Serotonin binding proteins in bovine retina: Binding of serotonin and catecholamines

Neurodwm. Int. Vol. 22, No. 2, pp. I 11 119, 1993

0197-0186/93 $6.00+0.00 Copyrighl ,c 1993Pergamon Press Lid

Primed in Great Britain. All rights reserved

S E R O T O N I N B I N D I N G P R O T E I N S IN BOVINE R E T I N A : B I N D I N G OF S E R O T O N I N A N D C A T E C H O L A M I N E S MARLENE JIMENEZ DEL RIO, ~* JEF PINXTEREN, 2 WERNER DE POTTER, 2 GUY EBINGER 3 a n d GEORGES VAUQUELIN t ~Department of Protein Chemistry, Institute of Molecular Biology, Free University Brussels, Paardenstraat 65, 1640 St. Genesius-Rode, Belgium 2Department of Neuropharmacology, Universitaire lnstelling Antwerpen, Universiteitsplein I, 2610 Wilrijk, Belgium ~Department of Neurology, University Hospital, Free University Brussels, Laarbeeklane 101, 1090 Brussels, Belgium ( Receit:ed 5 June 1992 ; accepted 9 July 1992)

Abstract--Serotonin binding proteins (SBP) are present in the soluble fraction of bovine retina homogenates. These proteins can be precipitated with 30% ammonium sulphate and their binding and physicochemical characteristics are very similar to those of SBP in bovine and rat brain. Binding of [~H]serotonin to bovine retina SBP requires Fe ~'+ but not Fe 3 ~. In the presence of an optimal concentration of Fe 2~ (0.1 mM), these proteins behave as a single class of non-cooperative sites for [~H]serotonin (Bm~,x= 242 + 10 pmol/mg protein, K D = 0.22 ± 0.44/~M). Competition binding studies reveal that serotonin analogs possessing an hydroxyl group on the mdole ring and catecholamine analogs possessing an intact catechol moiety are potent competitors (K, from 0.12 to 0.3/LM). In both cases, the afifnity is strongly decreased if aromatic hydroxyl groups are methoxylated. Catecholamine SBP interactions can also be demonstrated directly by binding experiments with [3H]dopamine. Binding of this catecholamine is greatly enhanced by Fe z ~, to a lesser extent by Cu =~ and Mn 2+, but not by Fe 3+. The Fe2+-dependent binding component is saturable (Bm~,~= 505+30 pmol/mg protein, KD -- 0.34_+0.04/~M). The SBP from bovine retina show the same physicochemical properties as SBP from bovine and rat brain: they elute immediately after the void volume on a Sephacryl S100 HR (1.6 x 140 cm) gel filtration column (reflecting aggregation) and they migrate with apparent molecular weights of respectively 43 kDa and 57 kDa on native polyacrylamide gel electrophoresis. The serotonin-storing role of SBP in serotonergic neurones has already been well documented. Since the bovine retina is rich in dopamine and almost devoid of serotonin, the present data make us speculate that SBP could also be involved in the housekeeping of dopamine in retinal neurons.

C a t e c h o l a m i n e - a n d serotonin-secreting cells have developed specific m e c h a n i s m s to reduce free conc e n t r a t i o n o f m o n o a m i n e and, therefore, also the osmotic pressure within their storage vesicles (Tamir a n d G e r s h o n , 1979 ; J o n a k a i t et al., 1979 ; Winkler et al., 1986). These m e c h a n i s m s are widely accepted to involve the f o r m a t i o n of m a c r o m o l e c u l a r complexes between catecholamines a n d A T P (Tamir et al., 1980 : Winkler et El/., 1981) a n d between serotonin a n d specific proteins ( G e r s h o n and Tamir, 1984). Such proteins with high affinity for serotonin, denoted as "serotonin binding p r o t e i n s " (SBP), were first found to be present in rat brain ( T a m i r a n d Huang, 1974) a n d

subsequently also in serotonin-secreting cells from neuroectodermal origin in the periphery: i.e. enteric serotonergic n e u r o n s ( G e r s h o n and Tamir, 1981: T a m i r a n d G e r s h o n , 1981) and parafollicular cells of the thyroid gland ( T a m i r a n d G e r s h o n , 1981 ; Barasch et al., 1987). Serotonin binding proteins from different species (rat, sheep, bovine a n d h u m a n ) comprise two comp o n e n t s with molecular weights close to 45 and 56 k D a (Liu et al., 1985; Barasch et al., 1987: T a m i r et al., 1989; Jimenez Del Rio et al., 1992). Both comp o n e n t s possess c o m p a r a b l e binding characteristics for serotonin (Liu et al., 1985), cross react with polyclonal as well as with m o n o c l o n a l antibodies (Liu et al., 1990a) and possess similar binding d o m a i n s (Liu et al., 1990b). This close structural similarity supports the proposal that the 56 k D a c o m p o n e n t might be the

*Author to whom all correspondence should be addressed. Abbre~,iations : PAGE, polyacrylamide gel electrophoresis :

SBP: serotonin binding proteins. 111

112

MARLENE JIMENEZ DEL RIO et al.

precursor o f the 45 k D a protein ( G e r s h o n et al., 1983 ; G e r s h o n a n d Tamir, 1984). Based on the o b s e r v a t i o n t h a t the binding o f serotonin to SBP is strictly d e p e n d e n t o n the presence o f Fe 2+, but n o t o n Fe 3+ ions, T a m i r a n d Liu (1982) presented the following two-step model to describe the s e r o t o n i n - S B P interaction. In this model, Fe 2+ would first b i n d to sulfhydryl groups o n SBP. Then, the t r a p p e d iron could a t t a c h up to four serotonin molecules via c o o r d i n a t i o n bonds. T h e i n v o l v e m e n t of such c o o r d i n a t i o n b o n d s explains why the binding of serotonin analogues to SBP requires the presence of at least one hydroxyl g r o u p o n their indole ring ( T a m i r a n d R a p p o r t , 1978; Jimenez Del Rio et al., 1992). In the past, little a t t e n t i o n has been paid to the possibility t h a t SBP m i g h t also be capable of interacting with n e u r o t r a n s m i t t e r s o t h e r t h a n serotonin. In a recent study, we have d e m o n s t r a t e d that this is indeed the case for SBP from bovine frontal cortex (Jimenez Del Rio et al., 1992). Catecholamine/[3H]serotonin c o m p e t i t i o n binding experiments revealed t h a t these SBP possess a b o u t the same affinity for catecholamines a n d for serotonin. Direct [3H]dopamine saturation binding experiments further indicated t h a t this c a t e c h o l a m i n e - S B P interaction is also d e p e n d e n t on the presence o f Fe 2+, but not o n Fe 3+ ions. The affinity of catecholamines a n d analogs for SBP was completely unrelated to the structure of, a n d even to the presence o f a n e t h a n o l a m i n e side chain. Instead, the structure-affinity relationship studies indicated that c a t e c h o l a m i n e analogs only possess high affinity for bovine SBP if they contain a n intact catechol moiety. Based on these criteria, we postulated that the c a t e c h o l a m i n e - S B P interaction implies the same twostep m e c h a n i s m as for serotonin ; i.e. binding of Fe 2+ to SBP, followed by c o o r d i n a t i o n b o n d f o r m a t i o n with the hydroxyl groups of the catechol ring. Since bovine SBP is capable at least o f binding catecholamines, as well as serotonin, we evoked the possibility t h a t SBP might also possess some catec h o l a m i n e - s t o r i n g function (Jimenez Del Rio et al., 1992). The m a m m a l i a n retina constitutes an interesting model to evaluate this possibility, since the a m o u n t o f e n d o g e n o u s serotonin is extremely low (Osborne, 1980b; O s b o r n e et al., 1982) while dopamine is a m a j o r n e u r o t r a n s m i t t e r in this tissue ( K a m p , 1985). In this study, we have detected the presence of SBP in soluble extracts from bovine retina. Their binding properties a n d physicochemical characteristics are similar to those of SBP from bovine frontal cortex. Their c o n c e n t r a t i o n in a m m o n i u m sulphate-precipitated extracts from bovine retina is a b o u t 2 times higher t h a n in c o r r e s p o n d i n g extracts

from the frontal cortex, a n d these findings indicate t h a t there is no positive correlation between the SBP a n d the serotonin content o f a tissue.

EXPERIMENTAL PROCEDURES

Materials 3-Hydroxy [G-3H]tryptamine creatinine sulphate ([3H[ serotonin, 8 Ci/mmol), and [7,8 3H]dopamine (49 Ci/ mmol) were obtained from Amersham (U.K.). (-)-Adrenaline (+)-bitartrate, dopamine hydrochloride, homovanillic acid, dihydroxyphenylacetic acid (DOPAC), 5-hydroxytryptamine hydrochloride (serotonin), 5-hydroxyindoleacetic acid (5-HIAA), noradrenaline bitartrate, 3-methoxytyramine, and melatonin were obtained from Sigma (U.S.A.). L-3,4-dihydroxyphenylalanine (L-DOPA) and 5-hydroxytryptophan were from Janssen Chimica (Belgium). Protein preparation Bovine eyes were collected in a local slaughterhouse. The retinae were rapidly dissected. All subsequent steps were performed at 0 4cC. The retinae were homogenized with an Ultraturax for 30 s in 0.32 M sucrose/10 mM potassium phosphate buffer, pH 7.5 (8 ml/per g of tissue). This suspension was further homogenized with a motor driven Potter Elvehjem homogenizer (10 strokes, maximum speed). The homogenate was centrifuged at 10,000 g for 30 min. All supernatants were pooled and centrifuged at 40,000 g for 60 min. The resulting supernatant was treated with ammonium sulphate (30% saturation) for 20 min and centrifuged at 15,000 g for 15 min. The pellets were resuspended in 20 mM potassium phosphate buffer, pH 7.5 (phosphate buffer) containing 10% glycerol and then dialyzed against an excess of the same buffer and stored frozen at - 2 0 ' C . Protein concentrations were determined according to Lowry et al. (1951), bovine serum albumin was used as the standard. Binding assay Samples of protein (0.1 mg/ml) were incubated with 0.1 mM FeSO4 at 2OC for 3 min in 400 id phosphate buffer. Then, 50 /d of [3H]serotonin or [3H]dopamine (final concentration: 10 nM 0.5/tM for saturation experiments and 0.2/~M for competition experiments) and 50 #1 of phosphate buffer (alone or containing competitors) were added and the incubation was continued for 15 rain. Then 300 ~1 of the mixture was applied to a small Sephadex G-50 column (0.7 x 15 cm) equilibrated with phosphate buffer, and eluted at 4 C with the same buffer. The void volume (1.5 ml) was discarded and the fraction containing labeled protein (1.8 ml) was collected and counted by liquid scintillation counting. The data are means of three experiments. Nonspecific binding was obtained in absence of Fe 2+. Binding isotherms were analyzed by nonlinear least square curve fitting with the program "LIGAND" (Munson and Rodbard, 1980) to determine maximum number of binding sites (Bm~0 and the equilibrium dissociation constants (KD for saturation binding and K~for competition binding). Control experiments revealed that binding of 0.2 #M [3H]serotonin and [3H]dopamine reached maximum after already 5 min incubation and that specific binding of both radioligands was decreased by less than 5% when the separations were done at room temperature instead of 4'C.

113

Serotonin and catecholamine binding proteins

Gel filtration and SDS- PA GE of SDP Samples of protein (2 mg/ml) were incubated with 0.1 mM FeSO4, then with 0.2/~M [3H]serotonin or [3H]dopamine, and eluted on a small Sephadex G-50 column (0.7 x 15 cm) as for the binding assay. The fractions containing bound radioligand were applied directly to a Sephacryl SI00 HR (1.6 x 140 cm) gel filtration column, and eluted with phosphate buffer. Fractions (2.8 ml) were collected and assayed for protein and radioactivity. Bound radioactivity eluted in one peak just after the void volume. The apparent molecular weight of the proteins present in the peak was assayed by sodium dodecyl sulphate polyacrylamide electrophoresis (SDS PAGE) on 12.5% gels, according to the method of Laemmli (1970). Molecular weight markers were: phosphorylase b, 94 kDa: bovine serum albumin, 67 kDa; ovalbumin, 43 kDa; carbonic anhydrase, 30 kDa; soybean trypsin inhibitor, 20.1 kDa and :t-lactalbumin, 14.4 kDa. Native PAGE of SBP Homogenates from retinae were centrifuged as above, but the subsequent ammonium sulphate precipitation step was replaced by dialysis of the supernatant against phosphate buffer (to reduce aggregation of SBP). Freshly dialyzed proteins were then preincubated with FeSO4, incubated with [3H]serotonin or [3H]dopamine and eluted on a small Sephadex G-50 column as above. Native PAGE Gels (7.5%) were run and the same molecular weight standards were used as for SDS PAGE. Gels were then sliced and the radioactivity present in the slices (1.6 mm) was counted.

~"

o

200 o

E

D. 1 5 0

,oo

S e r o t o n i n binding proteins are present in crude, soluble fractions prepared from bovine retina. Precipitation o f these proteins in 30% a m m o n i u m sulphate, followed by resolubilization a n d dialysis against fresh potassium p h o s p h a t e buffer, increases their binding capacity for serotonin a n d catecholamines. This partially purified p r e p a r a t i o n is used for the radioligand binding studies. W h e n the protein fractions are i n c u b a t e d for 15 min at 20°C with increasing c o n c e n t r a t i o n s o f [3H] serotonin (10 n M 500 n M ) , the a m o u n t o f b o u n d radioligand (i.e. the radioactivity t h a t co-elutes with protein on a small Sephadex G-50 column) is linearly p r o p o r t i o n a l to its free c o n c e n t r a t i o n (Fig. 1). Incub a t i o n o f the proteins with 0.1 m M FeSO4 for 3 rain prior to the addition of [3H]serotonin produces a m a r k e d increase in the a m o u n t of b o u n d radioligand (Fig. 1). This increase reflects the binding of [3H]sero t o n i n to SBP. This Fe2+-dependent binding comp o n e n t is saturable, a n d can be analyzed in terms of a single class o f non-cooperative sites : the Scatchard plot o f the s a t u r a t i o n binding data is linear (r = 0 . 9 8 + 0 . 0 1 , n = 2) a n d the slope of the Hill plot (nil = 1.00 _+0.04, n = 2) is close to one. The Ku value is 0.22+_0.04 ttM, n = 2, a n d the total a m o u n t of binding sites (B,,~,x) = 242_+ 10 p m o l / m g protein. As s h o w n in Fig. 2, Fe 2+ produces a dose-dependent increase in the a m o u n t of [3H]serotonin binding to SBP. M a x i m u m binding is attained at 100/~M Fe 2+.

1.2

A200

c

"g

RESULTS

E so

....

=O.Ol /

0

100

>., 200

o

150

300

.=_ 0 1 O0

200

300

400

500

c J~

100

o

50

[3H]serotonin concentration (nM) Fig. 1. Saturation binding of [3H]serotonin to soluble SBP from bovine retina: effect of Fe -'+. Serotonin binding proteins were partially purified from soluble extracts of bovine retina by precipitation in 30% ammonium sulphate. The proteins (0.1 mg/ml) were incubated for 3 min at 20'C in 20 mM potassium phosphate buffer (pH 7.5) either alone (A, non-specific binding) or in the presence of 0.1 mM Fe 2+ (C)) and then further incubated with increasing concentrations of ['H]serotonin (10 500 nM) for 15 min, after which proteinassociated radioactivity was measured. Binding is expressed in pmol/mg protein, and calculated parameters for specific (i.e. Fe -'~ dependent-)binding are: KD =0.22_+0.04 /~M, Bm~x= 242+__10 pmol/mg protein. Inset: Scatchard plot of the saturation binding data. Binding is expressed in pmol/mg protein and the free [3H]serotonin concentration in nM.

+

==

.E

~

+! +=++++

m

~

+ ¢J + +

+ "l" + + +

¢Z,., 1 O0

200

300

400

Fe 2÷ concentration (gM)

Fig. 2. Cation dependency of [3H]serotonin binding to SBP from bovine retina. SBP (0.1 mg/ml) were incubated for 3 min at 2@C with increasing concentrations of Fe 2~ (abscissa) and then further incubated for 15 min with 0.2 #M [~H]serotonin. Binding is expressed in pmol/mg protein. Inset." The proteins were preincubated for 3 min at 20'C with 0.1 mM of the indicated cations and then incubated for 15 min with 0.2 pM [~H]serotonin.

114

MARLENE JIMENEZ DEL RIO el al.

The inset in Fig. 2 compares the capability of 100 ILM of different metal ions to affect the binding of [3H]serotonin. In contrast to Fe 2+, Fe 3+ is unable to increase the binding. Except for Cu -'+, which produces an increase in binding up to 27% of the level induced by Fe z+ , all the other metal ions tested are ineffective. Competition binding studies reveal that serotonin, its precursor 5-hydroxytryptophan and its catabolite 5-HIAA possess high affinity for bovine retina SBP (K, values ranging between 0.15 and 0.30 ItM, Table

1). In contrast, melatonin produces no inhibition of [3H]serotonin binding, even at 10/*M (Table 1). These data confirm that presence of a hydroxyl group on the indole ring is essential for binding of serotonin analogs to SBP (Tamir and Rapport, 1978; Jimenez Del Rio et al., 1992). The catecholamine messengers dopamine, noradrenaline and adrenaline, possess equally high affinity for SBP (K~ values ranging betwccn 0.12 and 0.18 ttM, Table 1). Catabolites, where the hydroxyl group at the aromatic C3-position arc replaced by a methoxy group, display about 20-

Table 1. Binding of [~H]serotonin to SBP from bovine retina : competition by catecholamincs and serotonin-anulogs

]~ICH-(2

X

CH

NH--y

A

Substitution at position: Compound Dopamine Noradrenaline Adrenaline 3-Methoxytyramine

A

B

(7

X

OH OH OH OCH3

OH OH OH OH

H H H H

H OH OH H

Y H H CH~ H

Brain K~ 0.09 0. I 1 0.11 3.4

Retina {ItMI 0.13 0.12 0.18 0.93

Bg>_x A

Substitution at position : A

B

X

Bruin Ki

Retina (t;M)

OH OH OCH~

OH OH OH

CH2 COOH CH2 CH(NH2) COOH CH_, COOH

0.15 0.17 5.1

0.15 0.14 2.2

--

Compound DOPAC L-DOPA Homovunillic acid

c?_x

Substitution al position : A

X

Brain K~

Retina (/zM)

OH OH OH OCH3

CH2 NH2 CH(NH2) COOH COOH CH 2 NH COCH~

0.31 0.27 0.26 >200

0.15 0.30 0.30 >200

--

Compound 5-Hydroxytryptamine 5-Hydroxytryptophan 5-HIAA Melatonin

SBP were incubated with Fe 2~ and then further incubated with 0.2 #M [3H]serotonin and increasing concentrations of the listed compounds as shown in Fig. 3. K,-values of the competitors for SBP from bovine retina were calculated from the ICs~:values by the method of Cheng and Prusoff (1973). These values are compared to the K,-values for SBP from bovine brain (data from Jimenez del Rio et aL, 1992).

Serotonin and catecholamine binding proteins fold lower affinity for SBP (K~ values ranging between 0.93 #M and 2.2/,M, Table 1 and Fig. 3). Fe 2+ also markedly increases the binding of [3H] dopamine to the a m m o n i u m sulphate precipitated proteins from bovine retina. This FeZ+-dependent binding component is saturable and behaves as a single class of non-cooperative sites; the Scatchard plot is linear (r =- 0.87_+ 0.03, n = 2, Fig. 4 inset) and

o ~0o 8

"6

80

£ z~ .E

60

~

,0

p

2O

0 -7

-8

-6

-5

-4

Competitor concentration (Log M)

Fig. 3. Catecholamine/[3H]serotonin competition binding to bovine retina SBP. SBP were incubated for 3 rain at 20~C with 0.1 mM of Fe z+ and then further incubated for 15 rain with 0.2/~M [~H]serotonin and increasing concentrations of L-DOPA, dopamine, noradrenaline, adrenaline and homovanillic acid. Binding is expressed as percent of control binding, i.e. binding in absence of competitor.

400

E

"-o

300

E "0 " o .Q c "~ m Q. o "o "I*

200

100

0

0

100

200

300

400

S00

[3 H]dopamine concentration (nM)

Fig. 4. Saturation binding of [3H]dopamine to bovine retina SBP : effect of Fe 2 +. Samples of protein were incubated in buffer alone (•, non-specific binding) or with 0.1 mM Fe 2+ (O) and then further incubated with increasing concentrations of [3H]dopamine (10-500 riM) as described for Fig. 1. Specific binding ( 0 ) represents the Fe 2+ dependent binding component (KD = 0.31 + 0.04 /~M, Bm,~x= 505 + 30 pmol/mg protein). Inset : Scatchard plot of the saturation binding data. Binding is expressed in pmol/mg protein and the free [~H]dopamine concentration in nM.

115

the Hill coefficient ( n i l = 1.04+0.02, n = 2) is close to one. The Ko value (0.31 _+0.04 #M, n = 2) for [3H]dopamine saturation binding is similar to the K~ of unlabeled dopamine (0.13 /,tM, Table I) for competing with [3H]serotonin. The a m o u n t of Fe 2+dependent binding sites for [3H]dopamine is about twice the amount of sites for [3H]serotonin (i.e. 505_+30 vs 242_+10 pmol/mg protein, on the same protein preparation). Two different experiments indicate that the physicochemical properties of [3H]dopamine- and [~H] serotonin-binding proteins from bovine retina are similar to those found for SBP in rat and bovine brain (Gershon et al., 1983 ; Jimenez Del Rio et al., 1992). First, [3H]dopamine- and [3H]serotonin-labeled proteins elute in a sharp peak immediately after the void volume on a Sephacryl 100 H R gel filtration column. S D S - P A G E reveals that this peak contains four major protein bands, with molecular weights of respectively 38, 43, 48 and 56 kDa (data not shown). The gel filtration data thus indicate that the [~H] dopaminc- and [3H]serotonin-binding proteins form complexes with a high molecular weight. The aim of the second experiment is to determine the molecular weights of these binding proteins under non-denaturing conditions. In this context, we have noticed that protein aggregation is less pronounced when the a m m o n i u m sulphate precipitation step is omitted. Thus, soluble extracts from bovine retina are dialyzed against fresh phosphate buffer, and then immediately preincubated with Fe 2+ and incubated with radioligand. The labeled proteins are then separated from the excess free radioligand on a Sephadex G-50 column and subjected to native PAGE. Under these conditions, both radioligands are shown to remain associated with proteins migrating to apparent molecular weights of 43 and 57 kDa, respectively (Fig. 5). Radioactivity is also associated to components with high molecular weight, indicating that protein complex formation still takes place. The cation sensitivity of [3H]dopamine binding is shown in Figure 6. Fe 2+ produces a maximal increase in the binding of [3H]dopamine at concentrations ranging between 25 and 100/~M (Fig. 6). Fe ~+ (100 ~lM) does not affect binding of [~H]dopamine, but Cu 2+ and Mn 2+ increase the binding to more than 50% of the level induced by Fe 2+ (Fig. 6, inset). DISCUSSION

In this study, it is shown for the first time that the mammalian retina contains soluble SBP whose binding and physicochemical characteristics are very

116

MARLENE JIMENEZ DEL Rio et al. 3000

~--~ 2000

-----40

2

;'oo

10

,

i

,

20

30

40

50

slice number

Fig. 5. Native PAGE of [3H]dopamine SBP and [3H]serotonin SBP complexes. Soluble extracts from bovine retina were dialyzed against fresh buffer, preincubated with Fe 2+ and incubated with 0.5 pM radioligand as described in the legends of Figs 1 and 4. Labeled protein was separated from the excess free radioligand on a Sephadex G-50 column and run on 7.5% native polyacrylamide gels. The gels were then sliced and the radioactivity present in the slices was counted. [3H]Serotonin and [3H]dopamine-associated proteins showed two peaks (slices 23 and 30) with apparent molecular weights of respectively 43 kDa and 57 kDa. h~set: Calibration scale with protein standards and calculation of the molecular weights of slices 23 and 30.

similar to those of SBP in the brain. Catecholamine/ [SH]serotonin competition binding studies reveal that SBP in bovine retina (Fig. 3) and in bovine brain (Jimenez Del Rio et al., 1992) possess high affinity for catecholamines as well. Direct saturation binding

250

o

200

•--¢ 150

100

E ¢g

Oz. o 50 Z 0 0

1O0

200

300

400

Fe2+ concentration (gM)

Fig. 6. Cation dependency of [~H]dopamine binding to SBP from bovine retina. SBP were incubated with increasing concentrations of Fe 2+ (abscissa) and then with 0.2 #M [3H] dopamine as described for Fig. 2. Binding is expressed in pmol/mg protein. Inset: The proteins were incubated with 100 #M of the indicated cations and then with 0.2 ,uM [3H] dopamine.

studies indicate that these proteins behave as single class of binding sites, with about the same affinity for [3H]serotonin as for [3H]dopamine (KD = 0.22 #M and 0.31 /~M, respectively). Comparable affinity values were also found for SBP in bovine cortex (KD = 0.12 and 0.19/~M, respectively) (Jimenez Del Rio et al., 1992). [3H]Dopamine is capable of labeling about twice the amount of sites on SBP from bovine retina as compared to [SH]serotonin (B ..... = 505 and 242 pmol/mg protein, respectively). The same discrepancy between the number of binding sites was also observed for SBP from bovine cortex (Jimenez Del Rio et al., 1992), and this may be caused by a different accessibility of both radioligands to the binding sites. Crude preparations of SBP from rat, guinea pig, rabbit and sheep, as well as purified SBP from the rat have been described to possess sites with both " h i g h " (with KD's ranging from 0.4 to 9.7 nM) and " l o w " with KD's from 0.015 to 0.48 # M ) affinity for [3H]serotonin (Liu et al., 1985, 1987; Jonakait et al., 1977).When indicated in these studies, the high affinity sites did not represent more than 5 10% of the total binding capacity (Liu et al., 1985, 1987(Jonakait et al., 1977). However, such sites were not detected for SBP in the retina (Fig. 1) nor in the frontal cortex of the bovine (Jimenez Del Rio et al., 1992). Whether this difference is related to the species or to subtle differences in the binding procedure remains to be established. The cation dependency of [3H]serotonin and of [3H]dopamine binding is the same for SBP in bovine retina (Figs 2 and 6) and in brain (Jimenez Del Rio et al., 1992). Binding of [3H]serotonin requires Fe 2+, it is only slightly increased by Cu 2+, and it is not affected by Fe 3+ nor by any of the other metal ions tested. Binding of [3H]dopamine is also Fe2+-dependent, it is increased by Cu 2+ and Mn :+ to about 50% of the binding in the presence of Fe z+ , but not increased by Fe 3+. Competition binding studies with serotonin and catecholamine analogs (Table 1) provide a further indication for the similarity between the SBP in bovine retina and brain. These experiments also confirm that the presence of a hydroxyl group on the indole ring is essential for binding of serotonin analogs and that the affinity of catecholamine analogs decreases when the hydroxyl group at the aromatic C3-position is replaced by a methoxy group. Native P A G E of SBP from bovine retina reveals that binding of [3H]serotonin and of [3H]dopamine occurs to two peptides with apparent molecular weights of respectively 43 and 57 kDa (Fig. 5). These values are almost identical to those reported for SBP

Serotonin and catecholamifie binding proteins from rat brain (i.e. 45 and 56 kDa, Gershon et al., 1983; Gershon and Tamir, 1984). For bovine brain, labeling was reported to occur preferentially to a 5758 kDa peptide (Jimenez Del Rio et al., 1992), but more recent experiments reveal that the low molecular weight component can also be detected by native PAGE when the proteins did not undergo prior precipitation in ammonium sulphate (data not shown). The functional roles that have previously been attributed to SBP are limited to the housekeeping of serotonin (storage, protection from MAO, transport). In this context, a great deal of research has already been focussed on the exact subcellular localization of these proteins. It was initially found that isolated monoamine storage vesicles still contain SBP (Barasch et al., 1987: Tamir and Gershon, 1979) and these proteins are co-released with serotonin upon electrical stimulation of enteric neurons (Jonakait et al., 1979). These findings provided solid support for the involvement of SBP in the storage of serotonin. More recently, the possibility has also been raised that the 56 kDa form o1" SBP might be located at the cytosolic side of the vesicles or even be freely present in the cytosol. The putative role of such SBP might be to protect serotonin from degradation by monoamine oxidase, or even to participate in the transport of serotonin (Adlersberg eta/., 1987). The SBP have been proposed to be specifically located in serotonergic neurons from neuroectodermic origin (Gershon and Tamir, 1984). Serotonin is also stored in cells from other developmental origin (i.e. blood platelets, enterochromaffin cells and mast cells), but the implicated proteins are distinct from SBP (Tamir and Gershon, 1981 ; Tamir et al., 1980; Tamir and Liu, 1982). The above hypothesis also implies that SBP should be present in no other neurons than in serotonergic neurons. So far, this assertion has not been substantiated by conclusive evidence. Initial [3H]serotonin binding studies (Tamir and Kuhar, 1975) were performed in the absence of Fe 2+, so that the validity of their conclusions is questionable. Later on, it was shown that the appearance of SBP parallels the development of serotonergic neurons in the intestine of the rabbit, and precedes the ingrowth of the adrenergic innervation (Jonakait et al., 1977). These findings do not, however, prove that SBP is confined to serotonergic neurons only. More precise information concerning the cellular localization of SBP has recently been obtained by immunocytochemical studies wherein polyclonal and monoclonal antibodies were used against both molecular weight forms of SBP (Barasch et al., 1987: Kirchgessner et al., 1988: Liu et al., 1990a). The

I 17

presented data confirmed the occurrence of SBP in serotonin-secreting cells in sheep thyroid and in rat brain, but it was equally noticed in each of these studies that SBP may also be present in neurons which do not contain detectable amounts of serotonin. Although these latter observations could be attributed to a variety of experimental artefacts, including the limited specificity of the radioligands used (Liu et al., 1990a), the data presented in this study provide solid evidence for the occurrence of SBP in neurons wherein the concentration of serotonin is either nil, or at least too low to be detected. Indeed, the concentration of serotonin is extremely low in the retina from the bovine as well as from other mammalian species (Osborne, 1980b; Osborne et al., 1982) and no endogenous serotonin has, so far, been detected in individual cells by techniques such as aldehyde condensation histochemistry and immunohistochemistry (Ehinger and Flor6n, 1980: Ehinger et al., 1981; Osborne and Patel, 1984). Yet, SBP can readily be detected in the ammonium sulphate-precipitated soluble extracts of bovine retina, and its concentration is even about two times higher as in corresponding extracts of bovine cortex (Figs 1 and 4 in this study, and Jimenez Del Rio et al., 1992). Our findings therefore clearly indicate that there is no positive correlation between the SBP and the serotonin content of a tissue. High levels of SBP are thus present in the bovine retina and in neurones in the CNS which are apparently devoid of serotonin (Barasch et al., 1987; Kirchgessner et al., 1988; Liu et al., 1990a). These observations can only be reconciled with the classical view that SBP are confined to serotonergic neurons if one invokes the concept of "masked indoleamine cells" (Nishida et al., 1985). There is now compelling evidence for the occurrence of such serotonergic neurons in the retina. The ability of certain amacrine cells in the mammalian retina to take up and accumulate exogenous serotonin and related indoleamines was first demonstrated for the rabbit and the cat (Ehinger and Floren, 1976) and subsequently for several other species including the bovine (Osborne, 1980a). Although these amacrine cells may be of different morphology, they have the common characteristic to form reciprocal synapses with rod bipolar cells (Sandell et al., 1989; Daw et al., 1990). Despite the initial point of vicw that serotonin itself has no functional role in the retina (Ehinger and Floren, 1980), evidence is now accumulating in favor of its status as retinal neurotransmitter. Serotonin can be synthesized in the retina from exogenous tryptophan (Osborne. 1980b). and forskolin has recently been

118

MARLENE JIMENEZ DEL RIO et al.

s h o w n to induce the synthesis o f e n d o g e n o u s serotonin by part of the " i n d o l e a m i n e a c c u m u l a t i n g n e u r o n s " ( O s b o r n e a n d Barnett, 1989). In addition, serotonin receptors o f the 5-HT2 type have been shown to be present a n d functional in retina of the r a b b i t (Cutcliffe a n d Osborne, 1987). Based on these findings, serotonin has recently been proposed to provide a feedback at the rod bipolar terminal via these 5-HT~ receptors, resulting in an increased signal to noise ratio in rod pathways (Daw et al., 1990). The example of the retina has incited O s b o r n e and Barnett (1989) to conclude that serotonergic n e u r o n s do not necessarily need to contain detectable levels of e n d o g e n o u s serotonin. These new insights m a k e it still plausible for SBP to be confined to serotonergic neurons in the retina, where they would be implicated in the a c c u m u l a t i o n o f exogenous indoleamines. An alternative, more speculative view a b o u t the role of SBP in retina is based on the ability of (at least) bovine SBP to bind catecholamines with the same affinity, a n d by the same Fe2+-dependcnt mechanism, as serotonin (this study a n d Jimenez Del Rio et al., 1992) and on the fact that d o p a m i n e is the m a j o r c a t e c h o l a m i n e t r a n s m i t t e r in the retina ( K a m p , 1985). We are tempted to assume that SBP also plays a functional role in the storage or, more generally, in the housekeeping o f d o p a m i n e in its retinal neurons. The d o p a m i n e r g i c and serotonergic neurons are clearly distinct in the m a m m a l i a n retina (Ehinger and Floren, 1976, 1978) a n d they can also be distinguished from each o t h e r on the basis o f their m o r p h o l o g y (Wassle et al., 1987; Daw et al., 1990). The retina represents therefore a unique system to test the validity of the different functions which have been proposed for SBP in this report. In conclusion, we confirm in this study that bovine SBP bind serotonin and catecholamines with high affinity in the presence of Fe 2+. The presence of SBP in soluble extracts from bovine retina is in m a r k e d contrast with the present inability to observe endogenous serotonin in individual a m a c r i n e cells of the m a m malian retina. It may therefore be speculated that SBP is also involved in the housekeeping o f dopamine, the m a j o r c a t e c h o l a m i n e n e u r o t r a n s m i t t e r in the retina. However, SBP may also be present in " m a s k e d indoleaminc cells" in the retina.

REFERENCES

Adlersberg M., Liu K. P., Hsiung S.-C. and Tamir H. (1987) A Ca -,+ dependent protein kinase activity associated with serotonin binding protein. J. Neurochem. 49, 1105- I 115. Barasch J. M., Tamir H., Nunez E. A. and Gershon M. D. (1987) Serotonin-storing secretory granules from thyroid parafollicular cells. J. Neuroscience 7, 4017 4033. Cheng Y. C. and PrusoffW. H. (1973) Relationship between the inhibition constant (K,) and the concentration of inhibitor which causes 50 percent inhibition (150) of an enzymatic reaction. Biochem. Pharmac. 22, 3099 3108. Cutcliffe N. and Osborne N. N. (1987) Serotonergic and cholinergic stimulation of inositol phosphate formation in the rabbit retina. Evidence for the presence of serotonin and muscarinic receptors. Brain Res. 421, 95 104. Daw N. W., Jensen R. J. and Brunken W. J. (1990) Rod pathways in mammalian retinae. Trends" Neurol. Sci. 13, 110 115. Dowling J. E., Ehinger B. and Floren 1. (1980) Fluorescence and electron microscopical observations on the amineaccumulating neurons of the Cebus Monkey retina. J. comp. Neurol. 192, 665 685. Ehinger B. and Floren I. (1976) Indoleamine-accumulating neurons in the retina of rabbit, cat and goldfish. Cell Tiss. Res. 175, 37 48. Ehinger B. and Floren I. (1978) Quantitation of the uptake of indoleamines and dopamine in the rabbit retina. E.':p. Ere Res. 26, 1 11. Ehinger B. and Floren I. (1980) Retinal indoleamine accumulating neurons. Neurochemisto" !, 209 229. Ehinger B., Hansson C. and Tornqvist K. (1981) 5-Hydroxytryptamine in the retina of some mammals. Exp. Eye Res. 33, 663 672. Gershon M. D. and Tamir H. ( 1981 ) Release of endogenous 5-hydroxytryptamine from resting and stimulated enteric neurons. Neuroscience 6, 2227 2286. Gershon M. D. and Tamir H. (1984) Serotonectin and the family of proteins that bind serotonin. Biochem. Pharmac. 33, 3115 3118. Gershon M. D., Liu K. P., Karpiak S. E. and Tamir H. (1983) Storage of serotonin h7 rico as a complex with serotonin binding protein in central and peripheral serotonergic neurons. J. Neurosci. 3, 1901 1911. Jimenez Del Rio M., Pinxteren J., De Potter W. P., Ebinger G. and Vauquelin G. (1992) Serotonin-binding proteins in the bovine cerebral cortex : interaction wilh serotonin and catecholamines. Eur. J. Pharmac. 225, 225 234. Jonakait G. M., Tamir H., Rapport M. M. and Gershon M. D. 11977) Detection of a soluble serotonin-binding protein in mammalian myenteric plexus and other peripheral sites of serotonin storage. J. Neurochem. 28, 277 284. Jonakait G. M., Tamir H., Gintzler A. R. and Gcrshon M. D. (1979) Release of [3H]serotonin and its binding protein from enteric neurons. Brain Res. 174, 55 69. Kamp C. (1985) The dopamine system in the retina. In: Rethtal transmitters and modulators : models./~r the brain

The authors would like to thank JeanPaul De Backer and Wim Annaert for their expert technical assistance. G.V. is Onderzoeksdirecteur at the Nationaal Fonds voor Wetenschappelijk Onderzoek, Belgium. This work was supported by grants from the Fonds voor Geneeskundig Wetenschappelijk Onderzoek, the Koningin Elizabeth Sfichting and Lotto, Belgium. Acknowh, dgements

(Morgan, W. W., ed.), pp. I 32. CRC Press. Boca Raton. Kirchgessner A. L., Gerson M. D., Liu K. P. and Tamir H. (1988) Co-storage of serotonin binding protein with serotonin in the rat CNS. J. Neurosci. 8, 3879 3890. Laemmli U. K. (1971)) Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227, 680 685.

Serotonin and catecholamine binding proteins Liu K. P., Gershon M. D. and Tamir H. (1985) Identificatiom purification and characterization of two forms of serotonin binding protein from rat brain. J. Neurochem. 44,1289 1301. Liu K. P., Tamir H., Hsiung S., Adlersberg M. and Gershon M. D. (1987) Prenatal development of serotonin binding protein in rclation to other transmitter-related characteristics of central serotonergic neurons. Dev. Brain Res. 32, 31 41. Liu K. P., Yu P. Y., Hsiung S. H., Kirchgessner A. L., Gershon M. D. and Tamir H. (1990a) Preparation and characterization of monoclonal antibodies to serotonin binding protein. J. Neuroehem. 55, 1013 1021. Liu K. P., Tamir H. and Adlersberg M. (1990b) Photoaffinity labeling of two forms of serotonin binding protein : peptide mapping of the binding sites. J. Neuroehem. 54, 963 970. Lowry O. H., Rosebrough N. J., Farr A. C. and Randall R. J. (1951) Protein measurement with the Folin phenol reagent. J. hiol. Chem. 193, 265 275. Munson P. J. and Rodbard D. (1980) Ligand: a versatile computerized approach for characterisation of ligand binding systems. Analvl. Bioehem. 107, 220 239. Nishida K.. Ueda S. and Sano Y. (1985) Immunohistochcmical studies on masked indoleamine cells in the area prostrema of thc rat. Histoehemistry 82, 101 106. Osborne N. N. (1980a) Specificity of serotonin uptake by bovine retina : comparison with tryplamine. Exp. Fo'e Res. 31, 31 39. Osborne N. N. (1980b) h~ vitro experiments on the metabolism, uptake and releasc of 5-hydroxytryptamine in bovine retina. Brahl Res. 184, 283 297. Osborne N. N. and BarnEtt N. L. (1989) Serotonin levels in rabbit retina arc elevated following intraocular injection of l\~rskolin. J. NeurochenL 53, 1955 1958. Osborne N. N. and Patel S. (1984) Postnatal development of serotonin-accumulating neurones in the rabbit retina and an immunohistochemical analysis of thc uptake and release of scrotonin. Exp. Eye Res. 38, 611 620. Osborne N. N., Ncssclhut T., Nicholas D. A., Patel S. and Cuello A. C. (1982) Serotonin-containing neuroncs in vcrtebrate retinas. J. Neuroehem. 39, 1519 1528.

119

Sandell H. J., Masland H. R., Raviola E. and Dacheux F. R. (I 989) Connections of indoleamine-accumulating cells in the rabbit retina. J. Comp. Neurol. 283, 303 313. Tamir H. and Gershon M. D. (1979) Storage of serotonin and serotonin binding protein in synaptic vesicles. J. Neuroehem. 33, 35 44. Tamir H. and Gershon M. D. (1981) Intracellular proteins that bind serotonin in neurons, paraneurons and platelets. J. Physiol. (Paris) 77, 283 286. Tamir H. and Huang Y. L. (1974) Binding of serotonin to soluble protein from synaptosomes. Lflb Sei. 14, 83 93. Tamir H. and Kuhar M. J. (1975) Association of serotoninbinding protein to projections of midbrain raphe nuclei. Brahl Res. 83, 169 172. Tamir H. and Liu K. P. (1982) On the nature of the interaction between serotonin and scrotonin binding protein : effect of nucleotides, ions and sulfhydryl reagents. J. Neuroehem. 38, 135 141. Tamir H. and Rapport M. M. (1978) EffEcts of neurotoxins hi vitro on the binding of serotonin to serotonin binding protein. Ann. N. Y. Aead. Sei. 305, 85 95. Tamir H., Bebirian R., Muller F. and Casper D. (1980) Differences between intracellular platelet and brain proteins that bind scrotonin. J. Neuroehem. 35, 1033 1044. Tamir H., Liu K. P., Payette R. F., Hsiung S., Adlersberg M., Nunez E. A. and Gershon M. D. (1989) Human medullary thyroid carcinoma : characterization of serotonergic and neuronal properties of neuroectodermally derived cell line. J. Neurosei. 9, I 199 1212. Wassle H., Voigt T. and Patel B. (1987) Morphological and immnnocytochemical identification of indoleamineaccumulating neurons in the cat retina. J. Neurosei. 7, 1574 1584. Winkler H., Fischer-Colbrie R. and Wcbcr A. ( 198I ) Chemical Neurotransmission 75 years (Stjarne L., Hedqvist P., Lagercrantz H. and Wennmalm A., eds), pp. 57 73. Academic Press, London. Winkler H., Apps D. K. and Fischer-Colbrie R. (1986) The molecular function of adrenal chromaffin granules: established t~cts and unresolved topics. Neuroseienee 18, 261 290.