholinium-3.
Ligand binding assays (-)-[ 'HINicotine binding. Gradient fractions were washed a further two times in 10 ml of Krebs buffer to remove any residual traces of Percoll, which itself displays specific [3H]nicotine binding. The fractions were then resuspended in 3 ml of Krebs buffer and assayed essentially as previously described (MacAllan et al., 1988). Samples (0.25 ml; neat, twofold and fourfold dilutions) were incubated with (-)[3H]nicotine (20 nM) for 30 min at 20°C, in the presence and absence of excess unlabelled nicotine ( M ) , to determine nonspecific binding. Mitscurinic binding. Washed gradient fractions were resuspended in 1 ml of phosphate-buffered saline and assayed for ['HIQNB binding by the method of Yamamura and Snyder ( 1974). Samples (10-50-fold dilutions; 1 ml) were incubated with [3H]QNB(2 nM) for 60 min at 30"C, in the presM ) to determine nonence and absence of atropine ( specific binding. The samples were rapidly filtered through a double thickness of Whatman GF/C filter paper, using a Brandell cell harvester. Soaking the filter paper in 0.1% polyethylenimine reduced blank values (from 300 cpm to 150 cpm) and improved the reproducibility of replicates.
Protein determination Protein was measured using the method of Lowry et al. (195 I), using bovine serum albumin as standard. For samples containing detergent (ChAT assay), protein determination was camed out using the modification of the Lowry method as described by Markwell et al. (1978).
Statistical analyses Comparison of the specific activities of markers in S1 with their specific activities in the gradient fractions was made using Student's two-tailed, nonpaired t test. Statistical significance at the 95% level of confidence was accepted.
RESULTS Subcellular fractionation on Percoll gradients: general markers Fractionation of the low-speed supernatant (S 1) from the hippocampal homogenate on four-step Percoll density gradients (Dunkley et al., 1988) yielded five major fractions: interfacial fractions 1-4 and a pellet, fraction 5. T h e distribution of protein across the gradient (Table 1) showed that fraction 2 had two to five
CHOLINERGIC S YNAPTOSOMES FROM HIPPOCAMPUS TABLE 1. Distribution of hippocampal protein on Percoll gradients Subcellular fraction
mg of protein/g of tissue
S1
24.3 2 2.2
Fraction 1 Fraction 2 Fraction 3 Fraction 4 Fraction 5 Sum of fractions 1-5
1.7 f 0.4 6.4 f 0.4 3.4 f 0.2 2.0 f 0.3 1.1 f 0.1 14.6 f 1.4
% recovery
100
7.0 26.3 14.0 8.2 4.5 60.0
synaptosomes in fractions 2-4. Fraction 5, the mitochondrial fraction, had a lower level of LDH activity, and only 38% of this was occluded. Cholinergic markers Gradient fractions from rat hippocampus were assayed for cholinergic markers. Of the AChE activity measured in the S1 fraction, 68% was recovered in the combined gradient fractions. This enzyme was found in all the gradient fractions (Fig. 2a), and its activity
la
2.5-
S1 fractions were prepared from rat hippocampus and applied to
Percoll gradients. After centrifugation, fractions 1-5 were collected and washed twice with Kreb's buffer before determination of protein; values are calculated per gram of original tissue. Protein was also measured in the S1 fraction after two washes. The recovery of protein is expressed as a percentage of this value. Values are the means f SEM (n = 6).
L
.2" .> h c 0
