A novel glutamate agonist, TAN-950 A, isolated from streptomycetes

A novel glutamate agonist, TAN-950 A, isolated from streptomycetes

Europeun Journal of P~~aca~~~, 197 f 1991) 187- 192 @ 1991 Elsevier Science Publishers B.V. ~14-2999~91/$~3.5~ ADONIS ~14299~1~35~ 187 EJP 51878 To...

810KB Sizes 0 Downloads 31 Views

Europeun Journal of P~~aca~~~, 197 f 1991) 187- 192 @ 1991 Elsevier Science Publishers B.V. ~14-2999~91/$~3.5~ ADONIS ~14299~1~35~

187

EJP 51878

Tosbi Iwama ‘, Yasuo Nagai ‘, Norikazu Taxnura *, Setsuo Harada 3 and Akinobu ’ Biology Research ~~~at~es,

Nagaoka



’ Chemistry Research oratories and 3 ~ic~ob~~o~ Research L&oratories, Research and Deuelopment Division, Takeda Chemicui Industries, Ltd. Osaka, Japan Received 13 February 1991, accepted 5 March 1991

A novel antifungal ammo acid antibiotic, TAN-950 A ([S]-Z-amino-3-(2,5-dihydro-5-ox64isox~olyl)propanoic acid), was found to have affinity for three excitatory amino acid (EAA) receptors and to inhibit [3H]~-a~n~3-hy~oxy-5-methy~~x~l~ ~propionic acid ~~3H]AMPA), [~H]k~nate and f 3H]3~2-carbox~iper~n-~yl)propyl-l-phospho~e acid ((3H]CPP) binding ~mpetitively. It caused excitation of rat ~pp~ampal CA1 neurons in vitro, an effect that was antagonized by an AMPA~k~at~ antagonist, 6,7-di~tr~~noxaline--, 7 3-dione (DNQX). Chemical modi~cation of TAN-950 A brought about a large change in its ph~ma~lo~cal activity. Alkylation at the C-3 position of the isoxazolone ring markedly increased the ability to elicite neuronal firing. This agonistic effect was also antagonized by DNQX. The (R) enantiomer of TAN-950 A had increased selectivity for the N-methyl-D-aspartate (NMDA) receptor subtype. This selectivity was further enhanced by removal of the methylene group from the amino acid moiety. The most potent NMDA agonistic activity was observed with ~R]-~-a~n~2~2,5-~ydr~3-methyl-5-ox~ ~isox~lyl)a~tic acid. These derivatives of TAN-950 A might be useful agents for investigating the ph~a~o~o~~~ and physiolo~cal roles of EAA receptors. TAN-950 A ([S]-2-smino-3-(2,5-dihydro-5-oxo4-isox~olyl)propanoic Kainate; AMPA (u-a~n~3-hydroxy-5-methy~soxazol~~propio~c

1. tn~uction Excitatory amino acids (EAAs), especially L-glutamate, are regarded as one of the main neurotrans~tters in the vertebrate central nervous system ~Col~ng~dge and Lester, 1989). Recently, many reports have suggested that EAAs also play a key role in memos, learning (Cotman et al., 1988) and even neuronal damage (Mayer and Westbrook, 1987; Greenamyre et al., 1988). There are at least three subtypes of PAA receptors; a-a~n~3-hydroxy-5-methy~~xazole-~propionic acid (AMPA)-, kainate (I&%)- and N-methyl-~-aspartate (NM~A)-prefe~ng sites {Wat~ns and Evans, 1981), and each has a unique pharma~olo~cal profile (Monaghan et al., 1989). The AMPA and/or KA sites mediate fast synaptic responses, and NMDA sites are thought to play a fundamental role in synaptic plasticity (Collingridge and Singer, 1990). Selective agonists and antago~sts are useful for both investigation of the physiological functions of the EAA receptor subtypes

Correspondence to: T, fwama, Biology Research Laboratories, Research and Development Division, Takeda Chemical industries, Ltd., Jusohonmachi, Yodogawa-ko, Osaka 532, Japan.

acid); Excitatory amino acid: Hippocampus; Glutamate; acid}; NMDA (N-methyl-D-~partate); (Receptor binding}

and development of their functional m~ifiers (Johnson, 1989; Meld~m, 1985). TAN-950 A, [S]-2-a~no-3-(2,5-dihydro-5-oxo-~isox~zolyl~pro~a~oic acid, is a novel antifung~ antibiotic isolated from culture filtrates of ~f~e~fu~yces ~~~fensis A-136. In the present study, we investigated the af~~ty of this new compound for EAA receptors and its structure-activity relations~p by using receptor binding assays. We also cart-id out an electrophysiolo~cal evaluation using rat ~pp~ampal slices in vitro. A prel~~n~ des~~ption of this work has already been presented (Iwama et al., 1990).

2. patents

and methods

[ ‘H]ff-Amino-3-hydroxy-5~methylisox~ole-4-propionic acid (AMPA; 27.6 Ci/mmol), f3H]kainate (KA; 60.0 Ci/mmol} and ~3H]3-(2-carboxypiper~n~-yl~pr~ pyl-l-phosphonic acid {CPP; 24.0 Ci/~ol} were purchased from New England Nuclear Corp. (Boston, USA). CPP and 6,7-dinitroquinoxaline-2,3-dione (DNQX) were synthesized in the Chemistry Research Laborato~es of Takeda Chem. Ind. L-Glutamate and

A.

and dl-1-amino-?-phosphowere ~~r~~as~~ from Sigma

ate

were

used for

distilled water and then centriThe super~ata~t and the s&tam pelilet were g for 20 min. The

x g for 10 min. These ur times in order tu wash

8.

0.1 mh4 Giutamate

1mM

0.3 mM l-l i mio

H9SS).The CSMs were cent~f~ged at 48000 X g for 10 min. They were then wasbed three times with 50 mM CI buffer at room temperature and finally suspe~ded in ice-cold buffer. An afiquot (1 ml) of the s~s~x~ded CSMs (0.2 - 0.4 mg protei~~tube) was inboth in the presence and s incubation was terminated by the adtion of 3 ml of ice-cold The bound complexes ion throu~ Whatman then washed twice with 3 ml of ice-cold buffer. The rad~oacti~ty of these receptorligand c~~~~exes was counted by scintillation spectroscopy. For the adding study, the procedures were alsame as those of the AMPA birding study; .~~ Tutor X-100 nor 100 mM KSCN was 13H]KA was added instead before the period. binding study, 10 nM [ 3H]CPP was aided and the incubation was carried out for 15 min. her pr~~dures were si~~ar to those used for KA ding except for separa~on of the bound complexes. These were centrifuged at 30000 x g for 10 min, and e pellet was then washed with ice-cold buffer and cent~f~ged once more. ific binding was deter~ned in the presence ~-g~ut~ate. IC,, values of the test comestimated by compute~sed ~~~~robit analvalue represents the mean of thnee

Fig. 1. Unit discharge of a rat ~pp~~p~ CA1 ~eu~u in vitro. (A) A unit d~~h~~e was elicited on a po~~a~ioo spike in the rat ~p~ampal CAI area in vitro. The arrow head represents the artifacts of the stimuius. (B) Effect of glutamate on the firing rate of a neuron in the CA1 area. Ordinate: spike frequency; abscissa: perfusion time. The dice in the bath was perfused with L-glutamate for 2 min. every IO min. The dotted line represents I/s (20 spikes/20 s).

Male Wistar rats (250-350 g) were decapitated and ~ppocampa1 slices (400 pm thick) were prepared {AIger et al., 1984) using a brain slicer (No~yama Rika). The slices were incubated for at least 1 h at 32-33OC in a Krebs-Ringer solution of the following composition (in mN): NaCI 124; KCI 5; KH,FO, 1.24; MgSO, 1.3; CaCl, 2.4; NaHCO, 6; glucose 10; bubbSed with 95% OT5% CO2 gas. A single slice was transferred to a r~ording chamber and perfused with the Krebs.~nger solution at a rate of ca. 1.8 ml/min at 32-33” C, A glass pipette containing 2 M NaCl (5-10 MSZ)was inserted into the pyramidal layer of the CA1 area to record neuronal activity. A bipolar stimulating electrode was placed in the stratum rad~~tum to evoke a population spike. When the electrode monitoring the population spike was carefuully moved, a unit discharge from a single neuron was observed superimposed on the population spike {fig. 1). Stimuli were then stopped. Spontaneous firings were counted every 20 s with a spike counter (DSE-325A; Dia Medical System Corp.} and recorded with a lineur recorder (Recti Horiz 8K; San-ei). Neurons that fulfilled the following criteria were analyzed;

189

DNQX lD0 &I

TAN-QSOA

tAM450A

1

mM

0.3 m&4

TAN-@WA

TAN-SSOA

TAN-950~

? milt

TAN-950A

Fig. 2. Excitatory effect of TAN-950 A on rat ~pp~~pa~ CA1 neurons in vitro. The slice in the bath was perfused with TAN-950 A for ca 3 min. DNQX (A) was applied with TAN-950 A and antago~zed the excitatory effect of TAN-950 A, but CPP (B) did not.

TABLE 1 Receptor binding and neuronal excitatory effect of TAN-950 A and other excitatory ammo acids. Compound

Binding study a K&I. PM

TAN-950 A L-Glutamate Quisqualate AMPA

KA

NMDA

Excitation of neurons b i3HlKA

13H1CPP

MEC (llM)

Subtype

IsHlAMPA 0.28 0.24 0.0086 0.036 5.7 =-to0

3.6 0.11 0.028 87 0.0026 >lOO

19 0.98 26 1100 > 100 14

30e 300 1 1 0.1 10

Q/K Q/K Q/K Q/K Q/K

N

N

of the a Each value is the mean of three experiments. b Under ‘Excitation of neurons’, MEC represents mi~mum effective ~neentration compound to excite rat bippocampal CA1 neurons in vitro. Under ‘subtype’, Q/K and N indicate that the excitatory effect of the compound was antagonized by DNQX and AP’I/CPP, respectively.

TABLE 2 Structure-activity relationships of TAN-950 A derivatives with respect to EAA receptor binding in rat brain. No.

Chemical strncture B Ri

RZ

Binding study &I, FM [ 3H]AMPA

i31.rlK.A

[ 3H]CPP

TAN-9X?A -H Deriooriws of TAN-950 A a. 4 b. -H c. -CH, d. -H ’ Basic strttcture is

-CH,CHNH$OOH -CH~CHNH~COOC~~ -~H~CH(N(CH~)~)CG~~ -(CH,),CHNH+ZQOH -CHNH,CGOH

0.28 >lOO >lOO 100 ZlOO

3.6 >I00 >I0 89 Z-100

19 =-loo ,100 20 47

S~~~~~a~~~~~~~ ~~a~~onsbi~s of TAN-950 A derivatives with respect to EAA receptor binding and neuronal excitation in rat brain. Excitation of neurons

Binding study IC5,,. BM [~H]AMPA 0.28

3.6

19

0.30 0.67 0.11 0.31 3.8

17 17 3.8 57 67

40 35 13 >I00 >100 11

15

x Basic structtm is

MEC

Subtype

300

Q/K

IJH]CPP

[-‘H&4

26

1 1 O-3 0.3 3 30

Q/K Q/K Q/K Q/K Q/K Q/K

CH~CHNH~C~H

(I ) they must have a spontaneous firing rate of less than one per second (I/s), (2) they do not respond to less than 0.1 mM glutamate, and (3) they respond to more than 0.3 mM glutamate. Test compounds were then added to the perfusing solution. The minimum effective concen~atio~ (MEC) of a test compound was deTexas as the concentration ai which it elicited firing at a rate of more than I/2 (the dotted line in fig. lI3). At least three neurons in different slices were used for one test compound. To determine the preferred subtype site of a test combed for the excitation of neurons, the effects of ~tago~sts were investigated. When the firing rate elicited by the test compound was decreased to less than I/s by AP7 or CPP, we considered the agonistic action to be mediated by the NMDA site. Lack of effect by the NMDA receptor ~tagonists but depression to a similar extent by DNQX indicated that the agonistic action was mediated mainly by the ~~~PA/~ sites.

3. Results

TAN-950 A had relatively high affinity for all three subtypes of EAA receptors (table 1). It had an e:
TABLE 4

St~c~u~~cti~~~~ reiations~psof TAN-950 A derivatives with respect to EAA receptor binding and neuron& excitation in rat brain. No.

Chemical structure a

Binding study

RI

fCs,. FM

R2

[ 3H]AMPA

Excitation of neurons 13HKA

13HJCPP

MEC @MI

Subtype

T/&V-950A

k.

1. m. n. 0.

a Basic

-H

-CH~CHNH~C~H

-H -H -CH, -CH$IH3

-CH2C*HNHzCCX3H -C*HNH,COOH -C*HNH,COOH -C * HNH,COOH -C*HNH,COOH

-CHKH,),

0.28 14 >lOO >loo >lGQ >loB

3.6 37 >lflO >lOO >lOO 33

19 6.2 5.8 8.3 7.9 2.1

300 100 30 10 30 >3oO

Q/K N N N N

structureis the same as that in table 2. C*: the asymmetric carbon has the R con~~uration, while TAN-950 A has the S configuration.

191

2). Elongation (c in table 2) or shortening (d) of the chain at the C-4 position of the isoxazolone ring (R, side-chain in table 2) decreased the binding affinity for AMPA and KA sites, while the affinity for NMDA sites was not markedly affected. Alkylation at the C-3 position of the isoxazolone ring (R, side-chain in table 2) greatly increased the potency with which the compound elicited neuronal firing, without having a large effect on the binding affinity for EAA receptors (e-j in table 3). The most potent and selective AMPA agonist in this series was cyclopentylTAN-950 A (h in table 3). The enantiomer of TAN-950 A ((R)-TAN-950 A; k in table 4) had a slightly decreased binding affinity for AMPA and KA sites and a m~erately increased binding affinity for NMDA sites compared with TAN-950 A. The excitatory effect of (R)-TAN-950 A was mediated by NMDA sites. Removal of the methylene group from the amino acid moiety, ((R)-Nor-TAN-950 A; 1 in table 4), resulted in a decrease in binding affinity for AMPA and KA sites and an increase in potency to excite neurons via NMDA sites. Methylation (m) at the C-3 position of (R)-Nor-TAN-950 A (R, side-chain in table 4) increased the agonistic potency, but ethylation (n) did not. Isopropylation resulted in relatively high bin~ng affinity for NMDA sites (o in table 4}, but did not cause excitation of neurons even at a concentration of 300 FM.

4. Discussion TAN-950 A is a novel amino acid isolated as a microbial product. In the present study, we investigated its binding affinity at EAA receptors and its potency to excite neurons of the rat brain in vitro. TAN-950 A showed relatively high l~inding affinity for three EAA receptors, both a primary a-amino group and a free carboxyl group being essential for EAA agonist action (table 2). An R2 side-chain (the substituent at the C-4 position of the isoxazolone ring) with a three carbon chain length was the most effective for binding to AMPA/KA reco~tion sites (table 2). Despite the fact that TAN-950 A possessed a high binding affinity, it showed a markedly lower potency to excite neurons than the other EAAs. L-Glutamate also showed the same discrepancy between a high binding affinity for EAA receptors and a low potency to excite neurons (table 1). This discrepancy for L-glutamate is likely to be due to uptake inactivation because it shows high affinity for the uptake sites (Foster and Fagg, 1984). TAN-950 A might be similar, although further studies are needed to clarify this aspect. Alkylation at the C-3 position of the isoxazolone ring (R, side-chain in table 3) markedly potentiated the activity to excite neurons. Cyclopropyl-TAN-950 A (g)

had almost the same binding affinity as TAN-950 A, but its ability to excite neurons was 1000 times higher than that of TAN-950 A. Substitution with a sterically hindered group (i) or an aromatic group Q) resutted in a lower binding affinity than that of TAN-950 A, but the potency with which these compounds excited neurons was still high. A bulky R, side-chain potentiated the efficacy with which the compound excited neurons (Watts et al,, 1990). This suggests that a hydrophobic pocket might exist in tbe AMPA recognition site (Christensen et al., 1989; Keinanen et al., 1990). Cyclopentyl-TAN-950 A ([S]-2-amino-3-(2.5dihydro-3-cyclopentyl-5-oxo-4-isoxazolyl)propanoic acid) was the most potent and selective AMPA agonist in this series (h in table 3). (R) Configuration of the asymetric carbon in the R, side-chain is probably necessary for NMDA agonistic activity. The most effective length of the R2 side-chain was found to be two carbons (Hansen and KrogsgaardLarsen, 1990), and this correlates well with the report that NMDA is more potent than N-methyl-D-glut~ate (Watkins, 1962). As to alkylation at the C-3 position, only methylation potentiated the activity for NMDA sites. This alkylation greatly restricted activity at the NMDA site compared to the activity of tbe (S) enantiomer at the AMPA site (Olverman and Watkins, 1989). The most potent and selective NMDA agonist in this series was thus [RI-2-amino-3-(2,5-dihydro-3-methyl-5oxo-4-isoxazolyl)ace:ic acid (m in table 4). This compound is relatively rigid, compared with NMDA, and is very similar to an extended conformation of NMDA: a information in which NMDA-like molecules bind to the recognition site (Magnuson et al., 1988). Modification of the natural product, TAN-950 A, has yielded more selective and potent EAA agonists. These compounds should be useful agents for investigating EAA receptor subtypes. Acknowledgements The authors would like to thank Mr. Koyo Gomaibashi for his excellent technical assistance and Dr. Donal G. Murphy ar.3 Mr. Jeffry A. Hogan for their advice.

References AIger, B.E., S.S. Dhanjal, R. Dingledine, J. Garthwaite. G., Henderson, G.L. hing, P. Lipton, A. North, P.A. ~hw~~roin, T.A. Sears, M. Seagal, T.S. Whittingham and J. Williams, 1984, Brain slice methods, in: Brain Slices, ed. R. Diigiedine (Plenum Press, New York) p. 381. Christensen, LT., A. Reinhardt, B. Nielsen. B. Ebert, U. Madsen, E.O. Nielsen, L. Brehm and P. Krogsgaard-Larsen, 1989. Excitatory amino acid agonists and partial agonists, Drug Design Deliv. 557. Colling~dge, G.L. and R.A.J. Lester, 1989, Excitatory amino acid receptors in the vertebrate central nervous system. Pharmacol. Rev. 40. 143.

Kww~en. K.. W. VVisdcn. B. Sommer. P. Werner, A. Herb. T.A. Vrrdwwn. B. Sakrnann and P.H. Seehurg. 1990, A family of .AS1PA-sclwtiw plutamatc receptors. Science 249, S5h. Sfa~nuson. D.S.K.. K. Curry. M.J. Peet and McLcnnan. 198s. Structural ~u~~~~e~ts for activation of excitatory amino acid receptors in the rat spinal wrd in vitro. Exp. Brain Res. 73, 541.

Mayer. M.L. and G.L. Westhrook, 1987, Celiuiar m~~a~srns underlying excitotoxicity, Trends Neurosci. 10. 59. Mrfdrum, B.. 1985. Possible therapeutic applications of antagonists of excitatory amino acid neurotransmitters. Clin. Sci. 68. 113. Monaghan. D.T.. R.J. Bridges and C.W. Cotman. 1989, The excitntory amino acrd receptors: Their classes, pharmacology and distinct properties in the function of the central nervous system, Ann. Rev. Pharmacol. Toxicol. 29. 365. Murphy. D.E.. E.W. Snowhill and M. Williams, 1987. Characterization of quisqualale recognition sites m rat bram tissue using DL-I”Hju-amino-3-hydroxy-S-methyIisoxazole-~~propionic acid (AMPA) and a filtration assay, Neurochem. Res. 12, 775. Olverman, H.J. and J.C. Watkins. 1989. NMDA agonists and competitive antagonists, in: The NMDA Receptor. eds. J.C. Wa~ins and G.L. Col~in~ridge (Oirl Press. Oxford) p. 19. Watkins, J.C., 1962, The synthesis of some acidic amino acids possessing neuropharmacological activity, J. Med. Pharm. Chem. 5, 1187. Watkins. J.C. and R.H. Evans. 1981. Excitatory amino scid transmitters, Ann. Rev. Pharmacol. Toxicol. 21, 165. Watkins, J.C.. P. Krogsgaard-Larsen and T. Horurc 1990, Structureactivity relationships in the development of excitatory amino acid receptor agonists and competitive antagonists, Trends Pharmacol. Sci. 21, 25.