Development of a New Family of Conformationally Restricted Peptides as Potent Nucleators of β-Turns. Design, Synthesis, Structure, and Biological Evaluation of β-Lactam Peptide Analogue of Melanostatin

Published on Web 12/06/2003

Development of a New Family of Conformationally Restricted

Peptides as Potent Nucleators of â-Turns. Design, Synthesis,

Structure, and Biological Evaluation of a â-Lactam Peptide

Analogue of Melanostatin

Claudio Palomo,*^,†Jesus M. Aizpurua,*^,†Ana Benito,^†Jose´ Ignacio Miranda,^†

Raluca M. Fratila,^†Carlos Matute,^‡Maria Domercq,^‡Federico Gago,^§

Sonsoles Martin-Santamaria,^§and Anthony Linden^|

Contribution from the Departamento de Qu´ımica Orga´nica I. Facultad de Qu´ımica.

UniVersidad del Pa´ıs Vasco. Apdo. 1072, 20080 San Sebastia´n, Spain

Received August 28, 2003; E-mail: qoppanic@sc.ehu.es

Abstract: Novel enantiopure (i)-(â-lactam)-(Gly)-(i+3) peptide models, defined by the presence of a central

R-alkyl-R-amino-â-lactam ring placed as the (i+1) residue, have been synthesized in a totally stereocontrolled

way by R-alkylation of suitable N-[bis(trimethylsilyl)methyl]-â-lactams. The structural properties of these

â-lactam pseudopeptides have been studied by X-ray crystallography, Molecular Dynamics simulation,

and NOESY-restrained NMR simulated annealing techniques, showing a strong tendency to form stable

type II or type II′ â-turns either in the solid state or in highly coordinating DMSO solutions. Tetrapeptide

models containing syn- or anti-R,â-dialkyl-R-amino-â-lactam rings have also been synthesized and their

conformations analyzed, revealing that R-alkyl substitution is essential for â-turn stabilization. A â-lactam

analogue of melanostatin (PLG amide) has also been prepared, characterized as a type-II â-turn in DMSO-

d₆solution, and tested by competitive binding assay as a dopaminergic D₂modulator in rat neuron cultured

cells, displaying moderate agonist activity in the micromolar concentration range. On the basis of these

results, a novel peptidomimetic design concept, based on the separation of constraint and recognition

elements, is proposed.

Introduction and Objectives

based pharmacophore design has become a central topic in

medicinal chemistry.⁷

Turns, defined as sites where a peptide chain reverses its

overall direction, are common motifs in protein structures.¹

When direction reversal occurs over four residues in such a way

that the carbonyl oxygen atom of the first residue (i) and the

amide NH proton of the fourth residue (i+3) come close to

each other in space, a â-turn is formed.²Furthermore, in the

case of â-turns of types I, II and III, this conformation involves

formation of an intramolecular hydrogen bond between residues

(i) and (i+3) to give a pseudo-10-membered ring.³From a

structural chemistry viewpoint, â-turns are critical to protein

conformational stability⁴and many protein-protein interac-

tions.⁵In addition, turns are often a preferred site for degradation

by proteolytic enzymes.⁶Hence, it is not surprising that â-turn-

Synthesis of molecular templates for â-turns with precisely

oriented predetermined groups is a complex task, and a great

deal of effort has been dedicated to the development of such

peptidomimetics.^6,8Basically, a rational design of an efficient

mimetic requires the formal modification of the native bioactive

peptide to include two elements. First, a specific â-turn

(3) Four essential structural parameters define canonical â-turns: (a) donor/

acceptor distance _(i+3)NH‚‚‚O)C_(i)of the turn-stabilizing hydrogen bond

(typically, 1.8-2.5 Å for H‚‚‚O distance and 2.6-3.2 Å for N‚‚‚O distance);

(b) the CR_(i)-CR_(i+3)distance (<7 Å); (c) pseudodihedral angle δ

encompassing four consecutive CR atoms of the turn-forming amino acids

(ideal value: 50° > δ > -50°); (d) φ and ψ dihedral angle sets at the

-(i+1)-(i+2)- positions. See: (a) Robinson J. A. Synlett 2000, 4, 429-

441. (b) Mu¨ller, G.; Hessler, G.; Decornez, H. Y. Angew. Chem., Int. Ed.

2000, 39, 894-896.

(4) (a) Northtup, S. H.; Pear, M. R.; Morgan, J. D.; McCammon, J. A.; Karplus,

M. J. Mol. Biol. 1981, 153, 1087-1109. (b) Jasko´lski, M.; Tomasselli, A.

G.; Sawyer, T. K.; Staples, D. G.; Heinrikson, R. L.; Schneider, J.; Kent,

S. B. H.; Wlodawer, A. Biochemistry 1991, 30, 1600-1609.

(5) For a review, see: Burgess, K. Acc. Chem. Res. 2001, 32, 826-835.

(6) For detailed information on this subject: Arnold, U.; Hinderaker, M. P.;

Nilsson, B. L.; Huck, B. R.; Gellman, S. H.; Raines, R. T. J. Am. Chem.

Soc. 2002, 124, 8522-8523.

^†Universidad del Pa´ıs Vasco.

^‡Biological assays: Departamento de Neurociencias, Facultad de Me-

dicina, Universidad del Pa´ıs Vasco, 48940 Leioa, Spain.

^§MD calculations: Departamento de Farmacolog´ıa. Facultad de Me-

dicina. Universidad de Alcala´ de Henares. 28871 Madrid, Spain.

^|X-ray crystallography: Organisches-chemisches Institut der Universita¨t

Zu¨rich. Winterthurerstrasse 190, CH-8057, Zu¨rich, Switzerland.

(1) For a classification of turns, see: Wilmot, C. M.; Thornton, J. M. Protein

Eng. 1990, 3, 479-493.

(7) For a review, see: Souers, A. J.; Ellman, J. A. Tetrahedron 2001, 57, 7431-

7448.

(8) General reviews on peptidomimetic synthesis: (a) Synthesis of Peptides

and Peptidomimetics. In Houben-Weyl, Methods of Organic Chemistry;

Felix, A., Moroder, L., Toniolo, C., Eds.; Thieme: Stuttgart, New York,

2003; Vol. E22c. (b) Hanessian, S.; McNaughton-Smith, G.; Lombart, H.-

G.; Lubell, W. D. Tetrahedron 1997, 53, 12789-12854. (c) Gante, J.

Angew. Chem., Int. Ed. Engl. 1994, 33, 1699-1720. (d) Giannis, A.; Kolter,

T. Angew. Chem., Int. Ed. Engl. 1993, 32, 1244-1267.

(2) Reviews: (a) Rizo, J.; Gierasch, L. M. Annu. ReV. Biochem. 1992, 61,

387. (b) Rose, G. D.; Gierasch, L. M.; Smith, J. A. AdV. Protein Chem.

1985, 37, 1-109. (c) Schneider, J. P.; Kelly, J. W. Chem. ReV. 1995, 95,

2169-2187. (d) Serrano, L. AdV. Protein Chem. 2000, 53, 48-85.

9

J. AM. CHEM. SOC. 2003, 125, 16243-16260

16243

A R T I C L E S

Palomo et al.

Figure 1. Freidinger’s basic principle for â-turn peptidomimetic design.

constraining element (e.g., an intra- or interresidual bridge) is

needed to force the peptide backbone to overlay the desired

conformation as exactly as possible. Second, at least one

recognition group must be placed in a stereocontrolled and

synthetically easy way at the desired position for interaction

with the receptor or enzyme active site. Though formally simple

at a design level, practical access to peptidomimetic libraries

bearing a broad range of recognition groups, while retaining

the â-turn framework, often conflicts with the synthetic reli-

abilities arising from the structure of the scaffolds chosen.

Indeed, it is difficult or impossible to elucidate structural

elements exerting specific constraints or recognition functions

for most known â-turn surrogate molecules. As a consequence,

chemical alteration of such peptidomimetics, when synthetically

feasible, leads to unpredictable results in terms of â-turn motif

stability and/or proper spatial orientation of the recognition

groups. An important example of this phenomenon (Figure 1)

is the original approach of Freidinger⁹to â-turn mimetics,

involving the bridging of the betagenic -(i+1)-(i+2)- central

residues through the formation of five-, six-, seven-, and eight-

membered ring lactams 2. Accordingly, any structural modifica-

tion of the R¹bridge in pseudopeptide 2,¹⁰should force a global

change of conformation and recognition properties that, in

practice, involves a re-design of each derivative prepared, instead

of a single structural parameter variation.

The objective of this work is to propose a novel design

approach to â-turn peptidomimetics based on the principle of

the incorporation of groups, as small as possible, in the original

peptide to generate well-defined turns without affecting receptor

recognition. Herein we report in detail our investigation of the

design, synthesis, conformational characterization and biological

evaluation of a novel â-lactam peptide analogue of melanostatin

(PLG) that illustrates this concept. In addition, the identification

of the structural parameters involved in the stabilization of

â-lactam-based peptidomimetics as a novel family of â-turn

motif nucleators is also reported.

Figure 2. Principles for the design of a novel class of peptidomimetics as

nucleators of â-turn motifs: when the â-lactam methylene group mimics

simultaneously the CRH_(i+1)and HN_(i+2)protons of the native peptide, R¹

and R²groups can be designed specifically for recognition with the receptor.

simultaneously the CRH_(i+1)and HN_(i+2)protons of the native

peptide with a minimal structural divergence,¹¹thus retaining

the R¹R substituent of the original residue and facilitating a

closer recognition by receptors or enzymes. From this design,

a four membered ring with a quaternary stereogenic center at

the CR position results as a key element. In this approach, it

was also assumed that the presence of the R,R-disubstitution

pattern would enhance resistance to chemical and enzymatic

hydrolysis by proteases¹²and that the resulting â-lactam-peptides

3 would be attractive targets for pharmaceutical drug discovery.

The major difficulty in the development of â-lactam peptides

stems from the lack of methodology to create the R-quaternary

stereogenic center in â-unsubstituted azetidin-2-one rings.¹³We

recently addressed this issue¹⁴(Scheme 1) and found that the

asymmetric alkylation of the â-lactam 4 proceeds with control

of the configuration at the C₃position being dictated by the

stereogenic center of the oxazolidinone moiety to give the

R-amino-R-substituted products with high yields and excellent

(11) It has been reported that the CRH_(i+1)to HN_(i+2)interproton distance is an

essential parameter to differentiate type-I and type-II â-turns. For instance,

in Ac-Pro-Gly-Me model peptide the _(Pro)CRH‚‚‚HN_(Gly)distance is 3.5 Å

for a type-I â-turn, but only 2.1 Å for type-II conformation; see: Narasinga

Rao, B. N.; Kumar, A.; Balaram, H.; Ravi, A.; Balaram, P. J. Am. Chem.

Soc. 1983, 105, 7423-7428.

(12) Detailed information on this subject: (a) The Chemistry of â-Lactams; Page,

M. I., Ed.; Blackie Academic & Professional: New York, 1992. (b)

Chemistry and Biology of â-Lactam Antibiotics; Morin, R. B., Gorman,

M., Eds.; Academic: New York, 1982; Vols. 1-3. (c) The Organic

Chemistry of â-Lactams; Georg, G. I., Ed.; VCH: New York, 1993.

(13) For a general review on â-lactam synthesis, see: Backes, J. In Houben-

Weyl, Methoden der Organischen Chemie; Muller, E., Bayer, O., Eds.;

Thieme: Stuttgart, 1991; Band E16 B, p 31.

Results and Discussion

â-Lactam Peptide Design Plan. Our idea is based upon

consideration (Figure 2) that a methylene bridge between the

CR_(i+1)and the N_(i+2)atoms in peptide 1 should mimic

(9) (a) Freidinger, R. M.; Veber, D. F.; Perlow, D. S.; Brooks, J. R.; Saperstein,

R. Science 1980, 210, 656-658. (b) Freidinger, R. M.; Schwenk, D.; Veber,

D. F. J. Org. Chem. 1982, 47, 104-109. (c) Freidinger, R. M. J. Org.

Chem. 1985, 50, 3631-3633.

(10) For reviews on conformationally restricted peptides through short-range

cyclizations, see: (a) Toniolo, C. Int. J. Peptide Protein Res. 1990, 35,

287-300. (b) Liskamp, R. M. J. Recl. TraV. Chim. Pays-Bas 1994, 113,

1-19.

(14) (a) Palomo, C.; Aizpurua, J. M.; Galarza, R.; Benito, A.; Khamrai, U.K.;

Eikeseth, U.; Linden, A. Tetrahedron 2000, 56, 5563-5570. (b) Palomo,

C.; Aizpurua, J. M.; Benito, A.; Galarza, R.; Khamrai, U.K.; Vazquez, J.;

DePascual-Teresa, B.; Nieto, P. M.; Linden, A. Angew. Chem., Int. Ed.

1999, 38, 3056-3058.

9

16244 J. AM. CHEM. SOC. VOL. 125, NO. 52, 2003

Restricted Peptides as Potent Nucleators of â-Turns

A R T I C L E S

Moreover, the pseudodihedral angle¹⁷spanned by the four CR

atoms in the ideal type II â-turn (∼0°) was correctly mimicked

Scheme 1. (Top) Synthesis of â-Lactam Peptidomimetic 6.

(Bottom) Main Structural Parameters of â-Lactam Peptidomimetic

6 in the Solid State (X-ray Crystal Structure)^a

t

(9.3(2)°) by the sequence formed by the BuO oxygen atom,

the CR of the â-lactam ring, and the two CR atoms of the glycine

and Aib residues, respectively. Furthermore, the CdO‚‚‚H-N

hydrogen bond distance (2.37(3) Å) and the O‚‚‚H-N angle

(167(3)°) indicated the presence of an intramolecular interaction

which, on the basis of the φ and ψ dihedral angle sets at the

-(i+1)-(i+2)- positions, defined a type-II â-turn distorded by

∼15-25° at φ_(i+1), φ_(i+2), and ψ_(i+2). Solution conformational

analysis (NMR) of 6 in noncoordinating solvents (CDCl₃)

assessed the presence of a single type-II â-turn conformer at

room temperature, essentially similar to that observed in the

solid state.

In general, it is assumed that the betagenic bias of tetrapeptide

segments constrained by inter-residual lactams is primarily

determined by the central -(i+1)-(i+2)- pair and only partially

stabilized by N- and C-terminal residues. Consequently, with

the aim of establishing whether our â-turn design would be

general in scope, different enantiopure tetrapeptide models 7

(Figure 3) incorporating a central -(â-lactam)-(Gly)- pair and

achiral (i) and (i+3) termini were selected for development.

The insertion of glycine, the most flexible R-amino acid, in the

central (i+2) position was intentionally devised to prevent any

masking influence of the chiral residue on the constraining effect

of the azetidin-2-one ring. The first obvious question we

addressed was to determine whether the quaternary center at

the C₃position of the azetidin-2-one ring would be essential

for conformational restriction or not. To this end, the â-lactam-

based peptide 8 was envisaged for comparison with the above

C₃benzyl-substituted peptide 6. To simplify the solution NMR

analysis, Boc¹⁸and Aib groups were chosen, respectively, as

the (i) and (i+3) residues to provide singlet methyl signals in

the high field region, which would not interfere with the amide

NH region. Then, to rule out any possible specific influence of

the terminal Boc and Aib residues on the betagenicity of -(â-

lactam)-Gly- fragments, a â-lactam peptide 9 flanked by fully

flexible (i), (i+2), and (i+3) glycine residues was also devised.

In addition to the above issues, the substitution pattern at the

C₄position of the â-lactam ring is another concern relevant to

our design. In this instance, four diastereomeric â-lactams may

be formed and their relative configuration at the C₃and C₄

positions might influence the configurational behavior of the

corresponding tetrapeptide. To evaluate this question, â-lactam

peptides 10 and 11 with the 3,4-dialkyl substituents in a syn-

and anti-relationship, respectively, were selected as models for

the study.

^aKey: EDCl ) 1-(3-dimethylaminopropyl)-3-ethylcarbodiimide hydro-

chloride; HOBt ) 1-hydroxybenzotriazole; TsOH ) p-toluenesulfonic acid.

Ideal values indicated in parentheses.

diastereoselectivities. This approach, after deprotection of the

bis(trimethylsilyl)methyl moiety with cerium(IV) ammonium

nitrate in the resulting alkylated products and incorporation of

the carboxymethyl group at the nitrogen atom, allows access

to the required â-lactam scaffolds (e.g., 5), which can be coupled

at the N- and C-termini through conventional peptide synthesis.

At the outset of our investigation, however, it was not evident

which sort of conformation would adopt such â-lactam peptides.

Actually, the incorporation of â-lactam scaffolds into peptide

backbones has virtually been ignored¹⁵and structural informa-

tion regarding the conformation of the resulting peptide isosters

is very limited.¹⁶We were pleased to find that the first â-lactam

peptide (6) prepared using this approach presented some

remarkable structural features. For example, as deduced from

the X-ray crystal structure,^14bthe lactam carbonyl C-CO-N

angle was 91.7(2)°, considerably smaller than in parent peptides

(∼120°). The distance between the C₃and N₁atoms in the

â-lactam ring (2.085(5) Å) was also about 0.4 Å shorter than

in the ideal type II â-turn segment (CR_(i+1)‚‚‚N_(i+2)≈ 2.48 Å).

On the other hand, melanostatin or L-prolyl-L-leucylglyci-

namide (PLG) 12, an endogenous brain peptide of hypothalamic

origin possessing a type-II â-turn bioactive conformation, was

selected as a parent peptide to illustrate the “proof-of-principle”

of the proton mimic design strategy. This molecule is associated

with the modulation of dopaminergic D₂membrane receptors

in the central nervous system,¹⁹and owing to its neuroprotecting

properties, many conventional lactamic, spirolactamic, and

(15) â-Amino-desacetoxycephalosporanic acid oligopeptides have been show

to present weak in vitro inhibitory activity against porcine pancreatic

elastase, R-chymotrypsin and human immunodeficiency virus type 1

proteinase (HIV-1 PR): (a) Pitlik, J.; Bassogi, P.; Jeko, J.; Tozser, J.

Pharmazie 1996, 51, 700-704.

(16) Concurrent with our investigation, Alonso and Gonzalez have shown by

NMR studies that racemic spiro â-lactams derived from Boc-prolyl chloride

and benzaldehyde imines are able to adopt regular type II â-turn mimetics

which are stable in CDCl₃solution but collapse in DMSO-d₆. (a) Alonso,

E.; Lo´pez-Ortiz, F.; Del Pozo, C.; Peralta, E.; Mac´ıas, A.; Gonza´lez, J. J.

Org. Chem. 2001, 66, 6333-6338. See also: (b) Linder, M. R.; Podlech,

J. Org. Lett. 1999, 1, 869-871.

(17) Such a pseudodihedral angle has been proposed to be relevant to distinguish

type-II′ and type-I â-turns, see: Mu¨ller, G. Angew. Chem., Int. Ed. Engl.

1996, 35, 2767-2769.

(18) The Boc group has been proposed as a N-terminal residue mimic that does

not affect the conformational behavior of short peptides, see: Shaw, R.

A.; Perczel, A.; Fasman, G. D.; Mantsch, H. H. Int. J. Peptide. Res. 1996,

48, 71-78.

9

J. AM. CHEM. SOC. VOL. 125, NO. 52, 2003 16245

A R T I C L E S

Palomo et al.

Figure 3. Representative tetrapeptide models 8-11 as a function of the structural and stereochemical elements of the â-lactam core for â-turn induction.

Scheme 2. Synthesis of Model Peptides 8 and 9

Figure 4. Melanostatin 12 and peptidomimetic analogue 13 derived from

our proton mimic design strategy.

policyclic peptidomimetics²⁰of PLG have been developed

during the past decades as potential new drugs against Parkin-

son’s disease or tardive diskynesia. Application of the â-lactam

peptidomimetic approach²¹naturally leads to the mimetic 13.

An assessment of the type-II â-turn conformation for this peptide

model in water or highly coordinating solvents, together with

its biological activity retention with respect to PLG, would

provide evidence that the â-lactam methylene group can, indeed,

mimic CRH_(i+1)and HN_(i+2)protons.

â-Lactam Peptide Synthesis. Oligopeptides 8 and 9, con-

taining â-lactams unsubstituted at the C₄position of the ring,

were prepared uneventfully from a single N-[bis(trimethylsilyl)-

methyl]-â-lactam precursor 4 (Scheme 2). Accordingly, sequen-

tial deprotection of the 4-phenyloxazolidinone and bis-

(trimethylsilyl)methyl moieties under the same conditions used

for the preparation of 6 (Scheme 1) furnished the enantiomeri-

cally pure NH-â-lactam 14²²in 76% overall yield, which was

directly subjected to chemoselective carboxymethylation at the

hydroxybenzotriazole (HOBt), afforded peptidomimetic 8 as a

microcrystalline solid. Similarly, the 3-benzyl-N-carboxymethyl

substituted analogue 5, upon sequential double-glycine coupling

in the C- and N-terminal positions, provided 9 in 42% overall

yield.

Tetrapeptides 10 and 11 were also prepared in a similar way

as shown in Scheme 3. At the outset, however, it was not clear

which synthetic route would be best suited for the preparation

â-lactam nitrogen atom in the presence of cesium carbonate,

using the Miller’s method,²³followed by hydrogenolysis to the

free acid 15. Coupling of fragment 15 with R-aminoisobutiric

of the R-amino-R,â-dialkyl â-lactam core of each peptide

acid benzyl ester, carried out under homogeneous conditions

product. Apparently, the most direct access to these compounds

using the water-soluble carbodiimide method,²⁴(EDCl)/1-

would also be the asymmetric alkylation of R-amino â-lactams,

a technology established by Ojima almost 20 years ago.²⁵

(19) For a review of the biological activities of PLG: Mishra, R. K.; Chiu, S.;

However, from this strategy, only one isomer â-lactam product

Mishra, C. P. Methods Find. Exptl. Clin. Pharmacol. 1983, 5, 203-233.

(20) (a) Evans, M. C.; Johnson, R. L. Tetrahedron 2000, 56, 9801-9808. (b)

Webber, K.; Ohnmacht, U.; Gmeiner, P. J. Org. Chem. 2000, 65, 7406-

7416. (c) Baures, P. W.; Pradhan, A.; Ojala, W. H.; Gleason, W. B.; Mishra,

R. K.; Jonson, R. L. Bioorg. Med. Chem. Lett. 1999, 9, 2349-2352. (d)

Baures, P. W.; Ojala, W. H.; Costain, W. J.; Ott, M. C.; Gleason, W. B.;

Mishra, R. K.; Johnson, R. L. J. Med. Chem. 1997, 40, 3594-3600.

(21) The â-turn empirical assignment for a â-lactam analogue of Pro-Leu-Gly-

NH₂has been reported on the basis of competitive binding enhancement

on D₂dopaminergic receptors in the presence of agonist [³H]ADTN:

Sreenivasan, U.; K.Mishra, R.; Johnson, R. L. J. Med. Chem. 1993, 36,

256-263.

with the C₃and C₄groups in an anti-relationship is accessible.

We adopted this approach starting from â-lactam 17 (Scheme

3), easily prepared in 70% yield via the cycloaddition reaction

of Evans’ ketene, with the imine derived from isobutiraldehyde

and C,C-bis(trimethylsilyl)methylamine²⁶and have found that,

contrary to all expectations, the alkylation of the lithium enolate

of 17 with benzyl bromide provided the syn-3,4-dialkyl â-lactam

(22) (a) Palomo, C.; Aizpurua, J. M.; Legido, M.; Galarza, R. Chem. Commun.

1997, 233-234. Alternatively, this compound could also be prepared from

N-Boc-L-serine using Millers’ hydroxamate method in 67% yield; see: (b)

Mattingly, P. G.; Miller, M. J. J. Org. Chem. 1980, 45, 410-415.

(23) Guzzo, P. R.; Miller, M. J. J. Org. Chem. 1994, 59, 4862-4867.

(24) Carpino, L. A.; El-Faham, A. Tetrahedron 1999, 55, 6813-6830.

(25) (a) Ojima, I. Acc. Chem. Res. 1995, 28, 383-389. (b) Ojima, I. In The

Organic Chemistry of â-Lactams; Georg, G. I., Ed.; VCH: New York,

1992; p 197.

9

16246 J. AM. CHEM. SOC. VOL. 125, NO. 52, 2003

Restricted Peptides as Potent Nucleators of â-Turns

A R T I C L E S

Scheme 3. Synthesis of Peptidomimetic Models 10 and 11

18²⁷essentially as single diastereomer. Compound 18 was freed

of the oxazolidinone group by metal-reduction with lithium in

liquid ammonia and of the bis(trimethylsilyl)methyl group by

a Ce(IV) oxidation reaction, giving rise to the NH-â-lactam 19

which was submitted to an identical N-carboxymethylation

reaction and peptide coupling as compound 14 (Scheme 2),

thereby affording syn-R,â-dialkyl substituted â-lactam pepti-

domimetic 10 in 60% overall yield. Concurrently, it was found

that the preparation of the â-lactam 22 with the C₃and C₄

substituents in an anti-relationship, needed for peptidomimetic

11, could be realized upon exchange of the oxazolidinone ring

in 17 by a benzaldehyde-imine group. In this instance, the

alkylation of the intermediate enolate proceeded under stereo-

control dictated by the substituent at the C₄-position²⁸to give

21 in 60% yield and essentially as the sole diastereomeric

product. Hydrolysis of imine 21 was unusually difficult, but

upon exposure to strongly acidic Amberlyst resin in moist

diethyl ether solution, followed by basic treatment with ammonia

provided the desired R-amino-â-lactam quantitatively. Protection

of the latter as a tert-butyl carbamate also proved difficult under

ordinary conditions (ditertbutyl dicarbonate in acetonitrile), but

could finally be achieved by means of a variation of the method

of Kno¨lker using tributylphosphine as promoter.²⁹Further

elaboration to the desired peptidomimetic model 11 was carried

out similarly to its diastereomer 10.

(26) Direct access to enantiomerically pure 3-amino-4-alkyl-substituted â-lactams

from enolizable aldehyde-derived imines is not obvious. For a solution to

this problem through the ketene-imine Staudinger cycloaddition reaction

using N-bis(trimethylsilyl)methyl imines, see: (a) Palomo, C.; Aizpurua,

J. M.; Legido, M.; Galarza, R.; Deya, P. M.; Dunogues, J.; Picard, J. P.;

Ricci, A.; Seconi, G. Angew. Chem., Int. Ed. Engl. 1996, 35, 1239-1241.

(b) Palomo, C.; Aizpurua, J. M.; Legido, M.; Mielgo, A.; Galarza, R. Chem.

Eur. J. 1997, 3, 1432-1441 and references therein.

At this stage and, in view of the results noted above, it was

judged instructive to get some further insight into the scope of

(28) Ojima, I.; Chen, H.-J. C.; Nakahashi, K. J. Am. Chem. Soc. 1988, 110,

278-281.

(29) (a) Kno¨lker, H.-J.; Braxmeier, T.; Schlechtingen, G. Angew. Chem., Int.

Ed. Engl. 1995, 35, 2497-2500. (b) Kno¨lker, H.-J.; Braxmeier, T.

Tetrahedron Lett. 1996, 37, 5861-5864.

(27) The configuration of compound 18 was ascertained by X-ray crystal-

lographic analysis; see the Supporting Information.

9

J. AM. CHEM. SOC. VOL. 125, NO. 52, 2003 16247

A R T I C L E S

Palomo et al.

Scheme 4. R-Alkylation of C₄-Substituted N-[Bis(trimethylsilyl)methyl]-â-lactam Peptidomimetic

Scheme 5. Synthesis of PLG â-Lactam Peptidomimetic 13^a

^aTMEDA ) N,N,N′,N′-tetramethylethylenediamine.

Table 1. Syn-Stereoselective Alkylation of C₄-Substituted

N-[Bis(trimethylsilyl)methyl]-â-lactams 17, 24, and 25

the glycinamide residue could be performed by treatment of 30

with 1.2 equiv of preformed n-BuLi/TMEDA complex at -100

°C in THF and subsequent trapping of the lithiated intermediate

with trimethylsilyl isocyanate.³⁰Conventional peptide coupling

of the resulting 31 with Cbz-protected L-proline using the EDCl/

HOBt method proved inefficient, and only 25-35% yields of

the expected peptide were obtained after repeated runs. This

poor conversion, which was attributed to the combined steric

hindrance of the â-lactam amine and the activated Cbz-proline/

HOBt intermediate, was finally improved using Carpino’s amino

acid fluoride method.³¹Thus, coupling of the R-amino-â-lactam

derived from 31 with “in situ” formed Cbz-Pro-F in the presence

of N-methylmorpholine (NMM) followed by hydrogenolytic

debenzylation led to 13 in 52% overall yield. The PLG analogue

13 was a thick colorless oil, but proper single crystals for X-ray

crystallographic analysis were grown from solutions of its

N-Boc-protected form 32.

yield^a

(%)

[R]²⁵

D

1

2

product

R

syn/anti^b

(c ) 1.0, Cl₂CH )^c

2

26

27

28

29

ⁱBu CH₂dCHCH₂-

84

60

67

90

67:33

>98:2

93:7

-4.6

-6.8

Et

4-BrC₆H₄CH₂-

CH₂dC(Me)CH₂-

4-MeC₆H₄CH₂-

Ph

-54.5

-74.7

92:8

^aIsolated yield of the mixture of syn- and anti-R,â-dialkyl-â-lactams.

^bDetermined by ¹H NMR (500 MHz) analysis of the reaction crude.

Stereochemical assignments confirmed by X-ray crystal structure analyses

and/or NMR(NOE) spectroscopy. ^cData corresponding to the pure syn-

isomer.

the syn-selective alkylation of N-[bis(trimethylsilyl)methyl]-â-

lactams (Scheme 4, Table 1). It was found that, indeed, the

alkylation reaction of â-lactams 17, 24, and 25 with both allyl

and benzyl halides proceeds under high levels of stereocontrol

to give R-alkylated products in excellent syn/anti selectivity

ratios. Most notably, although the precise control elements that

govern these reactions are not clear at present, the results appear

to be quite regular regardless of the substituent at the C₄position

of the â-lactam ring. Since the stereochemical outcome of these

reactions may be reversed upon exchange of the oxazolidinone

ring, vide supra, our approach enables access to both syn- and

anti-R-amino-R,â-dialkyl-â-lactams starting from a single â-lac-

tam product, a feature of the approach that may also be of

general interest in the chemistry of â-lactams.

On the other hand, for the reasons that we have outlined

earlier, vide supra, the â-lactam peptide analogue of melanostatin

was prepared from the â-lactam 4 via alkylation with methallyl

bromide, subsequent catalytic hydrogenation and finally instal-

lation of the proline and glycinamide residues as shown in

Scheme 5. In this instance, it was found that introduction of

Solid-State Structures. After intensive trials, only com-

pounds 6,^14b11, and 32, the N-Boc derivative of 13, afforded

suitable single crystals for crystrallographic analysis (see the

(30) Only deprotonation at the CH(SiMe₃)₂methine group was observed.

Apparently, the known “R-silicon effect” in conjunction with a considerable

“complex-induced proximity effect” (CIPE) between the sterically R-blocked

â-lactam carbonyl oxygen lone pair and lithium cation are responsible for

the deprotonation at this position and not at the methylene group. For

reviews on the R-silicon effect, see: (a) Itami, K.; Kamei, T.; Mitsudo,

K.; Nokami, T.; Yoshida, J.-I. J. Org. Chem. 2001, 66, 3970-3976. (b)

Panek, J. S. Silicon Stabilization. In ComprehensiVe Organic Synthesis;

Trost, B. M., Fleming, I., Eds.; Pergamon: New York, 1991; Vol. 1, p

579. For surveys on CIPE, see: (c) Beak, P.; Meyers, A. I. Acc. Chem.

Res. 1986, 19, 356-363. (d) Snieckus, V. Chem. ReV. 1990, 90, 879-

933. (e) Beak, P.; Basu, A.; Gallagher, D. J.; Park, Y. S.; Thayumanyan,

S. Acc. Chem. Res. 1996, 24, 552-560. (f) Hoveyda, A. H.; Evans, D. A.;

Fu, G. C. Chem. ReV. 1993, 93, 1307-1370. An extention of the scope of

this reaction to other N-[bis(trimethylsilyl)methyl]-â-lactams is currently

underway in our laboratory.

(31) Carpino, L. A.; Mansour, E.-S. M. E.; Sadat-Aalaee, D. J. Org. Chem.

1991, 56, 2611-2614.

9

16248 J. AM. CHEM. SOC. VOL. 125, NO. 52, 2003

Restricted Peptides as Potent Nucleators of â-Turns

A R T I C L E S

Figure 5. ORTEP plots of the molecular structures of compounds 6, 11, and 32, showing â-turned conformations (only one of the two symmetry-independent

molecules in the asymmetric unit of 6 is shown).

Table 2. Selected Geometric Parameters for â-Lactam Peptidomimetics in the Solid State

compd

_(i)C)O‚‚‚HN_(i+3)(Å)

φ_(i+1);ψ_(i+1)(deg)

φ_(i+2);ψ_(i+2)(deg)

CR_(i)‚‚‚CR_(i+3)(Å)

δ (deg)^b

_(i)O‚‚‚H−N_(i+3)(deg)

6 (A)^a

2.32(3)

2.37(3)

2.18(2)

1.98

1.8-2.5

-40.9(5); 122.5(3)

-47.1(4); 122.5(3)

35.9(2); -120.5(1)

-37.9(4); 122.8(3)

-60; 120

100.5(4); -2.1(5)

105.4(4); -13.7(4)

-92.5(2); 18.0(2)

90.7(4); 21.7(4)

80; 0

5.957(5)

6.017(5)

5.885(2)

5.49^c

<7

17.5(2)

9.3(2)

4.5(8)

2.2

0

164(3)

167(3)

166(2)

152

160

6 (B)^a

11

32

II â-turn (ideal)

II′ â-turn (ideal)

60; -120

-80; 0

^aTwo symmetry-independent molecules in the asymmetric unit. ^bPseudodihedral angle formed by the four CR atoms (or equivalents) of the peptide

model. ^cThe value corresponds to the CR_(i)‚‚‚H^ZN

distance.

(i+3)

Supporting Information). To our delight (Figure 5), in each case,

intramolecular hydrogen bonds were formed between the Aib

or glycinamide HN_(i+3)amide donors and the Boc or Pro Cd

O_(i)oxygen atom acceptors, thereby creating a â-turn within

the molecule which has a graph set³²motif of S(10).

Type II′ â-turned conformation of the model 11 (anti-R,â-

dialkyl substituted at the â-lactam ring) is worth mentioning

since, owing to the steric congestion caused by the isobutyl

group, the -(â-lactam)-(gly)- segment in this compound was

expected to display a markedly less betagenic behavior than

models 6 or 10. Nevertheless, the isobutyl group at the â-lactam

ring C₄position, the bulky Aib_(i+3)residue, and the Boc_(i)NH

group are arranged on the same side of the plane defined by

the four atoms of the â-lactam ring. Finally, the melanostatin

Boc derivative 32 also displays a well-defined type II â-turn,

including a short hydrogen bond (1.98 Å) between the proline

CdO_(i)oxygen atom and the glycine H^EN_(i+3)hydrogen atom,

similar to that in the previously reported crystal structure

melanostatin hemihydrate (2.15 Å).³⁴

Despite these promising results, it is well-known that solid-

state secondary conformation of peptidomimetics is often

conditioned by “lattice effects”³⁵involving significant inter-

molecular hydrogen-bonding that, indeed, was clearly observed

in our molecules.³⁶Therefore, extension of the conclusions

outlined above to biologically meaningful media necessitated

assessment of the stability of the observed â-turns in highly

coordinating solvent solutions. Although water was considered

first, finally DMSO solvent was chosen for practical reasons

(model compounds 6, 8, 9, 10, and 11 were insoluble in water)

and also because the only NMR conformational study described

for melanostatin 12 in solution was conducted in this solvent.³⁷

Molecular Dynamics Simulations. MD simulations were

carried out in a DMSO box at 300 K and 1 atm using the

AMBER 6 force field (see the Experimental Section for details).

A more detailed analysis of the structural data, Table 2,

revealed the presence of well defined â-turned conformations

of type-II for 6 and 32 and of type-II′ for 11, depending upon

the configuration of the C₃stereocenter at the respective

â-lactam rings. In addition to the typical _(i)CdO‚‚‚HN_(i+3)

intramolecular hydrogen bonds, all of the four structural criteria³

used to define canonical type-II â-turns are clearly fulfilled by

the three peptidomimetics, suggesting that conformational

behavior observed previously for model 6 could be extended

to a variety of R-substituted R-amino-â-lactam peptides. Fur-

thermore, the hydrogen bond directionality defined by the _(i)O‚

‚‚H-N_(i+3)angles is almost ideal in all instances.

Slight dihedral angle distortions of about 15-20° for the

φ_(i+1), φ_(i+2), and ψ_(i+2)angles with respect to native ideal values

is also observed for 11 and 32, whereas the rigid ψ_(i+1)angle

remains, as expected, constant and very close to the ideal 120°

value. At the origin of this difference is the coplanar spatial

arrangement of the N_(i+1), CR_(i+1), N_(i+2), and CR_(i+2)atoms,

forced by the central â-lactam ring. This particular characteristic

of R-substituted R-amino-â-lactam peptides seems to favor the

formation of perfectly parallel peptide branches at the (i) and

(i+3) termini, as judged by the systematically small values of

the pseudodihedral angle δ, and anticipates the use of these

peptidomimetics, for instance, as promising antiparallel pleated

â-sheet nucleators.³³

(34) Reed, L. L.; Johnson, P. L. J. Am. Chem. Soc. 1973, 95, 7523.

(35) Steiner, T. Angew. Chem., Int. Ed. 2002, 41, 48-76.

(36) For details on intermolecular hydrogen bonding patterns of compounds 11

and 32, see the Supporting Information.

(37) Higashijima, T.; Tasumi, M.; Miyazawa, T.; Miyoshi, M. Eur. J. Biochem.

1978, 89, 543-556.

(32) For a detailed information on graph set definitions, see: Bernstein, J.; Davis,

R. E.; Shimoni, L.; Chang, N.-L. Angew. Chem., Int. Ed. Engl. 1995, 34,

1555-1573.

(33) Nowick, J. S.; Smith, E. M.; Pairish, M. Chem. Soc. ReV. 1996, 401-415.

9

J. AM. CHEM. SOC. VOL. 125, NO. 52, 2003 16249

A R T I C L E S

Palomo et al.

Figure 6. Values of the _(i)C)O‚‚‚HN_(i+3)inter-residual hydrogen bond distances (Å) from the MD 5 ns trajectories of â-lactam model compounds 6, 8, 9,

10, and 11 calculated in DMSO solution.

The peptidomimetic â-lactam molecules 6, 8, 9, 10, 11, and 13

were model-built using the bond lengths and angles taken from

the crystal structures of compounds 6, 11, and 23. The

melanostatin 12 model was generated similarly from the reported

crystal structure,³⁴and the prolyl residue was uncharged. Type-

II â-turn motif (type-II′ â-turn motif for compounds 8 and 11)

was reproduced, the restraint was removed thereafter, and the

system coordinates were collected every picosecond for a 5 ns

trajectory. The adoption of â-turn, γ-turn, or extended confor-

mations around the -(i+1)-(i+2)- segment was assumed for

average _(i)C)O‚‚‚H-N_(i+3)distances of 3.2 ( 0.3 Å, 5.5 ( 0.3

Å, and >6.0 Å, respectively. The relevant results are shown in

Figure 6.

great conformational heterogeneity. Peptide 6 only shows one

conformation corresponding to a â-turn motif which fits

reasonably well with the structural parameters of a slightly

deformed type II â-turn, while compound 10 exhibits two

deformed type II â-turns along the trajectory of the MD

simulation. Values of â-turn dihedral angles are shown in Table

4 (see below). On the other hand, the comparatively lower

stability of the â-lactam peptide model 9 with respect to 6

around the -(â-Lactam)-(Gly)- fragment partially shaded our

initial assumption concerning the stability of the â-turn motif

in the absence of bulky Boc- and quaternary Aib residues at

the N- and C- terminal positions.

Compound 11 quickly abandoned the type-II â-turn confor-

mation and visited several extended and a few γ-turned forms,

which was in contrast with the unique type-II′ conformation

that was found in the solid state. This effect can be attributed

Whereas the â-turn pattern remains stable over the 5ns

trajectory of the MD simulation for mimetics 6 and 10, analysis

of the trajectories of compounds 8, and especially 11, reveals a

9

16250 J. AM. CHEM. SOC. VOL. 125, NO. 52, 2003

Restricted Peptides as Potent Nucleators of â-Turns

A R T I C L E S

Figure 7. MD trajectories of melanostatin (PLG) 12 (10 ns, left) and â-lactam peptidomimetic 13 (5 ns, right) calculated in DMSO solution. Upper case:

the _(i)C)O‚‚‚HN_(i+3)inter-residual hydrogen bond distances. Lower case: the _(i)(proline)N‚‚‚HNCO_(i+1)γ-turn distances.

most likely to the destabilizing effect of the DMSO solvent. A

similar trend was observed for compound 8 as soon as the

hydrogen-bonding distance restraint was removed, resulting in

a remarkably homogeneous extended conformation. However,

in this case the long _(i)CdO‚‚‚H-N_(i+3)distances arose from the

180° rotation of the BocHN group around the φ_(i+1)dihedral

angle and not from the relative disposition of the Gly-Aib branch

with respect to the â-lactam ring. A comparison of the behavior

of compound pairs 6/8 and 10/11 reveals the paramount

importance of the presence of an alkyl substituent at the C₃

position as well as the relative syn/anti disposition of the

substituents of the â-lactam ring in controlling â-turn stability,

even in DMSO solution.

We also applied our MD protocol in DMSO to melanostatin

12 and the â-lactam peptidomimetic 13. As the involvement of

a five-atom loop, stabilized by a hydrogen bond between the

Pro amine nitrogen and the Leu amide NH proton, has been

suggested to explain the bioactivity of natural PLG amide 12,³⁸

we monitored both the _(i)C)O‚‚‚HN_(i+3)and the _(i)(Pro)N‚‚‚

HN_(i+1)distances along the trajectories of 12 and 13 (Figure

7). As can be inferred from these data, the â-turn for mel-

anostatin is very unstable in DMSO and coexists with several

extended conformations of the peptide. This behavior agrees

with previously reported observations showing that unprotonated

Pro-Leu-Gly-NH₂is not rigid in DMSO solution.³⁷

ately populated type-II â-turn structure (average _(i)C)O‚‚‚

HN_(i+3)distance of 3.12 Å). Two of these conformations

correspond to the intramolecular bridge between the terminal

glycinamide H^EN_(i+3)proton and the â-lactam N_(i+1)nitrogen,

giving rise to two possible five-membered rings: the first by

the “upper” re face of the â-lactam, and the second by the

“lower” si face. Finally, the high stability of the C₅loop around

the Pro residue was confirmed along all the trajectories of 12

and 13, showing average _(i)(Pro)N‚‚‚HN_(i+1)distances of 2.36

Å and constraining the (Pro) ψ_(i)dihedral angle close to 0°.

NMR Conformational Analysis in Solution. The study of

peptidomimetic models 6, 8, 9, 10, 11, and 13 was conducted

in DMSO-d₆, primarily by the temperature coefficient analysis

and short-range NOE inspection to identify intramolecular

hydrogen bonding and then using the procedure of Seebach and

van Gusteren,³⁹consisting of a cluster analysis of combined

experimental (NMR-500 MHz) and simulated annealing data.

Spin system identification (see Figure 8) and assignment of

individual resonances were carried out using a combination of

¹H, DQF-COSY, and NOESY spectra, according to the standard

procedures.⁴⁰Single sets of signals were observed for all

â-lactam pseudopeptides at 300 K under measurement condi-

tions.

The involvement of the NH_(i+3)amide protons in intramo-

lecular hydrogen bonding (Table 3) was first estimated from

The MD conformational analysis of derivative 13 led to the

identification of several conformations, apart from the moder-

(39) Daura, X.; Gademann, K.; Scha¨fer, H.; Jaun, B.; Seebach, D.; van

Gunsteren, W. F. J. Am. Chem. Soc. 2001, 123, 2393-2404.

(40) (a) Kessler, H.; Schmitt, W. In Encyclopedia of Nuclear Magnetic

Resonance; Grant, D. M., Harris, R. K., Eds.; J. Wiley & Sons: New York,

1995. (b) Gu¨ntert, P. Quart. ReV. Biophys. 1998, 31, 142-237.

(38) Valle, G.; Crisma, M.; Toniolo, C.; Yu, K.-L.; Johnson, R. L. Int. J. Peptide

Protein Res. 1989, 33, 181-187.

9

J. AM. CHEM. SOC. VOL. 125, NO. 52, 2003 16251

A R T I C L E S

Palomo et al.

strong NOEs of the respective HN_(i+3)amides of the Aib and

GlyOBn residues with the â-lactam H^Sâ_(i+1)protons, but not

with H^Râ_(i+1). In sharp contrast with this behavior, the HN_(i+3)

amide of 8 displayed interactions with both Hâ_(i+1)protons

which, in combination with the high-temperature coefficients

measured above, suggested an equilibrium between several

extended conformations. Finally, the PLG analogue 13 also

showed strong NOE signals of both glycinamide HN_(i+3)protons

exclusively with the â-lactam H^Sâ_(i+1)protons, providing

evidence of the presence of a â-turned conformation.

Figure 8. Definition of the substituent atoms of a general -(â-lactam)-

(Gly)- segment used for the conformational study.

Table 3. Temperature Coefficients (ppb/K) of the Amide NH

In the case of syn-/anti-R,â-disubstituted tetrapeptide models

10 and 11 (see Figure 10) only one Hâ_(i+1)proton was available

for observation and a supplementary region of the spectrum

involving isobutyl group protons (∼0.7-2 ppm) was examined

for evidence of close interproton interactions of HN_(i+3)with

the alkyl groups placed in the upper face of the â-lactam ring

plane. Again, peptide model 10 showed a strong bias for a

â-turned conformation, as judged by the HN_(i+3)/Hâ_(i+1)NOE

signal and the absence of any cross-peak between HN_(i+3)and

isobutyl or benzyl protons. On the other hand, anti-peptide 11

presented no NOE cross-peaks between HN_(i+3)and Hâ_(i+1)and

only a weak interaction of HN_(i+3)with the isobutyl methylene

protons, which was also consistent with a â-turn conformation.

This observation partially refuted the extended or γ-turned

conformations provided by MD calculations and agreed better

with the solid-state structure derived from X-ray crystallographic

analysis.

After qualitative identification of folded start conformations,

100 structures were generated for each â-lactam peptidomimetic

model 6, 8, 9, 10, 11, and 13 using the X-PLOR⁴³program,

and only acceptable conformers presenting no violations of

NOESY-derived interprotonic distance restraints were analyzed.

Clusters were created applying the same _(i)CdO‚‚‚HN_(i+3)

intervals used for MD analysis (see above), and the values of

the main structural parameters resulting from each cluster were

analyzed (Table 4, Figure 11).

In excellent agreement with the X-ray crystallography and

MD data, each peptide model 6 and 10 presented only one very

populated cluster mimicking a canonical type-II â-turn around

the -(â-lactam)-(Gly)- segment. Compounds 11 and 9 also

reproduced correctly folded â-turns, but they enjoyed an excess

of conformational freedom. In both instances, an appreciable

distortion of the pseudodihedral angle δ (25-35°) was observed,

which could be at the origin of the mobility observed during

MD trajectories. Nonetheless, it is worth mentioning that, neither

the anti-R,â-dialkyl substitution pattern of the â-lactam ring in

11, nor the presence of fully flexible glycine residues at the N-

and C-terminal positions of 9 sufficed to destabilize these

â-lactam peptide models to extended conformations in DMSO.

On the other hand, the R-nonalkylated pseudopeptide 8

presented, apart from the starting type II â-turn conformation

cluster (8-A), a second extended cluster (8-B) clearly defined

by the rotation of the BocHN group around the φ_(i+1)dihedral

angle. Finally, melanostatin â-lactam analogue 13 is also

distributed into two similarly populated clusters (13-A and 13-

Protons Measured for â-Lactam Peptidomimetics 6-13 (DMSO_D6

,

500 MHz) at 5 K Intervals between 300 and 325 K^a

compd

H−N

(i+3)

(i)

(i+1)

6

8

9

10

11

13

-8.3 (2.21) -2.3 (0.03)

-5.1 (-)^b-4.9 (-)^b

-6.4 (1.83)

-5.9 (1.50) -4.2 (0.25)

-10.0 (2.51) -2.9 (0.08)

-8.9 (2.48) -3.1 (0.03)

-2.6 (-)^b

-3.9 H^EN (-);^b-6.5 H^ZN (-)^b

^aData in parentheses correspond to chemical shift differences (ppm)

caused by the solvent effect (δ DMSO-d₆- δ CDCl₃) at 300 K.

^bCompound insoluble in CDCl₃solvent.

temperature coefficients (DMSO-d₆) and from chemical shift

differences in coordinating (DMSO-d₆) and noncoordinating

(CDCl₃) solvents. All data were recorded at 10-15 mM

concentrations. No chemical shift or line-width variation was

observed, compared with a 10-fold dilution, which confirmed

the absence of any noticeable intermolecular aggregation.

Significant small temperature coefficients (<3 ppb/K)⁴¹and

chemical shifts changes were found for the NH_(i+3)protons of

compounds 6, 10, and 11, thus confirming their participation

in intramolecular hydrogen bonding. The tetrapeptide analogue

9 showed slightly larger values, whereas for the R-unsubstituted

â-lactam pseudopeptide 8, the NH_(i+3)proton could not be

considered to form a stable intramolecular hydrogen bond.

Interestingly, the temperature coefficients of the glycinamide

H^EN_(i+3)and H^ZN_(i+3)protons differed for the melanostatin

analogue 13 but not for natural melanostatin 12 (H^EN_(i+3): -5.2

ppb/K; H^ZN_(i+3): -5.5 ppb/K).³⁷The coalescence temperature

was also found to be 30° higher for 13 (375 K) than for 12

(345 K), suggesting a relative stabilization of the hydrogen-

bonded structure caused by the presence of the â-lactam ring.

Finally, the low-temperature coefficients measured for HN_(i+1)

amide in 13 agreed fully with the C₅loop structure calculated

by MD simulation around the Pro_(i)residue.

Inspection of the NOESY cross-peak signals from the amide

HN_(i+3)and â-lactam ring Hâ1_(i+1)protons (Figure 9) proved

to be particularly informative as a qualitative indication of the

acceptors involved in the intramolecular hydrogen-bonded

conformations, since the diastereotopic protons H^Râ_(i+1)and

H^Sâ_(i+1)were easily distinguished from each other by the

³J(CHR-CHâ) coupling constants in 8 or using selective

NOESY experiments between the H^Râ_(i+1)and the benzyl or

isobutyl Hâ1_(i+1)protons in models 6, 9 and 13. This allowed

a tentative location of the -(Gly)-(i+3) branch in the lower

or upper plane of the â-lactam ring and gave an important

indication to establish the presence of â-turned or extended/γ-

turned conformations, respectively.⁴²Both 6 and 9 showed

(42) The formation of a seven-membered γ-turn structure in â-lactam pseudopep-

tides through hydrogen-bonding of HN_(i+3)amide to â-lactam CdO_(i+1)

oxygen has previously been suggested by our group, see ref 14(b). An

alternative five-membered γ-turn from HN_(i+3)amide to â-lactam N_(i+1)

has also been calculated; see ref 16a.

(43) Schwieters, C. D.; Kuszewski, J.; Tjandra, N.; Clore, G. M. J. Magn. Reson.

2003, 160, 66-74.

(41) Kessler, H. Angew. Chem., Int. Ed. Engl. 1982, 21, 512-523.

9

16252 J. AM. CHEM. SOC. VOL. 125, NO. 52, 2003

Restricted Peptides as Potent Nucleators of â-Turns

A R T I C L E S

Figure 9. Selected expansions of NOESY spectra (500 MHz) showing the amide NH and â-lactam Hâ regions of 6, 8, 9, and 13 in DMSO-d₆(10-15 mM),

400 ms mixing time, recorded at 300 K. Marked cross-peaks indicate a â-turn or extended conformations.

B) having, respectively, the type-II and extended conformations.

Apparently, the φ_(i+1)dihedral angle blocking exerted by the

isobutyl group in this case is less effective than R-benzyl

blocking (see pseudopeptides 6, 9, 10, or 11) and allows a partial

conformational freedom, as anticipated by MD calculations.

Biological Evaluation of Melanostatin (PLG) â-Lactam

Analogue 13. Several studies have shown that PLG and its

peptidomimetic analogues render the Gi-protein coupled dopam-

ine D₂-like receptors more responsive to agonists by maintaining

the high-affinity binding state of the receptor,⁴⁴albeit the exact

9

J. AM. CHEM. SOC. VOL. 125, NO. 52, 2003 16253

A R T I C L E S

Palomo et al.

Figure 10. Selected expansions of NOESY (500 MHz) experiments of 10 and 11 in DMSO-d₆(8.5 and 12 mM), 400 ms mixing time, recorded at 300 K.

â-Lactam Hâ_(i+1), glycine HR_(i+2)(top left) and isobutyl proton regions (top right) are compared to NH amide region (left). Marked cross-peaks indicate a

â-turn conformation.

Table 4. Averaged Structural Parameters of the Conformation Clusters of â-Lactam Peptidomimetic Models 6-13 Obtained with X-PLOR

from NOESY-Restrained NMR Data

cluster (n)^a

type

_(i)C)O‚‚‚HN ^b(Å)

(φ;ψ)_(i+1)^c(deg)

(φ;ψ)_(i+2)^c(deg)

CR₍₁₎‚‚‚CR₍₁₊₃₎(Å)

δ^d(deg)

(i+3)

6-A (83)

8-A (58)

8-B (42)

9-A (58)

9-B (37)

10-A (89)

11-A (33)

13-A (43)

13-B (40)

â-II turn

â-II′ turn

extended

â turn^e

2.9 (2.1)

3.1 (2.5)

6.3 (5.8)

3.6 (3.5)

4.1 (3.8)

3.1 (2.5)

3.2 (2.4)

3.0 (2.5)

5.9 (4.9)

-27; 114 (-20; 110)

21; -118 (25; -110)

177; -118

109; -40 (55; 25)

-96; 48 (-75; 0)

-96; 48

5.5

5.7

7.2

6.0

6.5

5.7

3.7

21

25

35

-17

34

24

7

-30; 116

82; 70

â turn^e

28; 115

86; 82

â-II turn

â-II′ turn

â-II turn

extended

-36; 113 (-20; 110)

30; -117 (20; -120)

-49; 120 (-47; 123)

-145; 120

56; 29 (50; 40)

-53; 34 (110; -40)

111; -39 (92; 22)

105; -37

5.3

5.9

- - ^[f]

^aNumber of structures with no violations of NOESY interprotonic distance restraints per cluster from 100 structures generated. ^bValues in parentheses

indicate average _(i)C)O‚‚‚HN_(i+3)distance. ^cValues in parentheses indicate average -(i+1)-(i+2)- dihedral angles calculated from MD simulations for

the whole trajectory of each compound. Ideal dihedral angles: type â-II: (φ; ψ)_(i+1)) -60; 120; (φ; ψ)_(i+2)) 80; 0, type â-II′ (φ; ψ)_(i+1)) 60; -120; (φ;

ψ)_(i+2)) -80; 0. ^dPseudodihedral angle formed by the four CR atoms (or equivalents) of the peptide model. ^eNoncanonical distorded â-turns. ^fHydrogen

atom placed at the CR

position.

(1+3)

nature of the molecular interactions of PLG in such a process

remains obscure. All of the research conducted thus far to

quantify the bioactivity of PLG has focused primarily on the

behavioral inspection of whole animal models and on radioli-

gand competitive binding assays of receptors obtained from

tissue homogenates. However, as the mechanism of action of

PLG may likely involve intraneuronal enzymes,⁴⁵direct com-

petitive binding assays on integral neurons should be highly

desirable to provide information more directly assignable to real

living systems.

conducted for the first time radioligand competitive binding

assays⁴⁶on cultured integer neuron cells⁴⁷from rat cerebral

cortex. The experiments involved the blocking of D₂receptors

in neurons with the specific tritiated antagonist [³H]spiroperidol,

followed by addition of increasing amounts of agonist N-

propylnorapomorphine (NPA), either in the absence (control)

or in the presence of fixed concentrations of PLG 12 and its

â-lactam analogue 13. Representative competitive binding

(45) Two putative mechanisms have been proposed. First, the PLG-mediated

allosteric stabilization of the dopamine/D₂receptor/Gi protein ternary

complex at the membrane and, second, the intraneuronal stimulation of

guanosine triphosphatase enzyme (GTPase) and simultaneous inhibition

of adenosyl cyclase enzyme (AC), see: Mishra, R. K.; Makman, M. H.;

Costain, W. J.; Nair, V. D.; Johnson, R. L. Neurosci. Lett. 1999, 269, 21-

24.

(46) Baures, P. A.; Ojala, W. H.; Costain, W. J.; Ott, M. C.; Pradhan, A.;

Gleason, W. B.; Mishra, R. K.; Johnson, R. L. J. Med. Chem. 1997, 40,

3594-3600.

(47) Abele, A. E.; Scholz, K. P.; Scholz, W. K.; Miller, R. Neuron 1990, 2,

413-419.

To test the relative activity of â-lactam analogue 13 versus

natural PLG 12 as a dopamine D₂receptor modulator, we

(44) The signal transduction mechanism triggered by dopamine in D₂receptors

leads ultimately to an intraneuronal cyclic adenosyl monophosphate (cAMP)

concentration reduction with the intermediacy of Gi proteins of “high” and

“low” affinity corresponding, respectively, to di- and triphosphorilated

species; see: Costain, W. J.; Gupta, S. K.; Johnson, R. L.; Mishra, R. K.

G Protein Methods Protocols 1997, 31, 119-138.

9

16254 J. AM. CHEM. SOC. VOL. 125, NO. 52, 2003

Restricted Peptides as Potent Nucleators of â-Turns

A R T I C L E S

Figure 11. Views of the backbone superposition of 10 structures of lowest energy with NOESY distance restraints derived from NMR data for conformational

clusters 6-A-13-B. Calculations performed in vacuo with X-PLOR.

13 with respect to PLG 12 suggests that the structural variation

arising from the replacement of CRH_(i+1)and HN_(i+2)protons

by a â-lactam methylene group does not hinder the maintenance

of a similar recognition pattern by D₂receptors.

Conclusions. A simplified peptidomimetic design based on

the concept of separation of constraint and recognition elements

has been developed to prepare short pseudopeptides containing

enantiopure R-substituted R-amino-â-lactam fragments that

mimic very stable â-turns of type II or type II′ with highly

predictable stereostructural properties. X-ray crystallography in

the solid state and MD or NMR conformational studies in

dimethyl sulfoxide solutions have shown that the R-alkyl-R-

amino substitution pattern of â-lactam rings placed as -(i+1)-

residues in tetrapeptide models is essential to fold the peptido-

mimetic to a stable â-turned conformation. syn-R,â-Disubsti-

tution seems not to provide additional stabilizaton and even anti-

R,â-disubstitution is compatible with â-turn formation in polar

solvents. The nature of the N- and C-terminal residues in

tetrapeptide models has little effect, albeit observable, on the

betagenic ability of the central -(â-lactam)-(Gly)- segment

since even models flanked by two glycine units afford rough

but correctly folded â-turned structures.

Figure 12. Representative competition curve of [³H]spiroperidol/N-

propylnorapomorphine (NPA) in cultured neurons from rat cerebral cortex

without treatment (control) and treated with PLG 12 (1 µM) or PLG

peptidomimetic 13 (1 µM). The concentration of [³H]spiroperidol in the

assay was 0.5 nM; previously determined K_dwas 0.99 nM. Inhibiton

constant (K_i) of agonist, calculated for the [³H]spiroperidol binding using

expressed as mean ( SEM from four separate experiments done in triplicate.

Nonspecific binding was determined in parallel assays in the presence of 1

µM (+)-butaclamol.

curves, see Figure 12, indicated that 13 slightly reduces the

dissociation constant of dopamine D₂receptors for NPA with

respect to PLG 12 at a micromolar concentration range. Despite

the low accuracy of the measurements, which not allow one to

distinguish between high- and low-affinity binding constants,

the full retention of bioactivity on living neurons observed for

Finally, conservation of the recognition requirements of the

binding pocket in the receptor for PLG neuropeptide â-lactam

analogue 13, despite the incorporation of a methylene bridging

in native PLG 12, seems to confirm the viability of a design

9

J. AM. CHEM. SOC. VOL. 125, NO. 52, 2003 16255

A R T I C L E S

Palomo et al.

129.7; 129.4; 81.2; 69.1; 58.9; 50.7; 43.7; 28.9. Reaction time

(hydrogenation): 8 h. Yield (overall): 0.49 g (40%).

separating constraint and recognition groups. Work to illustrate

the application of these novel â-lactam peptidomimetics to the

ligand-based â-turned agonists for G-protein-coupled receptors

is in progress in our laboratory.

(3R,4R)-3-Benzyl-3-tert-butoxycarbonylamino-1-carboxymethyl-

4-isobutylazetidin-2-one (20). The general procedure was followed

from (3R,4R)-3-benzyl-3-tert-butoxycarbonylamino-4-isobutyl azetidin-

2-one (19) (5 mmol, 2.23 g); eluant EtOAc/hexanes 1:3. Intermediate

benzyl ester: (3R,4R)-3-benzyl-1-(benzyloxycarbonyl methyl)-3-tert-

Experimental Section

butoxy carbonylamino-4-isobutyl-azetidin-2-one. [R]²⁵

) -11.79 (c

General Methods. All reactions were carried out under an atmo-

sphere of nitrogen in oven- or flame-dried glassware with magnetic

stirring. Solvents were distilled prior to use. Tetrahydrofuran (THF)

was distilled from sodium metal/benzophenone ketyl. Acetonitrile (CH₃-

CN) and dichloromethane (CH₂Cl₂) were distilled from calcium hydride.

Methanol (MeOH) was dried over magnesium metal and iodine. DMF

was purified by distillation on barium oxide. Purification of reaction

products was carried out by flash chromatography using silica gel 60

(230-400 mesh, from Merck 60F PF254). Analytical thin-layer

chromatography was performed on 0.25 mm silica gel 60-F plates.

Visualization was accomplished with UV light and phosphomolybdic

acid-ammonium cerium(IV) nitratesulfuric acid-water reagent, fol-

lowed by heating. Melting points were measured with a Bu¨chi SMP-

20 melting point apparatus and are uncorrected. Infrared spectra were

D

) 0.48, Cl₂CH₂). IR (cm^-1, KBr): 3263 (NH); 1758.7; 1732.0; 1720.9

(CdO). MS m/z (int): 91 (44.2); 129 (26.0); 146 (100.0); 189 (20.7);

1

215 (7.6); 234 (18.6); 324 (11.1); H NMR (δ, ppm, CDCl₃): 7.37-

7.20 (m, 10H, Ar); 5.22 (d, 1H, J ) 11.7 Hz, OCH₂Ph); 5.15 (d, 1H,

J ) 12.2 Hz, OCH₂Ph); 4.65 (s, 1H, NHBoc); 4.46 (d, 1H, J ) 18.1

Hz, NHCH₂CO); 3.99 (t, 1H, J ) 6.4 Hz, HC(ⁱBu)); 3.69 (d, 1H, J )

18.1 Hz, NHCH₂CO); 3.37 (d, 1H, J ) 13.7 Hz, CH₂Ph); 3.16 (d, 1H,

J ) 14.1 Hz, CH₂Ph); 1.55-1.49 (m, 1H, CH₂CH(CH₃)₂); 1.47-1.36

(m, 2H, CH₂CH(CH₃)₂); 1.42 (s, 9H, OCOC(CH₃)₃); 0.83 (d, 3H, J )

6.4 Hz, CH₂CH(CH₃)₂); 0.81 (d, 3H, J ) 6.4 Hz, CH₂CH(CH₃)₂). ¹³

C

NMR (δ, ppm, CDCl₃): 168.9; 168.1; 154.3; 135.2; 134.9; 129.9; 128.7;

128.6; 127.3; 80.1; 70.2; 67.3; 63.1; 41.6; 40.3; 37.1; 28.2; 25.7; 23.0;

22.5. Reaction time (hydrogenation): 4 h. Yield (overall): 1.69 g (87%).

(3S,4R)-3-Benzyl-3-tert-butoxycarbonylamino-1-carboxymethyl-

4-isobutylazetidin-2-one (23): The general procedure was followed

from (3R,4R)-3-benzyl-3-tert-butoxycarbonylamino-4-isobutyl azetidin-

2-one (22) (5 mmol, 2.23 g); eluant EtOAc/hexanes 1:10 then 1:5.

Intermediate benzyl ester: (3S,4R)-3-benzyl-1-(benzyloxycarbonyl-

recorded on a Shimadzu IR-435 spectrophotometer. ¹H NMR and ¹³

C

NMR spectra were recorded on a Bruker Avance DPX300 and Bruker

Avance500 spectrometers and are reported as δ values (ppm) relative

to residual CHCl₃δ_H(7.26 ppm) and CDCl₃δ_C(77.16 ppm) as internal

standards, respectively. Combustion analyses were performed on a Leco

CHNS-932 elemental analyzer. Analytical high-performance liquid

chromatography (HPLC) was performed on a Hewlett-Packard 1050

chromatograph equipped with a diode array UV detector, using the

analitycal columns (25 cm, phase Lichrosorb-Si60) and (25 cm, phase

Chiralcel OD) with flow rates of 1 and 0.5 mL/min, respectively, and

on preparative-scale columns (25 × 2.5 cm, phase Lichrosorb-Si60).

Optical rotations were measured at 25 ( 0.2 °C in methylene chloride

unless otherwise stated. Mass spectra were obtained on a Finnigan GCQ

mass spectrometer (70 eV) using GC-MS coupling (column: fused

silica gel, 15 m, 0.25 mm, 0.25 nm phase SPB-5). Compounds 4,^22a

17,²⁶24,²⁶and 25²⁶were prepared according to described procedures.

methyl)-3-tert-butoxycarbonylamino-4-isobutyl-azetidin-2-one. [R]²⁵

D

) +23.3 (c) 1.0, Cl₂CH₂). IR (cm^-1, KBr): 2946.0 (NH); 1761.1;

1751.3; 1748.3; 1707.8 (CdO). MS m/z (int): 91 (31.3); 129 (15.9);

1

146 (100); 189 (22.3); 324 (5.5). H NMR (δ, ppm, CDCl₃): 7.41-

7.30 (m, 10H, Ar); 5.23 (s, 2H, OCH₂Ph); 4.96 (s, 1H, NH); 4.13-

4.11 (m, 3H, CCH; NCH₂CO); 3.34 (d, 1H, J ) 13.4 Hz, CH₂Ph);

3.08 (d, 1H, J ) 14.0 Hz, CH₂Ph); 1.75-1.60 (m, 3H, CHCH₂(CH₃)₂);

1.46 (s, 9H, Boc); 0.98 (d, 3H, J ) 3.8 Hz, CH₃); 0.96 (d, 3H, J ) 3.8

Hz, CH₃). ¹³C NMR (δ, ppm, CDCl₃): 168.5; 168.0; 154.4; 135.4;

135.0; 130.4; 128.6; 128.5; 128.4; 128.2; 126.8; 79.9; 69.1; 67.3; 63.9;

42.4; 38.2; 35.2; 28.2; 25.9; 23.4; 21.8. Reaction time (hydrogena-

tion): 4 h. Yield (overall): 1.60 g (82%).

General Procedure for the Carboxymethylation of 3-tert-Bu-

toxycarbonylamino-NH-â-lactams. Synthesis of 15, 20, and 23. A

solution of the corresponding NH-â-lactam (5 mmol) in dry MeCN

(50 mL) was prepared under nitrogen in a flame-dried two-necked flask

fitted with a septum and a thermometer. The temperature was warmed

to 50 °C using a water bath, Cs₂CO₃(6.0 mmol, 1.96 g) and benzyl

iodoacetate (7.5 mmol, 1.39 g) were added, and the mixture was stirred

for 3 h at the same temperature. Water (50 mL) and EtOAc (100 mL)

were added to the mixture, the organic layer was separated, and the

aqueous phase was extracted with EtOAc (30 mL × 3). The combined

organic solutions were dried (MgSO₄) and evaporated under reduced

pressure. The intermediate N-[(benzyloxycarbonyl)methyl]-â-lactam

was purified by column chromatography (silica gel; eluant: EtOAc/

hexanes). This product was dissolved in dried EtOAc (30 mL),

10%Pd/C (0.2 g) was added, and the suspension was stirred under

hydrogen (1 atm) for 4-8 h. The reaction mixture was filtered through

Celite, washed with EtOAc (30 mL × 2), and evaporated in vacuo to

afford the pure product.

General Procedure for Peptide Coupling of 3-tert-Butoxycarbo-

nylamino-N-(carboxymethyl)-â-lactams with Benzyl 2-Aminoisobu-

tyrate Tosylate (TsOH‚AibOBn): Synthesis of 8, 10, and 11. To a

solution of TsOH‚AibOBn (3.0 mmol, 1.096 g) and NEt₃(10.0 mmol,

1.39 mL) in dry CH₂Cl₂(15 mL) cooled to 0 °C were added HOBt

(2.0 mmol, 0.27 g) DMF (0.5 mL), the corresponding 3-tert-butoxy-

carbonylamino-N-(carboxymethyl)-â-lactam (2.0 mmol), and EDC‚HCl

(4.0 mmol, 0.767 g). The reaction mixture was stirred at 0 °C for 1 h,

and then the bath was removed and the solution was stirred at room

temperature overnight. CH₂Cl₂(30 mL) was added, and the organic

solution was successively washed with 0.1 M KHSO₄(20 mL × 3;

until neutral pH) and H₂O (20 mL × 2), dried (MgSO₄), and evaporated

under reduced pressure. The intermediate benzyl ester was purified by

column chromatography (silica gel; eluant: EtOAc/hexanes). This

product was dissolved in dried EtOAc (10 mL), 10%Pd/C (0.10 g) was

added, and the suspension was stirred under hydrogen (1 atm) for 4-24

h. The reaction mixture was filtered through Celite, washed with EtOAc

(10 mL × 2), and evaporated in vacuo to afford the pure product.

(S)-1-(3-Aza-5-carboxy-4-methyl-2-oxopentyl)-3-tert-butoxycar-

bonylaminoazetidin-2-one (8). The general procedure was followed

from (S)-3-tert-butoxycarbonylamino-1-carboxymethyl-azetidin-2-one

(15) (2.0 mmol, 0.49 g); eluant EtOAc/hexanes 1:4. Intermediate benzyl

ester: (S)-1-(3-aza-5-benzyloxycarbonyl-4-methyl-2-oxo-pentyl)-3-tert-

butoxycarbonylaminoazetidin-2-one. IR (cm^-1, film NaCl): 3331 (NH);

2980; 2931; 1747; 1692 (CO); 1532. ¹H NMR (δ, ppm, CDCl₃): 7.62

(s, 1H, NHCMe₂); 7.27 (m, 5H, Ar); 5.28 (d, 1H, J ) 7.6 Hz, NHBoc);

5.13 (d, 2H, J ) 8.9 Hz, CH₂Ph); 5.08 (d, 2H, J ) 8.9 Hz, CH₂Ph);

4.40 (d, 1H, J ) 13.3 Hz, CH₂CONH); 4.34 (m, 1H, CHNHBoc); 3.59

(S)-3-tert-Butoxycarbonylamino-1-carboxymethylazetidin-2-

one (15). The general procedure was followed from (S)-3-tert-

butoxycarbonylaminoazetidin-2-one (14) (5 mmol, 0.93 g); eluant

EtOAc/hexanes 1:5, then 1:1. Intermediate benzyl ester: (S)-1-

[(benzyloxycarbonyl) methyl]-3-tert-butoxycarbonylamino-azetidin-2-

one. [R]²⁵) -3.4 (c ) 1.0, Cl₂CH₂). IR (cm^-1, KBr): 2976.6;

D

1

2928.4(NH); 1746.3; 1712.5 (CO). H NMR (δ, ppm, CDCl₃): 7.37

(s, 5H, Ar); 5.17-5.14 (sb, 2H, CH₂Ph, NHBoc); 4.93 (s_b, 1H,

CHNHBoc); 4.07 (s_b, 2H, CH₂CO₂Bn); 3.73 (t, 1H, J ) 5.4 Hz, CH₂-

â-lactam); 3.40 (dd, 1H, J ) 2.6, 5.4 Hz, CH₂-â-lactam); 1.44 (s, 9H,

C(CH₃)₃). ¹³C NMR (δ, ppm, CDCl₃): 168.8; 168.1; 155.5; 135.6;

9

16256 J. AM. CHEM. SOC. VOL. 125, NO. 52, 2003

Restricted Peptides as Potent Nucleators of â-Turns

A R T I C L E S

(dd, 1H, J ) 2.6, 5.6 Hz, CH₂lactam); 3.40 (m, 1H, CH₂lactam);

3.37 (d, 1H, J ) 13.3 Hz, CH₂CONH); 1.53 (s, 3H, CH₃); 1.48 (s, 3H,

CH₃); 1.36 (s, 9H, C(CH₃)₃). ¹³C NMR (δ, ppm, CDCl₃): 167.5; 167.2;

155.6; 136.7; 129.1; 128.7; 81.7; 67.6; 59.4; 57.3; 48.1; 45.9; 28.9;

26.6; 25.2. Reaction time (hydrogenation): 24 h. Yield (overall): 0.612

g (93%).

dissolved in CH₂Cl₂(10 mL). The organic solution was washed with

a saturated aqueous solution of NaHCO₃(20 mL × 2), dried (MgSO₄),

and evaporated under reduced pressure to afford crude intermediate

R-amino-â-lactam (0.29 mmol), which was dissolved in dry CH₂Cl₂

(3.0 mL). The mixture was cooled to 0 °C, and HOBt (0.50 mmol,

0.043 g), DMF (0.3 mL), Boc-glycine (0.50 mmol, 0.056 g), and EDC‚

HCl (0.69 mmol, 0.083 g) were added. The mixture was stirred at 0

°C for 1 h. After this time, the bath was removed and the solution was

stirred at room temperature overnight. The reaction was treated with

CH₂Cl₂(2 mL), and the organic solution was washed successively with

0.1 M KHSO₄(3 mL) and H₂O (3 mL), dried (MgSO₄), and evaporated

under reduced pressure. The product was purified by chromatography

(silica gel, eluant: EtOAc/hexanes 5:1). Yield (overall): 0.105 g (51%).

(3R,4R)-1-(3-Aza-4-carboxy-4-methyl-2-oxopentyl)-3-benzyl-3-

tert-butoxycarbonylamino-4-isobutylazetidin-2-one (10). The general

procedure was followed from (3R,4R)-3-benzyl-3-tert-butoxycarbonyl-

amino-1-carboxymethyl-4-isobutylazetidin-2-one (20) (2.0 mmol, 0.781

g); eluant EtOAc/hexanes 1:3. Intermediate benzyl ester: (3R,4R)-1-

(3-aza-4-benzyloxycarbonyl-4-methyl-2-oxopentyl)-3-benzyl-3-tert-bu-

toxiaminocarbonyl-4-isobutylazetidin-2-one. [R]²⁵_D) +66.2 (c ) 0.52,

Cl₂CH₂). IR (cm^-1, KBr): 3322.5; 2950.0 (NH); 1763.9; 1744.0;

1691.4; 1678.9 (CdO); ¹H NMR (δ, ppm, CDCl₃): 7.83 (s, 1H, NHCd

O); 7.39-7.28 (m, 10H, Ar); 5.17 (s, 2H, OCH₂Ph); 4.70 (s, 1H,

NHBoc); 4.42 (d, 1H, J ) 17.6 Hz, NCH₂CdO); 4.28 (t, 1H, J ) 6.45

Hz, CH-ⁱBu); 3.52 (d, 1H, J ) 17.4 Hz, NCH₂CdO); 3.22 (d, 1H, J )

14.7 Hz, CH₂Ph); 2.94 (d, 1H, J ) 14.8 Hz, CH₂Ph); 1.69-1.51 (m,

3H, CH₂CH(CH₃)₂); 1.58 (s, 3H, -NHC(CH₃)₂); 1.56 (s, 3H, -NHC-

(CH₃)₂); 1.38 (s, 9H, Boc); 0.95 (d, 3H, J ) 2.9 Hz, CH₂CH(CH₃)₂);

0.93 (d, 3H, J ) 2.9 Hz, CH₂CH(CH₃)₂). ¹³C NMR (δ, ppm, CDCl₃):

173.9; 168.7; 166.6; 154.6; 136.2; 133.8; 130.1; 128.9; 128.4; 128.0;

127.9; 127.6; 80.9; 68.8; 66.8; 60.8; 56.5; 43.6; 36.7; 28.1; 26.3; 25.6;

24.8; 23.2; 22.3. Reaction time (hydrogenation): 4 h. Yield (overall):

0.655 g (69%).

(3R)-1-[(Aminocarbonyl)methyl]-3-isobutyl-3-[(2S)-pyrrolidin-2-

ylcarbonylamino]azetidin-2-one (13). In a flame-dried three necked

round-bottomed flask fitted with a N₂-stream system, a condenser, and

an ammonia cylinder inlett was put Li (11.9 mmol, 0.082 g; finely

divided wire). Then, NH₃(40 mL) was condensed (acetone, CO₂bath)

at -78 °C until the solution turned to deep blue and all the lithium

was dissolved. A solution of (3R)-1-[(aminocarbonyl)methyl]-3-isobu-

tyl-3-[(4S)-4-phenyl-2-oxooxazolidin-3-yl]azetidin-2-one (31) (1.98

mmol, 0.684 g) in dry THF/t-BuOH (10:1) (12 mL) was added

dropwise. After the addition, the mixture was stirred at -78 °C for 5

min. Then, the flask was opened and solid NH₄Cl (11.9 mmol, 0.636

g) was added. The reaction mixture was left stirring at room temperature

for evaporation of the NH₃. The solvents were evaporated, and the

residue was dissolved in dry DMF (9 mL) and introduced in a flamed

round-bottomed flask under N₂. To this solution were added Cbz-Pro-

fluoride (2.97 mmol, 0.745 g) in dry CH₂Cl₂(3 mL) and N-

methylmorpholine (5.94 mmol, 0.69 mL). The mixture was stirred at

room temperature for 8 h, and CH₂Cl₂(15 mL) was added. The resulting

solution was successively washed with H₂O (10 mL), 1 M HCl (10

mL), and a saturated aqueous NaHCO₃(10 mL). The organic layer

was decanted and dried (MgSO₄), and the solvents were evaporated

under reduced pressure. The intermediate (3R)-1-[(aminocarbonyl)-

methyl]-3-isobutyl-3-[(2S)-1-(benzyloxycarbonyl)pyrrolidin-2-ylcarbo-

nylamino]azetidin-2-one was purified by column chromatography (silica

gel; CH₂Cl₂/MeOH 95:5). Yield (overall from 31): 0.443 g (52%).

(3S,4R)-1-(3-Aza-4-carboxy-4-methyl-2-oxopentyl)-3-benzyl-3-

tert-butoxycarbonylamino-4-isobutylazetidin-2-one (11): The general

procedure was followed from (3S,4R)-3-benzyl-3-tert-butoxycarbonyl-

amino-1-carboxymethyl-4-isobutylazetidin-2-one (23) (2.0 mmol, 0.781

g); eluant EtOAc/hexanes 1:2. Intermediate benzyl ester: (3S,4R)-1-

(3-aza-4-benzyloxycarbonyl-4-methyl-2-oxopentyl)-3-benzyl-3-tert-bu-

toxiaminocarbonyl-4-isobutylazetidin-2-one. [R]²⁵_D) -59.2 (c ) 0.52,

Cl₂CH₂). IR (cm^-1, KBr): 3322; 3276 (NH); 2960; 1772; 1762; 1698;

1

1655 (CdO). H NMR (δ, ppm, CDCl₃): 7.72 (s, 1H, NH); 7.39-

7.24 (m, 10H, Ar); 5.17 (d, 1H, J ) 12.2 Hz, OCH₂Ph); 5.12 (d, 1H,

J ) 12.2 Hz, OCH₂Ph); 4.66 (s, 1H, BocNH); 4.27 (d, 1H, J ) 17.6

Hz, NCH₂CO); 3.73 (t, 1H, J ) 6.8 Hz, NCH(ⁱBu)); 3.27 (d, 1H, J )

17.6 Hz, NCH₂CO); 3.20 (d, 1H, J ) 14.2 Hz, CCH₂Ph); 3.01 (d, 1H,

J ) 14.2 Hz, CCH₂Ph); 1.59 (s, 3H, NC(CH₃)₂(CO)); 1.60-1.52 (m,

1H, CH₂CH(CH₃)₂); 1.52 (s, 3H, NC(CH₃)₂(CO)); 1.46-1.42 (m, 1H,

CH₂CH(CH₃)₂); 1.42 (s, 9H, Boc); 1.33-1.30 (m, 1H, CH₂CH(CH₃)₂);

0.79 (d, 3H, CH₂CH(CH₃)₂); 0.78 (d, 3H, CH₂CH(CH₃)₂). ¹³C NMR

(δ, ppm, CDCl₃): 173.8; 167.9; 167.4; 154.8; 136.2; 134.1; 129.8;

129.1; 128.4; 128.0; 127.8; 80.7; 70.2; 66.7; 64.3; 56.3; 53.4; 44.9;

40.5; 36.6; 28.1; 25.6; 25.4; 24.6; 22.7; 22.6. Reaction time (hydrogena-

tion): 4 h. Yield (overall): 0.580 g (61%).

[R]²⁵) -44.8 (c ) 0.59, Cl₂CH₂). IR (cm^-1, KBr): 3392.7; 3181.2

D

(NH); 2955.2; 2922.7; 2869.0; 1750.0; 1687.9; 1667.0 (CO). MS m/z

(int): 91 (100); 160 (39.3); 288 (11.0); 330 (6.6); 429 (16.6); ¹H NMR

(δ, ppm, CDCl₃): 8.21 (s, 1H, -NHCO); 7.70 (s 1H, -CONH₂); 7.37

(s, 5H, Ar); 5.48 (s, 1H, -CONH₂); 5.17 (m, 2H, PhCH₂O); 4.47 (d,

1H, J ) 18.1 Hz, -NCH₂C-); 4.39 (s_b, 1H, Cbz-NCHCH₂-); 3.79

(s_b, 1H, -NCH₂CONH₂); 3.50 (s_b, 2H, Cbz-NCH₂CH₂-); 3.43 (d, 1H,

J ) 17.6 Hz, -NCH₂C-); 3.30 (s_b, 1H, -NCH₂CONH₂); 1.93 (m,

5H, CbzNCH₂CH₂-; CH₂CH(CH₃)₂); 1.57 (m, 2H, CbzNCHCH₂-);

1.00 (s_b, 3H, CH₂CH(CH₃)₂); 0.90 (d, 3H, J ) 6.4 Hz, CH₂CH(CH₃)₂).

¹³C NMR (δ, ppm, CDCl₃): 171.7; 170.3; 168.4; 156.8; 136.2; 128.6;

128.3; 127.9; 68.2; 67.6; 60.1; 53.2; 47.2; 44.9; 42.2; 26.9; 24.6; 23.8;

22.4. A suspension of this compound (1.0 mmol, 0.430 g) and (10%)

Pd/C (0.1 mmol, 0.10 g) in EtOH (10 mL) was stirred at room

temperature for 24 h under hydrogen (1 atm.). The solution was filtered

through a pad of Celite (washed several times with CH₂Cl₂), and the

solvents were evaporated under reduced pressure to afford pure product.

Yield (hydrogenolysis): 100%. Yield (overall from 31): 0.305 g (52%).

(R)-1-(3-Aza-4-benzyloxycarbonyl-2-oxobutyl)-3-benzyl-3-(2-tert-

butoxycarbonylaminoazetidin-2-one (16). To a solution of (3R)-3-

benzyl-1-carboxymethyl-3-tert-butoxycarbonylaminoazetidin-2-one (5)

(1.1 mmol, 0.371 g) in dry CH₂Cl₂(10 mL) was added EDC‚HCl (1.54

mmol, 0.30 g), p-toluensulfonate salt of glycine benzyl ester (1.5 mmol,

0.50 g) dissolved in CH₂Cl₂(5 mL), and Et₃N (3.0 mmol, 0.42 mL).

The mixture was stirred at room temperature for 18 h. After this time,

the solution was washed with 0.1 M HCl (6 mL × 3), saturated aqueous

solution of NaHCO₃(20 mL), and H₂O (20 mL). The organic phase

was dried (MgSO₄) and evaporated under reduced pressure, and the

resulting crude was purified by column chromatography (silica gel,

eluant: EtOAc/hexanes 3:1). Yield: 0.434 g (82%).

General Procedure for the r-Alkylation of 4-Alkyl(aryl)-1-[bis-

(trimethylsilyl)methyl]-3-[(4S)-4-phenyl-2-oxooxazolidin-3-yl]azeti-

din-2-ones. Synthesis of 18, 26, 27, 28, and 29. To a suspension of

1,10-phenantroline (2 mg, indicator) in anhydrous THF (16 mL) cooled

to -78 °C under nitrogen atmosphere was added 2.5 M n-BuLi

dropwise until a dark red color was observed (usually 2 or 3 drops).

Dry diisopropylamine (12.0 mmol, 1.68 mL) and 2.5 M n-BuLi (12.0

mmol, 4.8 mL) were added, and the mixture was stirred at -78 °C for

30min. A solution of the corresponding 4-alkyl(aryl)-1-[bis(trimethyl-

silyl)methyl]-3-[(4S)-4-phenyl-2-oxooxazolidin-3-yl]azetidin-2-one (8.0

(R)-1-(3-Aza-4-benzyloxycarbonyl-2-oxbutyl)-3-benzyl-3-(1-aza-

3-tert-butoxycarbonylamino-2-oxopropyl)azetidin-2-one (9). To a

solution of (R)-1-(3-aza-4-benzyloxycarbonyl-2-oxobutyl)-3-benzyl-3-

tert-butoxycarbonylaminoazetidin-2-one (16) (0.46 mmol, 0.222 g) in

dry CH₂Cl₂(4 mL) was added trifluoroacetic acid (TFA) (13 mmol,

1.0 mL). The mixture was stirred at room temperature for 90 min, TFA

was removed under reduced pressure, and the resulting residue was

9

J. AM. CHEM. SOC. VOL. 125, NO. 52, 2003 16257

A R T I C L E S

Palomo et al.

mmol) in anhydrous THF (24 mL) was added dropwise, and the stirring

was continued for 1 h at -78 °C. Freshly distilled alkyl bromide (40.0

mmol) was added slowly, and the dry ice/acetone bath was allowed to

reach room temperature overnight. The reaction mixture was taken up

over CH₂Cl₂(65 mL), washed successively with aqueous saturated NH₄-

Cl (32 mL × 2), 1 M HCl (20 mL), saturated aqueous NaHCO₃(20

mL), and H₂O (20 mL × 2), dried (MgSO₄), and evaporated at reduced

pressure. The products were purified by column chromatography (silica

gel, eluant: EtOAc/hexanes).

(3R,4R)-3-Benzyl-1-[bis(trimethylsilyl)methyl]-4-isobutyl-3-[(4S)-

4-phenyl-2-oxooxazolidin-3-yl]azetidin-2-one (18). The general pro-

cedure was followed from (3S,4R)-1-[bis(trimethylsilyl)methyl]-4-

isobutyl-3-[(4S)-4-phenyl-2-oxooxazolidin-3-yl]-azetidin-2-one (17) (8.0

mmol, 3.57 g) and benzyl bromide (40.0 mmol, 4.86 mL). Eluant:

EtOAc/hexanes 1:10. Yield: 3.56 g (83%).

(3R,4R)-3-Allyl-1-[bis(trimethylsilyl)methyl]-4-isobutyl-3-[(4S)-

4-phenyl-2-oxooxazolidin-3-yl]azetidin-2-one (26). The general pro-

cedure was followed using a 1/10 scale from (3S,4R)-1-[bis-

(trimethylsilyl)methyl]-4-isobutyl-3-[(4S)-4-phenyl-2-oxooxazolidin-3-

yl]azetidin-2-one (17) (0.8 mmol, 0.357 g) and allyl bromide (4.0 mmol,

0.352 mL). Eluant: EtOAc/hexanes 1:5. Yield: 0.194 g (50%).

(3R,4R)-1-[Bis(trimethylsilyl)methyl]-3-(4-bromobenzyl)-4-ethyl-

3-[(4S)-4-phenyl-2-oxooxazolidin-3-yl]azetidin-2-one (27). The gen-

eral procedure was followed using a 1/10 scale from (3S,4R)-1-

[bis(trimethylsilyl)methyl]-4-ethyl-3-[(4S)-4-phenyl-2-oxooxazolidin-

3-yl]azetidin-2-one (24) (0.8 mmol, 0.329 g) and 4-bromobenzyl

bromide (4.0 mmol, 0.586 mL). Eluant: EtOAc/hexanes 1:5. Yield:

0.282 g (60%).

2-one was purified by column chromatography (silica gel; eluant:

EtOAc/hexanes 1:200, then 1:100, and finally 1:50). Yield (overall from

18): 5.24 g (89%). [R]²⁵) +3.11 (c) 0.96, Cl₂CH₂). IR (cm^-1

,

D

KBr): 2943.0 (NH); 1742.2, 1717.0 (CdO). ¹H NMR (δ, ppm,

CDCl₃): 7.31-7.21 (m, 5H, Ar); 4.97 (s, 1H, NH); 3.76 (t, 1H, J )

6.5 Hz, CCHN); 3.36 (d, 1H, J ) 13.7 Hz, CH₂Ph); 2.93 (d, 1H, J )

13.7 Hz, CH₂Ph); 2.08 (s_b, 2H, CH(CH₃)₂, CH(SiMe₃)₂); 1.72-1.52

(m, 1H, CH₂CH(CH₃)₂); 1.50-1.41 (m, 1H, CH₂CH(CH₃)₂); 1.41 (s,

9H, Boc); 1.03 (t, 6H, J ) 6.6 Hz, CH(CH₃)₂); 0.22 (s, 9H, Si(CH₃)₂);

0.15 (s, 9H, Si(CH₃)₂). ¹³C NMR (δ, ppm, CDCl₃): 166.9; 154.5; 135.8;

130.4; 127.9; 126.5; 79.3; 67.1; 65.0; 37.9; 36.9; 35.0; 28.1; 25.2; 22.8;

22.7; 0.0. To a solution of this product (10.7 mmol, 5.24 g) in dry

acetonitrile (50 mL) cooled to 0 °C was added a solution of Ce(NH₄)₂-

(NO₃)₄(45.9 mmol, 25.2 g) in H₂O (30 mL), and the mixture was stirred

at 0 °C for 2 h (TLC). EtOAc (100 mL) and H₂O (50 mL) were added,

and the organic layer was separated, washed with saturated aqueous

NaHCO₃(50 mL), dried (MgSO₄), and evaporated under reduced

pressure. The crude intermediate (3R,4R)-3-benzyl-1-formyl-3-tert-

butoxycarbonylamino-4-isobutyl-3-azetidin-2-one was dissolved in a

mixture of methanol (33 mL), saturated aqueous NaHCO₃(33.5 mL),

and Na₂CO₃(5.3 mmol, 0.56 g) ,and the suspension was stirred at room

temperature for 2 h. Then, H₂O (40 mL) was added, the solution was

extracted with CH₂Cl₂(25 mL × 2), and the organic layer was dried

(MgSO₄) and evaporated in vacuo. The product was purified by column

chromatography (silica gel, eluant EtOAc/hexanes 1:3). Yield (overall

from 18): 4.07 g (76%).

(3S,4R)-3-Benzyl-3-benzylideneamino-4-isobutyl-1-[bis(trimeth-

ylsilyl)methyl]azetidin-2-one (21). In a flame-dried three-necked

round-bottomed flask fitted with a N₂-stream system, a condenser, and

an ammonia cylinder inlett, was put Li (30.6 mmol, 0.21 g; finely

divided wire). Then, NH₃(90 mL) was condensed (acetone, CO₂bath)

at -78 °C until the solution turned to deep blue and all the lithium

was dissolved. A solution of (3S,4R)-1-[bis(trimethylsilyl)methyl]-4-

isobutyl-3-[(4S)-4-phenyl-2-oxooxazolidin-3-yl]azetidin-2-one (17) (5.1

mmol, 2.03 g) in dry THF/t-BuOH (10:1) (28 mL), was added dropwise.

After the addition the mixture was stirred at -78 °C for 5 min, the

flask was opened and solid NH₄Cl (30.6 mmol, 1.63 g) was added.

The reaction mixture was left stirring at room temperature overnight

until the complete evaporation of ammonia. The solvents were

evaporated under reduced pressure and the residue was dissolved in

H₂O (85 mL), acidified with 6 M HCl to pH ) 3-4, and basified with

aqueous 40% NaOH to pH ) 10. The intermediate R-amino-â-lactam

was extracted with CH₂Cl₂(60 mL × 3), the organic phase was dried

(MgSO₄) and the solvents were removed under reduced pressure. To a

solution of this crude (5.0 mmol) in dry CH₂Cl₂(15 mL) under nitrogen

atmosphere and with molecular sieves (4 Å; 2 g) was added benzal-

dehyde (5.0 mmol, 0.5 mL), and the mixture was stirred at room

temperature for 1 h. After this time the molecular sieves were filtered

and the solvents were evaporated. The crude benzaldimine was used

without further purification. To a suspension of 1,10-phenantroline (2

mg, indicator) in anhydrous THF (40 mL) under nitrogen atmosphere

and cooled to -78 °C 2.5 M n-BuLi in hexane was added dropwise

until dark red color was observed (usually 2 or 3 drops). Dry DIPA

(5.44 mmol, 0.78 mL) and 2.5 M n-BuLi (5.40 mmol, 2.16 mL) were

added, and the mixture was stirred at -78 °C for 30 min. A solution

of (3S,4R)-3-benzylideneamino-4-isobutyl-1-[bis(trimethylsilyl)methyl]-

azetidin-2-one (4.53 mmol, 1.76 g) in dry THF (40 mL) was added

dropwise and stirring was continued for 1 h at -78 °C. After this time,

freshly distilled benzyl bromide (13.5 mmol, 1.61 mL) was added

slowly, and the reaction mixture was stirred to room temperature

overnight. The solution was diluted with CH₂Cl₂(45 mL), washed

successively with saturated aqueous NH₄Cl (25 mL), 1 M HCl (8 mL),

and saturated aqueous NaHCO₃(25 mL), dried (MgSO₄), and evapo-

rated. The product was used without further purification. Yield (overall

from 17): 1.295 g (53%).

(3R,4R)-1-[Bis(trimethylsilyl)methyl]-3-(2-methyl-2-propenyl)-4-

phenyl-3-[(4S)-4-phenyl-2-oxooxazolidin-3-yl]azetidin-2-one (28).

The general procedure was followed using a 1/10 scale from (3S,4R)-

1-[bis(trimethylsilyl)methyl]-4-phenyl-3-[(4S)-4-phenyl-2-oxooxazoli-

din-3-yl]azetidin-2-one (25) (0.8 mmol, 0.373 g) and methallyl bromide

(4.0 mmol, 0.416 mL). Eluant: EtOAc/hexanes 1:4. Yield: 0.242 g

(58%).

(3R,4R)-1-[Bis(trimethylsilyl)methyl]-3-(2-methyl-2-propenyl)-4-

phenyl-3-[(4S)-4-phenyl-2-oxooxazolidin-3-yl]azetidin-2-one (29).

The general procedure was followed using a 1/10 scale from (3S,4R)-

1-[bis(trimethylsilyl)methyl]-4-phenyl-3-[(4S)-4-phenyl-2-oxooxazoli-

din-3-yl]azetidin-2-one (25) (0.8 mmol, 0.373 g) and 4-methylbenzyl

bromide (4.0 mmol, 0.752 mL). Eluant: EtOAc/hexanes 1:4. Yield:

0.333 g (80%).

(3R,4R)-3-Benzyl-3-tert-butoxycarbonylamino-4-isobutyl-azetidin-

2-one (19). In a flame-dried three-necked round-bottomed flask fitted

with a N₂-stream system, a condenser, and an ammonia cylinder inlett

was put Li (72.0 mmol, 0.50 g; finely divided wire). Then, NH₃(190

mL) was condensed (acetone, CO₂bath) at -78 °C until the solution

turned to deep blue and all the lithium was dissolved. A solution of

(3R,4R)-3-benzyl-1-[bis(trimethylsilyl)methyl]-4-isobutyl-3-[(4S)-4-

phenyl-2-oxooxazolidin-3-yl]azetidin-2-one (18) (12.0 mmol, 6.44 g)

in dry THF/t-BuOH (10:1) (65 mL) was added dropwise. After the

addition, the mixture was stirred at -78 °C for 5 min, the flask was

opened, and solid NH₄Cl (72.0 mmol, 3.90 g) was added. The reaction

mixture was left stirring at room temperature overnight until complete

evaporation of ammonia. The solvents were evaporated under reduced

pressure, and the residue was dissolved in H₂O (190 mL), acidified

with 6 M HCl to pH ) 3-4, and basified with aqueous 40% NaOH to

pH ) 10. The product was extracted with CH₂Cl₂(130 mL × 3), the

organic phase was dried (MgSO₄), and the solvents were removed under

reduced pressure. The resulting crude (11.9 mmol) was dissolved in

CH₂Cl₂(50 mL), di-tert-butyl dicarbonate (23.8 mmol, 5.35 g) was

added, and the solution was stirred at room temperature for 22 h. It

was washed successively with H₂O (22 mL), 20% NaHSO₃(22 mL),

and saturated aqueous NaHCO₃(22 mL), dried (MgSO₄) _,and evaporated

under reduced pressure. The intermediate (3R,4R)-3-benzyl-1-[bis-

(trimethylsilyl)methyl]-3-tert-butoxycarbonylamino-4-isobutylazetidin-

9

16258 J. AM. CHEM. SOC. VOL. 125, NO. 52, 2003

Restricted Peptides as Potent Nucleators of â-Turns

A R T I C L E S

(3R, 4R)-3-Benzyl-3-tert-butoxycarbonylamino-4-isobutyl-azeti-

din-2-one (22). To a solution of crude (3S,4R)-3-benzyl-3-benzylide-

neamino-4-isobutyl-1-[bis(trimethylsilyl)methyl]azetidin-2-one (21) (5.0

mmol, 1.94 g) in Et₂O (50 mL) and H₂O (0.3 mL) cooled to 0 °C (ice

bath) was added strongly acidic Amberlist-15 resin (15 g; pH < 3),

and the mixture was stirred at room temperature for 1 h (monitored by

TLC until complete imine consumption). The resin was filtered and

suspended in CH₂Cl₂(60 mL), hexamethyldisilazane (20 mL) was

added, and the suspension was stirred overnight. The resin was filtered

off, and the solvents were evaporated under reduced pressure to afford

crude(3R,4R)-3-amino-3-benzyl-4-isobutyl-1-[bis(trimethylsilyl)methyl]-

azetidin-2-one which was dissolved in t-BuOH (40 mL) under nitrogen.

To this solution were added tri(n-butyl)phosphine (1.0 mmol, 0.200 g)

and (Boc)₂O (10 mmol, 2.30 g), the mixture was stirred at 70 °C for

24 h, and the t-BuOH was evaporated under reduced pressure. The

intermediate (3S,4R)-3-benzyl-1-[bis(trimethylsilyl)methyl]-3-tert-bu-

toxycarbonylamino-4-isobutylazetidin-2-one was purified by column

chromatography (silica gel; eluant: EtOAc/hexanes 1:50, then 1:20,

1H, HCHd); 4.72 (s, 1H, HCHd); 4.59 (t, 1H, J ) 8.4 Hz, HCPh);

4.40 (dd, 1H, J ) 8.6, 1.9 Hz, OCHHCHPh); 3.62 (d, 1H, J ) 6.5 Hz,

HCHN); 3.57 (d, 1H, J ) 6.5 Hz, HCHN); 2.73 (s, 1H, CHSi); 2.33

(d, 1H, J ) 13.7 Hz, CH₂Cd); 1.61 (d, 1H, J ) 13.7 Hz, CH₂Cd);

1.57 (s, 3H, CH₃); 0.15 (s, 9H, (CH₃)₃); 0.12 (s, 9H, (CH₃)₃). ¹³C NMR

(δ, ppm, CDCl₃): 164.8; 156.6; 140.1; 139.6; 129.1; 128.9; 127.8;

116.2; 71.1; 70.6; 59.6; 52.9; 39.4; 37.1; 23.8; -0.3. This product was

dissolved in EtOH (20 mL) containing 10% Pd/C (0.50 g), and the

mixture was stirred for 24 h under hydrogen atmosphere (1 atm) at

room temperature. After this time, the suspension was filtered through

a pad of Celite and washed with CH₂Cl₂(20 mL × 2), and the combined

organic solution evaporated under reduced pressure to afford pure

product. Yield (overall): 5.50 g (77%).

(3R)-1-[(Aminocarbonyl)methyl]-3-isobutyl-3-[(4S)-4-phenyl-2-

oxooxazolidin-3-yl]azetidin-2-one (31). A 0.60 M solution of n-BuLi/

TMEDA in THF was prepared under nitrogen by adding dried TMEDA

(12 mmol, 1.8 mL) and 2.5 M n-BuLi (12 mmol, 4.8 mL) to anhydrous

THF (13.0 mL) cooled to -100 °C (MeOH/N₂bath). In another flask,

a solution of (3R)-1-[bis(trimethylsilyl)methyl]-3-isobutyl-3-[(4S)-4-

phenyl-2-oxooxazolidin-3-yl]azetidin-2-one (30) (4 mmol, 1.79 g) was

dissolved in anhydrous THF (12 mL) and cooled to -100 °C under

nitrogen, and n-BuLi/TMEDA (0.60 M solution in THF, 4.8 mmol,

8.0 mL) was added dropwise. The mixture, which turned to a red-

orange color, was stirred at -100 °C for 30 min. Then, trimethylsilyl

isocyanate (12 mmol, 1.60 mL) was added, and the mixture was stirred

at -78 °C for 2 h. The reaction mixture was quenched at low

temperature (-78 °C) with saturated aqueous NH₄Cl (16 mL) and left

to reach room temperature by removing the cooling bath. Extraction

with CH₂Cl₂(20 mL × 3), drying (MgSO₄), evaporation, and

purification by column chromatography (silica gel, eluant: EtOAc

100%) afforded the product. Yield: 1.10 g (80%).

(3R)-1-Amidomethyl-3-isobutyl-3-[(2S)-1-(tert-butoxycarbonyl)-

pyrrolidin-2-ylcarbonylamino]azetidin-2-one (32). A solution of (3R)-

1-amidomethyl-3-isobutyl-3-[(2S)-pyrrolidin-2-ylcarbonylamino]azetidin-

2-one (13) (1 mmol, 0.296 g) and di-tert-butyl dicarbonate (2 mmol,

0.436 g) in THF (3.0 mL) was stirred at room temperature for 16 h

and then was poured into a mixture of CH₂Cl₂(20 mL) and H₂O (5

mL). The organic layer was washed successively with H₂O (10 mL),

20% NaHSO₃(10 mL), and saturated aqueous NaHCO₃(10 mL), dried

(MgSO₄), and evaporated under reduced pressure. The product was

purified by column chromatography (silica gel, eluants: EtOAc/hexanes

1:10, then, CH₂Cl₂/MeOH 90:10). Yield: 0.368 g (93%).

Molecular Dynamics of Compounds 6, 8, 9, 10, 11, 12, and 13.

MD simulations were carried out at 300 K and 1 atm using the

SANDER module in AMBER 6. The updated (parm99) second-

generation⁴⁸AMBER force field⁴⁹was used. Each compound was

immersed in a periodic box of ∼160 DMSO⁵⁰molecules that extended

up to 10 Å away from any solute atom. Periodic boundary conditions

were applied. A dielectric constant of 1 was used, and the cutoff distance

for the nonbonded interactions was 10 Å. The geometries of the

molecules were fully optimized with the ab initio quantum mechanical

program Gaussian 98⁵¹at the Hartree-Fock level using the 3-21G(d)

basis set. Partial atomic charges were then obtained using the RESP⁵²

methodology with the 6-31G(d) basis set. SHAKE⁵³was applied to all

bonds involving hydrogen atoms, and the integration time step was 1

fs. The simulation protocol consisted of a series of progressive energy

then 1:10 and finally 1:5). Yield (overall from 21): 1.34 g (60%). [R]²⁵

D

) +3.04 (c ) 0.92, Cl₂CH₂). IR (cm^-1, KBr): 1702.9; 1731.8 (Cd

1

O). H NMR (δ, ppm, CDCl₃): 7.34-7.20 (m, 5H, Ar); 4.84 (s, 1H,

NHBoc); 3.59-3.54 (d, m, 2H, CH₂Ph y HC(ⁱBu)); 3.29 (d, 1H, J )

13.7 Hz, CH₂Ph); 2.00 (s, 1H, CH(Si(CH₃)₃)₂); 1.80-1.78 (m, 1H,

CH₂CH(CH₃)₂); 1.60-1.55 (m, 1H, CH₂CH(CH₃)₂); 1.42 (s, 9H,

OCOC(CH₃)₃); 1.35-1.27 (m, 1H, CH₂CH(CH₃)₂); 0.94 (s, 3H, CH₂-

CH(CH₃)₂); 0.93 (s, 3H, CH₂CH(CH₃)₂); 0.17 (s, 9H, Si(CH₃)₃); 0.07

(s, 9H, Si(CH₃)₃). ¹³C NMR (δ, ppm, CDCl₃): 166.4; 154.4; 135.9;

130.6; 130.5; 128.4; 128.2; 126.7; 79.6; 69.3; 64.1; 39.0; 37.0; 36.9;

36.4; 28.2; 27.8; 25.4; 23.2; 22.7; 0.09; -0.13. To a solution of this

product (3.0 mmol, 1.34 g) in dry acetonitrile (20 mL) cooled to 0 °C

was added a solution of Ce(NH₄)₂(NO₃)₄(13.5 mmol, 7.40 g) in H₂O

(10 mL). and the mixture was stirred at 0 °C for 2 h (TLC). EtOAc

(100 mL) and H₂O (50 mL) were added, and the organic layer was

separated, washed with saturated aqueous NaHCO₃(50 mL), dried

(MgSO₄), and evaporated under reduced pressure. The crude intermedi-

ate (3S,4R)-3-benzyl-1-formyl-3-tert-butoxycarbonylamino-4-isobutyl-

3-azetidin-2-one was dissolved in a mixture of methanol (30 mL),

saturated aqueous NaHCO₃(30 mL), and Na₂CO₃(3 mmol, 0.32 g),

and the suspension was stirred at room temperature for 2 h. Then, H₂O

(40 mL) was added, the solution was extracted with CH₂Cl₂(20 mL ×

2), and the organic layer was dried (MgSO₄) and evaporated in vacuo.

The product was purified by column chromatography (silica gel, eluant

EtOAc/hexanes 1:2). Yield (overall from 21): 0.860 g (52%).

(3R)-1-[Bis(trimethylsilyl)methyl]-3-isobutyl-3-[(4S)-4-phenyl-2-

oxooxazolidin-3-yl]azetidin-2-one (30). To a suspension of 1,10-

phenantroline (2 mg, indicator) in anhydrous THF (32 mL) cooled to

-78 °C under nitrogen atmosphere was added 2.5 M n-BuLi dropwise

until a dark red color was observed (usually 2 or 3 drops). Dry

diisopropylamine (20.0 mmol, 2.70 mL) and 2.5 M n-BuLi (20.0 mmol,

8.0 mL) were added, and the mixture was stirred at -78 °C for 30

min. A solution of (3S)-1-[bis(trimethylsilyl)methyl]-3-[(4S)-4-phenyl-

2-oxo-1,3-oxazolidin-3-yl]azetidin-2-one (4) (16.0 mmol, 6.30 g) in

anhydrous THF (32 mL) was added dropwise, and the stirring was

continued for 30 min at -78 °C. After this time, freshly distilled

3-bromo-2-methylpropene (80.0 mmol, 8.35 mL) was added slowly

and the dry ice/acetone bath was allowed to reach room temperature

overnight. The reaction mixture was taken up over CH₂Cl₂(65 mL),

washed successively with aqueous saturated NH₄Cl (32 mL × 2) and

H₂O (20 mL × 2), dried (MgSO₄), and evaporated at reduced pressure.

The intermediate (3R)-1-[bis(trimethylsilyl)methyl]-3-[2-methyl-2-pro-

penyl]-3-[(4S)-4-phenyl-2-oxooxazolidin-3-yl]azetidin-2-one was puri-

fied by column chromatography (silica gel, eluant: EtOAc/hexanes 1:8).

(48) Cornell, W. D.; Cieplak, P.; Bayly, C. I.; Gould, I. R.; Merz, K. M.;

Ferguson, D. M.; Spellmeyer, D. C.; Fox, T.; Caldwell, J. W.; Kollman, P.

A. A second generation force field for the simulation of proteins, nucleic

acids, and organic molecules. J. Am. Chem. Soc. 1995, 117, 5179-5197.

(49) Case, D. A.; Pearlman, D. A.; Caldwell, J. W.; Cheatham, T. E., III; Ross,

W. S.; Simmerling, C. L.; Darden, T. A.; Merz, K. M.; Stanton, R. V.;

Cheng, A. L.; Vincent, J. J.; Crowley, M.; Tsui, V.; Radmer, R. J.; Duan,

Y.; Pitera, J.; Massova, I.; Seibel, G. L.; Singh, U. C.; Weiner, P. K.;

Kollman, P. A. AMBER 6; University of California, San Francisco, 1999.

(50) The molecular structure of DMSO was optimized until the experimentally

measured density and heat of vaporization of this liquid were produced,

thereby lending credence to the suitability of the parameters that were used

in the simulations (see the Supporting Information).

[R]²⁵) +0.8 (c ) 1.0, Cl₂CH₂). IR (cm^-1, KBr): 1740 (CO). MS

D

m/z (int): 45 (27.2); 73 (100); 80 (23.3); 104 (100); 165 (39.6); 229

1

(100); 244 (25.3). H NMR (δ, ppm, CDCl₃): 7.35 (m, 3H, Ph); 7.57

(m, 2H, Ph); 5.13 (dd, 1H, J ) 8.1; 1.9 Hz, OCHHCHPh); 4.80 (s,

9

J. AM. CHEM. SOC. VOL. 125, NO. 52, 2003 16259

A R T I C L E S

Palomo et al.

minimizations followed by a 20 ps heating phase and a 400 ps

equilibration period during which distance restraints (10 kcal mol^-1

Å^-2) were applied to maintain the _(i)C)O‚‚‚H-N_(i+3)distance (2.8 (

0.1 Å) as this hydrogen bond characterizes the â-turn. This restraint

was removed thereafter and the system coordinates were then collected

every picosecond for 5 ns which yielded an ensemble of 5000 structures

for each peptide for further analysis.

DNAse in Hank’s balanced salt solution during 5 min at 37 °C. After

stopping enzymatic reaction, tissue was further homogenized by

mechanical pippeting. Neurons were seeded into 24-well plates bearing

12-mm diameter coverslips coated with poly-^D-lysine (10 µg/mL) at a

density of 3 × 10⁵cells/well in Neurobasal medium (Gibco) supple-

mented with B27, 0.5 mM glutamine, 1000 U/mL of penicillin, and

100 µg/mL of streptomycin.

[³H]Spiroperidol/N-Propylnorapomorphine (NPA) Binding Com-

petition Assay. Cells were incubated with 0.5 nM [³H]spiroperidol and

varying concentrations of N-propylnorapomorphine (NPA) for 1 h at

room temperature in a total volume of 0.5 mL. The reaction was finished

by rapid wash in ice-cold buffer, and the retained radioactivity was

counted after lysing cells with 0.1 N NaOH plus 0.01% Triton X-100.

Nonspecific binding was determined by performing the assay in the

presence of 1 µM (+)-butaclamol. Concentration of [³H]spiroperidol

in competitive binding assay was near its affinity constant (K_d) 0.99

nM) as previously determined by saturation binding experiments. The

binding data were analyzed using the nonlinear curve-fitting Prism

program (Graph Pad Software, San Diego, CA).⁵⁶The IC₅₀values

obtained from the competition curves were converted to K_iusing the

Cheng and Prusoff equation. Data were expressed as mean ( sem.

Statistical differences between inhibitor constant were analyzed with

ANOVA with repeated measurements (p < 0.05; n ) 5 in triplicate).

RMN/X-PLOR Annealing Study of Compounds 6, 8, 9, 10, 11,

and 13. Upper-bonded interproton distances were calculated from

NOESY intensities (see the Supporting Information, Table S1), and a

hybrid distance geometry-simulated annealing method was applied using

the program X-PLOR-NIH 2.0.6.⁴³The MD protocol sa.inp⁵⁴was used

with modified parallhdg.pro and topallhdg.pro files with the following

settings: 10000 steps at 700 K (step time 1.5 fs) and subsequent cooling

in 5000 steps to 300 K. The NOE and DIHE scales were set,

respectively, to 50 and 5.0, and a soft-square potential was used. All

other parameters were left unchanged.

Biological Evaluation of Compounds 12 and 13 as Dopaminergic

D₂Receptor Agonists on Cultured Neurons. Neuronal Culture

Preparation. Rat cerebral cortex⁵⁵were obtained as previously

described⁴⁷with minor modifications. After cerebral cortex removal

from E17 rat brains, tissue was treated with 0.2% trypsin plus 0.02%

(51) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M.

A.; Cheeseman, J. R.; Zakrzewski, V. G.; Montgomery, J. A., Jr.; Stratmann,

R. E.; Burant, J. C.; Dapprich, S.; Millam, J. M.; Daniels, A. D.; Kudin,

K. N.; Strain, M. C.; Farkas, O.; Tomasi, J.; Barone, V.; Cossi, M.; Cammi,

R.; Mennucci, B.; Pomelli, C.; Adamo, C.; Clifford, S.; Ochterski, J.;

Petersson, G. A.; Ayala, P. Y.; Cui, Q.; Morokuma, K.; Malick, D. K.;

Rabuck, A. D.; Raghavachari, K.; Foresman, J. B.; Cioslowski, J.; Ortiz,

J. V.; Stefanov, B. B.; Liu, G.; Liashenko, A.; Piskorz, P.; Komaromi, I.;

Gomperts, R.; Martin, R. L.; Fox, D. J.; Keith, T.; Al-Laham, M. A.; Peng,

C. Y.; Nanayakkara, A.; Gonzalez, C.; Challacombe, M.; Gill, P. M. W.;

Johnson, B. G.; Chen, W.; Wong, M. W.; Andres, J. L.; Head-Gordon,

M.; Replogle, E. S.; Pople, J. A. Gaussian 98; Gaussian, Inc.: Pittsburgh,

PA, 1998.

Acknowledgment. We thank The University of the Basque

Country and Ministerio de Ciencia y Tecnologia (Spain) for

financial support (Project No. BQU 2002-01737). Grants to A.B.

from Gobierno Vasco and to R.M. . from European Commis-

sions (Marie Curie HMPT-CT-2000-00173) are acknowledged.

Supporting Information Available: Analytical and spectral

characterization data of all new compounds, NMR/NOESY

interprotonic distances, and crystallographic data (CIF). This

material is available free of charge via the Internet at

http://pubs.acs.org.

(52) Bayly, C. I.; Cieplak, P.; Cornell, W. D.; Kollman, P. A. A J. Phys. Chem.

1993, 97, 10269-10280.

(53) Ryckaert, J. P.; Ciccoti, G.; Berendsen, H. J. C. J. Comput. Phys. 1977,

23, 327-341.

(54) Nilges, M.; Kuszewski, J.; Brunger, A. T.; Hoch, J. C.; Poulsen, F. M.;

Redfield, C. Ed.; Plenum Press: New York, 1991.

(55) This region was selected because is highly enriched in D₂receptors and

yields sufficient cells for culture experiments: Khan, Z. U.; Koulen, P.;

Rubinstein, M.; Grandy, D. K.; Goldman-Rakic, P. S. Proc. Natl. Acad.

Sci. U.S.A. 2001, 98, 1964-1969.

JA038180A

(56) Cheng, Y. C.; Prusoff, W. H. Biochem. Pharmacol. 1973, 22, 3099-3108.

9

16260 J. AM. CHEM. SOC. VOL. 125, NO. 52, 2003

Article Doi

Development of a New Family of Conformationally Restricted Peptides as Potent Nucleators of β-Turns. Design, Synthesis, Structure, and Biological Evaluation of β-Lactam Peptide Analogue of Melanostatin

DOI: 10.1021/ja038180a

Source and publish data:

Authors:

Article abstract of DOI:10.1021/ja038180a

Full text of DOI:10.1021/ja038180a

Products guided by the article

R&D Labs maybe for 651749-59-6

Relevant to this article

Hot Product