Synthesis of new photoactivatable phenylalanine analogues and their incorporation into a model peptide - Phenylseleno derivatives as precursors of α,β-unsaturated ketones in peptide synthesis

Article abstract of DOI:10.1002/1099-0690(200205)2002:10<1677::AID-EJOC1677>3.0.CO;2-8

The asymmetric synthesis of (S)-Boc-p-(propanoyl)phenylalanine (10) was performed by alkylation of sultam N-(di- phenylmethylene)glycinate (4). Different pathways for the introduction of a phenylseleno moiety α to the ketone function were investigated. (S)-Boc-p-[2′-(phenylselenenyl)propanoyl]phenylalanine (11) was easily obtained from (S)-Boc-p-(propanoyl)phenylalanine (10). However, the phenylseleno moiety α to the carbonyl group was found to be unstable to the conditions required for solid-phase peptide synthesis. Therefore, (2S)-Boc-p-[2′-(phenylselenenyl)propanoyl]phenylalanine (11) was transformed into Boc-p-[3′-(phenylselenenyl)propanoyl]phenylalanine (13), which was successfully incorporated into the sequence of a model peptide. The corresponding enone function generated under mild acidic conditions was found to be stable in aqueous solution at pH = 7, but suitably reactive upon irradiation. Thus, these phenylalanine analogs bearing an unsaturated ketone represent new photoreactive probes. Wiley-VCH Verlag GmbH, 69451 Weinheim, Germany, 2002.

Full text of DOI:10.1002/1099-0690(200205)2002:10<1677::AID-EJOC1677>3.0.CO;2-8

FULL PAPER
Synthesis of New Photoactivatable Phenylalanine Analogues and Their
Incorporation into a Model Peptide ؊ Phenylseleno Derivatives as Precursors
of α,β-Unsaturated Ketones in Peptide Synthesis
[a]
[b]
[c]
[d]
´
´
´
Valerie Jullian, Veronique Monjardet-Bas, Celine Fosse, Solange Lavielle, and
[d]
´
Gerard Chassaing*
Keywords: Amino acids / Peptides / Photoaffinity / Selenium
The asymmetric synthesis of (S)-Boc-p-(propanoyl)phenylal- ylalanine (11) was transformed into Boc-p-[3Ј-(phenyl-
anine (10) was performed by alkylation of sultam N-(di-
phenylmethylene)glycinate (4). Different pathways for the
introduction of a phenylseleno moiety α to the ketone func-
selenenyl)propanoyl]phenylalanine (13), which was success-
fully incorporated into the sequence of a model peptide. The
corresponding enone function generated under mild acidic
tion were investigated. (S)-Boc-p-[2Ј-(phenylselenenyl)pro- conditions was found to be stable in aqueous solution at pH =
panoyl]phenylalanine (11) was easily obtained from (S)-Boc-
p-(propanoyl)phenylalanine (10). However, the phenylseleno
moiety α to the carbonyl group was found to be unstable to
the conditions required for solid-phase peptide synthesis.
7, but suitably reactive upon irradiation. Thus, these phenyl-
alanine analogs bearing an unsaturated ketone represent
new photoreactive probes.
( Wiley-VCH Verlag GmbH, 69451 Weinheim, Germany,
Therefore, (2S)-Boc-p-[2Ј-(phenylselenenyl)propanoyl]phen- 2002)
Introduction
lysis or chemical degradation, and Ϫ finally Ϫ high cross-
linking yields are required for development of this strategy
with picomole amounts of receptor. Aryl azides, aryldiazir-
ines, α-diazo-carbonyls, diazonium salts, enones, and
benzophenones are the chemical functions that are classic-
ally activated under irradiation conditions, giving rise to
nitrenes, carbenes, carbocations, and radicals.[6,7] All these
probes have more or less the required properties for photo-
affinity labeling. Four probes (aryl azide, aryldiazirine, α-
diazocarbonyl, and benzophenone) have been incorporated
into the same pentapeptide, thymopentin TP5, in order to
compare their cross-linking properties in 1-propanol/water
mixtures under identical irradiation conditions.[8] Carbenes
reacted with the same efficiency with 1-propanol and with
water, whereas nitrenes, less reactive towards water, afforded
rearrangement products. In this comparative study, p-
benzoylphenylalanine appeared to be the better
photoprobe. The diradical triplet state of benzophenone ab-
stracts a hydrogen either from a CϪH bond to produce a
stable diarylcarbinol or from an XϪH bond (X ϭ N, O, or
S) to afford unstable (reversible) species.[9] The benzo-
phenone moiety is generally introduced into a peptide as
p-benzoylphenylalanine, although this hindered amino acid
may hamper access of the biologically active peptide to its
receptor.
Photoaffinity labeling is one of the most powerful tech-
niques for receptor binding site mapping.[1Ϫ5] However,
generalization of this methodology for the analysis of pep-
tide/protein interactions is hampered by various difficulties.
Firstly, the precursor containing the intended photoreactive
species must be stable during peptide synthesis and should
not disturb the recognition step between the peptide and its
receptor. Secondly, the activated species generated by UV
irradiation must be highly reactive, except with water mole-
cules, sufficiently stable regarding eventual rearrangement
or elimination reactions, and also marginally selective to-
wards the different amino acid side chains that might be
present in the binding site. The covalent bond resulting
from the photolabeling must also be stable towards proteo-
[a]
´
UR 043 IRD Ϫ Faculte des Sciences Pharmaceutiques,
ˆ
35, Chemin des Maraıchers, 31062 Toulouse Cedex, France
Fax: (internat.) ϩ 33-5/62256852
E-mail: jullian@cict.fr
[b]
ˆ
CEA Saclay Ϫ DSV/DIEP/LCIP Ϫ CEA Saclay, batiment 152,
91191 Gif-sur-Yvette Cedex, France
Fax: (internat.) ϩ 33-1/69089071
E-mail: vmonjardet@club-internet.fr
[c]
´
UMR CNRS 8601 Ϫ Universite Paris V,
`
45, rue des Saints-Peres, 75006 Paris, France
Fax: (internat.) ϩ 33-1/42868387
[d]
´
Structure et Fonctions de Molecules Bioactives, UMR 7613
In order to develop new photoprobes, we were interested
in substitution of the p-benzoyl group attached to the aro-
matic nucleus with either a p-propanoyl or a p-propenoyl
group. We expected that such substituents should give a
´
CNRS Ϫ Paris VI, Universite P. et M. Curie, Case Courrier
no. 45,
4, place Jussieu, 75252 Paris Cedex 05, France
Fax: (internat.) ϩ 33-1/4427150
E-mail: chassain@ccr.jussieu.fr
Eur. J. Org. Chem. 2002, 1677Ϫ1684  WILEY-VCH Verlag GmbH, 69451 Weinheim, Germany, 2002 1434Ϫ193X/02/0510Ϫ1677 $ 20.00ϩ.50/0
1677

V. Jullian, V. Monjardet-Bas, C. Fosse, S. Lavielle, G. Chassaing
FULL PAPER
better probe than p-benzoylphenylalanine, at least with tional protection/deprotection steps. In this work, all these
sterically hindered binding pockets. syntheses required the preparation of the same electrophile
UV irradiation of the p-propenoyl group, must generate 3 (Scheme 2).
a 1,4-diradical that must provide stable cross-linking. Fur-
thermore, this enone group can also achieve cross-linking
as a Michael acceptor. Phenyl alkyl ketone and enone moi-
eties have been used as photoprobes in only a few cases.
Acetophenone derivatives of guanosine-5Ј-triphosphate
peptides for photoaffinity labeling of GTP-binding proteins
and human serum albumin have been reported, and steroids
bearing an enone function have been developed for photo-
affinity labeling of sex hormone binding globulin.[12,13]
In this study, we report the syntheses of three N-protected
phenylalanine derivatives: (S)-p-(propanoyl)phenylalanine
(10), (S)-Boc-p-[2Ј-(phenylselenenyl)propanoyl]phenylalan-
ine (11), and (S)-Boc-p-[3Ј-(phenylselenenyl)propanoyl]-
phenylalanine (13) (Scheme 1). The protected amino acids
11 and 13 were introduced into the sequence of the model
pentapeptide thymopentin TP5 (N-acetylated, C-terminal
Scheme 2. Synthesis of 2-[4-(bromomethyl)phenyl]-2-ethyl-1,3-di-
carboxamide), which, after HF cleavage, yielded peptides
bearing either a p-propanoyl or a p-propenoyl group on the
aromatic residue. We established that only the β-phenylse-
leno carbonyl derivative was usable as a precursor of an
enone function in solid-phase peptide synthesis, the genera-
tion of this enone function being achievable under mild
conditions. Finally, we demonstrated that this enone func-
tion, when incorporated into a peptide sequence, was stable
in aqueous solution at pH ϭ 7, but suitably reactive upon
irradiation.
oxolane (3)
Synthesis of 2-[p-(Bromomethyl)phenyl]-2-ethyl-1,3-
dioxolane (3)
The electrophilic reagent 3 was prepared from p-methyl-
propiophenone as starting material, after protection of the
ketone as a cyclic acetal (Scheme 2). Bromination under
classical conditions (N-bromosuccinimide/dibenzoyl perox-
ide in carbon tetrachloride) gave a complex mixture of
starting material 2, ketone 1, and the expected monobromi-
nated product 3 contaminated with the dibrominated prod-
uct, with brominated forms of ketone 1 also being identi-
fied. When this reaction was performed in cyclohexane the
formation of ketone was prevented, and after optimization
a mixture consisting of 3 (75%), dibrominated dioxolane
(15%), and 10% of starting material 2 was obtained (NMR
analysis of the crude mixture). After flash chromatography
and crystallization, 2-[p-(bromomethyl)phenyl]-2-ethyl-1,3-
dioxolane (3) was isolated in moderate yields (47%), but this
yield surprisingly dropped to 25% when the reaction was
tentatively scaled up to more than 5 mmol.
Scheme 1. Structures of the precursors 10, 11, and 13 of photoac-
tivatable analogues of phenylalanine
Synthesis of Boc-p-(Propanoyl)phenylalanine (10)
Results
Alkylation of N-(diphenylmethylene)glycinate 4[14] with
We have previously reported the synthesis of unnatural 2-[p-(bromomethyl)phenyl]-2-ethyl-1,3-dioxolane 3 was per-
amino acids starting fromsultam N-(diphenylmethylene)gly- formed after deprotonation by lithium diisopropylamide
cinate.[14] Strong asymmetrical induction was observed with (LDA) in a THF/DMPU mixture, resulting in compound 5
Oppolzer’s sultam,[15,16] the levorotatory enantiomer of the (Scheme 3). NMR analysis showed that the alkylation was
sultam yielding the S configuration at the α-carbon for the highly diastereoselective (de Ͼ 95%). Acidic hydrolysis of
monoalkylated amino acids. The difficulties encountered in compound 5 afforded 8, with a free amino group and a
the preparation of unnatural amino acids depend on the ketone, while removal of the sultam moiety under basic
nature of the side chain, incorporated as an electrophile in conditions yielded the free amino acid 9, which was then
the alkylation step, and are especially pronounced when the Boc-protected with (Boc)2O (Scheme 3). Boc-p-(propanoyl)-
electrophile possesses an acidic proton, which imposes addi- phenylalanine 10 was obtained in an overall yield of 60%.
1678
Eur. J. Org. Chem. 2002, 1677Ϫ1684

Model Peptide with New Photoactivatable Phenylalanine Analogues
FULL PAPER
Scheme 3. Syntheses of Boc-p-(propanoyl)phenylalanine (10) and Boc-p-[2Ј-(phenylselenenyl)propanoyl]phenylalanine (11)
Synthesis of Boc-p-[2Ј-(Phenylselenenyl)propanoyl]-
Deselenation Reaction
phenylalanine (11)
(S)-Boc-p-[2-(phenylselenenyl)propanoyl]phenylalanine
Different pathways for the introduction of the phenylse- (11) was incorporated into position 5 of the pentapeptide
leno group were investigated (Scheme 3). Starting from de- thymopentin, TP5
ϭ
Ac-Arg1-Lys2-Asp3-Val4-Phe5-
rivative 5, selective removal of the 1,3-dioxolane was per- CONH2 (see peptide synthesis). Matrix-assisted laser de-
formed by treatment with iron chloride hexahydrate in sorption ionization time-of-flight mass spectrometry
dichloromethane for 30 minutes.[17] After flash chromato- (MALDI-TOF) of the purified peptide indicated substitu-
graphy, compound 6 was obtained in 73% yield. The enolate tion of the phenylseleno group by one hydrogen. We there-
of 6 was generated at low temperature (LDA/THF, Ϫ78°) fore looked for the step at the origin of this deselenation
and phenylselenenyl chloride was added,[18] but instead of reaction, with α-phenylselenenyl-3Ј-methylpropiophenone
the expected selenation product, only aldol condensation as a model compound. At room temperature, the loss of a
products were observed. Selenation of compound 6 under phenylseleno group was found to be very slow under the
nonbasic conditions (EtOAc/PheSeCl) was examined,[19] conditions required for Boc deprotection (trifluoroacetic
but instead of the selenated product, compound 7 was isol- acid). The rate of this reaction was, however, faster (50% of
ated. This compound was the result of an unexpected halo- phenylseleno removal) in HF (Ϫ5 °C to 0 °C for 60 mi-
genation in the position α to propiophenone, as demon- nutes). Furthermore, in the presence of scavenger mixtures
strated by mass spectrometry, and consisted of a mixture (dimethylsulfide and anisole in HF), conditions required for
of two diastereoisomers in a 1:1 ratio (NMR analysis). We the cleavage of the peptide from the resin, complete loss of
suspected that steric hindrance of the sultam and/or N-di- the phenylseleno group was observed. These data suggested
phenylmethylene moieties was the origin of this opposite that nucleophilic species such as the fluoride or trifluo-
regioselectivity, since phenylselenyl chloride reacted as ex- roacetate ion or dimethyl sulfide may attack the selenium,
pected with Boc-p-(propanoyl)phenylalanine (10) under resulting in the cleavage of the seleniumϪcarbon bond
these conditions. The selenation product (S)-Boc-p-[2Ј- (Scheme 4). Formation of the stabilized enol of the ketone
(phenylselenenyl)propanoyl]phenylalanine (11) was isolated should be the driving force for this elimination. Phenylse-
in 56% yield after column chromatography, with (S)-Boc-p- leno removal has also been observed during the conversion
(2Ј-chloropropanoyl)phenylalanine being only the major of selenoesters to methyl ketones after addition of diazo-
by-product. This reaction was also performed under acidic methane followed by workup with aqueous HBr.[20]
conditions (acetic acid) with similar efficiency. Altogether,
We postulated that shifting the phenylseleneno group to
these results confirmed that regioselective attack by the β-position of the ketone function should provide a com-
phenylselenyl chloride depends strongly on steric effects.
pound stable under acidic conditions. Indeed, the model
1679
Eur. J. Org. Chem. 2002, 1677Ϫ1684

V. Jullian, V. Monjardet-Bas, C. Fosse, S. Lavielle, G. Chassaing
FULL PAPER
Syntheses of Peptides I, II, and III
(S)-Boc-p-[2-(phenylselenenyl)propanoyl]phenylalanine
(11) and (S)-Boc-p-[3Ј-(phenylselenenyl)propanoyl]phenyl-
alanine (13) were incorporated into position 5 of the pen-
tapeptide thymopentin (TP5 ϭ Ac-Arg1-Lys2-Asp3-Val4-
Phe5-CONH2).[7] This short peptide was used to test the
stability of the functions present in 11 and 13 towards the
conditions required for solid-phase peptide synthesis. The
peptide syntheses were carried out with p-methylbenzhyd-
rylamine resin, by a Boc protecting group strategy. N-α-
Boc-amino acids in tenfold excess (except for 11 and 13)
were used, with dicyclohexylcarbodiimide/1-hydroxyben-
zotriazole as coupling agents. After the peptide resins had
been treated with anhydrous HF, in the presence of scaven-
gers (anisole and dimethylsulfide), at Ϫ5 °C to 0 °C for
60 minutes, the HF was evaporated and the peptide was
extracted with 10% acetic acid. The lyophilized peptides
were purified by preparative reversed-phase HPLC, and
characterized by mass spectrometry and 2D NMR ana-
lyses. (S)-Boc-p-[2-(phenylselenenyl)propanoyl]phenylal-
anine (11) afforded a unique peptide I, whereas the incorp-
oration of (S)-Boc-p-[3Ј-(phenylselenenyl)propanoyl]phen-
ylalanine (13) resulted in a mixture of two peptides II/III.
The proton resonance assignments of peptides I, II, and
III (H2O) were accomplished in a standard manner, by me-
ans of COSY and TOCSY experiments.[24] 1H chemical
shift assignments are shown in Table 1; the chemical shift
values of the Hα protons were typical of unstructured pep-
tides.[25] According to MALDI-TOF and NMR spectra,
the phenyl ring of peptide I (m/z ϭ 761.58), obtained from
(S)-Boc-p-[2-(phenylselenenyl)propanoyl]phenylalanine
11, bore an ethyl ketone function para to the aromatic ring
[δ(H2Ј) ϭ 3.0, δ(H3Ј) ϭ 1.05]. Peptide I corresponds to
[(p-propanoyl)Phe5]TP5, a new type of photoreactive ana-
logue of thymopentin.
The MALDI-TOF spectrum of the crude product ob-
tained from (S)-Boc-p-[3Ј-(phenylselenenyl)propanoyl]-
phenylalanine (13) featured two ions at m/z ϭ 917.36 and
m/z ϭ 759.44. The isotopic distribution observed around
m/z ϭ 917.36 denoted the presence of selenium-containing
peptide. After purification, the peptides corresponding to
these ions (peptides II and III) were analyzed by NMR and
MALDI-TOF. The chemical shifts of peptide II (m/z ϭ
917.36), with δ(2Ј-H) ϭ 3.2, δ(3Ј-H) ϭ 3.36, indicated the
presence of a phenylselenenyl group in the 3Ј-position. In
aqueous solution at room temperature, peptide II, {[p-3Ј-
(phenylselenenyl)propanoyl]Phe5}TP5, rapidly lost its
phenylseleno group, resulting in peptide III (m/z ϭ 759.44).
The presence of peptide III in the crude product of peptide
II must result from oxidation and subsequent elimination
of the phenylselenenyl group from peptide II during the
workup (extraction of the peptide by 10% acetic acid). In-
deed, oxidation of the peptide II/III mixture by NaIO4 for
10 minutes, or air oxidation for 24 hours, exclusively pro-
vided peptide III [(p-propenoyl)Phe5]TP5. The enone func-
tion in the para position of the aromatic ring of phenyl-
alanine was characterized by NMR spectroscopy, especially
Scheme 4. Mechanism proposed for phenylselenyl substitution
compound
β-(phenylselenenyl)-4Ј-methylpropiophenone
was found to be stable under the different conditions listed
above for peptide synthesis by a Boc strategy. However,
β-(phenylselenenyl)-4Ј-methylpropiophenone was found to
be unstable to piperidine treatment, precluding a Fmoc
strategy for peptide synthesis. These results support the
postulated mechanism.
Synthesis of Boc-p-[3Ј-(phenylselenenyl)propanoyl]-
phenylalanine (13)
Alkylphenyl selenoxides undergo facile syn elimination to
form olefins,[21] this strategy having frequently been applied
in total synthesis for introduction of unsaturation.[22] The
one-pot strategy involves oxidation of the α-phenylseleno-
carbonyl compound to the corresponding selenoxide, which
eliminates at room temperature to afford the desired enone.
(S)-Boc-p-(propenoyl)phenylalanine 12 was cleanly ob-
tained by treatment of (S)-Boc-p-[2-(phenylselenenyl)pro-
panoyl]phenylalanine 11 with two equivalents of sodium
periodate in MeOH/H2O. The benzeneselenol generated by
action of sodium borohydride on diphenyldiselenide in the
presence of acetic acid[23] reacted by a 1,4-Michael addition
to afford (S)-Boc-p-[3Ј-(phenylselenenyl)propanoyl]phenyl-
alanine 13 (Scheme 5).
Scheme 5. Synthesis of Boc-p-[3Ј-(phenylselenenyl)propanoyl]phen-
ylalanine (13)
1680
Eur. J. Org. Chem. 2002, 1677Ϫ1684

Model Peptide with New Photoactivatable Phenylalanine Analogues
FULL PAPER
2
Table 1. Chemical shifts of peptides I, II, and III in H2O at 298 K
by three resonances at δ(2Ј-H) ϭ 7.19, δ(3Јtrans-H) ϭ 6.32, molecular insertion into a XϪH bond. ESI MS/MS sequ-
and δ(3Јcis-H) ϭ 6.02.
encing allowed the identification of the N-terminal residues
Ac-Arg and Lys; further sequencing failed.
Photolysis of [(p-Propenoyl)Phe5]TP5
The stability of [(p-propenoyl)Phe5]TP5 was followed by
recording HPLC and 2D NMR spectra over 48 hours, since
the amino group of the Lys moiety in the TP5 sequence
could give rise to 1,4-addition on the enone, which is an
efficient Michael acceptor. At pH ϭ 7, no 1,4-Michael addi-
Conclusion
Three optically active phenylalanine analogues Ϫ (S)-p-
tion was observed. UV irradiation (HPR 125-W lamp at (propanoyl)phenylalanine (10), (S)-Boc-p-[2Ј-(phenylse-
365 nm) of [(p-propenoyl)Phe5]TP5 peptide III and [(p- lenenyl)propanoyl]phenylalanine (11), and (S)-Boc-p-[3Ј-
benzoyl)Phe5]TP5 was carried out in water under the same (phenylselenenyl)propanoyl]phenylalanine (13) Ϫ have been
irradiation conditions (lamp, time, temperature, concentra- prepared by deprotonation/alkylation of the sultam N-(di-
tion) as used in a procedure previously[8] described for phenylmethylene)glycinate (4). The introduction of analogs
photolabeled analogs of TP5. The photolyzed peptides were 11 and 13 into a model pentapeptide thymopentin TP5 by
analyzed by reversed-phase HPLC and MALDI-TOF and/ solid-phase peptide synthesis with a Boc strategy has dem-
or ESI mass spectrometry. As already reported,[8] [(p- onstrated that an α-phenylseleno ketone moiety cannot be
benzoyl)Phe5]TP5 was completely transformed within used as a precursor of an enone function. However, we have
40 min into at least 15 photolysis products, with six major shown that an enone can be generated in a peptide sequence
products. Under the same conditions, [(p-propenoyl)- by incorporation of a β-phenylseleno carbonyl group on the
Phe5]TP5 was less reactive, about one third of the peptide aromatic nucleus of phenylalanine. A model peptide con-
being converted into a major compound after 40 min, while taining this enone function, [(p-propenoyl)Phe5]TP5,
extended irradiation (100 min) resulted in complete disap- proved to be sufficiently chemically stable to be used for
pearance of peptide III. The less hydrophobic peptide photoaffinity labeling. (p-Propenoyl)Phe-substituted TP5
formed upon irradiation {elution at 7.35 min, compared to appeared to be less reactive than (p-benzoyl)Phe-substituted
12.55 min for [(p-propenoyl)Phe5]TP5} had the same mass TP5 upon irradiation. This slightly higher stability might
(m/z: 759.32) as peptide III (m/z: 759.44), suggesting intra- be a benefit for high affinity peptide/protein complexes.
Eur. J. Org. Chem. 2002, 1677Ϫ1684
1681

V. Jullian, V. Monjardet-Bas, C. Fosse, S. Lavielle, G. Chassaing
FULL PAPER
moved under reduced pressure, and the aqueous layer was washed
three times with diethyl ether. The aqueous layer was concentrated
to give 2.8 g of a yellow oil (98% yield). Rf ϭ 0.45 (methanol/chlo-
roform/acetic acid, 90:10:1); mass spectrometry (DCI/NH3): MHϩ
m/z ϭ 419. 1H NMR (CDCl3, 400 MHz): δ ϭ 8.61 (m, 3 H, NH3),
7.92 (d, 2 H, aromatic), 7.53 (d, 2 H, aromatic), 4.88 [m, 1 H,
C(O)CHCH2], 4.10 (m, 1 H, CCHN), 3.47 (dd, 2 H, CH2SO2), 3.40
[m, 2 H, C(O)CHH2], 2.98 (dd, 2 H, CH2H3), 1.8Ϫ2.0 (m, 5 H,
Experimental Section
Syntheses of Phenylalanine Analogues. 2-Ethyl-2-tolyl-1,3-dioxolane
(2): 4Ј-Methylpropiophenone (15.2 g, 102 mmol), ethylene glycol
(12.56 g, 202 mmol), and p-toluenesulfonic acid (640 mg) in toluene
(200 mL) were placed in a round-bottomed flask fitted with a
DeanϪStark water separator and a reflux condenser. The reaction
mixture was heated until no more water could be collected (24
hours). The solution was washed twice with a 5% sodium bicarbon-
ate solution, dried (MgSO4), and concentrated to give 16.9 g of a
colorless oil (86%). Rf ϭ 0.44 (5% ethyl acetate/cyclohexane). 1H
NMR (CDCl3, 200 MHz): δ ϭ 7.35 (d, 2 H, ortho aromatic), 7.25
(d, 2 H, meta aromatic), 4.05 (m, 2 H, dioxolane), 3.80 (m, 2 H,
dioxolane), 2.35 (s, 3 H, C6H4CH3), 1.90 (q, 2 H, CH2-CH3), 0.90
(t, 3 H, CH2-CH3). C12H16O2: C,75.07, H 8.43; Found: calcd. C
74.97, H 8.39.
CH2CHHCHCHH or CHHCH2CHCHH), 1.31 (m,
2 H,
CH2CHHCHCHH or CHHCH2CHCHH), 1.23 (t, 3 H, CH2CH3),
1.07 (s, 3 H, CH3CCH3), 0.84 (s, 3 H, CH3CCH3).
(S)-Boc-p-(Propanoylphenyl)alanine (10): Compound 8 was dis-
solved in 1,4-dioxane (15 mL) in the presence of monohydrated
lithium hydroxide (440 mg, 10.5 mmol). The reaction mixture was
stirred at room temperature for 4 days. The resulting precipitate
was collected by filtration and washed twice with diethyl ether. This
solid (compound 9) was dissolved in water (4 mL). 1,4-Dioxane (2
mL) and triethylamine (840 µL, 6 mmol) was added to the mixture,
which was cooled at 0° C prior to the addition of Boc2O (1.20 g,
5.5 mmol) in 3 mL of 1, 4-dioxane. The solution was stirred at
room temperature overnight. After concentration under reduced
pressure to remove the dioxane, water (5 mL) was added. This basic
aqueous layer was washed twice with diethyl ether, and diluted with
10 mL of EtOAc. HCl (3 ) was then added slowly at 0° C until
pH ϭ 2 was reached. The aqueous layer was extracted twice with
EtOAc. The combined organic layers were dried (MgSO4) and con-
centrated to give an oil, which was purified by short flash chroma-
tography on silica gel (CH2Cl2/MeOH/AcOH, 98:2/0.1) to give
963 mg of a colorless oil (60% overall yield for the three steps).
Rf ϭ 0.49 (chloroform methanol/acetic acid, 95:5/0.5). [α]2D0 ϭ 50
(c ϭ 0.73, CHCl3); mass spectrometry (DCI/NH3): MNH4ϩ m/z ϭ
334, 1H NMR (CDCl3, 400 MHz, mixture of cis/trans isomers):
δ ϭ 7.93 (d, 2 H, ortho aromatic), 7.29 (d, 2 H, meta aromatic),
6.33 (m, 0.4 H, NH), 4.99 (d, 0.6 H, NH), 4.66 (m, 0.6 H,
C(O)CHCH2), 4.47 [m, 0.4 H, C(O)CHCH2], 3.29 [m, 1 H,
C(O)CHCHH], 3.15 [m, 1 H, C(O)CHCHH], 3.01 (2 H, CH2CH3),
1.43 (s, 5.4 H, Boc), 1.31 (s, 3.6 H, Boc), 1.24 (t, 3 H, CH2CH3).
13C NMR (CDCl3, 50 MHz): δ ϭ 200.9, 175.7, 155.4, 141.5, 135;7,
129.7, 128.4, 80.5, 54.1, 46.3, 31.8, 28.3, 8.3.
2-(p-Bromomethyl-phenyl)-2-ethyl-1,3-dioxolane (3): A mixture of
dioxolane 2 (546 mg, 2.83 mmol), N-bromosuccinimide (750 mg,
4.24 mmol), and benzoyl peroxide (100 mg) in 25 mL of cyclohex-
˚
ane in the presence of 3A molecular sieves was heated under reflux
for 4 hours. The reaction mixture was then filtered through a celite
pad, washed with a 5% sodium bicarbonate solution, dried
(MgSO4), and concentrated. After short chromatography on silica
gel (cyclohexane/EtOAc/NEt3: 95:5/few drops), 3 crystallized after
standing for 24 hours at 4 °C. Crystals were collected and were
washed twice with pentane (365 mg, 47%). Rf ϭ 0.44 (5% ethyl
acetate/cyclohexane); m.p. 65Ϫ67 °C (recrystallization from diethyl
ether/pentane). 1H NMR (CDCl3, 400 MHz): δ ϭ 7.45 (d, 2 H,
ortho aromatic), 7.35 (d, 2 H, meta aromatic), 4.50 (s, 2 H, CH2Br),
4.00 (m, 2 H, dioxolane), 3.80 (m, 2 H, dioxolane), 1.90 (q, 2 H,
CH2-CH3), 0.90 (t, 3 H, CH2-CH3). 13C NMR (CDCl3, 50 MHz):
δ ϭ 143.3, 137.5, 129.1, 126.7, 110.9, 65.0, 33.7, 33.6, 8.2.
C12H15BrO2: calcd. C 53.10, H 5.57; Found: calcd. C 53.02, H 5.56.
N-[NЈ-(Diphenylmethylene)-(S)-p-(2-ethyl-1,3-dioxolan-2-yl)-
phenylalanyl]-(؊)-10,2-bornane Sultam (5): Precursor 4 (2.18 g,
5 mmol, previously dried) was dissolved under argon in freshly dis-
tilled THF (15 mL) and DMPU (10 mL). The mixture was stirred
at Ϫ78° C, and LDA (2.75 mL, 2  solution in n-heptane/THF,
5.5 mmol) was added. After 15 min, the electrophile 3 (2.02 g,
7.5 mmol) in 3 mL of dry THF was added. The mixture was first
stirred for 1 hour at Ϫ78° C and allowed to warm up to room
temperature over one hour. Neat acetic acid (315 µL, 5.5 mmol)
and diethyl ether (10 mL) were added. The mixture was washed
three times with saturated NH4Cl, and the organic layer was dried
(MgSO4) and concentrated. The oily residue was purified by chro-
matography on silica gel (20% ethyl acetate/cyclohexane) to give a
colorless oil (1.7 g, 54%). Rf ϭ 0.16 (30% ethyl acetate/cyclohex-
ane). [α]2D0 ϭ Ϫ57 (c ϭ 1.14, CH2Cl2). 1H NMR (CDCl3,
400 MHz): δ ϭ 7.44 (d, 2 H, aromatic), 7.41 (d, 2 H, aromatic),
7.70Ϫ7.15 (10 H, aromatic benzophenone imine), 4.89 [dd, 1 H,
C(O)CHCH2], 4.05 (m, 2 H, dioxolane), 3.90 (m, 1 H, CCHN),
3.75 (m, 2 H, dioxolane), 3.45 [m, 2 H, C(O)CHCH2], 3.37 (dd, 2
H, CH2SO2), 2.0 (q, 1 H, CH2CH3), 2.10Ϫ1.72 (m, 5 H,
CHHCH2CHCHH or CH2CHHCHCHH), 1.4Ϫ1.5 (m, 2 H,
(S)-Boc-p-[2Ј-(Phenylselenenyl)propanoyl]phenylalanine (11): A so-
lution of Boc-amino acid 10 (341 mg, 1.05 mmol) in EtOAc (2 mL)
in the presence of phenylselenyl chloride (222 mg, 1.15 mmol) was
stirred under argon at room temperature for 72 hours. The precipi-
tate was filtered off and the filtrate was concentrated and purified
by careful flash chromatography on silica gel (CH2Cl2/MeOH/
AcOH, 96:4/0.1) to give 280 mg (56%) of a yellow oil. Rf ϭ 0.36
(chloroform/methanol/acetic acid, 95:5/0.5); mass spectrometry
ϩ
(DCI/NH3): MHϩ m/z ϭ478, MNH4 m/z ϭ 495. 1H NMR
(CDCl3, 400 MHz, mixture of two diastereoisomers): δ ϭ 7.8 (d, 2
H, ortho aromatic), 7.5 (d, 2 H, meta aromatic), 7.22 (m, 5 H, aro-
matic Se-Ph), 6.36 (d, 0.33 H, NH), 5.06 (m, 0.66 H, NH), 4.67 [m,
1.66 H, C(O)CHCH2, PhSeCHCH3], 4.40 [0.33 H, m,
C(O)CHCHH], 3.27 [m, 1 H, C(O)CHCHH], 3.13 [0.66 H, m,
C(O)CHCHH], 2.98 [0.33 H, m, C(O)CHCHH], 1.63 (d, 3 H,
PheSeCHCH3), 1.43 (s, 5.2 H, Boc), 1.35 (s, 3 H, Boc). 13C NMR
(CDCl3, 62.9 MHz): δ ϭ 175.2, 155.2, 136.5Ϫ126.6, 81.8, 80.3,
55.5, 53.9, 39.5, 39.2, 37.6, 28.1, 20.9, 17.0.
CHHCH2CHCHH or CH2CHHCHCHH), 0.93 (s,
3
H,
CH3CCH3), 0.90 (t, 3 H, CH2CH3), 0.83 (s, 3 H, CH3CCH3). 13C
NMR (CDCl3, 50 MHz): δ ϭ 170.5, 140.5, 139.3, 136.5, 135.5,
130.0Ϫ125.5, 110.5, 67.0, 64.9, 64.3, 52.9, 48.2, 47.5, 44.3, 40.7,
38.1, 33.2, 32.5, 26.2, 20.4, 19.7, 7.7.
(S)-Boc-p-(Propenoyl)phenylalanine (12): Sodium periodate
(118 mg, 0.55 mmol) was added to a solution of Boc-amino acid
11 (131 mg, 0.27 mmol) in MeOH (2.2 mL) and H2O (860 µL).
(S)-p-(Propanoyl)phenylalanyl-(؊)-10,2-bornane Sultam (8): Com-
pound 5 (3.53 g, 6.05 mmol) was stirred overnight at room temper- The reaction mixture was stirred for 2 hours at room temperature
ature in a mixture (1:3) of THF/1  HCl (30 mL). THF was re-
in the dark. The precipitate was removed by filtration and rinsed
1682
Eur. J. Org. Chem. 2002, 1677Ϫ1684

Model Peptide with New Photoactivatable Phenylalanine Analogues
FULL PAPER
with MeOH (2 ϫ 5 mL). The solution was concentrated and then
diluted with EtOAc (15 mL). The organic layer was washed with
KHSO4 solution (pH ϭ 2), dried (MgSO4), and concentrated. We
were unable to achieve complete purification of compound 12 by
chromatography, Rf ϭ 0.22 (dichloromethane/methanol/acetic acid:
protecting group was removed and the N-terminus was acetylated.
The peptidyl-MBHA resin was then treated with liquid HF at Ϫ20
°C (30 min) and 0 °C (30 min) in the presence of 1.5 mL of anisole,
0.25 mL of dimethyl sulfide per gram of peptidyl resin. The crude
peptides were purified by HPLC with an Applied Biosystems ap-
95:4:0.5). Thus, compound 12 was characterized by 1H NMR paratus, using a 10 ϫ 250 mm Brownlee column packed with Aqua-
(CDCl3, 250 MHz,), purity estimated around 90%, δ ϭ 7.86 (d, 2 pore Octyl, 20 µm pore size. The separation was accomplished with
H, ortho aromatic), 7.22 (d, 2 H, meta aromatic), 7.18 [dd, 1 H, acetonitrile gradients in aqueous trifluoroacetic acid (0.1%) at a
CC(O)CHCH2], 6.42 [dd, 1 H, CC(O)CHCHH], 5.91 [dd, 1 H, flow rate of 6 mL/min, with UV detection at 220 nm. Analytical
C(O)CHCHH], 5.05 (dd, 1 H, NH), 4.61 [m, 1 H, NC(O)CHCH2], monitoring was carried out by reversed-phase analytical HPLC on
3.30 [m, 1 H, NC(O)CHCHH] 3.18 [m, 1 H, N(O)CHCHH], 1.37 a LiChrosphor 100 RP-8e column (Merck) with acetonitrile gradi-
(br. s, 9 H, Boc).
ents in aqueous trifluoroacetic acid (A ϭ 0.1% TFA, B ϭ 40% A
and 60% CH3CN) at a flow rate of 1.5 mL/min with UV detection
at 210 nm. Yields of purified peptides were between 13% and 20%.
The purities of the collected fractions (between 95% and 98%) were
determined by HPLC in the gradient mode (from 20% to 80% B
in 30 min). Retention times: tR ϭ 12.3 min peptide I; tR ϭ 25.4 min
peptide II; tR ϭ 12.4 min peptide III.
(S)-Boc-p-[3Ј-(Phenylselenenyl)propanoyl]phenylalanine (13): The
above crude 12 (130 mg, 0.41 mmol) was dissolved in absolute eth-
anol (2.5 mL), and diphenyldiselenide (168 mg, 0.54 mmol) was ad-
ded. Sodium borohydride (41 mg, 1.07 mmol) was then added (cau-
tion, hydrogen evolution), and the reaction mixture was stirred at
0 °C for ten minutes. Neat acetic acid (109 µL, 1.89 mmol) was
added and the mixture was stirred for 3 hours at room temperature.
The reaction mixture was diluted with 10 mL of EtOAc and washed
with a saturated KHSO4 solution, and the combined organic layers
were dried (MgSO4). After concentration, the residual oil was puri-
fied by short column chromatography (CH2Cl2/MeOH/AcOH,
96:4:0.1) to afford 65 mg (50% from compound 11) of 13 as a yel-
low oil. Rf ϭ 0.36 (chloroform methanol/acetic acid: 95:5:0.5).
[α]2D0 ϭ 17 (c ϭ 1, CH2Cl2); mass spectrometry (DCI/NH3): MHϩ
Structural Determinations of Peptides
MALDI-TOF spectra were obtained on a Voyager Elite mass spec-
trometer (PerSeptive Biosystems), with α-cyano-4-hydroxycinnam-
inic acid as matrix, at concentrations around 5·10Ϫ2  in acetonitr-
ile/water (4:1, v/v). The positive MALDI mass spectrum gave a
unique ion at m/z ϭ 761.58 for peptide I, m/z ϭ 917.36 for peptide
II, and m/z ϭ 759.44 for peptide III.
ϩ
m/z ϭ 478, MNH4 m/z ϭ 495. 1H NMR (CDCl3, 250 MHz): δ ϭ
2
NMR experiments were carried out on 2 m solution in H2O at
7.83 (d, 2 H, ortho aromatic), 7.43 (d, 2 H, meta aromatic), 7.19
(m, 4 H, aromatic Se-Ph), 6.30 (m, 0.33 H, NH), 4.91 (d, 0.66 H,
NH), 4.61 (m, 0.66 H, C(O)CHCH2), 4.36 (m, 0.33 H,
C(O)CHCH2), 3.35 [m, 3 H, C(O)CHCHH, CH2CH2SePh or
CH2CH2SePh], 3.28 [m, 3 H, C(O)CHCHH, CH2CH2SePh or
CH2CH2SePh], 1.34 (br. s, 6 H, Boc), 1.23 (br. s, 3 H, Boc). 13C
NMR (CDCl3, 62.9 MHz): δ ϭ 135.4, 132.8, 129.7, 129.3, 128.4,
127.1, 77.3, 55.0, 39.4, 29.7, 28.3, 21.1.
298 K. NMR spectra were recorded on Bruker AM 500 or DMX
500 spectrometers and were processed with an Aspect 3000 com-
puter (AM 500) or with the Bruker UXNMR software (DMX 500),
Table 1. One-dimensional spectra were acquired over 16 K data
points with a spectral width of 5000 Hz. Solvent suppression was
achieved by presaturation during the relaxation delay (1.5 s). Two-
dimensional experiments were acquired in the absolute mode for
COSY and in phase-sensitive mode for TOCSY; TOCSY experi-
ments were carried out with mixing times of 40 ms. Typically, 400
to 600 increments were acquired over a spectral width of 5000 Hz.
Prior to Fourier transformation in t2 and t1, the free induction
decays were zero-filled and multiplied by a π/4-shifted sine-bell
function.
Deselenation Studies: The stability of model phenylselenyl com-
pounds was followed directly by NMR before and after acidic treat-
ment with trifluoroacetic acid, hydrogen fluoride, and hydrogen flu-
oride/anisole/dimethyl sulfide (10:2.5:0.25 per 0.5 mmol).
Preparation of the Model Compounds: α-Phenylselenenyl-4Ј-methyl-
propiophenone was prepared by the procedure previously re-
ported.[3] 1H NMR (CDCl3, 400 MHz,): δ ϭ 7.65 (d, 2 H, ortho
aromatic), 7.49 (d, 2 H, meta aromatic), 7.31 (m, 5 H, aromatic Se-
Ph), 4.70 (q, 1 H, PhSeCHCH3), 2.43 (s, 3 H, C6H4CH3), 1.67
(d, 3 H, PhSeCHCH3). β-Phenylselenenyl-4Ј-methylpropiophenone
was prepared by the procedure described for 13. 1H NMR (CDCl3,
400 MHz,): δ ϭ 7.83 (d, 2 H, ortho aromatic), 7.54 (d, 2 H, meta
Acknowledgments
We thank Dr. G. Bolbach and I. Correira for recording the
MALDI-TOF and NMR spectra, respectively.
[1]
J. Brunner, Ann. Rev. Biochem. 1993, 62, 483Ϫ514.
[2] [2a]
F. Kotzyba-Hibert, I. Kapfer, M. Goeldner, Angew. Chem.
Ed. Engl. 1995, 34, 1296Ϫ1312, [2b] T. Grutter, F. Kotzyba-Hib-
ert, M. Goeldner, Biochemistry 1999, 38, 7476Ϫ7484.
aromatic), 7.30 (m,
5 H, aromatic Se-Ph), 3.41 [m, 2 H,
[2c]
C(O)CH2CH2SePh], 3.26 [m, 2 H, C(O)CH2CH2SePh], 2.42 (s, 3
H, CH3).
F.
Kotzyba-Hibert, T. Grutter, M. Goeldner, Mol. Neurobiol.
1999, 20, 45Ϫ59.
G. J. Quellhorst, Jr., C. M. Allen, M. Wessling-Resnick, J. Biol.
Chem. 2001, 276, 40727Ϫ40733.
T. C. Turek, I. Gaon, M. D. Distefano, C. L. Stickland, J. Org.
Chem. 2001, 66, 3253Ϫ3264.
G. Dorman, G. D. Prestwich, Biochemistry. 1994, 33,
5661Ϫ5673.
G. D. Prestwich, G. Dorman, J. T. Elliott, D. M. Marecak, A.
Chaudhary, Photochem. Photobiol. 1997, 65, 222Ϫ234.
S. A. Fleming, Tetrahedron 1995, 51, 12479Ϫ12520.
P. J. A. Weber, A. J. Beck-Sickinger, J. Peptide Res. 1997, 49,
375Ϫ383.
[3]
[4]
[5]
[6]
Peptides Syntheses: Peptide synthesis was carried out manually on
a 0.1 mmol scale, with a p-methylbenzhydrylamine resin (MBHA
resin, typical substitution 0.77 mmol/g of resin) according to the
procedure previously described for the Boc strategy.[26] Standard
procedures for tert-butoxycarbonyl-protected amino acids were
used, with dicyclohexylcarbodiimide/1-hydroxybenzotriazole ac-
tivation (10 equiv. for conventional Boc amino acids and one equiv.
of Boc-amino acid 11 or 13). The activated Boc-amino acid 11 or
13 was coupled onto the resin for 24 hours and the residual un-
changed amino groups were acetylated until negative by the Kaiser
test. After completion of the peptide sequence, the last N-α-Boc
[7]
[8]
[9]
S. Sagan, personal communication.
Eur. J. Org. Chem. 2002, 1677Ϫ1684
1683

V. Jullian, V. Monjardet-Bas, C. Fosse, S. Lavielle, G. Chassaing
FULL PAPER
[10]
[18]
R. E. Galardy, L. C. Craig, M. P. Printz, Nature, New Biol.
H. J. Reich, I. L. Reich, J. M. Renga, J. Am. Chem. Soc. 1973,
95, 5813Ϫ5815.
K. B. Sharpless, R. F. Lauer, A. Teranishi, J. Am. Chem. Soc.
1973, 95, 6137Ϫ6139.
G. T. Back, G. R. Kerr, Tetrahedron 1985, 41, 4759Ϫ4764.
H. J. Reich, S. Wollowitz, Org. React. 1993, 44, 1Ϫ296.
M. Miyashita, A. Yoshikoshi, Synthesis 1980, 8, 664Ϫ666.
S. Uemura, K. Ohe, J. Chem. Soc., Perkin. Trans. 1 1990, 4,
907Ϫ910.
K. Wüthrich, NMR of proteins and nucleic acids, John Wiley,
New York, 1988.
1973, 242, 127Ϫ128.
T. Zor, I. Halifa, S. Kleinhaus, M. Chorev, Z. Selinger, Bio-
[19]
[11]
chem. J. 1995, 306, 253Ϫ258.
C. Grenot, A. De Montard, T. Blachere, M. R. De Ravel, E.
[20]
[21]
[22]
[23]
[12]
Mappus, C. Y. Cuilleron, Biochemistry 1992, 31, 7609Ϫ7621.
C. Grenot, T. Blachere, M. R. De Ravel, E. Mappus, C. Y.
[13]
Cuilleron, Biochemistry 1994, 33, 8969Ϫ8981.
[14] [14a]
H. Josien, A. Martin, G. Chassaing, Tetrahedron Lett.
[24]
[25]
[26]
[14b]
1991, 30, 6547Ϫ6550.
H. Josien, G. Chassaing, Tetrahe-
dron: Asymmetry 1992, 3, 1351Ϫ1354.
W. Oppolzer, R. Moretti, S. Thomi, Tetrahedron Lett. 1989,
30, 6009Ϫ6010.
W. Oppolzer, R. Moretti, C. Zhou, Helv. Chim. Acta 1994,
77, 2363Ϫ2380.
J. E. Lynh, S. E. Sen, J. Org. Chem. 1997, 62, 6684Ϫ6686.
D. S. Wishart, B. D. Sykes, F. M. Richards, J. Mol. Biol. 1991,
222, 311Ϫ333.
[15]
[16]
[17]
S. Girault, S. Sagan, G. Bolbach, S. Lavielle, G. Chassaing, Eur.
J. Biochem. 1996, 240, 215Ϫ222.
Received October 30, 2001
[O01528]
1684
Eur. J. Org. Chem. 2002, 1677Ϫ1684