Marine Toxins - American Chemical Society

Chapter 7 Nicotinic Acetylcholine Receptor Function Studied with Synthetic (+)-Anatoxin-a and Derivatives 1

2

3

K. L. Swanson , H. Rapoport , R. S. Aronstam , and E. X. Albuquerque 1

Downloaded by IOWA STATE UNIV on April 10, 2017 | http://pubs.acs.org Publication Date: January 29, 1990 | doi: 10.1021/bk-1990-0418.ch007

1

Department of Pharmacology and Experimental Therapeutics, University of Maryland School of Medicine, Baltimore, MD 21201 Department of Chemistry, University of California, Berkeley, CA 94720 Department of Pharmacology and Toxicology, Medical College of Georgia, Augusta, GA 30912 2

3

( + ) - A n a t o x i n - a is the most potent and most stereospecific nicotinic acetylcholine receptor agonist thus far identified. It is also highly selective for nicotinic receptors over muscarinic receptors. T h e molecular parameters which influence the binding affinity, channel activation, channel blockade, and receptor desensitization are being studied. Modifications o f the carbonyl and amine moieties can reduce o r nearly eliminate the receptor agonist potency o f the compounds and also determine the channel blocking characteristics. T h e physiology and pharmacology o f the nicotinic acetylcholine receptor ( A C h R ) have historically been beset with complications. T h e natural neurotransmitter acetylc h o l i n e ( A C h ) binds not only to the nicotinic receptor but also to many types o f muscarinic receptors. This has made it difficult to study the functional effects o f nicotinic receptors i n neuronal systems. O n e means to avoid the conflict o f m u l t i ple receptor types was to study the neuromuscular j u n c t i o n because the only postjunctional receptors present are nicotinic. Alternatively, A C h was used i n the presence o f a selective antimuscarinic agent such as atropine; unfortunately, atrop i n e also has noncompetitive antagonist effects at the n i c o t i n i c receptor (7). Furthermore, because A C h is rapidly hydrolyzed by acetylcholinesterase ( A C h E ) , a n t i - A C h E agents such as neostigmine were c o m m o n l y used, but these agents also have noncompetitive effects at the A C h R (2). T h e A C h R antagonist abungarotoxin ( a B G T ) was an excellent pharmacological t o o l for the localization o f receptors o n the muscle membrane because o f irreversible binding to a site overlapping the nicotinic agonist site. However, a B G T does not b i n d to the same site i n the central nervous system ( C N S ) as does (-)-nicotine (3). T h e debut o f the selective A C h R agonist (+)-anatoxin-a has provided a new t o o l for A C h R physiology and pharmacology. (-I-)-Anatoxin not o n l y has high affinity for the nicotinic A C h R but it also has high selectivity for nicotinic over muscarinic receptors i n the mammalian C N S . Recently, the use o f (+)-anatoxin-a was essential to the identification o f nicotinic receptors o n cultured neurons (4). W e are studying the features which allow it to bind with high affinity to the peripheral and central nicotinic receptors and the kinetic effects o n receptor conformational

0097-6156/90/0418-0107$06.00/0 o 1990 American Chemical Society

Hall and Strichartz; Marine Toxins ACS Symposium Series; American Chemical Society: Washington, DC, 1990.

108

MARINE TOXINS: ORIGIN, STRUCTURE, AND MOLECULAR PHARMACOLOGY

transitions. These studies are complicated by the need to assess noncompetitive A C h R antagonistic and a n t i - A C h E effects o f each analog. T h e single channel recording technique is particularly useful because the A C h E has been eliminated during isolation o f muscle fibers and because the kinetics can detect specific agonistic and antagonistic effects.


Discovery of Poisoning by Fresh Water Algal Blooms and Signs of Toxicity A l g a l blooms i n fresh water ponds occasionally poison livestock and waterfowl. A x e n i c cultures o f Anabaena flos-aquae N R C 44-1 were shown to produce the toxic principle (5) which can be present i n the algae and i n the water o f mature cultures (6). T h e discovery o f the toxin was fortuitous i n the sense that A C h R agonists do not have a (known) constructive function i n the algae; evolution o f the synthetic pathway was likely a by-product o f metabolic pathways i n the algae. T h e c o m p o u n d became evident only through its toxic effects o n other organisms. T h e toxin causes death rapidly i n mammals, birds and fish. T h e L D (i-P-> mouse) is only 0.2 mg/kg (6, 7). In each case paralysis by depolarizing blockade is the predominant sign: early muscle fasciculations i n mammals, prolonged opisthotonos i n birds, and muscular rigidity i n fish (5). D e a t h is due to respiratory paralysis. Poisoned calves were maintained by artificial ventilation for up to 30 hrs without sufficient recovery for independent survival. In vitro the toxin caused depolarization o f frog muscles and increased the spontaneous, vesicular release o f transmitter (8, 9). T h e active p r i n c i p l e o f the algal blooms was first extracted serially with ethanol, chloroform, and acidic water (5). In another method, a good recovery (73%) was achieved by freeze drying followed by acidic methanol and benzene—chloroform extractions; purification was by preparative thin-layer chromatography (6). T h e presence o f an enone i n the structure o f the toxin with absorpt i o n at 1670 c m ' was used i n final stages o f isolation (6). T h e chemical structure, 2-acetyl-9-azabicyclo[4.2.1]nonene, and its absolute configuration (Figure 1) were determined by X - r a y crystallography o f N-acetyl anatoxin-a (10). Subsequent synthesis from 1-cocaine confirmed the structure and stereochemistry (77). O t h e r c o m plex methods tackling the problem o f bridged alkaloids have synthesized racemic (7, 72, 75) o r optically active anatoxin-a (14). 5 0

1

Chemical Model of Nicotinic Agonists Structure-activity relationships form the basis o f receptor pharmacology and drug development research. A preliminary assay for nicotinic activation is to apply a drug to a muscle i n a bathing medium and measure the generation o f contractile forces. In o u r experiments we used the frog rectus abdominis muscle (Figure 2). S u c h an assay cannot be conclusive, i.e. it can also be influenced by desensitization and n o n competitive antagonism. This is due to m u l t i p l e drug binding sites o n the A C h R macromolecule (Table 1). M i x e d agonistic-antagonistic effects are characteristic o f many A C h R ligands. T h e agonist site o n the nicotinic receptor has i n general demonstrated only m i l d stereospecificity—in fact the agonist could be described by a small number o f characteristics. A positively charged group, usually an amine, is present w h i c h may be alkylated to varying degrees £75). A polar moiety, often a carbonyl group, forms a hydrogen b o n d at 5.9 or 6.0 A from the amine (16, 17). T h e polar group is also contained i n a planar region o f the agonist molecule (18). U s i n g synthetic enantiomers, we found that anatoxin-a is highly stereospecific w i t h the (+) isomer having 150-fold greater potency than the (-) isomer (Figure 2) (79). T h e semi-rigid nature o f anatoxin-a undoubtedly facilitates its stereospecificity.


SWANSON ET AL.

Nicotinic Acetylcholine Receptor Function

(+)Anatoxin—a

Anatoxin (R) — N—methyl— methylester anatoxinol


Figure 1. Structure o f (+)-anatoxin-a and two analogs w i t h and noncompetitive antagonist activity, respectively.

agonist

Log [Agonist] Figure 2. Potency o f anatoxin-a analogs to induce contracture i n frog rectus abdominis muscle. T h e data from two experiments are combined i n this figure. I n one, anatoxinmethylester was found to be equipotent w i t h carbamylcholine. I n the other, the anatoxin isomers were assayed against A C h [after cholinesterse i n h i b i t i o n by diisopropylfluorophosphate ( D F P ) followed by washing o f the preparation] (19). M a x i mal contracture was measured by depolarization w i t h K C 1 at the e n d o f each experiment.

Table I. Sites of Chemical Interaction on the Peripheral Nicotinic Receptor B i n d i n g Site and Specific Ligands

Site Localization o n Receptor

Pharmacological Effect

1.

agonist site A C h , (+)-anatoxin-a & aBGT

1 per alpha subunit, i n the extracellular domain

channel activation and depolarization

2.

phosphorylation site

delta subunit

desensitization

3.

high affinity site HTX, PCP

1 per A C h R i o n channel formed by a l l subunits

o p e n channel blockade desensitization

low affinity site chlorpromazine

many sites at proteinl i p i d interface

desensitization

4.

3

3

Desensitization is accompanied by increased agonist-receptor affinity and stimulation o f channel activity at l o w concentrations.



110


Stereospecificity suggests that there are three geometrically distinct determinants o f the structure which are important to agonist efficacy and potency, whereas previous models only utilized two features. A hydrophobic pocket may also be important (20); this could be supplied by the [4.2.1]nonene structure. O u r intentions are to explore the domains o f the agonist molecule to determine the specific determinants o f agonist activity. T w o regions o f the toxin are targets for this initial study: the carbonyl and the amine regions. T h e few compounds reported o n here do not provide a comprehensive evaluation o f structural parameters, but they do serve to introduce the variety o f questions which we anticipate w i l l be addressed by contrasts between a larger number o f compounds. In the first analog, anatoxin methylester, modification o f the carbonyl moiety altered the hydrogen bonding potential and the molecular dimensions, but may not have altered the amine—carbonyl separation o r the planarity o f the carbonyl region. Thus, by creating an ester, this relatively m i n o r modification could make the c o m p o u n d more similar to the natural transmitter than anatoxin. T h e other type o f alteration, reduction o f the carbonyl to an alcohol generated two analogs o f S and R configuration. T h e alcohol moiety w o u l d have significantly lower hydrogen bonding potential. Furthermore, the toxin loses planarity i n that region. A t the other target o f investigation, methylation o f the amine c o u l d also change the potency o f the toxin. In the related molecule, norferruginine, methylation increased the potency (21). Preliminary data indicate the N N - d i m e t h y l anatoxin may only have very weak agonistic properties along with antagonistic properties. A l s o , it is important that m u l t i p l e alkylation to form quaternary compounds could change the ability o f the analogs to penetrate C N S tissue. f

Structure and Function of the Nicotinic AChR T h e A C h R o f Torpedo o r immature muscle is a complex formed with 5 polypeptide chains, 2 alpha, and 1 each o f beta, gamma, and delta chains (22). E a c h alpha chain contains a binding site for A C h (and a B G T ) . A l t h o u g h the primary structures o f the alpha peptide chains are identical, they can be distinguished with antibodies because they have different 3-dimensional configurations in the protein complex (23). E a c h o f the other chains are similar, yet different enough to possess other binding sites (Table I). Several transmembrane alpha-helices o f each peptide line the i o n channel. It is necessary for two molecules o f agonist ( A ) to bind to the receptor ( R ) i n order to initiate a conformational shift from Jhe closed i o n channel configuration ( A ^ ) to the o p e n channel configuration (A2R ). In mature muscle the o p e n ionic channel has a conductance o f 3 0 - 3 2 pS; this is a constant property o f the receptor. [ A lower conductance state occurs with receptors found in immature o r denervated muscles (24-26).] T h e properties which depend u p o n the agonist are the rates o f binding and dissociation and the rates at which conformational shifts occur.

Agonist model:

2A + R ^ A + A R ^ A ^

A ^ *

U s i n g the cell-attached patch clamp technique o n frog muscle fibers (19), one can observe only two conditions: the open, conducting state o f the receptor and a nonconducting state o f u n k n o w n identity. T h e transitions behave according to stochastic principles; the lifetimes o f any particular c o n d i t i o n are distributed exponentially. T h e open state has a mean duration that is the inverse o f the rate o f channel closing. Because channel open time depends only u p o n a conformational shift, agonist concentration does not influence the parameter. It is, however, influenced


7. SWANSONETAL.


111

by the chemistry o f the two agonists which are b o u n d to the receptor and the transmembrane potential. T h e biophysical properties which determine the open duration are as yet to be determined.

Agonistic Effects of (+)-Anatoxin-a ( + ) - A n a t o x i n - a has high potency by virtue o f high affinity for the agonist binding site o f the receptor. Assay o f contracture potency demonstrated the overall relative potency o f a variety o f agonists (Figure 2). This data can be compared w i t h the concentrations o f the toxins which inhibit the binding o f the antagonist a B G T to Torpedo electroplaque receptors (Table II, a B G T I C ) and their ability to activate the i o n channel, reducing the barriers to the binding o f histrionicotoxin ( H T X ) and thus increasing specific binding (Table II, H T X E D ; see discussion o f H T X binding under Allosteric Antagonism of AChR Function by (+)-Anatoxin-a Analogs). W i t h agonists b o u n d to the receptor, the channel may o p e n and then close; after closure (while the agonist remains bound) the rates for channel reopening and agonist dissociation compete. W h e n A C h is the agonist, the dissociation rate is higher and therefore typical channel activity consists o f single, well-separated openings (Figure 3). W h e n (+)-anatoxin-a is the agonist, the dissociation rate is slower and the channel is likely to reopen; this c o n d i t i o n o f the receptor can be recognized experimentally by the presence o f short duration closures, "flickers". Thus with (+)-anatoxin-a, the flickers are significantly more frequent and channel activity consists o f several openings separated by only a fraction o f a millisecond. It is "pharmaco-gnomonic" o f agonists that the form o f the burst is independent o f agonist concentration. Thus, the average burst i n response to 20 o r 200 n M (+)anatoxin-a is the same; the individual open times (Figure 3), the total burst duration, the flicker durations, and the number o f openings per burst are all the same at these two concentrations (79). T h e rate o f i o n channel closure, which determines the mean open time, is more rapid for (+)-anatoxin-a than for A C h . Perhaps such comparison can facilitate understanding o f the biophysical properties which determ i n e the open duration. 5 0


5 Q

Desensitizing Properties of (-I-)-Anatoxin-a A n o t h e r characteristic o f agonists is the ability to desensitize the receptor, i.e., to make a p o p u l a t i o n o f receptors become non-responsive to further agonist. W h e n a high concentration o f agonist produces this effect it is k n o w n as depolarizing blockade. There are, i n addition to the agonist model drawn above, o n e or more desensitized states o f the receptor. Agonists, i n general, have a higher affinity for the alpha-subunit sites o f these desensitized states than o f the n o r m a l receptor. This effect occurs by the allosteric binding o f agonist o r desensitizing agent to allosteric sites o r by phosphorylation o f the delta-subunit protein (27). N o n c o m p e t i tive antagonists such as H T X and phencyclidine also p r o m o t e desensitization by allosteric mechanisms (25, 29). Because o f these allosteric mechanisms, desensitizing potency is not directly correlated with agonist potency. A l t h o u g h (-H)-anatoxin-a is m u c h more potent than A C h as an agonist, desensitization occurs more slowly with (-t-)-anatoxin-a than with an equipotent (as agonist) concentration o f A C h (79).

Allosteric Antagonism of AChR Function by (+)-Anatoxin-a Analogs Several drugs, the most well-known being local anesthetics and histrionicotoxin ( H T X ) (25), bind to an allosteric site o n the A C h R (relative to the agonist binding site). Biochemically, this site is identified by high affinity binding o f [ H ] H - H T X (Table I). It may be located at the i o n channel and coordinates between several o f 3

1 2



112


Table II. Binding Constants for (+)-Anatoxin-a and Related Molecules at Nicotinic and Muscarinic Receptor Sites Ligands

I C

(+)-Anatoxin-a Acetylcholine A n a t o x i n methylester (-)-Anatoxin-a (S)-N-Methylanatoxinol (7?)-iV-Methylanatoxinol

Rat Brain

Torpedo Electric Organ aBGT

HTX E D

50

0.085 0.30 2.4 4.4 >100 >100

HTX

*i

50

0.032 0.15 0.22 1.6

-

Scopolamine

—

-

320

—

117 8.5

9.3

-

0.33 4.4

N O T E : A l l values are i n / i M and are defined as follows: a B G T I C T Q = concentration producing 5 0 % i n h i b i t i o n at [ ^ I ] a B G T binding; H T X E D = concentration o f toxin producing 5 0 % stimulation o f [ H ] H T X binding; H T X K j = i n h i b i t i o n constant i n units o f toxin concentration w h i c h represent i n h i b i t i o n o f the binding o f [ H ] H T X which was enhanced by 1 / i M carbamylcholine; scopolamine K j = i n h i b i t i o n constant i n units o f toxin concentration which represent i n h i b i t i o n o f the binding o f [ H]scopolamine. 1 2

5 Q

3

3

3


7. SWANSON ET AL.



ACh

AnTX

"IT* XT

10 ms

Acetylcholine O 300 nM

(-f)Anatoxin—a •

20 nM •

200 nM

20.0

-160

-120 -80 Membrane Potential (mV)

-40

Figure 3. ( + ) - A n a t o x i n - a ( A n T x ) and A C h induced single i o n channel currents i n isolated frog muscle fibers. O p e n channels w i t h 32 pS c o n ductance are downward deflections (inward current at hyperpolarized potentials). T h e currents shown o n the left are all at o n e potential. T h e duration o f channel open events had a similar voltage-dependence for b o t h A C h and (+)-anatoxin-a. W i t h A C h , the events were most often singular, while with (+)-anatoxin-a the events were shorter and were m o r e frequently paired so that the mean duration o f the exponentially distributed open times and selected membrane h o l d i n g potentials was approximately one-half, independent o f the concentrat i o n o f the agonist applied.


113

114


the polypeptide chains (30, 31). This site was not thought to have stereospecificity (32, 33). T h e binding o f antagonists is dependent u p o n the presence o f agonist. I n the presence o f an agonist such as carbamylcholine, allosteric antagonists inhibit the binding o f H T X to the i o n channel site (Table II; H T X Kj). A sequential m o d e l o f i o n channel blockade is commonly used as a standard against w h i c h to characterize the effects o f blockers. O n c e i n the open configuration ( A ^ ), the drug ( D ) is able to b i n d and block the channel. T h e resulting state A ^ D has n o conductance; it can be distinguished from A ^ only by statistical analysis.


Sequential model:

A ^

A ^ * + D *=+ A ^ D

T h e primary characteristic o f a sequential blocker, as observed w i t h the patch clamp technique, is that the reciprocal o f the mean duration o f the lifetime equals the n o r m a l channel closing rate plus the rate constant o f channel blockade times the drug concentration. Therefore, increasing the drug concentration shortens the mean channel o p e n time. T h e (+)-anatoxin-a analogs (R)- o r (S)-N-methylanatoxinol, both possess the ability to block the i o n channel [the voltage-dependence o f their characteristics differ i n a way w h i c h is the topic o f another report (34)]. A s is often the case, the blocked state induced by (7?)-AT-methylanatoxinol and the closed state differed significantly i n duration, such that the blocked state was associated w i t h short closed periods w i t h i n groups o f openings called bursts (Figure 4). T h e number o f short closed periods per burst increased with the concentration o f the drug because the l i k e l i h o o d o f channel blockade increased relative to the l i k e l i h o o d o f channel closure. T h e duration for w h i c h the receptor remains i n the blocked state is a property o f the drug, because it is a simple dissociation reaction. It often depends u p o n v o l tage, although the strength o f the voltage relationship is a characteristic o f the drug. I n the case o f (i?)-iV-methylanatoxinol, the voltage dependence is insignificant; the mean short closed duration was a few milliseconds at a l l potentials. W e believe this may be due to a large degree o f hydrophobic binding utilizing van der Waals forces. I n contrast, the more polar S isomer binds rapidly but also dissociates rapidly due to reliance o n coulombic interactions. T h e slower dissociation o f the (/?)-N-methylanatoxinol isomer is the factor w h i c h determines the greater potency to inhibit H T X binding (Table II).

Neuronal Nicotinic AChR In the neuronal tissues, describing the function o f nicotinic A C h R has been considerably m o r e complicated than at the neuromuscular j u n c t i o n (35). T h e lack o f a suitable agonist w h i c h was selective for the A C h R was confounded by the difficulty that a B G T was also an unsuitable antagonist (3). Apparently, a B G T prefers to b i n d to a different receptor site i n the C N S than does nicotine. O n l y recently has optically active [ H ] n i c o t i n e become available to make selective agonist binding studies feasible. T h e Kj o f (+)-anatoxin-a for the i n h i b i t i o n o f (-)-[ H ] n i c o t i n e binding to rat brain membranes was 0.34 n M (36) and the I C for i n h i b i t i o n o f [ H ] A C h binding was 4.5 n M (37). (+)-Anatoxin-a was a 2- to 20-fold better ligand than (-)-nicotine, i n these respective studies. It also has high selectivity for nicotinic over muscarinic receptors; the Kj for i n h i b i t i o n o f a muscarinic antagonist, [ H]scopolamine, binding to rat brain synaptosomes was 9.3 / i M (Table II). ( + ) - A n a t o x i n - a acts presynaptically as an agonist i n the C N S , as evidenced by 3

5 0

3

3



^

j

^

5 0 fiU

^wn

( R ) - N - m e t h y l a n a t o x i n o l + A C h (400 nM) 200 /xM

>2S msec

IP

F i g u r e 4. Concentration-dependent i o n channel blockade by (R)-JV-methylanatoxinol. T h e patterns identified as bursts and separated by l o n g (>8 msec) closed intervals are indicated w i t h a bar, the figure was designed to show approximately 2 bursts per trace. T h e dose-related decrease i n mean channel o p e n time resulted from the blockade o f the o p e n channel by the (R)-N-methylanatoxinol. T h e channel amplitude is related to membrane voltage (as was given i n F i g u r e 3) by the slope conductance such that 1 p A is equivalent to 30 m V . Continued on next page.

n

Control




( o a S U l ) 9UIIJ, U 9 d Q

U^3J^


7. SWANSON ET AL.


117


the stimulation o f neurotransmitter release (38). In this case, 1 uM anatoxin was as effective as 10 uM nicotine. Aracava et al. (4) disclosed the presence o f nicotinic A C h R o n cultured, neonatal brain stem and hippocampal neurons. Single i o n channel currents were recorded from the base o f the apical dendrites o f pyramidal cells using the patch clamp technique with A C h o r (+)-anatoxin-a as the agonists. These nicotinic A C h R had l o w conductance and the openings were brief, with characteristics which are similar to those observed for embryonic muscle receptors. Also found i n retinal ganglion cells, the neuronal nicotinic receptors responded w i t h the kinetics o f i o n channel blockade, including enhancement u p o n hyperpolarization as seen w i t h peripheral receptors; the single channel currents were sensitive to phencyclidine and also showed a possible blockade by m i c r o m o l a r concentrations o f (+)anatoxin-a (39). W h i l e many details o f the function o f the central nicotinic receptors remain to be elucidated, the initial results suggest a large degree o f functional homology between peripheral and central receptors (40). Conclusion ( + ) - A n a t o x i n - a has already proven its usefulness as a research t o o l i n o u r laboratories. It is facilitating the understanding o f the biophysical properties o f the A C h R and o f the localization o f the A C h R i n the C N S . T h e toxin o r derivatives o f it c o u l d be useful therapeutically i n diseases o f nicotinic receptor pathology (myasthenia gravis or Alzheimer's disease), because as a secondary amine (+)anatoxin-a can penetrate into the C N S . Acknowledgments This w o r k was supported by N I H G r a n t N S 25296 and U . S . A r m y Research and Development C o m m a n d Contract D A M D 1 7 - 8 8 - C 8 1 1 9 .

Literature Cited 1. Adler, M.; Albuquerque, E. X.; Lebeda, F. J. Mol. Pharmacol. 1978, 14, 514. 2. Aracava, Y.; Deshpande, S. S.; Rickett, D. L.; Brossi, A.; Shonenberger, B.; Albuquerque, E. X. New York Acad.Sci.1987, 505, 226. 3. Clarke, P. B. S Trends in Pharmacological Sciences, 1987, 8, 32. 4. Aracava, Y.; Deshphande, S. S.; Swanson, K. L.; Rapoport, H.; Wonnacott, S.; Lunt, G.; Albuquerque, E. X. FEBS Letters 1987, 222, 63. 5. Carmicael, W. W.; Biggs, D. F.; and Gorham, P. R. Science (Washington, D.C.) 1975, 187, 542-544. 6. Devlin, J. P.; Edwards, O. E.; Gorham, P. R.; Hunter, N. R.; Pike, R. K.; Stavric, B. Can. J. Chem. 1977, 55, 1367. 7. Bates, H. A.; Rapoport, H. J. Amer. Chem. Soc. 1979, 101, 1259. 8. Biggs, D. F.; Dryden, W. F. Proc. West. Pharmacol. Soc. 1977, 20, 461. 9. Spivak, C. S.; Witkop, B.; Albuquerque, E. X. Mol. Pharmacol. 1980, 18, 384. 10. Huber, C. S. Acta. Cryst. 1972, B28, 2577. 11. Campbell, H. F., Edwards, O. E.; and Kolt, R. Can. J. Chem. 1977, 55, 1372. 12. Campbell, H. F.; Edwards, O. E.; Elder, J. W.; Kolt, R. J. Polish J. Chem. 1979, 53, 27. 13. Vernon, P.; Gallagher, T. J. Chem. Soc., Chem. Commun. 1987, 245. 14. Petersen, J. S.; Fels, G.; Rapoport, H. J. Am. Chem. Soc. 1984, 106, 4539. 15. Spivak, C. C.; Albuquerque, E. X. In Progess in Cholinergic Biology: Model Cholinergic Synapses; Hanin, I.; Goldberg, M., Eds.; Raven Press: New York, 1982, 323. 16. Beers, W. H.; Reich, E. Nature (Lond.) 1970, 228, 917.



118


17. Koskinen, A. M. P.; Rapoport, H. J. Med. Chem. 1985, 28, 1301. 18. Chothia, C.; Pauling, P. Proc. Natl. Acad.Sci.USA 1970, 65, 477. 19. Swanson, K. L.; Allen, C. N.; Aronstam, R. S.; Rapoport, H.; Albuquerque, E. X. Mol.Pharmacol.1986, 29, 250. 20. Waksman, G.; Changeux, J. -P.; Roques, B. P. Mol. Pharmacol. 1980, 18, 20. 21. Albuquerque, E. X.; Spivak, C. E. In Natural Products and Drug Development; Krogsgaard-Larsen, P.; Brøgger Christensen, S.; Kofod, H., Eds.; Alfred Benson Symposium 20; Munksgaard, Copenhagen, 1984; pp 301-321. 22. Stroud, R. M. Neuroscience Commentaries 1983, 1, 124. 23. Maelicke, A ; Fels, G.; Plümer-Wilk, R.; Wolff, E. K.; Covarrubias, M; Methfessel, C. In Ion Channels in Neural Membranes, Allan R. Liss, 1986, 275. 24. Allen, C. N.; Albuquerque, E. X. Exp.Neurol.1986, 91, 532. 25. Akaike, A; Ikeda, S. R.; Brookes, N.; Pascuzzo, G. J.; Rickett, D. L.; Albuquerque, E. X. Mol. Pharmacol. 1984, 25, 102. 26. Aracava, Y.; Ikeda, S. R.; Daly, J. W.; Brookes, N.; Albuquerque, E. X. Mol. Pharmacol. 1984, 26, 304. 27. Steinbach, J. H.; Zempel, J. Trends in Neuroscience 1987, 10, 61. 28. Changeux, J. -P.; Nevillers-Thiéry, A ; Chemouilli, P. Science 1984, 225, 1335. 29. Oswald, R. E.; Heidmann, T.; Changeux, J. -P. Biochem. 1983, 22, 3128. 30. Heidmann, T.; Oswald, R. E.; Changeux, J. -P. Biochem. 1983, 22, 2112. 31. Changeux, J. -P.; Revah, F. Trends in Neuroscience 1987, 10, 245. 32. Spivak, C. E.; Maleque, M. A.; Takahashi, K.; Brossi, A.; Albuquerque, E. X. FEBS Lett. 1983, 163, 189. 33. Henderson, F.; Prior, C.; Dempster, J.; Marshall, I. G. Mol. Pharmacol. 1986, 29, 52. 34. Swanson, K. L; Aracava, Y.; Sardina, F. J.; Rapoport, H.; Aronstam, R. S.; Albuquerque, E. X. Mol.Pharmacol.1989, 35, 223. 35. Rovira, C.; Ben-Ari, J.; Cherubini, E.; Krnjevic, K.; Roper, N. Neurosci. 1983, 8, 97. 36. Macallan, D. R. E.; Lunt, G. G.; Wonnacott, S.; Swanson, K. L.; Rapoport, H.; Albuquerque, E. X. FEBS Lett 1987, 226, 357. 37. Zhang, X.; Stjernlof, P.; Adem, A.; Nordberg, A. Eur. J. Pharmacol. 1987, 135, 457. 38. Lunt, G.; Wonnacott, S.; Thome, B. Neurosc. Abst. 1987, 13, 940. 39. Albuquerque, E. X.; Alkondon, M.; Lima-Landman, M. T.; Deshpande, S. S.; Ramoa, A. S. In Neuromuscular Junction; Fernström Foundation Series Vol. 13; Sellin, L. C.; Libelius, R.; Thesleff, S.; Eds.; Elsevier Science Publ.: Cambridge, U.K., 1988, pp. 273-300. 40. Aracava, Y.; Swanson, K. L.; Rozental, R.; Albuquerque, E. X. In Neurotox Elsevier Science Publ: Cambridge, U.K., 1988, pp. 157-84. RECEIVED May 17,

1989


Marine Toxins - American Chemical Society

Recommend Documents