The binding of aluminium to [Leu5]-enkephalin an investigation using 1H, 13C and 27Al NMR spectroscopy

Biochimica et Biophysica Acta, 717 (1982) 465-472 Elsevier Biomedical Press

465

BBA 21199

THE BINDING OF ALUMINIUM TO [LEUS]-ENKEPHALIN AN INVESTIGATION USING 1H, 13C AND Z~A!NMR SPECTROSCOPY HONORI~ MAZARGUIL a RAYMOND HARAN b and JEAN-PIERRE LAUSSAC b., a Laboratoire de Pharmacologie et Toxicologie Fondamentales du CNRS, 205, route de Narbonne, 31400 Toulouse and b Laboratoire de Chimie de Coordination du CNRS, associ~ h l'Universitb Paul Sabatier, 205, route de Narbonne, 31400 Toulouse (France) (Received April 5th, 1982)

Key words: Aluminum binding," Enkephalin; NMR; Leu'~-enkephalin

The binding of aluminium ion to [LeuS]-enkephalin has been investigated by tH, 13C and 27Ai NMR spectroscopy in dimethyl sulphoxide solution at different peptide/metai ratio. Analysis of the spectra suggests that A! 3+ binds at two metal-binding sites. The binding of A! 3+ at the first site involves the Tyr t CO, and LeuSCOO - groups to give a 2:1 species in a tetracoordinated structure; whereas the binding of AI 3+ at the second site utilizes the NH 2 terminal groups of the tyrosine moiety in a 2:2 species. 27A! chemical shift values strongly suggests that the second type of aluminium atom displays an octahedral environment. On this basis, we discuss our data in terms of the coordination of aluminium with [LeuS]-en kephalin.

Introduction A number of metal ions are known to be vital in the normal functioning of the nervous system. They can act as cofactors for many enzymes; they may also be involved in neurotransmitter synthesis, storage and transport [1-3]. But there is considerable evidence that certain trace metal ions could be hazardous for the central nervous system and may play a key role in the aetiology of a variety of neurological disorders [4]. Among these metals, it is now well established that aluminium is harmful to the central nervous system. Indeed, elevated concentrations of aluminum have been observed in the brains of patients with dialysis dementia [5-6]; in senile and presenile dementia of the Alzheimer type [7] and in amyotrophic lateral sclerosis-Parkinson dementia [8] and have

* To whom correspondence should be addressed. 0304-4165/82/0000-0000/$02.75 © 1982 Elsevier Biomedical Press

been implicated in the aetiology of these diseases. Little is known about the effects of aluminium on the central nervous system and how these effects may be related to the overall neurotoxicity of aluminum observed in vivo. A plausible mechanism of action may imply aluminium interference with some essential reactions [9-10]. On the other hand, the biological activity of endogenous central nervous system peptides (enkephalins) with morphine-like action has been shown to be modified by metal ions. Thus, manganese increases the apparent affinities of both natural enkephalins for the opiate receptor [11]. Previous studies of the interactions of the [Leu5] and [Met 5] enkephalin and its analogues with metal ions have focused at the Mn2+ , Cu 2+ , Zn 2+ , Ca 2+ and Mg 2+ complexes [12-15]. The introduction of Cu 2+ into the peptide structure of both natural peptides does not perturb the binding of the polypeptide with its receptor. Moreover, the biological activity of [MetS]-enkephalin in the

466 guinea pig ileum assay is somewhat enhanced as a consequence of complexation with Cu 2+ . This could be explained by the strengthening of the biologically active fl-turn configuration upon complexation [15]. The binding of Zn 2+ also shows a rigidification and a stabilizing effect of these peptides in a particular conformation. Because of the apparently important role of metals in the activity of enkephalins, we have investigated, by ~H, 13C and 27A1 NMR spectroscopy, the binding of AI 3+ with [LeuS]-enkephalin. The spectra provide information on the nature of the binding sites, and we now report results showing two major type of [LeuS]-enkephalin/A13+ interactions. Materials and Methods

The synthesis of [LeuS]-enkephalin has been described previously [16]. The peptide was lyophilized twice from 99.8% 2H20, then dissolved in (C2H3)2SO (99.8%) obtained from Commissariat /~ rEnergie Atomique (CEA, France) at a concentration range of 10 -2-10 -3 M. AI(NO3) 3 • 9 H20 of highest purity was purchased from Merck. Metal salt titration of the ligand NMR spectra were prepared by adding the appropriate amount of the metal ion as its nitrate to the NMR sample. N M R parameters. 1H and 13C NMR spectra were obtained with a Bruker WH-90 spectrometer operating in the Fourier transform mode using a Bruker-Nicolet data system, model B-NC 12. Field stabilization was provided by an internal deuterium lock signal. The temperature of the probe was controlled within ±2°C with a Bruker B-ST 100/700 temperature control unit. 13C (22.63 MHz) sample volumes were generally 1 ml, contained in a 10 mm NMR tube with a teflon plug for vortex suppression. All spectra were recorded under conditions of full proton decoupling at an ambiant probe temperature of 27°C. 1H and 13C chemical shifts are reported in ppm downfield from 3-(trimethylsilyl)-l-propanesulphonic acid. 27A1 NMR spectra were recorded on a Bruker WH 90 spectrometer operating at 22.63 MHz using ~3C equipment and on a Bruker WM-250 spectrometer operating at 65.14 MHz using a sweepwidth of 50 kHz and 16K memory. At 22.63

MHz, the magnetic field was lowered down to 2.04 T to find the 27A1 resonance. No field-frequency lock could be used under these conditions; consequently chemical shifts were measured with an error of ± 2 ppm. Chemical shifts are expressed in ppm downfield from the resonance of AI(H20)63+. Line widths in Hz were measured at half-height (vl/2)-

Results and Discussion i H N M R spectroscopy

The 1H NMR spectra of metal-free [Leu~]-enkephalin and its aluminium complexes have been obtained; the values of the chemical shifts are given in Table I. The resonances have been assigned by homonuclear spin decoupling experiments and are consistent with previous work [17]. Gradual additon of A13+ to enkephalin results in changes throughout the whole 1H NMR spectrum. The magnitudes of the chemical shift perturbations induced by addition of A13+ increases with increasing metal ion concentration. From this observation, we infer that the ligand is in fast exchange on the 1H NMR chemical shift time scale between its free and complexed states. This fact exchange makes possible the unequivocal assignment of the proton resonances for the aluminium complex. In order to establish the stoichiometry of the complex, we determined the relative magnitude of the chemical shifts at various ligand/metal ratios. A large change in chemical shift is observed up to a ratio of 2:1; at higher ligand/metal ratios, the NMR spectrum remains unchanged (Table I). These data clearly point to the formation of a 2:1 complex (see 13C results). The protons NMR chemical shifts are dependent on the electron density at the nucleus, and the coordination of a diamagnetic metal center at or near the proton results in inductive electron withdrawal. This effect is similar to the shifts resulting from protonation which causes downfield shifts of the adjacent proton resonances. However, conformational modifications can produce either shielding or deshielding effects. But, generally the results can be reasonably interpreted by assuming that changes in electron density alone are responsible for the observed effects. As expected, down-

467 TABLE I IH CHEMICAL SHIFTS AND TEMPERATURE COEFFICIENT (dS/dT) OF [LEUS]-ENKEPHALINand [LEUS]-ENKEPHALIN/AI3+ IN (C2H3)2SO. n.o., not observed. Residues

Proton Atoms

2: 1 a

2 :2

dS/dT (ppm/°C)(× 103) free peptide

2: 1

12.4

4.66

NH 2 OH

3.43 2.59 2.86 n.o. fLo.

n.o. 2.76 2.78 n.o. 9.38 (30)b

3.95 2.76 2.78 8.04 9.36 (3)

t~

3.63

NH

8.39

3.78 3.79 8.70

3.78 3.79 8.75

Gly3

a

3.67

NH

7.80

3.60 3.61 8.09

3.60 3.61 8.12

2.66

4.33

Phe 4

a j8 NH

4.39 2.79 3.07 8.31

4.54 3.02 3.04 8.1 !

4.57 3.02 3.04 8.06

5.66

5.33

a fl 8

3.97 1.55 0.86

NH

7.72

4.17 1.54 0.86 0.91 8.26

4.19 1.54 0.84 0.90 8.37

2.33

6.66

Tyr

Gly2

Leu5

a

Free peptide

a Pcptide/metal ratio. b Values incidentally correspond to the line width in Hz. field shifts are observed in the resonances arising from a - C H groups in the ligand. The largest changes observed for the 2:1 complex occur for the proton resonance of the leucine residue, where the chemical shift difference is A = 0.20 ppm. It is difficult to give a value for the ATyra-CH resonance because the T y r a - C H peak overlaps with the ABX multiplet of Gly 2 and Gly 3. But a dramatic change in chemical shift occur in the 2: 2 system (A = 0.52 ppm). Both giycine protons are of A 2 type in the free ligand whereas Gly 2,3 C H 2 protons are of the ABX type upon addition of the metal. The nonequivalent nature of the two glycine (Gly 2 and Gly 3) methylene protons is an indication of a constrained conformation [18]. The rather large nonequivalence of the Leu 5 ~-CH 3 resonances are due similarly to a change in the conformation upon aluminum binding. As shown in Table I, all of the exchangeable

primary amide resonances can be accounted f o r , as well as the resonances from the O H phenolic group at approx. 9.36 ppm, and the Tyr I N H 2. But the latter resonances are not observed in metal-free peptide because they undergo fast exchange with water contamination in (C2H3)2SO. In the 2:1 system the Tyr I N H 2 protons are not observed, but they are seen in the 2 : 2 system indicating that the terminal N H 2 groups are not exchanged rapidly by deuterium. This observation is consistent with coordinated N H 2 groups [19]. It can be outlined that the TyrlOH, which in the free peptide also exchanges with the TyrlNH2, comes out sharply when the terminal amino groups are metal-bound, particularly in the 2 : 2 system.

Temperature dependence of proton chemical shifts The behaviour of the peptide N H resonance as a function of the temperature was then investi-

468

At3÷

o

t,.66

o%

Giy z

%. 12./~

o ~

o-"~'~....~.~,~ '~o

~ •~ 0 °~°~.

2.33

6.66

LeuS

, ~

•

~

~ o ~

e

4.33

2.66 Gly3

posed peptide NH. Addition of aluminium induces large changes in the temperature dependence of chemical shifts for the different NH proton resonances (Fig. 1). But, the most obvious change in temperature coefficients occurs in the LeuSNH resonance. Thus dS/dT-~ 2.33- 10 - 3 p p m / ° C in the free peptide and 6.66. 10 - 3 p p m / ° C upon metal binding. These results may be explained by a process in which the hydrogen bond between the CO of Gly 2 and NH of Leu 5 is disrupted by interaction with aluminium. Furthermore, the Gly 2 d S / d T is reduced from 12.4- 10 - 3 t o 4.66. 10 - 3 p p m / ° C upon complexation. Thus, the breakdown of the fl-turn which exposes the GlyZNH proton in the free peptide, leads to an apparently more solvent-protected resonance in the complex. 13C data agree well with this interpretation.

~~C NMR spectroscopy

o . ~

5.33 T'(C)

3b

~b

s'0

6'0

7~

Fig. 1. Temperature dependences of the chemical shifts of the NH proton resonances in (C2H3)2SO for the free peptide (O O) and its 2 : 1 complex (© ©); in this graphic representation, 8complex --8freepoptid e =0pp m at 2 0 ° C .

gated. Temperature dependence of the peptide NH proton resonances provide information in order to distinguish between 'exposed' and 'intramolecularly hydrogen-bonded' amide protons. Exposed protons exhibit larger temperature coefficients than do hydrogen bonded protons [20-21]. The temperature coefficients ( d S / d T ) for all the NH protons of [LeuS]-enkephalin and the 2:1 complex were obtained and are given in Table I. A plot of the chemical shifts of the peptide NH protons versus temperature in the range 25-70°C is shown in Fig. 1. In the free peptide, the leucine NH proton exhibits a small temperature coefficient (2.33 • 10-3 ppm/°C), as expected for NH proton involved in intermolecular hydrogen bonding. In agreement with earlier studies [17] which suggest a ill-turn between the CO of Gly 2 and NH of Leu5 (vide infra). However, the d S / d T value for the Gly2NH is 12.4.10 -3 ppm/°C, i.e., the most solvent ex-

The results obtained by IH NMR are supported by the 13C NMR spectra of [LeuS]-enkephalin and its complex with ml 3+ in (C2H3)2SO. Both negative and positive shifts upon metal binding are possible; in contrast to the proton resonance, the complexation induced t3C shift does not necessarily reflect the electron density change [22]. Fig. 2 shows the relationship between the [Leu5]-enkephalin/A13+ ratio to the change in chemical shift of the LeuSCOO - , Tyr]CO, P h e n c o and corresponding a-C. The change in chemical shift for all the carbons is linear up to a 2:1 ratio. Addition of excess AI 3+ to the 2:1 ratio does not affect the 13C NMR spectrum of [LeuS]enkephalin-Al 3+ . The tyrosine carbons show a slight shift variation as far as a 2:2 ratio (vide infra). This experiment is consistent with the conclusion that: (i) the system is in fast exchange on the ]3C NMR chemical shift time scale between its free and complexed states, since continuous shifts in peak positions are observed, rather than intensity changes; (ii) this behaviour enables us to conclude that [LeuS]-enkephalin forms a 2:1 complex with aluminium metal. Since the titrations indicate that the metal-bound and free peptide are in fast exchange on the NMR time scale, it is possible to unequivocally assign the carbon resonances for the aluminium complex. The 13C NMR spectrum of [LeuS]-enkephalin and its 2:1 complex has been obtained, and the

469 Chemical

shift (ppm)

~,

'\o

\

Leu $ CO0"

0 PI~ t CO

~ o

"1"

C a - Tyr 1

~ O ' "

o

C " - Phe t

Cc¢ - Leu s

SS,

[Lcu s] - cn kephatin I A I "

4~1

211

2~2

,e

Fig. 2. Variation of the 13C chemical shift in (C2H3)2SO as a function of the [LeuS]-enkephalin/A13+ ratio.

values of the chemical shifts resonances are illustrated in Table II. The most important changes in chemical shift occur in the tyrosine residue. The largest one is observed for the carbonyl resonance which shows a very important upfield shift of A - - - 4 . 9 5 ppm. In the same way the carbon resonances due to the methine and methylene group experience large upfield shifts -2.08 and -0.84 ppm respectively, upon aluminium binding. Finally, all the tyrosine ring carbons shift downfield upon addition of AI a+ , while the y*C resonance moves up field. The distance from the metal center to the carbon atom seems to be the major factor affecting the position of resonance, so that the shifts provide strong evidence in order to assign the carbonyl group of the tyrosine moiety as the metal-ligating group. In the data shown in Table II, perturbations are

observed also in the resonances arising from the Leu residue. Both leucine 8-CH3 and y-CH are shifted upfield in the presence of A13+ . It is also clear that the fl-CH 2 resonance is shifted upfield but in this case it is difficult to assign this particular resonance in the free peptide because of the (C2H3)2SO signal. However, the largest perturbations induced by A13+ are the upfield shifts seen for the resonances of the carboxylate ( - 1.40 ppm) and the methine (-1.92 ppm) carbons of the leucine moiety. In light of these observations, it seems that the Leu5COO- residue may be directly involved in metal complexation. The shifts of both 8-CH 3 are remarkably large upon complexation. Considering that the nearest ligand atom is four bonds away, conformadonal effects may be largely responsible for the observed shifts of these carbon atoms. Analysis of the 13C NMR spectra in the phenylalanine region indicates that the complexation of aluminum ion does not result in any chemical shifts changes for the ring carbons, except for the Phe4y-C which showns an upfield shift of -0.45 ppm. But the most striking ~3C NMR observation is that A13+ causes a low field chemical shift in the Phe4CO carbon resonance (A = 1.30 ppm). These results are opposite to those observed for the carbonyl and carboxylate groups of the tyrosine and leucine residues. There is one plausible mechanism to account for the observed Phe4CO shift. It has been suggested that [LeuS]-enkephalin in (C2H3)2SO solution exists in a well defined compact conformation stabilized by a fl-turn involving an intramolecular hydrogen bond between Leu5 NH and Gly2CO [17,23]. Our results also agree with these data (vide supra). Furthermore, Llinas et al. [24] showed that the resonance of a carbonyl carbon in a CO-NH group is shifted upfield if its NH moiety becomes hydrogen bonded. Thus, we can imagine disruption of this hydrogen bond upon complexation implying a downfield shift for the P h e 4 C O r e s o n a n c e . Finally, two of the carbonyl and methylene carbon resonances (Gly 2'3) are essentially unaffected b~( aluminium binding. From these results, it is interesting to note that aluminium ion causes a high field chemical shift in the Tyr~CO and LeuSCOO - carbon resonances and a lowfield chemical shift in the Phe4CO residue due to disruption of hydrogen bonding.

470

T A B L E II

13C N M R

D A T A F O R [ L E U S ] - E N K E P H A L I N A N D [ L E U S I - E N K E P H A L I N / A I 3+ I N ( C 2 H 3 ) 2 S O

n.o., not observed. Residues

Tyr i

Carbon Atoms

Free peptide

16:1 a

4:1

2:1

Ab

2:2

A' c

a

60.66

60.38

59.75

58.58

- 2.08

58.38

- 0.20

fl

41.87

41.87

41.87

41.03

- 0.84

40.60

- 0.43

y

130.75

130.90

131.46

129.77

- 0.98

129.38

- 0.39

6

134.59

134.59

134.72

134.91

0.32

134.97

c

119.63

119.63

119.70

119.83

0.20

119.96

,~

160.59

160.59

160.73

160.98

0.39

161.10

0.12

CO

178.87

178.48

176.68

173.92

- 4.95

173.14

- 0.78

Gly 2

a CO

46.55 173.51

46.57 173.46

46,36 173,40

46.30 173.04

- 0.25 0.47

46.30 173.00

-

Gly 3

a CO

Phe 4

Leu 5

46.55

46.57

46.36

46.30

- 0.25

46.30

173.12

173.12

173.08

173.04

0.08

173.00

a

59.11

59.06

58.71

58.20

-0.91

58.13

fl ~,

n.o. 142.71

n.o. 142.68

42.90 142.45

n.o. 142.26

- 0.45

n.o. 142.20

0.13

8

133.67

133.67

133.67

133.74

0.07

133.74

c

132.51 130.75

132.52 130.75

132.57 130.75

132.51 130.75

0 0

132.57 130.81

CO

174.31

174.44

175.00

175.61

1.30

175.61

a fl

56.86 n.o.

56.52 n.o.

55.65 45.38

54.94 44.60

- 1.92 -

54.94 44.58

-

~, 8

28.93 27.63

28.93 27.60

28.80 27.44

28.74 27.30

- 0.19 - 0.33

28.73 27.25

-

26.72 179.71

26.62 179.58

26.27 179.06

25.94 178,31

- 0.78 - 1.40

25.90 178.22

-

COO-

-

a P e p t i d e / m e t a l ratio b A ppm ~---8pprn 2 : 1 -- 8ppm free peptide c Appm~_ ~ppm 2 : 2 -- ~ppm 2 : 1

It is also noteworthy that the addition of a second half-mole of A13+ to produce a 2 : 2 species causes further modifications. All of the shifted resonances are in the tyrosine portion of the peptide while the four remaining residues are not affected. Thus, as shown in Table II, the largest changes in chemical shifts occur for the carbon resonances of the following groups; Tyr~CO, --0.78 ppm; Tyrlfl, - 0 . 4 3 ppm and Tyrnp, - 0 . 3 9 ppm. On this basis, we suggest that [LeuS]-en kephalin binds the first aluminium via TyrnCO and Leu 5C O 0 - carbons to give a 2:1 compound. In addition, [LeuS]-enkephalin is able to bind a second aluminium ion. The second site apparently involves the terminal N H 2 of the tyrosine residue as suggested by IH NMR.

27.4l N M R spectroscopy 27A1 is an attractive nucleus on account of its high sensitivity and a 100% isotopic abundance. However, its spin number, I = 5/2, results in a nuclear quadrupolar moment which strongly interacts with electric field gradients originating from an asymmetrical arrangement of the ligands around the A13+ cation. Thus, if the aluminium atom is located in a region of high molecular symmetry, its N M R absorption line can be less than 10 Hz at half-height [Al(H20)6~+ , Apl/2 approx. 10 Hz]. In less symmetrical environments, a broader absorption line is observed [25-26]. At 22.63 MHz, this later situation occurs and no resonance was detected for solutions of either the 2:1 or 2:2 species in ( C 2 H 3 ) 2 S O . But, according to the

471

Stokes-Einstein relation, ~"= V~I/KT, where ~" is the rotational correlation time, V is the molecular volume, ~ the viscosity and T the absolute temperature, the line widths show a significant dependence on viscosity. Thus warming the sample first sharPens the absorPtion by decreasing the viscosity and accordingly, the correlation time (T21= ~rvl/2 a: ~). In these conditions, the solutions were examined at 60°C, where it has been shown that both resonances are narrowed. By comparison, under similar temperature conditions, the 2:1 species exhibits a 27A1resonance with a line width of approx. 500 Hz, at approx. 50 ppm relative to the hexa-aquo 27A1 resonance, here taken as the 0 ppm reference line. On the other hand, in the 2:2 system, the 27A1results suggest that A13+ chelates the pentapeptide in two different sites. The first one carries about the same line width and shift as observed in the 2:1 species, while the second one displays a line width of approx. 600 Hz and a chemical shift of - 1 0 ppm. 27A1 NMR spectroscopy at 65.14 MHz opens the possibility of much better dispersion of the resonances arising from quadrupolar nuclei. Thus the 27A1 spectra consist of two overlapping broad lines (Fig. 3) at room temperature with much better resolution than that observed at 60°C (22.63 MHz). Both species exhibit a line width of approx. 3.5 kHz and a chemical shift of approx. 50 and - 10 ppm, respectively. Although no general statements regarding the factors which determine 27A1chemical shift have been formulated, Haraguchi and al. [27] observed that the position of the 27A1 resonance correlates well

with the type of coordination of the A13+ . Thus, the chemical shifts of octahedral complexes extend from about - 5 0 to +20 ppm while tetrahedral derivatives span the region from about +20 to + I I0 ppm. Our own observations agree well with these data [28]. The above results strongly suggest that the signal observed for the 2:1 species arise from aluminium complex where the metal is in a tetrahedral configuration, while in the 2:2 system, the second type of aluminium atom displays an octahedral environment. In this complex, the coordination around metal might be completed by water molecules.

Conclusions These combined IH, ]3C and 27A1NMR studies indicate that [LeuS]-enkephalin form 2:1 complex with AI 3+ , involving the TyrtCO, and the Leu5COO- groups in a tetracoordinated structure. [LeuS]-enkephalin, however, can bind additional A13+ in a 2:2 species, via the NH 2 terminal groups of the tyrosine residue. It is suggested that an intramolecular hydrogen bond which occurs between the LeuSNH and CO of Gly 2 in the free peptide is disrupted upon metal binding. The present study demonstrates that [LeuS]-enkephalin is an appropriate ligand for conducting a comprehensive investigation by NMR method of the interaction of 'm0rPhine-like' molecule ~vith diamagnetic metal ion; such investigations should prove of significant interest in the elucidation and understanding of aluminium ion complexation and in the role played by specific trace metal ions in the central nervous system.

Acknowledgements

",.~2

We are grateful to Mr. G. Commenges for excellent technical assistance with the NMR studies. This research was supported by the Centre National de la Recherche Seientifique.

References ,

i

i

i

i

i

i

i

i

w

w

200

i

r

wl

1 0

I l l

0

Jt

I ' l l l

-100

I I I

I

-200

I 1 1 1

I1

ppm

Fig. 3. 27A1 NMR spectrum at 65.14 MHz and 22°C of a solution of [LeuS]-enkephalin/Al3+ = 1 in (C2H3)250.

1 Donalson, J. (1981) Trends Pharmacol. Sci. 3, 75-78 2 Rajah, K.S., Colburn, R.W. and Davis, J.M. (1976) in Metal Ions in Biological Systems (Sigel, H. ed.), Vol. 6, pp. 291-321, Dekker, New York

472 3 Roques, B.P., Foum~-iS-Zaluski, M.C., Soroca, E., Lecomte, J.M., Malfroy, B., Llorens, C. and Schwartz, J.C. (1980) Nature 288, 286-288 4 Barbeau, A., Inove, N. and Clouter, T. (1976) Adv. NeuroL 14, 339-351 5 Alfrey, A.C. (1978) Annu. Rev. Med. 29, 93-98 6 Arieff, A.I., Cooper, J.D., Armstrong, D. and Lazarowitz, V.C. (1979) Annu. Intern. Med. 90, 741-747 7 Crapper, D.R., Quitkat, S., Krishnan, S.S., Dalton, A.J. and De Boni, U. (1980) Acta Neuropathol. 50, 19-24 8 Yase, Y. (1977) Proc. llth World Congress Neurol. pp. 413-427 9 Womack, F,C. and Colowick, S.P. (1979) Proc. Natl. Acad. Sci. U.S.A. 76, 5080-5084 10 Lanssac, J.P. and Laurent, J.P, (1980) C.R. Acad. Sci. 291C, 157-159 11 Simantov, R. and Snyder, S.H. (1976) Mol. Pharmacol. 12, 987-998 12 Holl6si, M., Dobolyi, Z. and Bajusz, S. (1980) FEBS Lett. 110, 134-140 13 Gairin, J.E., Haran, R., Mazarguil, H., Lemaire, S., SaintPierre, S. and Sharrock, P. (1980) XXI ICCC, p. 83, Toulouse 14 Haran, R., Gairin, J.E. and Mazarguil, H. (1980) C.R. Acad. Sci. 291D, 685-688 15 Sharrock, P., Day, R., Lemaire, S., Saint-Pierre, S., Mazarguil, H., Gairin, J.E. and Haran, R. (1982) Inorg. Chim. Acta 66, L91-L95

16 Audigier, Y., Mazarguil, H., Gout, R. and Cros, J. (1980) Eur. J. Med. Chem., 15, 173-177 17 Garbay-Jauraguiberry, C., Roques, B.P., Oberlin, R., Anteunis, M. Combrisson, S. and Lallemand, J.Y. (1977) FEBS Lett. 76, 93-98 18 Khaled, M.A., Renugopalakrishnan, V. and Urry, D.W. (1976) J. Am. Chem. Soc. 98, 7547-7551 19 Evans, E.J., Grice, J.E., Hawkins, C.J. and Heard, M.R. (1980) Inorg. Chem. 19, 3496-3502 20 Kopple, K.D., Ohnishi, M. and Go, A. (1969) Biochemistry 8, 4087-4095 21 Ohnishi, M. and Urry, D.W. (1969) Biochem. Biophys. Res. Commun. 36, 194-202 22 Horsley, W.J. and Sternlicht, H.J. (1968) J. Am. Chem. Soc. 90, 3738-3742 23 Stimson, E.R., Meinwald, Y.C. and Scheraga, H.A. (1979) Biochemistry 18, 1661-1671 24 Llinas, M., Wilson, D.M. and Klein, M.P. (1977) J. Am. Chem. Soc. 99, 6846-6850 25 Petrakis, L. (1968) J. Phys. Chem. 72, 4182-4188 26 Akitt, J.W. (1972) Annu. Rep. NMR Spectrosc. 5A, 466-556 27 Haraguchi, H. and Fujiwara, S. (1969) J. Phys. Chem. 73, 3467-3473 28 Laussac, J.P. (1976) Th6se de Doctorat d'Etat, Toulouse, France