036L9230/92 $5.00 + .OO
Brain Research Bulletin,Vol. 28, pp. 789-197, 1992 Printed in the USA. All rights reserved.
Copyright0 1992Pergamon Press Ltd.
BRIEF COMMUNICATION
Regional and Subcellular Distribution of [‘251]EndothelinBinding Sites in Rat Brain GORDON
T. BOLGER,’
ROBERT
BERRY
AND
JORGE
JARAMILLO
Department of Pharmacology, Bio-Mega Inc., Laval, Quebec, Canada H7S 2G5 Received
6 November
1990
BOLGER, G. T., R. BERRY AND J. JARAMILLO. Regional and subcellular distribution of[‘2sl]endothelin binding sites in rat brain. BRAIN RES BULL 28(5) 789-797, 1992.-THE binding of [‘251]endothelin-l (‘*51-ET-1)to membranes from whole rat brain, from individual brain regions, and derived from subcellular fractionation of whole rat brain was investigated. ‘251-ET-1 binding to whole rat brain membranes was rapid, concentration-dependent, saturable, and characterized as irreversible because it was not displaced by unlabeled endothelin-1 (ET- I) and different concentrations of ligand produced, with time, a similar magnitude of binding. The maximum binding site capacity and second-order forward rate association constant of binding were 1,946 f 147 fm/mg protein and 5.53 + 1.72 X IO6 M-’ s-l. Removal of either extramembranal calcium or membrane-bound calcium and calcium binding proteins did not affect the binding of ‘251-ET-l to whole rat brain membranes. The brain stem and cerebellum contained the highest levels of ‘251-ET-Ibinding sites, whereas the cerebral cortex, striatum, and hippocampus contained binding site levels three- to fourfold less. Subcellular fractionation of whole rat brain and subsequent analyses of the distribution of ‘251-ET-1binding demonstrated a twofold enrichment of binding sites in the synaptosomal fraction compared to the homogenate. The myelin fraction contained a similar density of binding sites compared to the homogenate, while the mitochondrial and microsomal fractions contained considerably less binding sites. The ribosomal fraction did not contain any ‘251-ET-1binding sites. The subcellular distribution of ‘*%ET- I binding sites did not correlate with the distribution of S-nucleotidase, cytochromeC ox&se, phosphodiesterase, and alkaline phosphatase. Depletion of extracellular calcium increased ‘251-ET-I binding in the synaptosomal fraction but not in the myelin and mitochondrial fractions. The second-order association rate constant for ‘2sI-ETI binding in the mitochondrial fraction was significantly different from that calculated for the synaptosomal and myelin fractions. These results demonstrate the presence of substantial densities of high-affinity binding sites for ET-1 in rat brain and suggestthat both plasmalemmal and intracellular ‘*‘I-ET-l binding sites with different properties may exist in the brain. Endothehn
Binding sites
Brain
Subcellular localization
Consistent with the identification of mRNA and high-affinity binding sites for ET- 1 in the CNS, significant levels of ET- 1 have been found in the rat spinal cord (44), although relatively high levels of ET-3 compared to either ET- 1 or ET-2 were identified in the intestine, lung, and pituitary in rat brain (38). ET- 1, presumably through its high-affinity binding site, is able to increase intracellular calcium levels in human and rat glioma cell lines and neuroblastoma cells ( 15,47,52) and stimulate the release of substance P from the rat pituitary and hypothalamus and aspartate from cultured cerebellar granule cells ( 10,33). ET- 1, ET2, and ET-3 stimulated phosphoinositide turnover in rat and guinea pig brain and cultured glioma and cerebellar granule cells ( 1,13,37,47). ET- 1 was also found to stimulate biosynthesis and mitogenesis in glioma cells (47). Consistent with the presence of high-affinity functionally linked “%ET- 1 binding sites in the brain, direct administration
NUMEROUS studies have characterized the biological effects of the potent vasoconstrictor endothelin peptides (endothelin- 1, ET1; endothelin-2, ET-2; endothelin-3, ET-3) in the periphery. The role of the endothelins in brain function, in particular ET-l, has recently gained considerable interest. mRNA coding for ET-l has been identified in both the rat and human CNS (20,2 1,3 1,38). Speciftc high-affinity binding sites for ET-1 have been identified in both tat and human brain ( 1,2,17,26,29,43) that interact with ET1 in an apparently irmversible manner (50). In the rat, ET- 1 binding sites am mostly local&d to the brainstem, basal ganglia, cerebellum, and choroid plexus ( 18,26,29) with similar patterns of binding site densities being observed in human brain (26). Comparative studies indicate that the neuronal [ tzI]endothelin- 1 (‘%ET- 1) binding site has a molecular weight of -50 kDa while ET-3 binds to sites with molecular weights of - 50 and - 38 kDa, suggesting multiple neuronal binding sites for the endothelins (2,43).
’ Requests for reprints should be addressed to Gordon T. Bolger, Ph.D., Research Scientist, Department of Pharmacology, Bio-Mega, Inc., 2100 rue Cunard, LavaI, Quebec, Canada H7S 2G5.
789
790
BOLGER, BERRY AND JARAMILLO
of ET- 1 into the CNS produced several effects depending on the area of injection and any accompanying physiopathologic conditions. In the spinal cord, intrathecal administration of ET-l produced spinal lesions (24) whereas intrastriatal injection of ET- 1 was also found to produce lesions ( 18) that were exacerbated during hypoglycemia (27). Centrally administered ET-l produced apnea in anesthetized rats (16). In conscious rats and at low doses, centrally administered ET-l produced decreases in motor activity, ataxia, and barrel rolling and at high doses wild running fits, convulsions, and death (32). Together, the rather marked biochemical and behavioral actions of the endothelins in the CNS warrant further biochemical characterization of their neuronal binding sites. To date, little, if any, information is available on the subcellular localization of ET-l binding sites in the brain. Thus, we studied the brain region and subcellular distribution and properties of high-affinity binding sites for the most potent of the endothelins, ET-l, in rat brain. METHOD
Preparation of Membranes
The binding of 12’I-ET-1 was investigated either in a washed homogenate of whole rat brain, washed homogenates of rat cerebral cortex, striatum, hippocampus, cerebellum, and brain stem, or washed subcellular fractions of whole rat brain. Male Sprague-Dawley rats (250-400 g; Canadian Breeding Farms, St. Constant, Quebec) were used for all studies. Rats were killed by decapitation. Brains were rapidly removed and placed into an ice-cold physiologic saline solution. All brain regions were isolated by blunt dissection on a glass stage cooled on ice. Either homogenates of whole rat brain or brain regions were prepared by homogenizing the brain tissue in 100 vol ice-cold buffer A using two bursts ofa polytron (Brinkmann Instruments, Toronto, Ontario). Buffer A had the following composition (mM): TrisHCl(25), NaCl(l35) KC1 (2.7) CaC12, (1.8) MgSOg (1. I), pH 7.4, at 22°C. The homogenate was centrifuged at 24,000X g for 15 min and the supernatant discarded. The membrane pellet was resuspended in 100 vol buffer A. This membrane preparation was used for radioligand binding. Under conditions were the calcium dependence of ‘251-ET-1 binding was to be studied, two protocols were employed. The first of these only removed extracellular calcium from brain membranes and employed the protocol above, replacing the CaC12component of buffer A with 5 PM EDTA (buffer B). The second one removed both membrane-bound calcium and certain calcium binding proteins from brain membranes (4,41). Following homogenization, centrifugation, and resuspension of brain membranes with buffer B, they were subsequently incubated with a buffer containing 1.5 mM EDTA (same buffer composition as buffer B but the 5 PM EDTA was replaced by 1.5 mM EDTA) at 4°C for 30 min. The membrane suspension was subsequently centrifuged at 24,000X g for 15 min and the pellet resuspended in 100 vol buffer B. Subcellular fractions of whole rat brain were prepared according to the method of Whittaker (5 1). All membrane fractions were washed free of excess sucrose by suspension in 100 vol buffer A and centrifugation at 100,000X g for 1 h. Following centrifugation, supematants were discarded and the membrane pellets resuspended in 100 vol buffer A. In experiments investigating the calcium dependence of ‘251-ET-l binding, the sucrose was removed from the subcellular fractions as just noted but by substituting buffer B for buffer A. Radioligand Binding
‘251-ET-i binding was performed in polypropylene tubes in a total assay volume of 1 ml consisting of 100 ~1 radioligand,
400 ~1 either buffer A or buffer B, and 500 ~1 brain membranes (washed homogenates or subcellular fractions suspended in either buffer A or B), 5 U aprotinin, and 1 PM phenylmethyl-sulfonylfluoride. The total amount of membrane protein in the assay ranged from 50-100 pg. Binding was initiated by addition of radioligand and the tubes were incubated either for 120 min or for the times indicated in the Table and Figure legends at 25°C. Binding was terminated by vacuum filtration of the membrane suspension through Whatman GF/B glass-fiber filters, followed by three 5-ml washes of ice-cold buffer A using a Brandel Cell Harvester (Model M24-R, Brandel Instruments, Gaithersburg, MD). Before filtering, filters were first soaked in a 0.1% v/v mixture of polyethyleneimine/distilled water for 15 min and subsequently in a 1% w/v solution of albumin/distilled water for a further 15 min. The nonspecific binding of “‘I-ET- 1 was determined by inclusion of unlabeled ET- 1 (lo-’ M) in the binding assay before addition of ‘“‘I-ET-l. Protein was determined by the method of Bradford (9). Marker Enzyme Studies
The subcellular distribution of the activities of S-nucleotidase, cytochromeC oxidase, alkaline phosphatase, and phosphodiesterase type I were determined spectrophotometrically (HewlettPackard Model 8452A Diode Array Spectrophotometer) using the following substrates (in parentheses) and protocols: 5’-nucleotidase [5’-adenosine monophosphate; (49)], cytochrome-C oxidase [oxidation of reduced cytochrome-C; (1 l)], alkaline phosphatase [paranitrophenylphosphate; (19)], and phosphodiesterase type I [paranitrophenyl-5’-thymidylate; (49)]. Kinetic Analysis of Radioligand Binding
The results of the radioligand binding studies were analyzed using the kinetic eq. (2) describing the irreversible second-order interaction between two molecules to yield product(s) (1): .4+B+CorC+D
[ l/(a - b)]ln{ [b(a
- x)]/a(b - x)} = k,t,
111 PI
where a and b are the initial concentrations of A and B at time zero, x is the amount of A and B that has reacted at time t, C and D are products of the reactions, and k2 is the second-order forward rate association constant [for a further description of the derivation of this equation, see (39)]. Adapting the above equation to the interaction of ‘251-ET-1 with its binding site, the following substitutions were made: a = the total “‘I-ET-1 binding site concentration (B-), this value being calculated from maximum binding obtained from the association plot; b = the total ligand concentration at time zero (L,,,,); x = the amount of ligand bound to the receptor at time f (BOUND); b - x = the amount of free ligand at time t (Lf,). Substituting these values into eq. (2) the following was obtained: { Wd&,,,
-
BOUJWII[B~,(~F,)I I/ Vkm - Lo,) = kd.
[31
If in eq. (3) we allow the left side to equal the arbitrary variable “RATIO,” we obtain: RATIO = k2t.
[41
Thus, a plot of RATIO vs. time should give a linear plot with the slope equivalent to the second-order association rate constant of irreversible binding. The preceeding equations and the unconstrained linear regression of the plot of RATIO vs. time to yield slope k2 were used to determine the second-order forward
ENDOTHELIN
association
BINDING
rate constant
791
IN BAT BRAIN
of ‘251-ET-l binding to brain mem-
branes with time. Materials ET-l (synthetic/porcine sequence) was obtained from Nova Biochem (Laufelfingen, Switzerland). ‘251-ET-1(specific activity: 63 Ci/mM) was prepared by labeling ET-l according to the method of Bolger et al. (7). All reagents for the enzyme marker studies were obtained from Sigma Chemical Co. (St. Louis, MO). All other reagents were obtained from standard commercial sources and were of the highest purity grade. Statistics For comparison of two groups of data, an unpaired Student’s t-test was used, accepting thep < 0.05 levels as significant. Analysis of variance (ANOVA) was used for multiple-group comparisons, accepting p < 0.05as the level of significance. Linear regression of the kinetic plots was accomplished using the linear regression statistical package contained in BMDP (BMDP Statistical Software, Los Angeles, CA). RESULTS
The ability of “‘I-ET-1 to bind to rat brain membranes was initially investigated using a washed homogenate of whole rat brain and employing assay conditions previously determined for ‘2SI-ET-l binding to non-CNS rat tissues (5). Specific Iz51ET-l binding to whole rat brain membranes was concentration dependent and saturable (Fig. 1) and was also found to be irreversible since addition of unlabeled ET-l to membranes prelabeled with ‘2SI-ET-1 for 60 min did not result in displacement of ‘2sI-ET-1 (data not shown). The nonspecific binding of 12’1ET-l was on average as a percentage of the total binding 2382% within the radioligand concentration range of 0.10-5.2 nM. The majority of nonspecific ‘251-ET-l binding (>95%) was to the filters (7). Thus, the amount of nonspecific binding appearing on the filter was not related to the amount of membrane protein used in the binding assay. ?-ET-l binding to brain membranes
0
2
4
CONCENTRATION
was rapid and displayed kinetic characteristics ofthe interaction of a ligand irreversibly with its binding site, concentrations of 0.44, 1.8 1, and 5.19 nM ‘251-ET-1 producing a similar magnitude of binding over time (Fig. 2). Utilizing kinetics that describe a second-order irreversible interaction between binding site and ligand (see the Methods section), the maximum binding site capacity and second-order association rate constant of binding were found to be 1,946 * 147 fmol/mg protein and 5.53 + 1.72 X lo6 M-’ s-‘, respectively (values are presented as the mean f. SEM of four determinations). The calcium dependence of “‘I-ET-1 binding to whole rat brain membranes was studied in two ways: 1) washing and subsequent exposure of brain membranes to 5 PM EDTA and 2) preincubation of brain membranes with 1.5 mM EDTA and subsequent washing with 5 PM EDTA. Both treatments affected neither the saturation plot (Fig. 3) nor the association rate (results not shown) of ‘251-ET-1 binding to brain membranes. Investigation of the distribution of ‘251-ET-l binding in rat brain demonstrated high levels of binding to membranes from the cerebellum and brain stem with lower and approximately similar levels in the cerebral cortex, hippocampus, and striatum (Fig. 4). Seven subcellular fractions of whole rat brain prepared according to the procedure of Whittaker (51) were assessed for enzyme marker activity (the plasmalemmal membrane marker enzyme 5’-nucleotidase, the mitochondrial membrane marker enzyme cytochrome-C oxidase, and the general membrane enzyme markers phosphodiesterase type I and alkaline phosphatase) (Table 1) and ‘251-ET-1binding (Figs. 5a and b and Tables 2 and 3). The following subcellular fractions were investigated: the homogenate (Ho), the cell debris and nuclei fractions (Pl), the microsomal fraction (P3), the ribosomal fraction (S3), the myelin fraction (MY), the synaptosomal fraction (SYN), and the mitochondrial (MITO) fraction. Of the three subcellular fractions (MY, SYN, and MITO) isolated from the discontinuous sucrose gradient, 5’-nucleotidase activity was enriched sixfold in the MY and lowest in the MIT0 (Table I), with substantial activity also being observed in the Pl (threefold enrichment compared to Ho) (Table 1). Cytochrome-C oxidase activity was
8
6
OF
10
12
‘“I-ET-1 (nM)
FlG. 1. Concentration dependence of ‘2JI-ET-l binding to a washed preparation of whole rat brain homogenate. Whole rat brain homogenate was incubated for 2 h with ‘*‘I-ET-I in buffer A over the concentration range illustrated. Under these conditions, the concentration of ‘2’I-ET-1 producing half maximal saturation was 0.1 f 0.05 nM (value presented k SEM of four determinations). The line through the data points is handdrawn.
792
BOLGER, BERRY AND JARAMILLO
l
0.44nM
A 1.81 nM .
5.19 nM
FIG. 2. Association of different concentrations of ‘2SI-ET-I with a washed preparation of whole rat brain homogenate. The time course for the binding of ‘251-ET-1binding to whole rat brain homogenate was investigated in buffer A using three different concentrations of “‘I-ET- 1: 0.44, I .8I, and 5.19 nM. The maximum binding obtained at each concentration of radioligand was: 0.44 nM. 108 + 10 fmol; 1.8 I nM. I2 1 k 12 fmol; 5.19 nM, I3 I + I8 fmol (these values were not significantly different, ANOVA accepting the p < 0.05 level of significance).The second-order association rate constants calculated from the association curves were: 5.83 X lo6 M-’ SK’at 0.44 nM, 8.41 X IO6Mm1s-’ at 1.81 nM, and 8.41 X IO6M-l s-l at 5.19 nM. The lines through the data points were handdrawn.
highest in the MIT0 (threefold enrichment compared to Ho) with little or no activity observed in the S3, P3, and MY (Table 1). The distribution of both alkaline phosphatase and phosphodiesterase activities among the subcellular fractions was similar, with the highest amount of enzyme activity observed in the PI. Among the MY, SYN, and MITO, alkaline phosphatase and phosphodiesterase type I activities were highest in the MY and
somewhat lower in the SYN and MIT0 (Table I), the enrichment of enzyme activity within these fractions compared to Ho ranging from two- to fourfold. The highest density of ‘251-ET-l binding sites (3,583 f 320 fmol/mg protein) was observed in the SYN, whereas no “‘I-ET1 binding could be detected in the S3 fraction (Table 2). The MY contained the second highest density of ‘*‘I-ET-l binding
2S00T
0
--_b_
0
__+
2
__~----+_
4
CONCENTRATION
,-
6
OF
?-ET-l
EDTA TREATED (5 uM)
l
EDTA TREATED (1.5 mM)
1
---___--+
6
*
10
12
(nM)
FIG. 3. Calcium dependence of ‘251-ET-I binding to whole rat brain homogenate. Calcium dependence was investigated either by washing brain membranes with buffer B or incubating brain membranes with 1.5 mM EDTA for 30 min prior to washing with buffer B (see the Method Section) and subsequently investigating the binding of ‘*‘I-ET- I in buffer B. The comparative control was “‘I-ET- I binding to brain membranes conducted in buffer A. The maximum binding obtained under each condition was similar (no significant differences noted, ANOVA accepting the p = 0.05 level of significance). The lines through the data points were handdrawn.
ENDOTHELIN
BINDING
793
IN BAT BRAIN
BOUND
(lmollmg protein)
+
+
7-
CORTEX
STRIATUM
+
+
HIPPOCAMPUS CEREBELLUM
BRAINSTEM
FIG. 4. Distribution of iz51-ET-Ibinding in rat brain. Binding of “‘I-ET- I to crude brain membrane preparations from the cerebral cortex, striatum, hippocampus, cerebellum, and brain stem was investigated in buffer A. The binding to membranes obtained from each brain region was (fmol/ mg protein): cerebral cortex, 850 + 67; striatum, 1,093 f 177; hippocampus, 1,143 f 83; cerebellum, 4,378 -C532; and brain stem, 3,104 + 356. The results are presented as the mean f SEM of four experiments.
sites (2,104 + 57) fmol/mg protein), with the P3 and Pl containing a similar density of “51-ET-l binding sites (1,225 -t 58 and 1,305 + 9 1 fmol/mg protein, respectively) (Table 2). We decided to study the binding properties of “‘I-ET- I in the P 1 fraction because it contained a relative enrichment of 5’-neucleotidase, alkaline phosphatase, and phosphodiesterase activities compared to the other subcellular fractions (Table 1). The MIT0 fraction contained the lowest density of “‘I-ET- 1 binding sites (834 f 90 fmol/mg protein) (Table 2). ‘251-ET-1 binding to the SYN fraction represented only a twofold enrichment of binding compared to Ho (Table 2). The subcellular distribution of ‘251ET-1 binding did not correlate with the distribution of any of the enzyme markers studied. As was the case for ‘251-ET-1 binding to whole rat brain membranes, association of 12’I-ET- I binding with subcellular
fractions from rat brain was very rapid. For example, in the SYN fraction (Fig. 5a) maxima1 binding at 25°C was attained in 10 min. In general, the second-order association rate constants (for illustration of their determination, see Figs. 5a and b) for “‘I-ET- 1 binding to the subcellular fractions were similar, with the exception that the association rate constant for ‘251-ET-1 binding to the MIT0 fraction was significantly different (Table 2). Two properties of “‘I-ET- 1 binding to subcellular fractions of rat brain were investigated: 1) the reversal of ‘251EET-1binding by exposing subcellular fractions previously incubated with lz51ET- I for 30 min to unlabeled ET- 1 (lo-’ M) for 60 min and 2) the effect of removal of extracellular calcium. The results are presented in Table 3. In none of the subcellular fractions was a significant reversal of ‘251-ET-l binding noted following 60-min exposure to a 1OO-fold excess of unlabeled ET- 1. With regard
TABLE I SUBCELLULAR
Subcellular Fraction
Ho Pl s3 P3 MY SYN MIT0
S-Nucleotidase (PM inorganic phosphate/pg protein)
0.000638 0.002190 0.000621 0.001433 0.003799 0.001702 0.000662
f O.OOOOlS f O.OOOl19 + 0.000016 + 0.000058 f 0.000201 l?r0.000046 + 0.000052
DISTRIBUTION
OF ENZYME
Cytochrome-C Oxidax (U cytochrome-C oxidase activity/g protein)
0.00578 f 0.000130 0.00478 + 0.000584 0 0.000092 f 0.000039 0.000460 + 0.000027 0.004332 f 0.000162 0.017760 f 0.003330
ACTIVITIES
IN RAT BRAIN
Alkaline Phosphatase (PM paranitrophenol/pg protein)
0.000509 ? 0.001219 f 0.000 163 f 0.00047 I + 0.000970 f 0.000855 + 0.000529 +
0.000028 0.000024 0.000044 0.000036 0.000024 0.000036 0.00006 I
PhosphodiesteraseType I (aM paranitrophenol/pg protein)
0.0000786 0.0003 15 0.00003 1 0.000139 0.000272 0.000143 0.000107
+ + * f f f +
0.0000022 0.000060 o.OOOOO34 0.000012 O.OoOO1O 0.000013 O.OOOOO8O
For 5’-nucleotidase, enzyme activity was measured by the liberation of inorganic phosphate from 5’-adenosine monophosphate. For cytochrome-C oxidase, enzyme activity is expressed as units of activity and was calculated from a standard curve employing purified cytochrome-C oxidase. For alkaline phosphatase and phosphodiesterase type I, enzyme activity was measured by the liberation of paranitrophenol from paranitrophenol phosphate and paranitrophenyl-5’-thymidylate, respectively. The results are presented as the mean +_SEM of four determinations.
794
BOLGER. BERRY AND JARAMILLO
t 400
i
600
a00
IMK)
+ 1200
1400
1600
law
TIME (seconds) 0.004
0.0025 0.0035 !
0
100
200
200
400
SW
600
700
800
FIG. 5. Association of ‘251-ET-1binding and corresponding plot for determination of the second-order rate association constant in the synaptosomal fraction of rat brain. (a) The time-dependent association of ‘Z51-ET-Iwith its binding site on synaptosomal membranes was investigated using a radioligand concentration of 1.95 nM. (b) Plot of the RATIO (for definition of the RATIO, see the Method Section) vs. time. Unconstrained linear regression of this plot yielded a slope equivalent to the second-order forward association rate constant of “‘I-ET-1 binding (see Table I). The data used for this plot were taken from Fig. 5a. The correlation coefficient of linear regression for this plot was >0.98. The lines through the points were handdrawn.
to the calcium-dependent component of binding, a significant effect of calcium was only noted for the SYN fraction, in which the omission of calcium resulted in a 19% increase in ‘251-ET-1 binding (Table 3). DISCUSHON
This study clearly demonstrates the presence of a substantial density of high-affinity binding sites for 1251-ET-1in rat brain and is consistent with previous studies (1,17,26). The density of ‘251-ET-l binding sites observed in whole rat brain membranes is similar to that observed in rat brain using radioautography (40). Two lines of evidence indicated that ‘*‘I-ET-l interacted with its binding site irreversibly. First, unlabeled ET- 1 was unable to displace ‘*-?-ET-l from its binding site. Second, the kinetics of association of different concentrations of ‘*‘I-ET-I with its binding site produced, with time, similar maximum binding
characteristic of the irreversible interaction of a ligand with its binding site (Fig. 2). These observations are consistent with several studies indicating ET- 1 interacts with its binding site either in an irreversible or slowly reversible manner both in brain and peripheral tissues (7,12,14,23,36,50). In fact, it has been demonstrated that, dependent on the system investigated, only exposure to extremes of pH (35), T&on-X 100 (12), or the calcium chelator EDTA (5) substantially dissociated ET- 1 from its binding site within a reasonable time () OF ‘2s1-ET-l BINDING TO SUBCELLULAR FRACTIONS OF WHOLE RAT BRAIN
AssociationRate constantt (M-‘s-l)
Maximum Binding Site Capacity* (fmol/mg protein)
Subcell&r Fraction Homogenate MY SYN MIT0 P3 s3 Pl
1,873 + 2,104 f 3,583 + 834 + 1,225 + 0 1,305 +
134 57 320 90 58
4.32 3.23 5.32 9.59 6.72
+ 1.72 X f 0.39 X + 1.94 x & 0.76 X f 0.59 X 2.62 + 0.18 x
91
IO6 lo6 lo6 106# lo6 lo6
* Values are presented as the mean k SEM of four experiments and were determined as the maximum binding obtained from saturation plots of ‘*rI-ET-l binding in buffer A (see Fig. 1). t Values were determined from linear regression analysis of the type of kinetic plot illustrated in Fig. 5b and are presented as the mean + SEM of three determinations. $ Significantly different from all other values presented, p < 0.05 ANOVA.
irreversible properties of ‘%ET- 1 binding in rat brain prompted us to use an irreversible ligand binding analysis to describe the kinetic properties of binding. The second-order association rate constant calculated for ‘2SI-ET-l binding to whole brain membranes (5.53 X lo6 M-’ s-‘) is in excellent agreement with the value calculated for irreversible ‘2SI-ET-l binding to rat cere-
bellum (8.3 X lo6 M-’ s-‘) by Waggoner et al. (50) but did not agree with the association rate constant of 8.0 X lo5 M-l SC’ calculated assuming that ‘251-ET-1 binding is reversible (23). The removal of either extracellular calcium or membranebound calcium and certain calcium binding proteins did not affect ‘251-ET-l binding to a homogenate of brain membranes. A similar observation was made by Ambar et al. (1) in rat brain membranes where an independence from calcium was noted both for “‘I-ET- 1 binding and ET- l-mediated phosphoinositide hydrolysis. Furthermore, Kohzuki et al. (28) also noted that 12’1ET-l binding to brain slices was insensitive to the chelation of
it was observed that dissociation of ‘251-ET-l from its binding site on rat glomerular mesangial cells could be precipitated by the addition of excess unlabeled ET- 1 in the presence of EDTA. The lack of effect of calcium depravation on ‘251-ET-1 binding to brain membranes as compared to rat mesangial cells may be due to either tissue differences or the obvious difference, a broken cell vs. intact cell preparation. However, recent studies suggest that the properties of the endothelin binding sites that recognize ET- 1 may differ in a tissue-dependent manner (48) making tissue differences the more likely explanation. The brain stem (mostly white matter) and the cerebellum contained the highest densities of ET-l binding sites. This observation is consistent with radioautographic and membrane studies of “‘I-ET-1 binding in rat brain in which rat forebrain was found to contain considerably less binding sites for ET-l than the cerebellum and brain stem (6,17,30). A similar distribution of ET- 1 binding sites was also observed in human brain (26). The correlation between membrane binding and radioautographic studies in rat brain suggest that following either direct injection into the brain or incubation with brain slices ET-l binds directly to its high-affinity site rather than being sequestered nonspecifically into the neuron. Previous studies have demonstrated that the autoradiographic distribution of ET-l binding sites neither correlates with binding sites for such neurally active substances as w-conotoxin, vasopressin, endorphin, neurotensin, calcitonin, and receptors for corticotrophin-releasing factor, dopamine, thyrotropin-releasing hormone, substance-P, and insulin [(29) and references cited therein] nor with the distribution of cerebral blood vessels (28,29). The wide distribution of neuronally associated ET- 1 binding sites (28) and immunoreactive ET1 (23) suggest a critically important role for ET-l in the CNS. Several lines of evidence from the study of the subcellular distribution of ‘2SI-ET-l binding sites in rat brain suggest that ET-l binding sites are probably associated with the neuronal plasmalemmal membrane (cell soma, axonal, and synaptosomal) and certain membranes from intracellular organelles. First, while the highest density of ET- 1 binding sites was found in the SYN
TABLE 3 REVERSIBILITY AND CALCIUM DEPENDENCE OF ‘*51-ET-I BINDING TO SUBCELLULAR FRACTIONS OF RAT BRAIN ‘2’1-ET-I Bound (fmol/mg protein)
Dissociation*
CalciumDcpendencet
FollowingET-I Subcellular Fraction
MY SYN MIT0 P3 PI
Control
2,435 + 3,137 f 793 + 1,225 + 1,234 +
125 222 25 58 79
lo-‘M for 60 min
2,398 + 3,146 + 816 + 1,218 + 1,213 +
242 240 32 58 61
In the Presence of Ca+*
2,589 ? 3,583 + 905 + ND 1,126 f
418 320 99 250
In the Absence of Ca++
2,229 f 252 4,437 f 348* 886 f 120 ND 1,034 + 180
Values are presented as the mean * SEM of four subcellular membrane fraction preparations. ND, not determined. * Membranes were incubated with “‘I-ET-1 (2.0 nM) for 180 min in the control case. For dissociation, membranes were initially incubated with lZ51-ET-1 for 120 min followed by 60-min exposure to lo-’ M ET- 1. 7 The calcium dependence of “‘I-ET- 1 binding to subcellular membrane fractions was investigated in sucrose-resuspended membrane fractions washed once and resuspended in a buffer B (see the Method section). The concentration of ‘*‘I-ET-l in the binding assay was 8.5 nM.
796
BOLGER, BERRY AND JARAMILLO
fraction it was only twofold compared to the homogenate, whereas other subcellular fractions had lower or comparable densities of ‘251-ET-l binding sites to that in the homogenate. This observation was made despite the fact that S-neucleotidase and cytochrome-C oxidase displayed substantial enrichments in certain subcellular fractions. Second, the subcellular distribution of ‘251-ET-1 binding sites did not correlate with the distribution of, most notably, S-neucleotidase and cytochrome-C oxidase and phosphodiesterase type I and alkaline phosphatase. These observations are consistent with a general distribution of ‘251ET- 1 binding sites throughout neuronal membrane subtypes rather than an enrichment of binding in, for example, either the plasmalemmal or mitochondrial membrane. Thus, ‘251-ET-l binding sites are likely associated with both the plasmalemmal membrane of the neuron as well as membranes of various intraneuronal organelles, with higher densities associated with synaptic membranes. The protocol employed for the subcellular distribution, that of Whitakker (5 I), fractionates the brain homogenate into different components largely based on their morphological characteristics rather than their specific membrane properties (i.e., separation of synaptosomes and mitochondria). Thus, to more clearly define the subcellular localization of ‘251-ET-1 binding sites a more suitable subcellular fractionation technique might have to be employed. However, it was of interest to note that the second-order association rate constant of ‘251-ET-1 binding to the mitochondrial fraction was significantly different from other subcellular fractions and that synaptosomal ET-l binding sites were sensitive to removal of extramembranal calcium. These observations strengthen the argument for an intracellular localization of ‘251-ET-l binding sites but in addition, suggest that the properties of ‘251-ET-1 binding sites may also be dependent on their neuronal localization. For instance, the unique calcium dependence of ‘251-ET-1 binding to the synaptosomal fraction may reflect either a perturbation of the ‘251-ET-l binding site by the binding of calcium to the synaptic
membrane or the regulation of binding by calcium binding proteins, synaptic membranes being particulary rich in calcium binding proteins (42). If the ‘251-ET-l binding site should be associated with synaptic membrane calcium binding proteins, then ET-l may play a critical role in modulating synaptic activity. The behavioral (32) and biochemical effects of ET- 1 (1,13,37,47) clearly support such speculation. Although the exact subcellular localization of “‘I-ET- 1 binding sites in the CNS remains equivocal, it is clear that their cellular localization favors both neurons and astrocytes. Both cell culture and lesion studies with ibotenic acid confirmed the cellular localization of ‘*‘I-ET- 1 binding sites (18,34,37,52). Detailed studies of ‘251-ET-1 binding sites in rat brain revealed heterogeneity, ‘*%ET-1 binding to two sites with molecular masses of -38,000 and -50,000 Da (2,25,43,48). Whether these reflect ET- 1 binding sites that either have a different subcellular localization or whose properties depend on their brain region localization remains to be determined. The function of endothelin binding sites in the CNS is unknown. One possibility is that they regulate neuronal ion channel activity directly or through proposed allosteric regulatory sites such as peripheral benzodiazepine receptors and psychotomimetric drug binding sites (38) and channel phosphorylation sites (45). Endothelin has been shown to initiate calcium currents in cultured neuronal and glial cell lines (15,22,47,52). However, preliminary studies
in our laboratory did not clearly indicate that ET-l could evoke calcium influx or alter depolarization-induced calcium influx in a preparation of brain neurons (A. Jodoin and G. T. Bolger, unpub lished observations). More recent evidence suggests that endothelin can open potassium channels in glial cells (46). These observations, coupled with the possibility that intracellular binding sites exist for ET- I, implies that ET- 1 may have as yet undetermined far-teaching neuroregulatory actions. Clearly, further studies are neces%uy to elucidate the mechanism(s) of action of ET-l in the CNS.
REFERENCES 1. Ambar, I.; Kloog, Y.; Schvartz, I.; Hazum, E.; Sokolovsky, M. Competitive interaction between endothelin and sarafotoxin: Binding and phosphoinositides hydrolysis in rat atria and brain. Biochem. Biophys. Res. Comm. 158:195-201; 1989. 2. Ambar, I.; Kloog, A.; Sokolovsky, M. Cross-linking of endothelin I and endothelin 3 to rat brain membranes: Identification of the putative receptor(s). Biochemistry 29:6415- 1418; 1990. 3. Anholt, R. R. H.; Murphy, K. M. M.; Mack, G. E.; Snyder. S. H. Peripheral-type benzodiazepine receptors in the central nervous system: Localization to olfactory nerves. J. Neurosci. 4:593-603; 1984. 4. Badar, M. F.; Hikata, T.; Trifario, J. M. Calcium dependent calmodulin binding to chromaffin granule membranes: Presence of a 65-kilodalton calmodulin-binding protein. J. Neurochem. 44:526539; 1985. 5. Baldi, E.; Dunn, M. J. Endothelin binding and receptor down regulation in rat glomerular mesangial cells. J. Pharmacol. Exp. Ther. 256:581-586; 1991. 6. Banasik, J. L.; Hosick, H.; Wright, J. W.; Harding, J. W. Endothelin binding in brain of normotensive and spontaneously hypertensive rats. J. Pharmacol. Exp. Ther. 257:302-306; 1991. 7. Bolger, G. T.; Liard, R.: Krogsrud, R.; Thibeault, D.; Jaramillo, J. Tissue specificity of endothelin binding sites. J. Cardiovasc. Pharmacol. 16:367-375; 1990. 8. Bolter. G. T.: Raffertv, M. F.; Skolnick, P. Phencyclidine increases the-affinity of dihydropyridine calcium antagonist binding in rat brain. Naunvn-Schmied. Arch. Pharmacol. 330:227-234: 1985. 9. Bradford, M: M. A rapid and sensitive method for quantitation of microgram quantities of protein utilizing the principle of proteindye binding. Anal. Biochem. 72:248-254; 1976.
IO. Calve. J. J.: Gonzalez, R.: De Carvalho, L. F.; Takahashi, K.; Kanse,
11.
12. 13.
14. 15.
16. 17. 18.
S. M.; Hart, G. R.; Ghatei, M. A.; Bloom, S. R. Release of substance P from rat hypothalamus and pituitary by endothelin. Endocrinology 126:2288-2295; 1990. Cooperstein, S. J.; Lazarow, R. A microspectrophotometric method for the determination of cytochrome oxidase. J. Biol. Chem. 189: 665-670; 1951. Courad, P.-O.; Durieu-Trautmann, 0.; Le Ngukyen, D.; Marin, P.; Gilbert, F.; Strosberg, A. D. Functional endothelin-1 receptors in rat astrocytoma C6. Eur. J. Pharmacol. 206:191-198; 1991. Crawford, M. L. A.: Hiley, C. R.; Young, J. M. Characteristics of endothelin- 1 and endothelin-3 stimulation of phosphoinositide breakdown differ between regions of guinea-pig and rat brain. Naunyn-Schmied. Arch. Pharrnacol. 341:268-271; 1990. Fischlie, W.; Clozel, M.; Guilly, C. Specific receptors for endothelin on membranes from human placenta. Characterization and use in a binding assay. Life Sci. 44:1429-1436; 1989. Fu. T.: Chang, W.; Ishida, N.; Saida, K.: Mitsui, Y.; Okano, Y.: Nozawa, Y. Effects of vasoactive intestinal contractor (VIC) and endothelin on intracellular calcium level in neuroblastoma NG IOS15 cells. FEB Lett. 257:35 l-353: 1989. Fuxe, K.; Andbier, B.; Kalia, M.; Agnati, L. F. Centrallv administered endothelin- I produces apnoea in the cu-chloralose-anaesthetized male rat. Acta Phvsiol. Stand. 137:157-158: 1989. Fuxe, K.; Anggard, E.; Lundgren, K.; Cintra, A.; Agnati, L. F.; Galton. S.; Vane, J. Localization of [‘zsI]endothelin-3 binding sites in the rat brain. Acta Physiol. Stand. 137:563-564; 1989. Fuxe, K.; Cintm, A.: Andbjer, B.; Anggard, E.; Goldstein, M.; Agnati, L. F. Centrally administered endothelin-I produces lesions in the brain of the male rat. Acta Physiol. Stand. 137:155- 156; 1989.
ENDOTHELIN
BINDING IN RAT BRAIN
19. Garen, A.; Levinthal, C. A fine-structure genetic and chemical study of the enzyme alkaline phosphatase. Biochim. Biophys. Acta 38: 470-476; 1960. 20. Giaid, A.; Gibson, S. J.; Herrero, M. T.; Gentleman, S.; Legon, S.; Yanagisawa, M.; Masaki, T.; Ibrahim, N. B. N.; Roberts, G. W.; Rossi, M. L.; Polak, J. M. Topographical localisation of endothelin mRNA and peptide immunoreactivity in neurones of the human brain. Histochemistry 95:303-3 14; 199 1. 21 Giaid, A.; Gibson, S. J.; Ibrahim, N. B. N.; Legon, S.; Bloom, S. R.; Yanagisawa, M.; Masaki, T.; Vamdell, 1. M.; Polak, J. M. Endothelin 1, an endotheliumderived peptide, is expressed in neurons of the human spinal cord and dorsal root ganglia. Proc. Natl. Acad. Sci. USA 86:7634-7638; 1989. 22. Goldman, R. S.; Finkbeiner, S. M.; Smith, S. J. Endothelin induces a sustained rise in intracellular calcium in hippocampal astrocytes. Neurosci. Lett. 123:4-8; 199 1. 23. Hiley, C. R.; Jones, C. R.; Pelton, J. T.; Miller, R. C. Binding of ‘251-endothelin-l to rat cerebellar homogenates and its interactions with some analogues. Br. J. Pharmacol. 101:319-324; 1990. 24. Hokfelt, T.; Post, C.; Freedman, J.; Lundberg, J. M.; Terenius, L. Endothelin induces spinal lesions after intrathecal administration. Acta Physiol. Stand. 137555-556; 1989. 25. Jones, C. R.; Hiley, C. R.; Pelton, J. T.; Miller, R. C. Endothelin receptor heterogeneity: Structure activity, autoradiographic and functional studies. J. Receptor Res. 11:299-310; 199 1. 26. Jones, C. R.; Hiley, C. R.; Pelton, J. T.; Mohr, M. Autoradiographic visualization of the binding sites for [‘251]endothelinin rat and human brain. Neurosci. Lett. 97:276-279; 1989. 27. Kataoka, Y.; Koizumi, S.; Kumakura, K.; Kurihara, M.; Niwa, M.; Ueki, S. Endothelin-triggered damage under hypoglycemia evidenced by real-time monitoring of dopamine release from rat striatal slices. Neurosci. Lett. 107:75-80; 1989. 28. Kohzuki, M.; Chai, S. Y.; Paxinos, G.; Karavas, A.; Caseley, D. J.; Johnston, C. I.; Mendelsohn, F. A. 0. Localization and characterization of endothelin receptor binding sites in the rat brain visualized by in vitro autoradiography. Neuroscience I :245-260; 199 1. 29. Koseki, C.; Imai, M.; Hiarata, Y.; Yanagisawa, M.; Masaki, T. Autoradiographic distribution in rat tissues of binding sites for endothelin: A neuropeptide? Am. J. Physiol. 256:R858-R866; 1989. 30. Koseki, C.; Imai, M.; Hiarata, Y.; Yanagisawa, M.; Masaki, T. Autoradiographic localization of [‘2SI]endothelin binding sites in rat brain. Neurosci. Lett. 106581-585; 1989. 31. Lee, M.-E.; de la Monte, S. M.; Ng, S.-C.; Bloch, K. D.; Quertermous, T. Expression of the potent vasoconstrictor endothelin in the human central nervous system. J. Clin. Invest. 86:141-147; 1990. 32. Lecci, A.; Maggi, C.-A.; Rovero, P.; Giachetti, A.; Meli, A. Effect of endothelin- 1, endothelin-3 and C-terminal hexapeptide, endothelin (16-2 1) on motor activation in rats. Neuropeptides 16:2 l-24; 1990. 33. Lin, W. W.; Lee, C. Y.; Chuang, D.-W. Endothelin-1 stimulates the release of preloaded [3H]Daspartate from cultured cerebellar granule cells. Biochem. Biophys. Res. Comm. 167:593-599; 1990. 34. Lin, W. W.; Lee, S. Y.; Chuang, D.-M. Comparative studies ofphosphoinositide hydrolysis induced by endothelin-related peptides in cultured cerebellar astrocytes, C&ioma and cerebellar granule cells. Biochem. Biophys. Res. Comm. 168512-519; 1990.
797
35. Liu, J.; Kelly, J.; Casley, D. J.; Johnston, C. I.; Nayler, W. G. Solubilization and characterization of endothelin- I receptors in rat cardiac tissue. Biochem. Biophys. Res. Comm. 166:299-307; 1990. 36. Lysko, P. G.; Feuerstein, G.; Pullen, M.; Wu, H.-L.; Nambi, P. Identification of endothelin receptors in cultured cerebellar neurons. Neuropeptides 18:83-86; 199 1. 37. MacCumber, M. W.; Ross, C. A.; Snyder, S. H. Endothelin in brain: Receptors, mitogenesis, and biosynthesis in glial cells. Proc. Natl. Acad. Sci. USA 87:2359-2363; 1990. 38. Matsumoto, H.; Suzuki, N.; Onda, H.; Fujino, M. Abundance of endothelin-3 in rat intestine. pituitarv eland and brain. B&hem. Biophys. Res. Comm. 164:74-80; 1989: 39. Moore, W. J. Physical chemistry (fourth ed.). Englewood Cliffs, NJ: Prentice Hall; 1972. 40. Nambi, P.; Pullen, M.; Feuerstein, G. Identification of endothelin receptors in various regions of rat brain. Neuropeptides 16:195- 199; 1990. 41. Sahyoun, N.; Le Vine, H., III; Bronson, D.; Cuatrecasas, P. Ca”calmodulin-dependent protein kinase in neuronal nuclei. J. Biol. Chem. 259:9341-9344: 1984. 42. Sahyoun, N.; Le Vine,‘H., III; MacDonald, B. 0.; Cuatrecasas, P. Specific postsynaptic density proteins bind tubulin and calmodulindependent protein kinase type II. J. Biol. Chem. 261:12339-12344; 1986. 43. Schvartz, I.; Ittoop, 0.; Hazum, E. Identification of endothelin receptors by chemical cross-linking. Endocrinology 126:I829- 1833; 1990. 44. Shinmi, 0.; Kimura, S.; Yoshizawa, T.; Sawamura, T.; Uchiyama, Y.; Sugita, Y.; Kanazawa, I.; Yanigisawa, M.; Goto, K.; Masaki, T. Presence of endothelin-1 in porcine spinal cord: Isolation and sequence determination. Biochem. Biophys. Res. Comm. 162:340346; 1989. 45. Strong, J. A.; Fox, A. P.; Tsien, R. W.; Kaczmarek, L. K. Stimulation of protein kinase C recruits covert calcium channels in aplasia bag cell neurons. Nature 235:7 14-7 17; 1987. 46. Supattapone, S.; Ashley, C. C. Endothelin opens potassium channels in glial cells. Eur. J. Neurosci. 3:349-355; 1991. 47. Supattapone, S.; Simpson, A. W. M.; Ashley, C. C. Free calcium rise and mitogenesis in glial cells caused by endothelin. Biochem. Biophys. Res.Comm. 165:1115-l 122; 1989. 48. Takayanagi, R.; Ohnaka, K.; Takasaki, C.; Masao, 0.; Hawata, H. Multiple subtypes of endothelin receptors in porcine tissues: Characterization by ligand binding, affinity labelling and regional distribution. Regul. Peptides 32:23-37; 199 1. 49. Touster, 0.; Aronson, N. N.; Dulaney, J. T.; Hendrickson, H. Isolation of rat liver plasma membranes. J. Cell Biol. 47:604-618; 1970. 50. Waaaoner, W. G.: Brent. S. L.: Rash. V. A. Scatchard analysis of endzhelin (ET) binding data in invalid. FASEB J. Abstr. 4983(5a): 1229; 1991. 51. Whittaker, V. P. The application of subcellular fractionation techniques to the study of brain function. Prog. Biophys. Mol. Biol. 15: 41-60; 1965. 52. Zhang, W.; Sakai, N.; Yamada, H.; Fu, Y.; Nozawa, Y. EndothelinI induces intracellular calcium rise and inositol 1,4,5,-triphosphate formation in cultured rat and human ghoma cells. Neurosci. Lett. 112:199-204; 1990.