Neurochemical Research, VoL 17, No. 10, 1992, pp. 991-996
[3H]Serotonin Binding Sites in Goldfish Retinal Membranes Lucimey Lima 1, Isabelle Radtke 1, and Boris Drujan 1. (Accepted January 27, 1992)
Serotonin (5HT) binding sites were studied in goldfish retinal membranes by radioligand experiments. The binding site of [3H]5HT was sensitive to pre-treatment of the membranes at 40~ or 60~ C. 5HT and 5-methoxy-N,N-dimethyltryptaminewere the best inhibitors of [3H]5HT binding to retinal membranes. The 5HT2 agonist, 1-(-naphtyl)piperazine,was also a potent inhibitor, however, (+)-l-2,5-dimethoxy-4-iodophenyl-2-aminopropanewas less efficient. The catecholaminergicagents haloperidol and clonidine did not display an important inhibition. Propranolol, also reported as 5HTm antagonist, was a relatively potent blocker. Monoamine uptake blockers did not show potent inhibition. The GTP analog, GppNHp, inhibited the binding. The iterative analysis of saturation curves revealed two classes of binding sites, a high affinity component (Bm~,2.45 pmol/mg of protein, kd 6.86 riM), and a low affinity component (B~.~,53.46 pmol/mg of protein, I~ 232.07 nM). Analysis of the association and dissociation kinetics suggested a binding site (I~ 2 nM). The semilogarithmic plot of the dissociation kinetics gave curves concave to the upper side. The selectivity of the binding and the inhibition by GppNHp suggest the existance of 5HT1 receptors in goldfish retina. The low affinity interaction probably represents the transporter of 5HT or a suptype of receptor expressed in glial cells. KEY WORDS: Goldfish; 5HT receptors; retina.
by birds and lizards (3). The content of 5HT in bovine and rabbit retina has been more difficult to demonstrate, however, by high performance liquid chromatography, low concentration of this monoamine has been evident (1,4). By radioligand experiments 5HT receptors have been demonstrated in bovine (5) and rabbit retina (4). Moreover, 5HT has been reported to evoke an increase in the levels of cyclic AMP (6,7) and in the accumulation of inositol phosphates (8,9,10) in rabbit retina. The influence of agonists and antagonists selective for 5HTIA and 5HT2 receptors supports a physiological function of these receptors on the activity of ON and OFF ganglion cells throughout the rod pathway (11,12). 5HT-inmunoreactive ceils have been demonstrated in teleost retina (13,14). Autoradiography of goldfish retinas with [3I-I]SHT displayed three types of cells, two amacrine and one bipolar (14). 5HT-like immunoreactivity was only ob-
INTRODUC~ON Endogenous serotonin (5HT) has been detected in the retina of several species (1,2). Frogs and teleosts contain the highest levels of the indolamine (1), followed 1 Laboratorio de Neuroquimica, Centro de Biofisicay Bioquimica, Instituto Venezolanode InvestigacionesCientificas,Apdo. 21827, Caracas 1020AVenezuela. fin memoryof Dr. Boris Drujanwho died on Dee. 24, 1991. Abbreviations used: Bm~, maximumbindingcapacity;CPP, 1-(3 chlorophenyl)piperazine;CLN, clonidine;DMI,desimipramine;DMT, 5methoxy-N,N-dimethyltryptamine;DOI, (+)-l-(2,5-dimethoxy-4-iodophenyl-2-aminopropane; DPAT, (+)-8-hydroxy-2-(D1-N-propylamino)tetralin; GppNHp, 5-guanylylimidodiphosphate; HAL, haloperidol; 5HT, serotonin; ICso, concentrationof drug producing 50% inhibitionof binding; IMI, imipramine;K,~,equilibriumdissociationconstant;M1AN,mianserin;NOM, nomifensin;NP, 1-(1-napthyl)piperazine;PRP, propranolol. 991
0364-3190/92/1000-0991506.50/09 1992PlenumPublishingCorporation
992 served in one population of small amacrine cells (14). Despite the great amount of information concerning the role of 5HT in teleost retina the characterization of 5HT binding sites has not been carried out in detail. In the present report we demonstrate the existence of specific interaction of 5HT with goldfish retinal membranes as a basis work to further characterize 5HT receptors and transporter.
EXPERIMENTAL PROCEDURE Materials: 5-[1,2-3H(N)]-hydroxytryptamine, creatinine sulfate (NEN, Boston, MA, specific activity 25.4 Ci/mmol); 5HT, creatinine sulfate, 5-methoxy-N,N-dimethyltryptamine DMT, haloperidol HAL, propranolol PRP, clonidine CLN, imipramine IMI, desipramine DMI, 5-guanylylimidodiphosphate GppNHp, Tris-[hydroxymethyl]-aminoethane (Sigma Chemical Co., St. Louis, MO); p-aminophenylethyl-mtrifluoromethylphenylpiperazine, (+)-8-hydroxy-2-(D1-N-propylamino)tetralin hydrobromide DPAT, mianserin hydrochloride MIAN, 1(3-chlorophenyl)piperazine dihydrochloride CPP, 1-(1-naphtyl)pipe~:azine hydrochloride NP, (+)-l-2,5-dimethoxy-4-iodophenyl-2-aminopropanehydrochlodde DOI (Research BiochemicalsInc., Natick, MA); nomifensin was a gift of Dr. L. Cubeddu. Membrane Preparation. Goldfish, (Carassius auratus) 4-7 cm were adapted to darkness for 30 min prior to dissection of the retina. The retina was homogenized in a Tissumizer (Tekmar Company, Cincinnati, OH) in 50 mM Tris-HC1 pH 7.7, centrifuged at 22,000 g for 10 rain, and washed twice for removing endogenous 5HT. The final pellet was resuspended in the incubation buffer (Tris-HCl 50 mM at pH 7.7, pargyline 10 ~M, CaC12 4 mM, ascorbic acid 0.1%) to a protein concentration of 700-1200 I~g/ml. For the determihation of the appropriate concentration of membranes a curve of dilution of the preparations was carried out. Radioligand Binding Assay. The drugs were dissolved in the incubation buffer except the first dilution of MIAN and CPP (0.1 N HC1), and HAL (50% ethanol). Membranes, 100 I~1 (0.7-1.2 ~g per tube), were added to the incubation medium, the reaction was started by the addition of 50 I~1of ligand, final volume 500 I~1.The selectivity of the interaction was determined by inhibition experiments in the presence of 15 concentrations (0.1 nM to 1 mM) of the drags. Saturation experiments were performed with concentrations of 5HT from 0.1 to 50 nM. The association and dissociation (by addition of 100 IxM 5HT) constants were determined by kinetic experiments (4 nM [3H]5HT). The specific binding was defined by the substraction of the binding in the presence of 10-100 IxM 5HT from the total binding (in the absence of the inhibitor). After incubating at 25~ for 30 min the separation of the ligand-binding site complex from free ligand was performed by rapid filtration over Whatman GF/A glass fiber filters. The filters were washed twice with 5 ml of cold buffer, placed in vials and dried. Radioactivity was determined in a Packard liquid scintillation counter Model 3385 (efficiency 30-45%). Protein concentration was determined by the method of Lowry et al. (15). Analysis of Data. The statistical analysis was carried out by two ways analysis of variance in indicated experiments (16). Kinetic and saturation curves were analysed by the iterative procedures LIGAND (17,18), and ENZFITTER (19). Inhibition experiments were calculated by INHIBITION (20).
Lima, Radtke, and Drujan RESULTS
Dilution of the Membranes. The concentration of protein in the preparation of goldfish retinal membranes was linear to the amount of [ a H ] f H T bound to the membranes (Figure 1). Effect of the Temperature. Binding of [3H]fHT to retinal membranes occured by incubation at 25~ for 30 min or at 4~ for 4 h. Preincubation of the tissue preparation at 40 and 60~ for 10 min decreased [3H] binding (Figure 2), supporting the interaction with membrane protein components. Corresponding control values of the two experiments were not statistically different. Selectivity of Binding. Serotonergic agents inhibited the interaction of [3H]fHT (2 nM) to retinal membranes in a selective manner. The most potent of the serotonergic agents was DMT followed by NP, 5HT, DOI, and DPAT (Table I). The catecholaminergic agent HAL was not a potent inhibitor, however PRP displayed a relative low IC5o (Table I). Monoamine uptake blockers were not inhibitors of 5HT binding, except DMI, although the IC5o of this drug was high (Table I). The GTP analog GppNI-Ip had a IC50 of 711 I~M (Table I). The analysis for two sites was not possible, however,
500
E
"-"
Q
9
Q. O
o
300
Z 0 FI I
I00 150
70
MembrQnes ( pg prot) Fig. 1. Correlation between different dilution of goldfish retinal membrane prepartion and the binding of 2 nM [3H]5HT. The analysis was carried out by LINEFIT (20).
[3H] S e r o t o n i n B i n d i n g Sites in G o l d f i s h R e t i n a
10
993
L]
6
nM (Figure 3 and 4). The analysis of the data by Bound/ Free vs. Bound plots (20,21) suggested the existance of two recognition sites (Figure 4). The analysis of one site gave values reported in Table II ( B ~ , maximum binding capacity; Kd, equilibrium dissociation constant). The Scatchard plots were analysed by LIGAND (17) and best fitted to a two sites model (Figure 4, F = 3-4, Table II). The Hill coefficient was near 1 (Table II). The removal of photoreceptors did not modify the interaction of 5HT with the retinal membranes in a significant man-. per.
x
o "
4
E p c 0.05
E
Q.
nn
2 T
0
C
Pre-incubotion
40%
C
Pre-incubation
2.8
60%
Fig. 2. Effect of temperature on the binding of [3H]5HT to goldfish retinal membranes. The membrane preparation was pre-incubated at 40 or 60~ for 10 min prior to the incubation in the presence of 2 nM pH]5HT. Values are mean • SEM, n = 3.
Bma x 6.74 pmol /mg prot Kd 25.9l nM
2.4
-~
210
E
~ 9
z
1.2
O m
08
0-4 I Bmax 9.02prnol/mg prot
9
:
i
,~o_o~%~ . .
~ 0.4 Table I. Inhibition of pH]5HT Binding to Goldfish Retinal
I
Membranes lCso (~M)
n.
Serotonergic agents 5HT DMT DPAT NP DOI MIAN
18.48 8.13 100.46 10.21 54.54 >
• 2.53 --- 2.25 • 7.75 --- 3.28 • 17.21 1000
0.39 0.35 0.80 0.43 0.61
• • • • •
0.04 0.005 0.19 0.10 0.02
215.98 - 100.61 16.36 --- 13.00 > 1000
5
0
I0"12
10 15 FREE [3H1-SHT (nM}
20
Fig. 3. Saturation curve of specifc [3H]5Ht binding to goldfish retinal membranes. Nonspecific binding was defined in the presence of 10 ixM 5HT. The membranes were incubated with 20 concentrations of pH]5HT. The curve were constructed and the values of saturation parameters calculated by the non-linear iterative analysis ENZFITIER (19). B,,~, and Kd were also calculated by the Scatchard plot (21).
Catecholaminergic agents HAL PRP CLN
ound
I
if)
0.248
0.43 --- 0.16 0.18 - 0.14 0.206
Bmox H 2.45 pmol/rng prot
Bmax L 53.46 pmol/mg prot
Kd H 6.86 nM
Kd L 232.07 nM
Monoamine uptake blockers DMI IMI NOM
51.13 --- 20.23 > 1000 > 1000
0.52 • 0.33 0.t65 t.o cr u_
Nucleotide Gpp
711.67 • 59.46
0.93 • 0.14
""
0324
Z
The potency of inhibition was determined by incubating the membranes with 2 nM [3H]5HT in the presence of 15 concentrations of the corresponding drug. Values are mean -+- SEM, n = 3. IC~o were obtained by the iterative analysis INHIBITION (20). Pseudo-Hill coefficients (ha) were determined according to Chou (31).
9
0
0.083
0041 .~
~
~
~\
\l
the pseudo-nil was less than 1 for all the drugs (Table
1). Saturation Experiments. All the saturation curves performed presented a component not saturable up to 50
0
5.00E-IO
i I.OOE-09
I
....
1.50E-09
i 2 00E-09
BOUND (M) Fig. 4. Saturation analysis of a representative experiment calculated by the iterative program LIGAND (17). The values correspond to the data in Figure 3.
994
Lima, Radtke, and Drujan
Table H. Saturation Parameters of the Specific Binding of [3H]5HT Defined by 5HT to Goldfish Retinal Membranes
One site Two sites
Kd (nM)
B~,, (pmol/mg prot)
Hill coefficient
50.06 -4- 11.13 6.86 - 2.08 232.07 --- 95.34
23.78 --- 9.35 2.45 --- 0.60 53.46 --- 11.35
0.96 --- 0.01
The membranes were incubated with 20 concentrations of [3H]5HT, unspecific binding was defined by 10 ~M 5HT. Values are mean --SEM, n = 10. The parameters were calculated by the iterative analysis LIGAND (17).
I0
o
"o v
o
8
6
Z 0,-n T tO
4
~3
m -0.55
k~ Q.
0
L 0
30
60 Time
I
I
90
120
(rain)
Fig. 5. Association and dissociation kinetics of specific binding of [3H]5HT to goldfish retinal membranes. The association constant (-o-) was determined by incubating the membranes with 4 nM pH]5HT, unspecific binding was the counts in the presence of 10 ~M 5HT. The dissociation constant (-e-) was determined by the displacement of 4 nM [3H]5HT at equilibrium with 100 I-LM5HT. Inset: semilogatihmic plot of dissociation data.
Kinetic Experiments. The association of [3H]5HT reached the equilibrium within 30 min and the binding was stable up to 90 min (Figure 5). The apparent association constant (kobs) was 0.73 min-L The dissociation constant was determined by the addition of 100 ~M 5HT as displacer at 60 min of incubation, the dissociation rate constant (k2) calculated by the slope of the semilogarithmic plot was 0.24 min -1. The dissociation was not completed up to 120 min and remained in approximately 50% of the association. The equilibrium dissociation constant (I~) calculated by the association and dissociation rate constants was 2 nM (Figure 5). The semilogarithmic transformation of the dissociation data corresponded to a concave curve to the upper side (Figure 5, inset).
DISCUSSION The characterization of 5HT interaction with goldfish retinal membranes was not carried out previously, however, various reports suggest the presence of 5HT recognition sites in the retina of several species by physiological approaches (6,10,14,22,23). In addition, resuits obtained by radioligand experiments in mammalian retina suggested the presence of 5HT 1 and 5HT2 receptors in rabbit retina (4), and of 5HT1 receptors in bovine retina (5). As have been consistently reported the concentration of 5HT in teleost retina is elevated (1,2). A recent report (14) demonstrated the existance of three population of cells in goldfish retina able to uptake [3H]5HT, but only one subtype of amacrine cells containing 5HT, according to other authors (13). The high affinity component of [3H]5HT binding site in goldfish retina in the present report (Table 2) might be the representation of 5HT receptors expressed at the postsynaptic interaction site of the neuron that contain 5HT endogenously, and the low affinity site might be representing the uptake sites expressed in the three types of cells that transport 5HT. Another possibility is that Miiller cells could express a 5HT receptor with different affinity from the subtype present in retinal neurons. The fact that a concentration of 5HT of 10 izM, used to block 5HTx receptors in order to define specific binding (24,25), only blocked 50% of the binding of [3H]5HT to goldfish retinal membranes is indicative of the existance of a limited number of sites probably representing the population of 5HT receptors. The component of low affinity and high capacity demonstrated by the saturation experiments is also observed by the kinetic analysis of the present work (Figure 5). The inhibition of [3H]5HT binding by the serotonergic agents reported in Table 1 as IC5o are higher than the values reported for 5HT~ and 5HT2 receptors in the brain (26). This might be the result of the dual interaction of 5HT in goldfish retina and the impossibility of these agents to block the binding to the carrier of 5HT to membranes of other cell populations of the retina (14). The presence of a second recognition site for 5HT in goldfish retinal membranes is not according with the results reported by Osborne (5) in bovine retina in which he identified one site by saturation experiments in concentrations up to 20 nM of [3H]5HT. Mitchell and Redbum (4) also demonstrated the existance of a single binding site for [3H]LSD and [3H]spiperone in rabbit retina, however, the possibility of observation of two components is difficult due to exploration of a small range of ligand concentrations. It might also be possible that the interaction sites of 5HT in teleost retina differ from
[3H] Serotonin Binding Sites in Goldfish Retina
mammalian retina, in which the population of cells that specifically transport 5HT is lower than in fish (14). Selective agonists and antagonists of 5HT1A and 5HT2 receptors support a physiological function in the regeneration of receptive field properties (7,22). The selective 5HTIA agonist, DPAT (27) was not a potent inhibitor of [3H]5HT binding to retinal membranes (Table 1) probably because the number of 5HTIA receptors in the retina is very low and the selectivity of the inhibitor is masked by the component of low affinity. The lower IC5o observed for 5ttT and DMT suggests the presence of 5HT1 receptors ('Fable 1). It is interesting that PRP exhibited a relatively low IC5o, this drug has been proposed as a 5HT m antagonist (28). The monoamine uptake blockers used in this study were not able to inhibit [3H]5HT binding (Table 1), suggesting that the ligand is not interacting with these uptake systems or that the monoamine carriers in goldfish retina are insensitive to these inhibitors. However, in certain manner IMI is not the best inhibitor of 5HT transporter, and express a very complex interaction with chickens and pigs retinal membranes (29). The pseudo-Hill coefficients of the inhibitory curves were less than one, suggesting either negative cooperativity or the binding of the ligand to high and low affinity sites (Table I). However, the competition curves were all best fitted to a one-site model. The inhibition displayed by GppNHp indicates the presence of 5HT1 receptors, which are regulated by guanine nucleotides (30). The kinetic analysis of the binding shows the dissociation constant of the high affinity complex because the values of Kd calculated by these experiments are close to the values obtained by saturation experiments, and also because the conditions of inhibition only blocks the 50% of the total binding (Figure 5). The semilogarithmic plot of the dissociation analysis are suggestive of a complex interaction, such as two sites. The described results are evidence of a low affinity interaction of 5HT in the retina of goldfish and might suggest the presence of 5HT1 receptors. However, the observation of a second component of the binding makes difficult the analysis by simple competition experiments and by using an unspecified ligand such as the endogenous molecule, w h i c h was necessary to start the kinetic characterization of 5 H T b i n d i n g sites in the retina of goldfish.
ACKNOWLEDGMENTS This work was partially supported by IVIC. Isabelle Radkte was an Assistant Student of Centro de Estudios Avanzados, IVIC, from
995 the University of Basilea, Switzerland. We appreciate the help of Lic. P. Matus in the dissection of the retina. We also thank Mrs. M. Florez for secretarial work and Mrs. Dhuwya Otero for making the figures.
REFERENCES 1. Osborne, N.N. 1982. Evidence for serotonin being a transmitter in the retina pages 401--430,/n Osborne, N.N., (ed.) Biology of serotonergic transmission. John Wiley & Sons, Chichester. 2. Tornquist, K., Hansson, C., and Ehinger, B. 1983. Immunohistochemical and quantitative analysis of 5-hydroxytryptamine in the retina of some vertebrates. Neurochem. Int. 5:299-305. 3. Ehinger, B. 1982. Neurotransmitters systems in the retina. Retina 2:305-321. 4. Mitchell, C.K., and Redbum, D.A. 1985. Analysis of pre- and postsynaptie factors of the serotonin system in rabbit retina. J. Cell Biol. 100:64--73. 5. Osborne, N.N. 1981. Binding of [31-I]serotonin to membranes of the bovines retina. Exp. Eye Res. 33:371-380. 6. Blanzinsky, C., Ferrendelli, J.A., and Cohen, A.I. 1985. Indolamine-sensitive adenylate cyclase in rabbit retina: Characterization and distribution. J. Neurochem. 45:440--447. 7. Osborne, N.N. 1990. Effects of GTP, forskolin, sodium fluoride, serotonin, dopamine, and carbachol on adelynate cyclase in teleost retina. Neurochem. Res. 15:523-528. 8. 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, 1987. 9. Ghazi, H., and Osborne N.N. 1988. Agonist-induced stimulation of inositol phosphates in primary rabbit retinal cultures. J. Neurochem. 50:1851-1858. 10. Osborne, N.N., and Ghazi, H. 1989. Serotonin receptors and their involvement in Ca* § movilization and glycogenolysis in the rabbit retina Pages 159-175, in Redburn, D., and Pasantes-Morales H., (eds.) Extracellular and intracellular messengers in the vertebrate retina. Alan R. Liss, Inc., New York. 11. Brunken, W.J., and Daw, N.W. 1987. The actions of serotonergic agonists and antagonist on the activity of brisk ganglion celia in the rabbit retina. J. Neurosei. 7:4054-4065. 12. Daw, N.W., Brunken, W.J., and Parkinson D. 1989. The function of synaptic transmitters in the retina. Ann. Rev. Neurosci. 12:205-225. 13. Jaffe, E.H., Urbina, M., Ayala, C. and Chemello, M.E. 1987. Serotonin containing neurons in the retina of the teleost Eugerres plumieri. Vision Res. 27:2015-2026. 14. Marc, R.F., Liu, W-L, S., Scholz, K., and Muller, J.F. 1988. Serotonergic and serotonin-accumulating neurons in the goldfish retina. J. Neurosci. 8:3427-3450. 15. Lowry, O.H., Rosebroughh, N. J., Farr, A.L., and Randall, R.J. 1951. Protein measurements with folin phenol reagent. J. Biol. Chem. 193:265-275. 16. Sokal, R., and Rohlf, F. 1979. Biometria. 1st. Edition. Blume Edieiones, Madrid. 17. McPherson, G.A. 1985. Analysis of radioligand binding experiments: A collection of computer programs for the IBM PC. J. Pharmacol. Methods 14:213-228. 18. Munson, P.J., and Roadbard, D. 1980. LINGAND: A versatile computerized approach for characterization of ligand-binding systerns. Ann. Bioehem. 107:220-239. 19. Leatherbarrow, R.J. 1987. A non-linear regression data analysis program for the IBM PC (and true compatibles) pp. 1-91 Elsevier Science Publishers, Amsterdam. 20. Barlow, R.B. 1983. Biodata Handlig with Microcomputers. pp. 125-143. Elsevier Science, Amsterdam.
996 21. Scatchard, G. 1949. The attraction of proteins for small molecules and ions. Ann. N. Y. Acad. Sci. 51:660-672. 22. Daw, N.W., Jensen, R.J., and Brunken, W.J. 1990. Rod Pathways in mammalian retina. Trends Neurosci. 13:110-115. 23. Osborne, N.N., and Barnett, N.L. 1989. Serotonin levels in the rabbit retina are elevated following intraocular injection of forskolin. J. Neurochem. 53:1955-1958. 24. Peroutka, S.J., and Snyder, S.H. 1979. Multiple serotonin receptors: Differential binding of 3H 5-hydroxytrypamine, 3H Lysergic acid diethylamide and 3H spiroperidol. Mol. Pharmacol. 16:687699. 25. Salazar, M., and Lima, L. 1990. Serotonin receptors modulation in rat hippocampus by clonazepam treatment. Acta Cientifica Venezolana 41:134. 26. Peroutka, S.J., Schmitdt, A.W., Sleight, A.J., and Harrington, M. 1990. Serotonin receptor "Families" in the central nervous system: An overview in Pages 104-113, Whitaker-Axmitia, P.M., and Peroutka, S.J., (eds) The neuropharmacology of serotonin. The New York Academy of Sciences, New York.
Lima, Radtke, and Drujan 27. Gozlan, H., Mestikawy, S. El, Pichat, L., Glowinsky, J., and Hamon, M. 1983. Identification of presynaptie serotonin autoreceptors using a new ligand: 3H-PAT. Nature 305:140-142. 28. Hoyer, D., Engel, G., and Kalkman, H.O. 1985. Characterization of the 5HT1B recognition site in rat brain: Binding studies with (-) [125I] iodocyanopindolol. Eur. J. Pharmacol. 118:1-1112. 29. Karl, E., Tuomisto, L , Airaksinen, M.M., and Lcino, M. 1988. 3H Imipramine binding in the retinas of chickens and pigs. Exp. Eye Res. 47:679-688. 30. Nelson, D.L. 1988. Biochemistry and Pharmacology of the 5HT1 sertonin binding sites in: The serotin receptors. (Sanders-Bush, E. ed.) pp. 23-58. The Humana Press Publishers New Jersey. 31. Chou, T.C. 1974. Relationships between inhibition constants and fractional inhibition in enzyme-catalyzed reactions with different numbers of reactants, different reaction mechanisms, and different types and mechanisms of inhibition. Mol. Pharmacol. 10:235247.
[3H]Serotonin Binding Sites in Goldfish Retinal Membranes Lucimey Lima 1, Isabelle Radtke 1, and Boris Drujan 1. (Accepted January 27, 1992)
Serotonin (5HT) binding sites were studied in goldfish retinal membranes by radioligand experiments. The binding site of [3H]5HT was sensitive to pre-treatment of the membranes at 40~ or 60~ C. 5HT and 5-methoxy-N,N-dimethyltryptaminewere the best inhibitors of [3H]5HT binding to retinal membranes. The 5HT2 agonist, 1-(-naphtyl)piperazine,was also a potent inhibitor, however, (+)-l-2,5-dimethoxy-4-iodophenyl-2-aminopropanewas less efficient. The catecholaminergicagents haloperidol and clonidine did not display an important inhibition. Propranolol, also reported as 5HTm antagonist, was a relatively potent blocker. Monoamine uptake blockers did not show potent inhibition. The GTP analog, GppNHp, inhibited the binding. The iterative analysis of saturation curves revealed two classes of binding sites, a high affinity component (Bm~,2.45 pmol/mg of protein, kd 6.86 riM), and a low affinity component (B~.~,53.46 pmol/mg of protein, I~ 232.07 nM). Analysis of the association and dissociation kinetics suggested a binding site (I~ 2 nM). The semilogarithmic plot of the dissociation kinetics gave curves concave to the upper side. The selectivity of the binding and the inhibition by GppNHp suggest the existance of 5HT1 receptors in goldfish retina. The low affinity interaction probably represents the transporter of 5HT or a suptype of receptor expressed in glial cells. KEY WORDS: Goldfish; 5HT receptors; retina.
by birds and lizards (3). The content of 5HT in bovine and rabbit retina has been more difficult to demonstrate, however, by high performance liquid chromatography, low concentration of this monoamine has been evident (1,4). By radioligand experiments 5HT receptors have been demonstrated in bovine (5) and rabbit retina (4). Moreover, 5HT has been reported to evoke an increase in the levels of cyclic AMP (6,7) and in the accumulation of inositol phosphates (8,9,10) in rabbit retina. The influence of agonists and antagonists selective for 5HTIA and 5HT2 receptors supports a physiological function of these receptors on the activity of ON and OFF ganglion cells throughout the rod pathway (11,12). 5HT-inmunoreactive ceils have been demonstrated in teleost retina (13,14). Autoradiography of goldfish retinas with [3I-I]SHT displayed three types of cells, two amacrine and one bipolar (14). 5HT-like immunoreactivity was only ob-
INTRODUC~ON Endogenous serotonin (5HT) has been detected in the retina of several species (1,2). Frogs and teleosts contain the highest levels of the indolamine (1), followed 1 Laboratorio de Neuroquimica, Centro de Biofisicay Bioquimica, Instituto Venezolanode InvestigacionesCientificas,Apdo. 21827, Caracas 1020AVenezuela. fin memoryof Dr. Boris Drujanwho died on Dee. 24, 1991. Abbreviations used: Bm~, maximumbindingcapacity;CPP, 1-(3 chlorophenyl)piperazine;CLN, clonidine;DMI,desimipramine;DMT, 5methoxy-N,N-dimethyltryptamine;DOI, (+)-l-(2,5-dimethoxy-4-iodophenyl-2-aminopropane; DPAT, (+)-8-hydroxy-2-(D1-N-propylamino)tetralin; GppNHp, 5-guanylylimidodiphosphate; HAL, haloperidol; 5HT, serotonin; ICso, concentrationof drug producing 50% inhibitionof binding; IMI, imipramine;K,~,equilibriumdissociationconstant;M1AN,mianserin;NOM, nomifensin;NP, 1-(1-napthyl)piperazine;PRP, propranolol. 991
0364-3190/92/1000-0991506.50/09 1992PlenumPublishingCorporation
992 served in one population of small amacrine cells (14). Despite the great amount of information concerning the role of 5HT in teleost retina the characterization of 5HT binding sites has not been carried out in detail. In the present report we demonstrate the existence of specific interaction of 5HT with goldfish retinal membranes as a basis work to further characterize 5HT receptors and transporter.
EXPERIMENTAL PROCEDURE Materials: 5-[1,2-3H(N)]-hydroxytryptamine, creatinine sulfate (NEN, Boston, MA, specific activity 25.4 Ci/mmol); 5HT, creatinine sulfate, 5-methoxy-N,N-dimethyltryptamine DMT, haloperidol HAL, propranolol PRP, clonidine CLN, imipramine IMI, desipramine DMI, 5-guanylylimidodiphosphate GppNHp, Tris-[hydroxymethyl]-aminoethane (Sigma Chemical Co., St. Louis, MO); p-aminophenylethyl-mtrifluoromethylphenylpiperazine, (+)-8-hydroxy-2-(D1-N-propylamino)tetralin hydrobromide DPAT, mianserin hydrochloride MIAN, 1(3-chlorophenyl)piperazine dihydrochloride CPP, 1-(1-naphtyl)pipe~:azine hydrochloride NP, (+)-l-2,5-dimethoxy-4-iodophenyl-2-aminopropanehydrochlodde DOI (Research BiochemicalsInc., Natick, MA); nomifensin was a gift of Dr. L. Cubeddu. Membrane Preparation. Goldfish, (Carassius auratus) 4-7 cm were adapted to darkness for 30 min prior to dissection of the retina. The retina was homogenized in a Tissumizer (Tekmar Company, Cincinnati, OH) in 50 mM Tris-HC1 pH 7.7, centrifuged at 22,000 g for 10 rain, and washed twice for removing endogenous 5HT. The final pellet was resuspended in the incubation buffer (Tris-HCl 50 mM at pH 7.7, pargyline 10 ~M, CaC12 4 mM, ascorbic acid 0.1%) to a protein concentration of 700-1200 I~g/ml. For the determihation of the appropriate concentration of membranes a curve of dilution of the preparations was carried out. Radioligand Binding Assay. The drugs were dissolved in the incubation buffer except the first dilution of MIAN and CPP (0.1 N HC1), and HAL (50% ethanol). Membranes, 100 I~1 (0.7-1.2 ~g per tube), were added to the incubation medium, the reaction was started by the addition of 50 I~1of ligand, final volume 500 I~1.The selectivity of the interaction was determined by inhibition experiments in the presence of 15 concentrations (0.1 nM to 1 mM) of the drags. Saturation experiments were performed with concentrations of 5HT from 0.1 to 50 nM. The association and dissociation (by addition of 100 IxM 5HT) constants were determined by kinetic experiments (4 nM [3H]5HT). The specific binding was defined by the substraction of the binding in the presence of 10-100 IxM 5HT from the total binding (in the absence of the inhibitor). After incubating at 25~ for 30 min the separation of the ligand-binding site complex from free ligand was performed by rapid filtration over Whatman GF/A glass fiber filters. The filters were washed twice with 5 ml of cold buffer, placed in vials and dried. Radioactivity was determined in a Packard liquid scintillation counter Model 3385 (efficiency 30-45%). Protein concentration was determined by the method of Lowry et al. (15). Analysis of Data. The statistical analysis was carried out by two ways analysis of variance in indicated experiments (16). Kinetic and saturation curves were analysed by the iterative procedures LIGAND (17,18), and ENZFITTER (19). Inhibition experiments were calculated by INHIBITION (20).
Lima, Radtke, and Drujan RESULTS
Dilution of the Membranes. The concentration of protein in the preparation of goldfish retinal membranes was linear to the amount of [ a H ] f H T bound to the membranes (Figure 1). Effect of the Temperature. Binding of [3H]fHT to retinal membranes occured by incubation at 25~ for 30 min or at 4~ for 4 h. Preincubation of the tissue preparation at 40 and 60~ for 10 min decreased [3H] binding (Figure 2), supporting the interaction with membrane protein components. Corresponding control values of the two experiments were not statistically different. Selectivity of Binding. Serotonergic agents inhibited the interaction of [3H]fHT (2 nM) to retinal membranes in a selective manner. The most potent of the serotonergic agents was DMT followed by NP, 5HT, DOI, and DPAT (Table I). The catecholaminergic agent HAL was not a potent inhibitor, however PRP displayed a relative low IC5o (Table I). Monoamine uptake blockers were not inhibitors of 5HT binding, except DMI, although the IC5o of this drug was high (Table I). The GTP analog GppNI-Ip had a IC50 of 711 I~M (Table I). The analysis for two sites was not possible, however,
500
E
"-"
Q
9
Q. O
o
300
Z 0 FI I
I00 150
70
MembrQnes ( pg prot) Fig. 1. Correlation between different dilution of goldfish retinal membrane prepartion and the binding of 2 nM [3H]5HT. The analysis was carried out by LINEFIT (20).
[3H] S e r o t o n i n B i n d i n g Sites in G o l d f i s h R e t i n a
10
993
L]
6
nM (Figure 3 and 4). The analysis of the data by Bound/ Free vs. Bound plots (20,21) suggested the existance of two recognition sites (Figure 4). The analysis of one site gave values reported in Table II ( B ~ , maximum binding capacity; Kd, equilibrium dissociation constant). The Scatchard plots were analysed by LIGAND (17) and best fitted to a two sites model (Figure 4, F = 3-4, Table II). The Hill coefficient was near 1 (Table II). The removal of photoreceptors did not modify the interaction of 5HT with the retinal membranes in a significant man-. per.
x
o "
4
E p c 0.05
E
Q.
nn
2 T
0
C
Pre-incubotion
40%
C
Pre-incubation
2.8
60%
Fig. 2. Effect of temperature on the binding of [3H]5HT to goldfish retinal membranes. The membrane preparation was pre-incubated at 40 or 60~ for 10 min prior to the incubation in the presence of 2 nM pH]5HT. Values are mean • SEM, n = 3.
Bma x 6.74 pmol /mg prot Kd 25.9l nM
2.4
-~
210
E
~ 9
z
1.2
O m
08
0-4 I Bmax 9.02prnol/mg prot
9
:
i
,~o_o~%~ . .
~ 0.4 Table I. Inhibition of pH]5HT Binding to Goldfish Retinal
I
Membranes lCso (~M)
n.
Serotonergic agents 5HT DMT DPAT NP DOI MIAN
18.48 8.13 100.46 10.21 54.54 >
• 2.53 --- 2.25 • 7.75 --- 3.28 • 17.21 1000
0.39 0.35 0.80 0.43 0.61
• • • • •
0.04 0.005 0.19 0.10 0.02
215.98 - 100.61 16.36 --- 13.00 > 1000
5
0
I0"12
10 15 FREE [3H1-SHT (nM}
20
Fig. 3. Saturation curve of specifc [3H]5Ht binding to goldfish retinal membranes. Nonspecific binding was defined in the presence of 10 ixM 5HT. The membranes were incubated with 20 concentrations of pH]5HT. The curve were constructed and the values of saturation parameters calculated by the non-linear iterative analysis ENZFITIER (19). B,,~, and Kd were also calculated by the Scatchard plot (21).
Catecholaminergic agents HAL PRP CLN
ound
I
if)
0.248
0.43 --- 0.16 0.18 - 0.14 0.206
Bmox H 2.45 pmol/rng prot
Bmax L 53.46 pmol/mg prot
Kd H 6.86 nM
Kd L 232.07 nM
Monoamine uptake blockers DMI IMI NOM
51.13 --- 20.23 > 1000 > 1000
0.52 • 0.33 0.t65 t.o cr u_
Nucleotide Gpp
711.67 • 59.46
0.93 • 0.14
""
0324
Z
The potency of inhibition was determined by incubating the membranes with 2 nM [3H]5HT in the presence of 15 concentrations of the corresponding drug. Values are mean -+- SEM, n = 3. IC~o were obtained by the iterative analysis INHIBITION (20). Pseudo-Hill coefficients (ha) were determined according to Chou (31).
9
0
0.083
0041 .~
~
~
~\
\l
the pseudo-nil was less than 1 for all the drugs (Table
1). Saturation Experiments. All the saturation curves performed presented a component not saturable up to 50
0
5.00E-IO
i I.OOE-09
I
....
1.50E-09
i 2 00E-09
BOUND (M) Fig. 4. Saturation analysis of a representative experiment calculated by the iterative program LIGAND (17). The values correspond to the data in Figure 3.
994
Lima, Radtke, and Drujan
Table H. Saturation Parameters of the Specific Binding of [3H]5HT Defined by 5HT to Goldfish Retinal Membranes
One site Two sites
Kd (nM)
B~,, (pmol/mg prot)
Hill coefficient
50.06 -4- 11.13 6.86 - 2.08 232.07 --- 95.34
23.78 --- 9.35 2.45 --- 0.60 53.46 --- 11.35
0.96 --- 0.01
The membranes were incubated with 20 concentrations of [3H]5HT, unspecific binding was defined by 10 ~M 5HT. Values are mean --SEM, n = 10. The parameters were calculated by the iterative analysis LIGAND (17).
I0
o
"o v
o
8
6
Z 0,-n T tO
4
~3
m -0.55
k~ Q.
0
L 0
30
60 Time
I
I
90
120
(rain)
Fig. 5. Association and dissociation kinetics of specific binding of [3H]5HT to goldfish retinal membranes. The association constant (-o-) was determined by incubating the membranes with 4 nM pH]5HT, unspecific binding was the counts in the presence of 10 ~M 5HT. The dissociation constant (-e-) was determined by the displacement of 4 nM [3H]5HT at equilibrium with 100 I-LM5HT. Inset: semilogatihmic plot of dissociation data.
Kinetic Experiments. The association of [3H]5HT reached the equilibrium within 30 min and the binding was stable up to 90 min (Figure 5). The apparent association constant (kobs) was 0.73 min-L The dissociation constant was determined by the addition of 100 ~M 5HT as displacer at 60 min of incubation, the dissociation rate constant (k2) calculated by the slope of the semilogarithmic plot was 0.24 min -1. The dissociation was not completed up to 120 min and remained in approximately 50% of the association. The equilibrium dissociation constant (I~) calculated by the association and dissociation rate constants was 2 nM (Figure 5). The semilogarithmic transformation of the dissociation data corresponded to a concave curve to the upper side (Figure 5, inset).
DISCUSSION The characterization of 5HT interaction with goldfish retinal membranes was not carried out previously, however, various reports suggest the presence of 5HT recognition sites in the retina of several species by physiological approaches (6,10,14,22,23). In addition, resuits obtained by radioligand experiments in mammalian retina suggested the presence of 5HT 1 and 5HT2 receptors in rabbit retina (4), and of 5HT1 receptors in bovine retina (5). As have been consistently reported the concentration of 5HT in teleost retina is elevated (1,2). A recent report (14) demonstrated the existance of three population of cells in goldfish retina able to uptake [3H]5HT, but only one subtype of amacrine cells containing 5HT, according to other authors (13). The high affinity component of [3H]5HT binding site in goldfish retina in the present report (Table 2) might be the representation of 5HT receptors expressed at the postsynaptic interaction site of the neuron that contain 5HT endogenously, and the low affinity site might be representing the uptake sites expressed in the three types of cells that transport 5HT. Another possibility is that Miiller cells could express a 5HT receptor with different affinity from the subtype present in retinal neurons. The fact that a concentration of 5HT of 10 izM, used to block 5HTx receptors in order to define specific binding (24,25), only blocked 50% of the binding of [3H]5HT to goldfish retinal membranes is indicative of the existance of a limited number of sites probably representing the population of 5HT receptors. The component of low affinity and high capacity demonstrated by the saturation experiments is also observed by the kinetic analysis of the present work (Figure 5). The inhibition of [3H]5HT binding by the serotonergic agents reported in Table 1 as IC5o are higher than the values reported for 5HT~ and 5HT2 receptors in the brain (26). This might be the result of the dual interaction of 5HT in goldfish retina and the impossibility of these agents to block the binding to the carrier of 5HT to membranes of other cell populations of the retina (14). The presence of a second recognition site for 5HT in goldfish retinal membranes is not according with the results reported by Osborne (5) in bovine retina in which he identified one site by saturation experiments in concentrations up to 20 nM of [3H]5HT. Mitchell and Redbum (4) also demonstrated the existance of a single binding site for [3H]LSD and [3H]spiperone in rabbit retina, however, the possibility of observation of two components is difficult due to exploration of a small range of ligand concentrations. It might also be possible that the interaction sites of 5HT in teleost retina differ from
[3H] Serotonin Binding Sites in Goldfish Retina
mammalian retina, in which the population of cells that specifically transport 5HT is lower than in fish (14). Selective agonists and antagonists of 5HT1A and 5HT2 receptors support a physiological function in the regeneration of receptive field properties (7,22). The selective 5HTIA agonist, DPAT (27) was not a potent inhibitor of [3H]5HT binding to retinal membranes (Table 1) probably because the number of 5HTIA receptors in the retina is very low and the selectivity of the inhibitor is masked by the component of low affinity. The lower IC5o observed for 5ttT and DMT suggests the presence of 5HT1 receptors ('Fable 1). It is interesting that PRP exhibited a relatively low IC5o, this drug has been proposed as a 5HT m antagonist (28). The monoamine uptake blockers used in this study were not able to inhibit [3H]5HT binding (Table 1), suggesting that the ligand is not interacting with these uptake systems or that the monoamine carriers in goldfish retina are insensitive to these inhibitors. However, in certain manner IMI is not the best inhibitor of 5HT transporter, and express a very complex interaction with chickens and pigs retinal membranes (29). The pseudo-Hill coefficients of the inhibitory curves were less than one, suggesting either negative cooperativity or the binding of the ligand to high and low affinity sites (Table I). However, the competition curves were all best fitted to a one-site model. The inhibition displayed by GppNHp indicates the presence of 5HT1 receptors, which are regulated by guanine nucleotides (30). The kinetic analysis of the binding shows the dissociation constant of the high affinity complex because the values of Kd calculated by these experiments are close to the values obtained by saturation experiments, and also because the conditions of inhibition only blocks the 50% of the total binding (Figure 5). The semilogarithmic plot of the dissociation analysis are suggestive of a complex interaction, such as two sites. The described results are evidence of a low affinity interaction of 5HT in the retina of goldfish and might suggest the presence of 5HT1 receptors. However, the observation of a second component of the binding makes difficult the analysis by simple competition experiments and by using an unspecified ligand such as the endogenous molecule, w h i c h was necessary to start the kinetic characterization of 5 H T b i n d i n g sites in the retina of goldfish.
ACKNOWLEDGMENTS This work was partially supported by IVIC. Isabelle Radkte was an Assistant Student of Centro de Estudios Avanzados, IVIC, from
995 the University of Basilea, Switzerland. We appreciate the help of Lic. P. Matus in the dissection of the retina. We also thank Mrs. M. Florez for secretarial work and Mrs. Dhuwya Otero for making the figures.
REFERENCES 1. Osborne, N.N. 1982. Evidence for serotonin being a transmitter in the retina pages 401--430,/n Osborne, N.N., (ed.) Biology of serotonergic transmission. John Wiley & Sons, Chichester. 2. Tornquist, K., Hansson, C., and Ehinger, B. 1983. Immunohistochemical and quantitative analysis of 5-hydroxytryptamine in the retina of some vertebrates. Neurochem. Int. 5:299-305. 3. Ehinger, B. 1982. Neurotransmitters systems in the retina. Retina 2:305-321. 4. Mitchell, C.K., and Redbum, D.A. 1985. Analysis of pre- and postsynaptie factors of the serotonin system in rabbit retina. J. Cell Biol. 100:64--73. 5. Osborne, N.N. 1981. Binding of [31-I]serotonin to membranes of the bovines retina. Exp. Eye Res. 33:371-380. 6. Blanzinsky, C., Ferrendelli, J.A., and Cohen, A.I. 1985. Indolamine-sensitive adenylate cyclase in rabbit retina: Characterization and distribution. J. Neurochem. 45:440--447. 7. Osborne, N.N. 1990. Effects of GTP, forskolin, sodium fluoride, serotonin, dopamine, and carbachol on adelynate cyclase in teleost retina. Neurochem. Res. 15:523-528. 8. 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, 1987. 9. Ghazi, H., and Osborne N.N. 1988. Agonist-induced stimulation of inositol phosphates in primary rabbit retinal cultures. J. Neurochem. 50:1851-1858. 10. Osborne, N.N., and Ghazi, H. 1989. Serotonin receptors and their involvement in Ca* § movilization and glycogenolysis in the rabbit retina Pages 159-175, in Redburn, D., and Pasantes-Morales H., (eds.) Extracellular and intracellular messengers in the vertebrate retina. Alan R. Liss, Inc., New York. 11. Brunken, W.J., and Daw, N.W. 1987. The actions of serotonergic agonists and antagonist on the activity of brisk ganglion celia in the rabbit retina. J. Neurosei. 7:4054-4065. 12. Daw, N.W., Brunken, W.J., and Parkinson D. 1989. The function of synaptic transmitters in the retina. Ann. Rev. Neurosci. 12:205-225. 13. Jaffe, E.H., Urbina, M., Ayala, C. and Chemello, M.E. 1987. Serotonin containing neurons in the retina of the teleost Eugerres plumieri. Vision Res. 27:2015-2026. 14. Marc, R.F., Liu, W-L, S., Scholz, K., and Muller, J.F. 1988. Serotonergic and serotonin-accumulating neurons in the goldfish retina. J. Neurosci. 8:3427-3450. 15. Lowry, O.H., Rosebroughh, N. J., Farr, A.L., and Randall, R.J. 1951. Protein measurements with folin phenol reagent. J. Biol. Chem. 193:265-275. 16. Sokal, R., and Rohlf, F. 1979. Biometria. 1st. Edition. Blume Edieiones, Madrid. 17. McPherson, G.A. 1985. Analysis of radioligand binding experiments: A collection of computer programs for the IBM PC. J. Pharmacol. Methods 14:213-228. 18. Munson, P.J., and Roadbard, D. 1980. LINGAND: A versatile computerized approach for characterization of ligand-binding systerns. Ann. Bioehem. 107:220-239. 19. Leatherbarrow, R.J. 1987. A non-linear regression data analysis program for the IBM PC (and true compatibles) pp. 1-91 Elsevier Science Publishers, Amsterdam. 20. Barlow, R.B. 1983. Biodata Handlig with Microcomputers. pp. 125-143. Elsevier Science, Amsterdam.
996 21. Scatchard, G. 1949. The attraction of proteins for small molecules and ions. Ann. N. Y. Acad. Sci. 51:660-672. 22. Daw, N.W., Jensen, R.J., and Brunken, W.J. 1990. Rod Pathways in mammalian retina. Trends Neurosci. 13:110-115. 23. Osborne, N.N., and Barnett, N.L. 1989. Serotonin levels in the rabbit retina are elevated following intraocular injection of forskolin. J. Neurochem. 53:1955-1958. 24. Peroutka, S.J., and Snyder, S.H. 1979. Multiple serotonin receptors: Differential binding of 3H 5-hydroxytrypamine, 3H Lysergic acid diethylamide and 3H spiroperidol. Mol. Pharmacol. 16:687699. 25. Salazar, M., and Lima, L. 1990. Serotonin receptors modulation in rat hippocampus by clonazepam treatment. Acta Cientifica Venezolana 41:134. 26. Peroutka, S.J., Schmitdt, A.W., Sleight, A.J., and Harrington, M. 1990. Serotonin receptor "Families" in the central nervous system: An overview in Pages 104-113, Whitaker-Axmitia, P.M., and Peroutka, S.J., (eds) The neuropharmacology of serotonin. The New York Academy of Sciences, New York.
Lima, Radtke, and Drujan 27. Gozlan, H., Mestikawy, S. El, Pichat, L., Glowinsky, J., and Hamon, M. 1983. Identification of presynaptie serotonin autoreceptors using a new ligand: 3H-PAT. Nature 305:140-142. 28. Hoyer, D., Engel, G., and Kalkman, H.O. 1985. Characterization of the 5HT1B recognition site in rat brain: Binding studies with (-) [125I] iodocyanopindolol. Eur. J. Pharmacol. 118:1-1112. 29. Karl, E., Tuomisto, L , Airaksinen, M.M., and Lcino, M. 1988. 3H Imipramine binding in the retinas of chickens and pigs. Exp. Eye Res. 47:679-688. 30. Nelson, D.L. 1988. Biochemistry and Pharmacology of the 5HT1 sertonin binding sites in: The serotin receptors. (Sanders-Bush, E. ed.) pp. 23-58. The Humana Press Publishers New Jersey. 31. Chou, T.C. 1974. Relationships between inhibition constants and fractional inhibition in enzyme-catalyzed reactions with different numbers of reactants, different reaction mechanisms, and different types and mechanisms of inhibition. Mol. Pharmacol. 10:235247.
