Nuclear-Magnetic-Resonance Study on Met-enkephalin and Met-enkephalin : Molecular Association and Conformation

Eur. J. Biochem. 97, 43 - 57 (1979)

Nuclear-Magnetic-Resonance Study on Met-enkephalin and Met-enkephalinamide Molecular Association and Conformation

Tsutomu HIGASHIJIMA, Junichi KOBAYASHI, Ukon NAGAI, and Tatsuo MIYAZAWA

Department of Biophysics and Biochemistry, Faculty of Science, University of Tokyo, and Mitsubishi-Kasei Institute of Life Sciences, Tokyo

(Received December 1, 1978)

The 270-MHz 'H and 68-MHz 13C nuclear magnetic resonance spectra of Met-enkephalin (Tyr-Gly-Gly-Phe-Met) and Met-enkephalinamide are analyzed in a variety of solvents. For the dipolar form of Met-enkephalin in (C2H3)2S0 solution, significant concentration dependences are found of C-a proton chemical shifts, indicating an aromatic ring-current effect in molecular aggregates. An anomalous temperature dependence is observed of the amide proton chemical shift of the Met-5 residue. Furthermore, the chemical shifts of C-cc protons of the dipolar form are found to depend appreciably on temperature. From the analyses of the temperature dependences together with concentration dependences of C-cc proton resonances, the dipolar form of Met-enkephalin is found to be in an equilibrium of folded and extended conformations at low concentration in (C2H3)2S0 solution. Solvent-composition dependences of the amide and C-cc proton resonances and carbonyl and cc-carbon resonances of the dipolar form in 2H20/(C2H3)2S0 solution are consistent with the conformation equilibrium and the association equilibrium. The folded conformation of the dipolar form in (C2H3)2S0 solution is stabilized by the intramolecular attraction between the positively charged N-terminal group and negatively charged C-terminal group. The presence of the folded conformation is confirmed by the measurements of Gd(II1)-induced relaxation enhancements of C-cc protons. Nuclear Overhauser effects on the dipolar form are not consistent with the predominant formation of the B-turn structure with the intramolecular hydrogen bond (Gly-2) C = O . H - N (Met-5). For the dipolar form of Met-enkephalin in 2H20 solution and for the cationic form of Met-enkephalinamide in (C2H3)2S0 solution and in 'H20 solution there is no evidence for the formation of folded conformations

Enkephalins are endogenous brain peptides with morphine-like activity; they are analgesics and operate at the same receptor as natural opiates and their antagonists [I]. The amino acid sequences of enkephalins are Tyr-Gly-Gly-Phe-Met (Met-enkephalin) and Tyr-Gly-Gly-Phe-Leu (Leu-enkephalin) [ 11. A number of enkephalin analogues have been synthesized and more potent peptides have been found [2 - 61, including [~-Ala~]Met-enkephalin [3 - 51, [MePhe4]Met-enkephalin [4], and Met-enkephalinamide (Tyr-Gly-Gly-Phe-Met-NH2) [6]. These active analogues as well as enkephalins possibly bind to the

Abbreviations. N M R , nuclear magnetic resonance; Met-enkephalin, enkephalin with C-terminal methionine (Tyr-Gly-Gly- Phe-Met) ; Met-enkephalinamide, amide derivative of the C-terminal methionine in Met-enkephalin; [MePhe4]Met-enkephalin, Met-enkephalin with an N-methyl group on Phe-4.

~~ ~- .~

same receptor as do morphine and other opiates, and are expected to be similar in conformations at the binding site [7,8].

The molecular conformations of enkephalins in solution have been studied by 'H NMR [9-171, I3C NMR 115-191, and circular dichroism [20]. As for the dipolar form of enkephalin in (C2H3)2S0 solution and in aqueous solution, a type-I p-turn conformation has been proposed [9 - 11,14, 15, 17, 181, with an intramolecular hydrogen bond (Gly-2) C = O . H-N (Met-5). This p-turn conformation is primarily suggested from the small temperature coefficient of the chemical shift [21,22] of the amide proton of the Met-5 residue. The small temperature coefficients of amide protons are taken as suggesting the participation in intramolecular hydrogen bond, from studies on cyclic model peptides such as

44 Association and Conformation of Met-enkephalin

gramicidin S and valinomycin [22]. Furthermore for Met-enkephalin, the pH-dependences of 'H chemical shifts in H20/(C2H3)2S0 (2/1) solution (87 mM) [l 11 and of 13C chemical shifts in H20/(CzH3)2SO ( l / l ) solution (100mM) [18] have been observed; the chemical shifts of the C-a proton of Met-5 and C-p proton of Tyr-1 are sensitive to the pH titrations of the terminal NHz and COOH groups respectively. This observation has been considered to suggest the proximity of the N and C-terminal groups in a /j-turn conformation [9- 11,15,18].

For the cationic form of Met-enkephalin and Leu- enkephalin, however, all the amide proton resonances show large temperature coefficients, suggesting that the cationic form does not take fi-turn conformations with intramolecular hydrogen bonds [12,13,16]. Accordingly the intramolecular electrostatic interactions between the charged groups of N- and C-terminal residues have been considered to play an important role for stabilizing the p-turn conformation of the dipolar form of enkephalins.

Recently an X-ray diffraction study has been made on Leu-enkephalin crystallized from an aqueous methanol solution ; the molecule takes the type-I1 fi-turn structure with two intramolecular hydrogen bonds, (Tyr-1) C = 0 . H ~ N (Phe-4) and (Phe-4) C = 0 . H3Nf (Tyr-1) [23]. However, this 0-turn structure may not necessarily be the unique active conformation of enkephalins, since this structure is not possible for the highly potent analogue of [MePhe4]Met-enkephalin. In fact, the circular dichroism and absorption spectra of Met-enkephalin in aqueous solution may be inter- preted without assuming that Met-enkephalin pre- ferentially takes a single conformation in dilute solution [20].

Our previous studies on melanostatin and related peptides indicate the importance of taking the molecular association and ionization into account before analyzing molecular conformations of linear oligopeptides [24,25]. In fact, large concentration dependences of the amide proton and carbonyl carbon resonances of Met-enkephalin and Leu-enkephalin have been observed, indicating that intermolecular association is significant in (C2H3)2SO solution [17].

For a clear-cut understanding of the molecular association and conformation of Met-enkephalin in solution, Met-enkephalinamide was also investigated in the present study as well as Met-enkephalin. A variety of NMR measurements were systematically carried out including (a) pH dependences of chemical shifts of amide and C-a proton resonances, (b) concentration dependences, (c) temperature dependences, and (d) solvent-composition dependences of chemical shifts of amide and C-a proton resonances and carbonyl and x-carbon resonances, (e) Gd(II1)-induced enhancements of spin-lattice relaxation rates

of C-a protons, and (0 nuclear Overhauser effects on amide and C-a proton resonances.

MATERIALS AND METHODS

Met-enkephalin was synthesized by the liquid- phase method and was confirmed, by 'H NMR, to be identical with the sample purchased from Protein Re- search Foundation. Met-enkephalinamide was synthesized from Boc-Tyr(Bzl)-Gly-Gly-Phe-Met-OCH3 (where Boc = tevt-butyloxycarbonyl and Bzl = ben- zyl), which was prepared by the liquid-phase method. After the treatment of this protected peptide with methanolic ammonia, the protecting groups (Boc- and Bzl-) were removed with H F in the presence of anisole. Gel filtration of the deprotected peptide on a column of Sephadex LH-20 (2.0 x 100 cm) equilibrated with 1 % acetic acid gave Met-enkephalinamide, Tyr-Gly- Gly-Phe-Met-NH2. Thin-layer chromatography was carried out on silica gel plates (F245 from Merck) in butan-1-ol/acetic acid/water (65/15/25 by volume, solvent A) and chloroform/methanol/ammonia water (60/45/20, solvent B). The purity of Met-enkephalinamide was established by this thin-layer chromatography ( R F = 0.40 in solvent A and RF = 0.82 in solvent B) and by amino acid analysis of the 6 M HCl hydrolysis product (Tyr 0.95, Gly 1.99, Phe 1 .OO, and Met 0.98). Amino acid analyses were performed with a Durrum D-500 amino acid analyzer.

The solvent (C2H3)2S0 (99.8 %) was obtained from Commissariat a I'Energie Atomique and 'H20 (99.75%) and C2HC13 (99.8%) from Merck. (C2H3)2S0 was stored with synthetic zeolite to remove water contamination.

pH values (direct pH meter readings) were directly measured in NMR tubes at 23 "C, with a Radiometer PHM 26 pH meter equipped with a long, thin Nisshin- Rika combination electrode.

NMR spectra were recorded on a Bruker WH-270 pulse FT spectrometer(270 MHz for 'H and 67.9 MHz for 13C resonances), equipped with a Bruker B-ST- 100/700 temperature control unit. However, 'H NMR spectra of H20 solutions were also measured with the Bruker Correlation Digital Sweep NMR Program. Internal reference for chemical shifts were 2,2-di- methyl-2-silapentane 5-sulfonate for 'H and dioxane for I3C. For the measurements of Gd(II1)-induced relaxation enhancements, a small amount of Gd(II1) nitrate solution was added successively to the solution of Met-enkephalin in (C2H3)2S0 solution or in 'HzO solution. For better separation of the C-a proton resonances of Tyr-1 and Met-5 residues, Pr(II1) nitrate was added to 'HZ0 solution of Met-enkephalin. Spin-lattice relaxation rates (1 /TI) were measured by the inversion recovery method. Nuclear Overhauser effects (NOE) were measured by the selective gated

T. Higashijima, J. Kobayashi, U. Nagai, and T. Miyazawa 45

Fig. 1. in (C2

270-MHz ' H MMR spectra of Met-enkephalin (the dipolurform, 5 m M ) ( A ) and Met-enkephalinamide (the cationicform, 33 mM) H3)2SO solution at 23 ' C

decoupling method. Circular dichroism spectra were observed using a Jasco J40A spectrophotopolarimeter.

RESULTS Assignments of ' H and I3C Resonances

The 270-MHz 'H NMR spectra of Met-enkephalinamide as well as of Met-enkephalin were assigned with the aid of the spin-decoupling method and pH and solvent effects. The 'H spectra of Met- enkephalin and Met-enkephalinamide in (C2H3)2S0 solution are shown in Fig.1, together with peak assignments. Chemical shifts and spin-coupling constants for amide and C-a protons are listed in Table 1. These assignments are consistent with those reported previously for the dipolar form [I41 and for the cationic form [16]. The C-p proton signals of the Phe-4 residue of Met-enkephalin (Fig. 1) have been un- ambiguously assigned to the pro-S and pro-R protons, by comparison with the C-fi proton signal of Met-enkephalin containing (2S,3S)-[2,3-2H2]Phe as the fourth residue [26].

The assignments of 13C resonances of Met-enkephalin have been made with the aid of pH and solvent effects and are consistent with those reported previously [ 171.

p H Dependence of Chemical Shijts

The 'H20 solution of Met-enkephalin (about 10 mM) as synthesized showed a pH value of 5.9. The pH dependences of chemical shifts were measured of the amide protons in H 2 0 solution and of C-a protons in 'HzO solution (Fig.2). The pK, values were determined as 3.50 for the carboxyl group of C-terminal methionine and 7.75 for the amino group of N-terminal tyrosine. Accordingly Met-enkephalin, as synthesized, is in the dipolar form. The cationic form of Met-enkephalin was prepared by lyophilizing the aqueous solution after the adjustment to pH 1 with HCl. The pH dependences of proton chemical shifts were also measured of Met-enkephalin in 'HzO/ (C2H3)2S0 (2/1) solution at the concentration of 3 mM (Fig. 3).

The aqueous solution of Met-enkephalinamide (about 10mM) as synthesized showed a pH value of 5.3. The pH dependences of chemical shifts were measured of the Tyr-1 C-a and C-0 protons and the pK, values were determined as 7.50 for the amino group and as 10.25 for the hydroxyl group of Tyr-1 residue. Accordingly Met-enkephalinamide as synthesized is in the cationic form.

Concentration Dependence of Proton Chemical Shifts The concentration dependences of the chemical

shifts of amide and C-a protons are shown in Fig. 4A,


Table 1 . Chemical shift and spin-coupling constants Concentration 5 mM

Form of Solvent Chemical shift Spin-coupling constant .. _ _ ._ enkephalin ~~

Tyr-1 Gly-2 Gly-3 Phe-4 Met-5 J H N C ~ H

C-a N-H C-cc N-H C-a N-H C-a N-H C-a Gly-2' Gly-3' Phe-4 Met-5 ~~ ~ ~~~~ - . - -

'HzO(pH 6)" 4.24 8.62 1;:;: 8.04 1 i::: 8.06 4.66 7.97 4.24 broad 10.9 8.0 7.4 Dipolar Met-enkephalin

(CZH,),SOb 3.69' 8.62 1:::; 7.92 3.71' 8.34 4.43 7.91 4.07 broad 11.4 8.3 7.2

3.70 8.79 1;::; 7.86 [:::: 8.12 4.51 7.63 4.19 11 11.0 8.3 6.8 C'HCI, (CZH3)2SOC

3.94 3.88 Cationic Met-enkephalin 'HzO (pH 1.5)" 4.24 8.59 [3.88 8.00 13.85 8.12 4.63 8.45 4.49 12.5 11.7 7.8 7.8

(C2H3)2S0 4.01 8.78 1 8.15 1::;; 8.13 4.59 8.45 4.35 10.8 10.7 8.8 7.7

3.89 4.59 4.38 3.94 Cationic Met-enkephalin- 'HzO (pH 5) 4.12 amide

13.90

(CZH3)'SOd 3.39 8.24 3.73 8.13 [i::: 8.15 4.51 8.05 4.27 broad 11 7.7g 8.3

* Amide proton resonances were observed in HzO. 6.5 mM. Mixing ratio 2: 1. Carboxamide proton resonance of Met-5 at 7.10 ppm. For ['Hz-Gly']Met-enkephalin (prepared by J. Kobayashi).

' Sum of J for two methylene protons. At 53'C.

10 - 9 -

8 -

7 -

g:: 4 -

3 -

2 -

1 -

I z N - a

8

n U L

8.5 8.0 4.5 4.0 3 5 Chemical s h i f t (pprn)

Fig. 2. p H dependences of the chemical shifts ofamide protons (in H z 0 solution, 20 m M ) and C-a protons of Met-enkephalin (in ' H 2 0 solution, I0 m M ) at 23'C

for the dipolar form of Met-enkephalin in (C2H3)2S0 solution at 23 "C. As the solution is diluted, the amide proton resonances of Gly-2, Gly-3 and Met-5 residues are shifted to higher field while that of Phe-4 is shifted

to lower field. This is in contrast to the previous observation on Leu-enkephalin [17] that the amide proton resonance of Phe-4 is shifted to higher field but that of Gly-3 is shifted to lower field upon dilution.


I I I 6 I I , I I I I I I 4.5 4.0 3.5 3.0

Chemical s h i f t (Ipm)

Fig. 3. pH dependences of the chemical shifts of the C-cc and C-p protons of Met-enkephalin in 'H20/ (C2H3)2S0 (211) solution ( 3 m M ) at 23 -C

200

100

- 2 0 E - 8.5 8.0 4.5 4.0

9.0 8.5 4.5 4.0 a 200

1 0 0

0 8.5 8.0 ZO 4.5 35

Chemical shift (ppm)

Fig. 4. Concentration dependences of the chemical shifts of the amide and Ccc protons q f ' ( A ) the dipolarform and (%) the cationicform of Met- enkephalin and (C) the cationicform of Met-enkephalinamide in (C2H3)2S0 solution at 23 'C

Accordingly in the present study NMR measurements were made of the two samples of different origin but exactly identical results were obtained.

It is remarkable to note that the chemical shifts of C-a protons of the dipolar form of Met-enkephalin in (C2H3)2S0 solution are also concentration-dependent (Fig. 4A). Accordingly, the concentration dependences of the chemical shifts of amide proton resonances may not necessarily be taken as the direct indication of intermolecular hydrogen bonding. The concentration

dependences of the chemical shifts were also observed for the cationic form of Met-enkephalin and Met- enkephalinamide in (C2H3)2S0 solution (Fig. 4B, C).

Concentration Dependence of Carbon Chemical Shifts

The concentration dependences of the chemical shifts of carbonyl and a-carbons are shown in Fig. 5, for the dipolar and cationic forms of Met-enkephalin and for Met-enkephalinamide in (C2H3)2S0 solution.


Form of enkephalin Solvent Concn lo3 x Temperature coefficient of chemical shift for

Gly-2 Gly-3 Phe-4 Met-5

Dipolar Met-enkephalin

mM

13 220 150 75 45 25 6

C2HC13/(C2Hs)zSO" 5

Cationic Met-enkephalin H20 (PH 1.5) (C2Hs)2S0

Cat ionic Met-enkephalinamide (C2H3)2S0

10 5

5

PPm/'C ~~~~~~~~ ~ ... ~

- 10.5 -6.5 -6.3 - 8.7 -5.1 - 6.5 - 9.3 -5.1 -6.5 - 8.4 -4.1 -7.5 - 8.3 - 3.8 -7.2 - 8.0 -3.3 - 7.7 - 8.1 - 2.9 - 8.2 - 12.2 -4.3 -8.7

- 7.4 -5.1 - 6.7 - 5.2 -5.1 - 6.2

- 3.9 - 4.6 - 4.9

-7.1 - 2.4 -2.1 -2.0 -1.8

Oto -1.9 0 to -1.2

- 1.3 to -2.6

- 7.7 -6.8

- 5.6

a Mixing ratio, 2/1. ' Temperature coefficient of carboxamide proton is -4.6.

The concentration-dependent shifts for the carbonyl carbons of the dipolar form agree, in the direction but not in the magnitudes, with those reported previously for Leu-enkephalin [17].

For the dipolar form of Met-enkephalin (Fig. 5A), concentration-dependent shifts are largest for the carbonyl carbon of the Tyr-1 residue and for the carboxylate carbon of the Met-5 residue. It is remarkable to note that the chemical shifts of cc-carbons are as large as those of carbonyl carbons for the dipolar form of Met-enkephalin in (C2H3)2S0 solution.

Temperature Dependence of Proton Chemical Shijts

The temperature dependences of chemical shifts of amide proton resonances are useful for distinguishing between 'exposed' and 'intramolecularly or intermolecularly hydrogen-bonded' amide protons [21,22] ; exposed protons exhibit larger temperature coefficients than do hydrogen-bonded amide protons. The temperature coefficients of amide proton chemical shifts are shown in Table 2, for Met-enkephalin and related peptides.


Table 3. Temperature coej'icients of chemical shifts of C-x protons

Form of enkephalin Solvent Concn lo3 x Temperature coefficient of chemical shift for ~ ~ -~ ~ . .

Tyr-1 Gly-2 Gly-3 Phe-4 Met-5

Dipolar Met-enkephalin 'H20 (PH 5.9) 5 -0 -0 -0 -0 -0 (C2H3)zS0 220 0.6 1.9

150 - 3.4 0.7 2.0 15 1.1 2.4 45 1.2 2.6 25 - 2.8 -0.4b 1.7 2.8 6 -1.7 - O S b 2.3 3.4

C2HC13/(C2H3)2SO" 5 1.4 3.1

(CZH3)2S0 5 0.3 0.3 0.4 Cationic Met-enkephalinamide (CZH3)2S0 5 1.1 -0.5 -0.5 0.3 0.3

Cationic Met-enkephalin 'Hz0 (PH 1.5) 5 -0 0 0 -0 -0

a Mixing ratio, 2/1. For ['Hz-GlyZ]Met-enkephalin (prepared by J. Kobayashi).

r I

-5.2 "I

8.5 8 I) 4.5 Chemical s h i f t (ppm)

4.1 3.7

Fig. 6. Temperature dependences of' the chemical sh~f t s of' the amide and C-x protons of the dipolar form of Met-enkephalin ( A ) at the concentration of 6 mM and ( B ) at the concentration of25 m M , in the presence ofNH4Cl04 ( 1 M ) in (C2H3)zS0 solution

In contrast to the amide proton chemical shifts, the chemical shifts of C-a proton resonances of gramicidine S [27], melanostatin [24] and related peptides [25] are independent of temperature. In the present study, however, significant temperature dependences were found for the first time for the C-a proton chemical shifts of the dipolar form of Met- enkephalin in (C2H3)2S0 solution and in C2HC13/ (C2H3)2S0 (2/1) solution, as shown in Table 3.

For the dipolar form of Met-enkephalin in (C2H3)2S0 solution (6 mM), the temperature de-

pendences of the chemical shifts of amide and C-a protons are also shown in Fig.6A. The temperature dependences for the amide protons of the Gly-2 and Phe-4 residues are much larger than that for the amide proton of the Gly-3 residue. However, it is remarkable to note that the amide proton resonance of the Met-5 residue does not show a linear temperature dependence and furthermore the C-a proton resonances of the Tyr-1, Phe-4 and Met-5 residues exhibit significant temperature dependences. For analyzing these anomalous temperature dependences, the

50

-

-

a 25 - I

E I I I I I I

Association and Conformation of Met-enkephalin

a E J;i: 50 \ /!

concentration dependences of the temperature coefficients of chemical shifts were also observed for the dipolar form of Met-enkephalin in (C2H3)2S0 solution (Tables 2 and 3). As the concentration is raised, the negative temperature coefficient for the amide proton of the Met-5 residue becomes larger, whereas the temperature coefficients of the Tyr-1 , Phe-4 and Met-5 residues become smaller. The concentration dependence of the temperature coefficient for the amide proton of the Met-5 residue observed in the present study is opposite to that reported previously [ 171. However, the temperature coefficient (0 to - 1.2 x ppm/"C) of the amide proton of the Met-5 residue measured at 6 m M in the present study agrees with that (-0.6 x ppm/"C at 7 mM) reported by Jones et al. [13].

The anomalous temperature dependences of amide and C-a proton resonances of the dipolar form of Met-enkephalin in (C2H3)2S0 solution (Fig. 6A), however, were eliminated by the addition of NH4C104 as shown in Fig. 6 B. In the presence of NH4C104 (1 M), the amide proton resonance of the Met-5 residue now exhibits significant negative temperature dependence (with the temperature coefficient of -4.6 x ppm/'C) and the chemical shifts of the C-a proton resonances of the Tyr-1, Phe-4 and Met-5 residues depend little on temperature.

Normal temperature dependences of amide and C-a proton resonances are observed also for the dipolar form of Met-enkephalin in 'H20 solution, the cationic form of Met-enkephalin in (C2H3)2S0 solution and in 'H20 solution, and for Met-enkephalinamide in (CZH3)2S0 solution (see Tables 2 and 3). For reference, the temperature dependences of C-a proton chemical shifts were also measured for N-acetylphenylalanine methylamide and of AT-acetyltyrosine

methylamide and the temperature coefficients were found to be 0.5 x and 0.4 x ppm/"C respectively.

Temperature Dependence of Carbon Chemical Shifts

The temperature dependences of chemical shifts of the carbonyl and a-carbons are shown in Fig.7, for the dipolar form of Met-enkephalin in (C2H3)2S0 solution at the concentration of 75 mM. Largest temperature dependences are observed for the carbonyl and a-carbons of the Tyr-1 residue and for the carboxylate and a-carbons of the Met-5 residue. However, significant temperature dependences are also observed for the carbonyl and/or a-carbons of the Gly-2, Gly-3 and Phe-4 residues.

Solvent-Composition Dependence of Proton Chemical Shijts

Solvent-composition dependences of amide proton chemical shifts have been used for distinguishing between 'free' and 'intramolecularly hydrogen bonded' NH groups of peptides [28,29]. For the dipolar form of Met-enkephalin, the dependences of chemical shifts on solvent composition are shown in Fig. 8, for amide proton resonances in HzO/(C'H~)~SO solution and for C-a protons in 2H20/(C2H3)2S0 solution. It is remarkable to note that C-a proton chemical shifts as well as amide proton chemical shifts depend significantly on solvent composition. Furthermore, the solvent-compositon dependences are not monotonous, suggesting the possibility of changes in molecular conformations as well as molecular association induced by changes in solvent compositions [30].

T. Higashijima, J. Kobayashi, U. Nagdi, and T. Miyazawa 51

8.5 8.0 4.5 4 .O Chemical shift (ppm)

Fig. 8. Solvent composition dependences of’ the proton chemical shifts of’ the dipolar form of Met-enkephalin: amide protons in H z 0 ( p H 5 . 5 ) / ( C z H 3 ) 2 S 0 solution (10 m M ) , C-a protons in ’ H z 0 ( p H S . S ) / ( C 2 H 3 ) z S 0 solution (5 m M ) at 23’,C

100

75

50

25 - 5 0 I- 100 0

N

75

50

25

0

111 110 109 108 107 106 105 104 103 lo2

- 10 -11 -1 2 -13 -14 - 24 -2 5 -2 6 -15 -23 Chemical shift ( p p m )

Fig.9. Solvent composition dependences of the chemical shifts of the carbonjl and a-carbons o j the dipolar ,form of Met-enkephalin in Z H 2 0 / I C 2 H 3 ) 2 S 0 solution (20 m M ) at 23’C

Solvent-Composition Dependence of’ Carbon Chemical Shifts

Solvent-compositon dependences of chemical shifts of carbonyl carbons are useful distinguishing between ‘free’ and ‘intramolecularly hydrogen-bonded’ carbonyl groups of peptide molecules; the latter carbonyl groups show little solvent dependence [31, 321. The solvent-composition dependences of the chemical shifts of carbonyl and a-carbons are shown in Fig.9, for the dipolar form of Met-enkephalin in 2H20/(C2H3)2S0 solution at the concentration of 20 mM. It is remarkable to note that a carbon chemical shifts as well as carbonyl carbon chemical shifts depend significantly on solvent composition. Further- more, the solvent-composition dependences are not monotonous for the carbonyl carbon of Tyr-1 and %-carbons of all the residues.

Gd(I1I)-Induced Proton Relaxation Enhancements

Gd(II1)-induced relaxation enhancements are pro- portional to the average values of the inverse sixth power of the distances from the Gd(II1) ion to the observed nuclei [33]. Before the analyses of Gd(II1)- induced relaxations, it is important to examine con-

formational changes of ligand molecules upon the binding of lanthanide ions. To the dipolar form of Met-enkephalin (10 mM) in 2H20 solution at pH 5.2, La(II1) nitrate was added to the concentration of 200 mM so that about 70% of Met-enkephalin molecules were bound to the La(II1) ion. However, vicinal spin-coupling constants JNH -CnH did not change upon the addition of La(II1) ion. Further, the circular dichroism spectrum (210-240 nm) of the dipolar form of Met-enkephalin (1 mM) in H2O solution at pH 5.9 did not change by the addition of La(II1) nitrate [50 o/, of Met-enkephalin molecules are bound to the La(II1) ion]. These observations indicate that there are no significant conformational changes upon binding with lanthanide ions. Accordingly, the lanthanide-ion-probe method may be applied to the conformational studies on the dipolar form of Met- enkephalin in solution.

The ratios of Gd(II1)-induced relaxation enhancements of the C-a protons of the dipolar form of Met- enkephalin in 2H20 solution are shown in Fig.10A. For comparison, Gd(II1)-induced relaxation measurements were made of the cationic form of Met-enkephalinamide in ’H20 solution (pH 5.5) with the same concentration of Gd(II1) nitrate as for the


OH

0 y z H Y $Hz FHz H:N-C-CONKC-CONH-C-CONH-C-CONH-C-CC~ m3* A k k A k

A. 0.04 0.18 0.23 1.0 B. 0.39 0.29 0.29 0.36 1.0 C. 010 0.11 0.18 0.49 1.0

Fig. 10. Gd(II1)-induced relaxation enhancement of the C-a protons irrlutive to that of Met-5 residue) of the dipolarform of Met-enke- phrilin ( A ) in ' H 2 0 solution I18 m M , pH 5.9, with Pr(III) ions] Lit 23 C, ( B ) in (CLH-,I2S0 solution (20 m M ) at 5V"C,and (C) in IC2H3)2SO~rolution (20 m M ) with hr2H4C104 ( 2 M ) at 5V'C

dipolar form of Met-enkephalin. The ratios of relaxation enhancements of C-a protons of Met-enkephalinamide, relative to that of the Met-5 residue of Met-enkephalin, were as small as 0.02 (Gly-2) and 0.01 (Gly-3, Phe-4 and Met-5); these observations indicate that the Gd(II1) ion is certainly bound to the carboxylate group of the Met-5 residue of Met- enkephalin but not specifically to any group of the cationic form of Met-enkephalinamide.

Gd(II1)-induced relaxation measurements on the (C2H3)2S0 solution (Fig. 10B) were made at 50 "C so that the C-a proton resonance of Tyr-1 was separated from those of Gly-2 and Gly-3 residues. The ratios of Gd(II1)-induced relaxation enhancements of the dipolar form in (C2H3)2S0 solution were also measured in the presence of NH4C104, for elucidating the effect of salt addition on the conformation of the dipolar form of Met-enkephalin.

h'uclear Overhauser Effects

Nuclear Overhauser effects were measured for the amide and C-a protons of the dipolar form of Met- enkephalin and of the cationic form of Met-enkephalinamide in (C2H3)2S0 solution. The temperature of the solution was raised to 50 "C so as to reduce the correlation time (z,) and thus to increase nuclear Over- hauser effect values [34,35].

DISCUSSION

Molecular Conformation of Met-enkephalin in Aqueous Solution

The pH dependences of proton chemical shifts are useful for determining pK, values of ionizable groups and also for studying interactions between the N-terminal and C-terminal groups of linear oligopeptides in solution. For Met-enkephalin in aqueous solution, pH dependences of the chemical shifts of

amide protons and C-a protons are shown in Fig.2. The chemical shift of the C-a proton of the Tyr-1 residue is affected by the titration of the N-terminal amino group but not by the titration of the C-terminal carboxyl group. Similarly the chemical shift of the amide and C-a protons of the Met-5 residue are affected by the titration of the C-terminal carboxyl group but not by the titration of the N-terminal amino group. These observations indicate that the N-terminal Tyr-1 and C-terminal Met-5 residues are not in close proximity to each other and accordingly unassociated extended conformation is predominant at the concentration of 10 mM in 2 H z 0 solution (Fig. 11).

This conclusion is consistent with the large temperature coefficients of the amide proton chemical shifts at pH 2.6 [12,16] and at pH 1.5 (cationic form) (Table 2). In the present study, however, temperature coefficients of amide proton chemical shifts were also measured for Met-enkephalin in HzO solution at pH 5.9 (dipolar form), with the Correlation Digital Sweep Program (Table 2). For both the dipolar form and the cationic form, the large negative temperature coefficients of the amide proton chemical shifts indicate that the amide protons of Met-enkephalin in aqueous solution are exposed to solvent and are not involved in strong intramolecular or intermolecular hydrogen bonding. In these solutions the chemical shifts of the C-a protons of Met-enkephalin are practically independent of temperature (Table 3).

Molecular Association and Conformation of Met-enkephalin in 2H20/(CzH3)2S0 Solution

The pH dependences of C-H proton resonances of Met-enkephalin have been observed previously in 'H20/(C2H3)2S0 (2/1) solution at the concentration of 87 mM [ l l ] and those of I3C resonances in 'H20/ (C2H3)2S0 ( l / l ) solution at the concentration of 100mM [18]. The chemical shifts of proton and carbon resonances of the Tyr-1 residue are affected by the titration of the C-terminal carboxyl group while those of the Met-5 residue are affected by the titration of the N-terminal amino group. These observations indicate the proximity of the N-terminal and C-terminal residues and have been considered to be evidence for the formation of a folded conformation.

Actually, however, these pH dependences at the high concentrations of about 100 mM are found, in the present study, to be due to the intermolecular association rather than to molecular folding of Met- enkephalin. At the low concentration of 3 mM in 'H20/(CZH3)2SO (2jl) solution (Fig. 3) the proton chemical shifts of the Tyr-1 and Met-5 residues are not affected by the titration of the C-terminal and N-terminal groups, respectively. These observations indicate that, at this low concentration, the N-termi-

T. Higashijima. J. Kobdyashi, U. Nagai, and T. Miyazawa 53

nal and C-terminal residues are not in proximity and accordingly the molecule of Met-enkephalin does not appreciably form folded conformations.

Molecular Association of the Cationic Form of Met-enkephalin and Met-enkephalinamide in ( C2 H3) 2SO Solution

For the cationic form of Met-enkephalin in (C2H3)2S0 solution, large temperature coefficients have been reported previously for the chemical shifts of amide proton resonances [12,13,16]. These are confirmed in the present study, as shown in Table 2, and furthermore large temperature coefficients are also observed for the cationic form of Met-enkephalinamide in (C2H3)2S0 solution at the concentration of 5 mM. These observations indicate that the amide protons of the cationic form of Met-enkephalin and Met-enkephalinamide are exposed to solvent and are not involved in strong intramolecular or intermolecular hydrogen bonding at the concentration of 5 mM. In these solutions the chemical shifts of the C-cr protons depend little on concentration (Table 3).

As the concentration is raised in (C2H3)2S0 solution, however, concentration dependences of chemical shifts are observed as shown in Fig.4B and C, similar to the case of melanostatin. In our previous studies on melanostatin (Pro-Leu-Gly-NH2) [24,25], the cationic form prepared with HCl has been found to form molecular aggregate in (C2H3)2S0 solution. The chemical shifts of the leucine amide proton and the glycine carboxamide carbon resonances exhibit significant concentration dependences. This observation indicates the formation of an intermolecular hydrogen bond between the amide proton of the second residue Leu-2 and the carboxamide oxygen atom. Also in the present case of the cationic form of Met-enkephalin and Met-enkephalinamide, the chemical shift of the amide proton of the Gly-2 residue shows the largest concentration dependences (Fig. 4B, C), suggesting that again the amide proton of the second residue is involved in intermolecular hydrogen bonding. On the other hand, no carbonyl carbon resonances show significant concentration dependence of chemical shifts, in contrast to the case of the cationic form of melanostatin. However, in the present case of Met-enkephalin and Met-enkephalinamide, the chemical shift changes of amide proton and carbonyl carbon resonances are possibly due to the ring-current effects in addition to the direct effects of intermolecular hydrogen bonding. These effects appear to be additive for the amide proton resonances of the Gly-2 residue but subtractive for the carbon resonances of the carbonyl group involved in intermolecular hydrogen bonding. For analyzing the concentration dependences of chemical shifts of linear peptide molecules bearing aromatic side-chains, it is

thus important to take the ring-current effects into account.

Molecular Association of the Dipolar Form of Met-enkephalin in (C2H3)2S0 Solution; Ring- Curren t Eflects

For the dipolar form of Met-enkephalin in (C2H3)2S0 solution, significant molecular association has been found from concentration dependences of the chemical shifts of amide proton resonances [17]. In fact, as the concentration of Met-enkephalin is raised from 0.3 mM to 220 mM (Fig.4A), the amide proton resonances of the Gly-2 and Gly-3 residues show larger low-field shifts (0.17 ppm and 0.20 ppm respectively), than those of the Phe-4 (-0.06 ppm) and Met-5 residues (0.05 ppm). These observations appear to suggest that the amide protons of the Gly-2 and Gly-3 residues are involved in intermolecular hydrogen bonding in (C2H3)2S0 solution. If this is the case, the temperature coefficients of the chemical shifts of these amide protons will be decreased as the concentration is raised, as observed previously for melanostatin [24] and related peptides [25].

Actually, however, the negative temperature coefficient for the amide proton of the Gly-3 residue is appreciably increased as the concentration is raised (Table 2). Thus, the concentration dependences of the amide protons of Met-enkephalin in (C2H3)2S0 solution are ‘anomalous’ and may not be explained simply by ‘normal’ intermolecular hydrogen bonding.

Surprisingly, concentration dependences are also found, in the present study, for the chemical shifts of C-a protons of Met-enkephalin (dipolar form) in (C2H3)2S0 solution. As the concentration of Met- enkephalin is raised from 0.3 mM to 220 mM (Fig. 4A), significant low-field shifts are observed for the Tyr-I ( > 0.1 1 ppm), Gly-2 and Gly-3 (0.1 1 pprn), Phe-4 (0.08 ppm) and Met-5 residues (0.05 ppm). These concentration dependences may not be explained by intermolecular N - H . 0 = C hydrogen bonding but are possibly due to the ring current effects of Tyr-I and/or Phe-4 residues in molecular aggregates of Met-enkephalin.

The concentration dependences of the chemical shifts of amide protons (Fig.4A) may also be due, at least in part, to the aromatic ring-current effects; especially the ‘anomalous’ upfield shift of the amide proton resonance of the Phe-4 residue, on molecular association, may not be explained by intermolecular hydrogen bonding and accordingly is largely due to the ring-current effects in molecular aggregates. These significant aromatic ring-current effects on amide and C-a protons suggest that the molecular aggregates of the dipolar form of Met-enkephalin is not of a ‘linear chain’ type but rather is of ‘antiparallel cross-8’ type [17]; molecular aggregates of this type are pos-


Folded conformer

W

2 Y

Extended conformer

A..A

Fig. 1 t . Schematic pictures of conformation (main chain) and as- sociaiion equilibria of the dipolar form of Met-enkephalin, ( A ) in I C H 3 ) z S 0 solution and ( B ) in HzO solution

sibly stabilized by the attraction, at either terminal of adjacent molecules, between the positively charged N-terminal amino group and the negatively charged C-terminal carboxylate group (Fig. 11).

In fact, because of the molecular association by the interaction of the N-terminal amino group and C-terminal carboxylate group, the concentration- dependent shifts of the dipolar form of Met-enkephalin are largest for the carbonyl and a-carbons of Tyr-1 and Met-5 residues (Fig.5A). However, the concentration-dependent shifts are also large for the carbonyl and a-carbon resonances of the Gly-2, Gly-3 and Phe-4 residues; these shifts are also possibly due to the ring-current effects from the aromatic residues in molecular aggregates of the dipolar form at higher concentration in (C2H3)2S0 solution.

Anomalous Temperature Dependences of Proton Chemical Shifts of the Dipolar Form ojhilet-enkephalin ; Conformational Equilibrium

For the dipolar form of Met-enkephalin in (C2H3)2S0 solution, a small temperature coefficient has been observed of the chemical shift of the Met-5 amide proton resonance [9,10,13,14,17]. However, the temperature dependence of the amide proton of the Met-5 residue is not linear, as shown in Fig. 6, and furthermore C-a protons of the Phe-4 and Met-5 residues show significant positive temperature dependences while the C-a proton of the Tyr-1 residue show negative temperature dependence (Fig. 6 and Table 3). These anomalous temperature dependences are also observed for the dipolar form in C2HC13/ (C2H3)2S0 (2/1) solution (Table 3).

The anomalous temperature dependences of the proton chemical shifts of the dipolar form are possibly due to the association equilibrium and to the conformational equilibria of the main-chain and side- chains of Met-enkephalin. If only the former equilibrium is the case, the anomalous temperature dependences should be consistent with the concentration dependences of proton chemical shifts. As

shown in Fig.4A, the C-a proton resonances of Tyr-1, Phe-4 and Met-5 residues are all shifted upfield as the solution is diluted (in other words, as the degree of molecular association is reduced). Accordingly, the temperature coefficients of all these protons are expected to be negative. Actually, however, the C-a proton resonances of the Phe-4 and Met-5 residues show appreciable positive coefficients of 2.3 x lop3 and 3.4 x ppm/"C, respectively (Fig. 6 and Table 3). Furthermore, if the temperature dependences of C-a proton chemical shifts are due to the association equilibrium only, the temperature coefficients should be maximum for the equimolar fractions of Met-enkephalin molecules in the free and associated species (at the concentration range of 60 - 80 mM, as seen in Fig. 4A). Actually, however, the temperature coefficients of the chemical shifts of the C-a protons become larger as the (C2H3)2S0 solution is diluted. Accordingly the anomalous temperature dependences of proton chemical shifts of the dipolar from at the concentration of about 10 mM are more effectively due to the conformational equilibria than to the association equilibrium (Fig. 11).

As for the conformational equilibrium about the C-a and C-8 bond of the side-chain of the Phe-4 residue, the temperature dependences of the vicinal spin coupling constants (JabR and Jnps) were measured previously [14,27]. As the temperature is raised, the fractional population of the rotamer I (gauche-gauche) is decreased while those of the rotamers I1 (trans- gauche) and I11 (gauche-trans) are increased. However, similar temperature dependences of fractional po- pulations of the rotamers of the Phe-4 side-chain are also observed for the cationic form of Met- enkephalin in (C2H3)2S0 solution [36], which does not show anomalous temperature dependences of C-a proton chemical shifts. Accordingly, the conformational equilibria of the aromatic side-chains do not appear to contribute directly to the anomalous temperature dependences of the dipolar form of Met-enkephalin. In fact, for the model molecules of N-acetylphenylalanine methylamide and N-acetyltyrosine methylamide in (C2H3)2S0 solution, the temperature coefficients of C-a proton chemical shifts are as small as 0.5 x lop3 and 0 . 4 ~ lop3 ppm/"C respectively.

Accordingly the anomalous temperature dependences of the C-a proton chemical shifts of the dipolar form at low concentration are largely due to the conformational equilibria of the main chain. Since the molecule of Met-enkephalin has two aromatic residues, Tyr- 1 and Phe-4, the aromatic ring-current effects from these two residues will be different for different conformers. Accordingly, if the enthalpy differences between various conformers are of the order of the thermal energy (k7') around the temperature of NMR measurements, appreciable tem-

T. Higashijirna, J. Kobayashi, U . Nagdi, and T. Miyazawa 55

perature dependences of chemical shifts may be expected for the C-a proton resonances of the dipolar form. The non-linear temperature dependence of the amide proton of the Met-5 residue may also be explained as being due to the change in conformational equilibrium of the dipolar form at low concentration.

The conformational equilibrium of the dipolar form is also supported from the anomalous solvent- composition dependences of the amide and C-a proton chemical shifts (Fig.8) and carbonyl and r-carbon chemical shifts (Fig. 9). At the concentration as low as about 10 mM, the effect of intermolecular association is not very significant. Accordingly, if it were not for conformational changes, the chemical shifts of these resonances should vary monotonously with the solvent composition changes from 100% (C2H3)2S0 to 100% 'H20. Actually, however, the solvent-composition dependences of chemical shifts are far from being monotonous for the amide protons of Gly-2, Gly-3, and Met-5, the C-cc proton of Met-5 (Fig.8) and carbonyl carbon of Tyr-1 and a-carbons of the Tyr-1, Gly-2 and Gly-3 residues (Fig. 9). These observations support the changes in conformational equilibrium as induced by the solvent composition changes of 2H20/(C2H3)2S0 solution [30].

Gd( I I I ) -Induced Relaxation and Molecular Conformations of the Dipolar Form of Met-enkephalin in ( C2H3)2S0 Solution

Gd( 111)-induced proton relaxation enhancements are useful for studying molecular conformations in solution. In our previous study [37] such relaxation measurements were made of the molecule of Z-Gly- Pro-Leu-Gly-Pro (Z = benzyloxycarbonyl) in which the Gd(II1) ion is bound to the C-terminal carboxylate group. The relaxation enhancement of the CH2 protons of the benzyloxycarbonyl group is much larger than those of the C-a protons of Gly-1 and the C-0 protons of the Leu-3 residue, giving strong support to the predominance of the folded conformation in 'H20 solution [37].

Similarly for the dipolar form of Met-enkephalin, the Gd(II1) ion is bound to the C-terminal carboxylate group and the relaxation enhancements are pro- portional to the average values of the inverse sixth power of the distances from this Gd(II1) ion to observed nuclei. In (C2H3)2S0 solution (Fig. lOB) the relaxation ratio of the C-a proton of the Tyr-1 residue is larger than those of the C-cc protons of the Gly-2, Gly-3 and Phe-4 residues. This observation directly indicates that there is an appreciable fraction of folded conformation(s) in which the Tyr-1 residue is in the proximity of the C-terminal carboxylate group. The folded conformation of the dipolar form in (C2H3)2S0 solution is presumably stabilized by the interaction between the positively charged N-ter-

minal group and the negatively charged C-terminal group (Fig. 11).

On the other hand, for the dipolar form of Met- enkephalin in 2H20 solution (Fig. lOA), the relaxation enhancement of the C-cc proton of the Tyr-1 residue is much smaller than those of the Gly-2 and Gly-3 residue. This observation directly indicates that the dipolar form predominantly takes extended conformation(s) in 2H20 solution. In 2H20 solution the attraction between the positively charged N-terminal group and the negatively charged C-terminal group of the dipolar form is much weakened by intervening water molecules.

This conclusion is further supported from the analysis of the effect of NH4C104 on Gd(II1)-induced relaxation enhancements of the dipolar form in (C2H3)2S0 solution. In the presence of 2 M NH4C104, the relaxation enhancements of the C-cc protons are monotonously decreased: Tyr-1 < Gly-2 < Gly-3 < Gly-4 < Met-5 (Fig.10C). This observation indicates that the fraction of folded conformation(s) is much decreased by the addition of NH4C104 to the solution of the dipolar form in (CzH3)zSO solution. For elucidating the effect of NH4C104 on intramolecular hydrogen bonding of peptide molecules, the temperature dependences of chemical shifts were measured, in the present study, for the amide protons of valine, leucine, phenylalanine and ornithine residues of gramicidin S. However, the temperature coefficients of all the amide protons were hardly changed by the addition of NH4C104 (1 M), indicating that the presence of NH4C104 does not appreciably affect intramolecular hydrogen bonding. Instead, the presence of NH4C104 is expected to weaken the attraction between the N-terminal group and C-terminal group of the dipolar form in solution and then the extended conformation is predominant. This is consistent with the effect of NH4C104 on the temperature dependence of proton chemical shifts; on addition of NH4C104, the anomalous temperature dependences of amide and C-a proton chemical shifts are eliminated (Fig. 6B). and the amide proton resonances of all residues show large negative temperature coefficients. These observations indicate that, on addition of NH4C104, the folded conformers of the dipolar form are largely converted to extended conformers.

Folded Conformer of' the Dipolar Form of Met-enkephalin in (C2H3)2S0 Solution

Main-chain conformations of the dipolar form of Met-enkephalin may also be studied from the analyses of nuclear Overhauser effect enhancements [35,38, 391. On irradiation of the amide proton of the Met-5 residue, the intensity of the C-a proton resonance of the Phe-4 residue is enhanced by 7%, which is larger


than the enhancement (4%) of the intensity of the C-r proton of the Met-5 residue. These nuclear Overhauser effect enhancements are consistent with the local conformation of the (Phe-4)CHa-CO-NH- CH,(Leu-5) moiety found in the crystal of Leu- enkephalin [24].

For the dipolar form of Met-enkephalin, the small temperature coefficient of the amide proton chemical shift of the Met-5 residue (0 to - 1.2 x ppm/"C) has been taken as the evidence for the strong intramolecular hydrogen bond (Gly-2) C = 0 . H - N (Met-5) of the type-I p-turn structure [9,10,13,14,17]. Actually, however, the temperature coefficients of proton chemical shifts of the dipolar form are subject to temperature-dependent ring-current effects. Ac- cordingly for the purpose of approximately correcting for the ring-current effects, the temperature coefficients (3.4 x ppm/"C) of the C-a proton may be subtracted from that of the amide proton of the Met-5 residue. Then, the 'intrinsic' temperature coefficient of this amide proton is estimated to be as large as (- 3.4 to -4.6) x lop3 ppm/*C, indicating that this amide proton is not primarily involved in intramolecular hydrogen bonding of the p-turn structure in (C2H3)2S0 solution. In fact, the large nuclear Overhauser effect enhancement of the C-a proton of Phe-4 upon irradiation of the amide proton of Met-5 is not consistent with the p-turn structure with the intramolecular hydrogen bond (Gly-2) C = 0 . H - N (Met-5) ; in this p-turn structure, the distance between the C-r proton of Phe-4 and amide proton of the Met-5 residue is as long as 0.5 nm and the maximum nuclear Overhauser effect value is expected to be smaller than 3 % [41].

For the cationic form of Met-enkephalinamide in (C'HJ)ZSO solution, the intensity of the C-u proton of the Phe-4 residue is enhanced by as much as 11 yo, upon irradiation of the amide proton of the Met-5 residue, much larger than the nuclear Over- hauser effect enhancement (7 %) of the C-a proton of the Met-5 residue. This large enhancement is not consistent with the predominance of p-turn structures. In fact, no evidence is found, in the present study, for the presence of folded conformation(s) of Met- enkephalinamide in (C2H3)2S0 solution.

CONCLUSION

Linear oligopeptides with free N and C-terminal groups are in the dipolar form in 'H20 solution at physiological pH. The dipolar forms of linear oligopeptides appreciably form molecular aggregates at high concentrations in (CLH3)2S0 solution, because of the intermolecular attraction between the positively charged N-terminal group and negatively charged C-terminal group. Accordingly for studying the molec-

ular conformation of free molecules of the dipolar form, it is important to analyze NMR spectra at sufficiently low concentrations (in a physiological environment, the concentration of physiologically active peptides is extremely low).

In the present study significant concentration dependences of chemical shifts were found for the C-a protons of the dipolar form of Met-enkephalin in (C2H3)2S0 solution. For such oligopeptide molecules with aromatic residues, the effects of the aromatic ring-current need be taken into account in the analyses of the concentration-dependent chemical shifts of amide protons as well as of C-a protons. Furthermore, in the present study significant temperature dependences were found for the chemical shifts of C-r protons of the dipolar form of Met-enkephalin in (C2H3)2S0 solution at low concentrations. From the analyses of temperature dependences together with concentration dependences of C-a proton resonances, the dipolar form of Met-enkephalin was found to be in an equilibrium of folded and extended conformations, although the folded conformation is not predominant. For the dipolar form of oligopeptides with aromatic residues, the temperature coefficients of amide proton chemical shifts are possibly affected by conformation-dependent ring-current effects, so that a small temperature coefficient of amide proton chemical shift is not necessarily a direct indication of the formation of intramolecular hydrogen bonds. For such cases, Gd(II1)-induced proton relaxation enhancements and nuclear Overhauser effect enhancements provide important information of overall conformations and local conformations respectively.

For comparison, in the present study NMR analyses were also made of the cationic form of Met- enkephalinamide. No evidence was found for the formation of a folded conformation. Accordingly the opiate activity of Met-enkephalinamide is not associated with the formation of a specific conformation in solution but is perhaps due to the ability to take a specific conformation at the receptor site. Similarly, the molecule in the dipolar form of Met-enkephalin does not predominantly take a unique conformation in solution. This is consistent with the recent finding that Met-enkephalin associates and dissociates from the opiate receptor with 8- 10-fold slower kinetics than opiates [41].

REFERENCES 1. Hughes, J., Smith, T. W., Kosterlitz, H. W., Fothergill, L. A..

Morgan, B. A. Sr Morris, R. H. (1975) Nature (Lond.) 258.

2. Szekely, J. L., Ronai, A. Z., Dunai-KovBcs, Z., Miglecz, E., Berzetri, I., Bajusz, S. B Graf, L. (1977) Euu. J . Plzarrnacol.

3. Pert, C. B., Pert, A., Chang, J.-K. Sr Fong, B. T. W. (1976)

511 - 579.

43,293 - 295.

Science (Wash. D.C.) 194, 330-332.

T. Higashijima, J . Kobayashi, U. Nagai, and T. Miyazawa 51

4. Roemer, D., Buescher, H. H., Hill, R. C., Pless, J . , Bauer, W.. Cdrdinaux, F., Closse, A,, Hauser, D. & Huguenin, R. (1977) h’ature (Lond.) 268, 547 - 549.

5. Coy, D. H., Kastin, A. J., Schally, A. V., Morin, O., Caron, N. G., Labrie, F., Walker, J. M., Fertel, R., Berntson, G. G. & Sandman, C. A. (1976) Biochem. Biophys. Res. Commun. 73, 632 - 637.

6. Chang. J.-K., Fong, B. T. W., Pert, A. &Pert , C . B. (1976) Lifi.

7. Bradbury, A. F., Smyth, D. G. & Snell, C. R. (1976) Nature (Lond.) 260, 165-166.

8. Horn, A. S. & Rodgers, J . R. (1976) Nature (Lond.) 260, 795 - 797.

9. Roques, B. P., Jaureguiberry, C . G., Oberlin, R., Anteunis, M. & Lala, A. K. (1976) Nature (Lond.) 262, 778-779.

10. Jaureguiberry, C. G., Roques, B. P., Oberlin, R., Anteunis, M. & Lala, A. K. (1976) Biochem. Biophys. Res. Commun. 71, 558 - 565.

11. Anteunis, M., Lala, A. K., Jaureguiberry, C. G. & Roques, B. P. (1977) Biochemistry, 16, 1462- 1466.

12. Bleich, H. E., Day, A. R., Freer, R. J. & Glasel, J . A. (1977) Biochem. Biophys. Res. Commun. 74, 592 - 598.

13. Jones, C. R., Garsky, V. & Gibbons, W. A. (1977) Bioc,hem. Biophys. Res. Commun. 76, 619-625.

14. Jones, C. R., Gibbons, W. A. & Garsky, V. (1976) Nature (Lond.) 262, 779-782.

15. Jaureguiberry, C . G., Roques, B. P., Oberlin, R., Anteunis, M. & Combrisson, S. (1977) FEBS Lett. 76, 93-98.

16. Bleich, H. E., Cuttnell, J. D., Day, A. R., Freer, R. J., Glasel, J . A. & McKelvy, J. F. (1976) Proc. Natl Acad. Sci. U.S.A.

17. Khaled, M. A., Long, M. M., Thompson, W. D., Bradley, R. J., Brown, G. B. & Urry, D. W. (1977) Biochem. Biophys. Res. Commung. 76, 224-231.

18. Combrisson. S., Roques, B. P. & Oberlin, R. (1976) Tetra- hedron Lett. 38, 3455 - 3458.

19. Tancrede, P., Deslauriers, R., McGregor, W. H., Ralston, E., Sarantakis, D., Somorjai, R. L. & Smith, I . C. P. (1978) Bio- chemistry, 17, 2905-2914.

20. Spirtes, M. A., Schwartz, R. W., Mattice, W. L. & Coy, D. H. ( I 977) Biochem. Biophys. Res. Commun. 81. 602 - 609.

21. Kopple, K. D., Ohnishi, M. & Go, A. (1969) Biochemistry, 8 ,

Sci. 18, 1473- 1482.

73,2589 - 2593.

4087 - 4095.

22. Ohnishi, M. & Urry, D. W. (1969) Biochem. Biophys. Res. Commun. 36, 194-202.

23. Smith, G. D. &Griffin, J . F. (1978) Science (Wash. D.C.) 199, 1214-1216.

24. Higashijima, T., Tasumi, M. & Miyazdwa, T. (1975) FEBS Lett. 57, 175-178.

25. Higashijima, T., Tasumi, M., Miyazawa, T. & Miyoshi, M. (1978) Eur. J . Biochem. 89, 543- 556.

26. Kobayashi, J . & Nagai, U. (1978) in Peptide Chemistry IY77 (Shiba, T., ed.) pp. 91 -96, Protein Research Foundation, Osaka.

27. Higashijima, T., Tasumi, M., Miyazawa, T., Kawai, M. & Nagai, U. (1978) in Pepride Chemistry 1977 (Shiba, T., ed.) pp. 97- 102, Protein Research Foundation, Osaka.

28. Pitner, T. P. & Urry, D . W. (1972) J . Am. Chem. Soc. 94,

29. Kumar, N. G. & Urry, D. W. (1973) Biocltemistry, 12, 4392-

30. Glickson, J. D., Urry, D. W. & Walter, R. (1972) Proc. h’atl

31. Sogn, J . A., Craig, L. C. & Gibbons, W. A. (1974) J . An7.

32. Urry, D . W., Mitchell, L. W. s( Ohnishi, T. (1974) Proc. Nut1

33. La Mar, G. N. & Faller, J. W. (1973) J . Am. Chem. Soc. 95,

34. Noggle, J . H. & Schirmer, R. E. (1971) The h’uclear 0verhau.w

35. Glickson, J. D., Gordon, S. L., Pitner, T. P., Agresti, D. G.

36. Kobayashi, J., Nagdi, U., Higashijima, T. & Miyazawa. T.

37. Higashijima. T., Tasumi, M. & Miyazawa. T. (1977) in Peptide Chemistry 1976 (Nakajima, T., ed.) pp. 89-92, Protein Research Foundation, Osaka.

38. Khaled, Md. A. & Urry, D. W. (1976) Biochem. Biophys. Rrs. Commun. 70,485 - 491.

39. Rae, I . D., Stimson, E. R. & Scheraga, H. A. (1977) Biochem. Biophys. Res. Commun. 77, 225-229.

40. Bell, R. A. & Saunders, J. K. (1970) Can. J . Chem. 48, 1114- 1122.

41. Simantov, R., Childers, S. R. & Snyder, S. H. (1978) Eur. J . Pharmacol. 47, 319-331.

1399 - 1400.

4398.

Atad. Sci. U.S.A. 69, 2566-2569.

Chem. Soc. 96, 3306-3309.

Acud. Sci. U.S.A. 71, 3265 - 3269.

3817 - 3818.

&fleet, Academic Press, New York.

& Walter, R. (1971) Biochemistry, 15, 5721 -5729.

(1 979) Biochim. Biopliys. Acta, 577, 196 - 207.

T. Higashijima and T. Miyazawa*, Department of Biophysics and Biochemistry, Faculty of Science, University of Tokyo, Hongo, Bunkyo-ku, Tokyo-to, Japan 11 3

J. Kobayashi and U. Nagai, Mitsubishi-Kasei Institute of Life Sciences, 11 Minamiooya, Machida-shi, Tokyo-to, Japan 194 ~ ~~

* To whom correspondence should be addressed.

Documents

Nuclear-Magnetic-Resonance Study on Met-enkephalin and Met-enkephalin : Molecular Association and Conformation