Highly stereoselective peptide modifications through Pd-catalyzed allylic alkylations of chelated peptide enolates

Article abstract of DOI:10.1002/chem.200700084

Deprotonation of peptides in the presence of zinc chloride gives rise to highly reactive nucleophiles that can be subjected to palladium-catalyzed allylic alkylation reactions. Excellent diastereoselectivities are obtained that are nearly independent of the allylic substrate used. By using this protocol, highly functionalized side chains can also be incorporated in excellent yields and selectivities. The stereochemicaloutcome of the reaction is exclusively controlled by the peptide chain as long as terminal π-allyl-palladium complexes are involved. Probably, there is a threefold coordination, at least, ofthe deprotonated peptide chain to the chelating zinc ion. In such metal peptide complexes, one face of the generated enolate is shielded by the side chain of the adjacent amino acid, thus directing the electrophilic attack onto the opposite face. This behavior explains why an S amino acid always generates an R amino acid (and the other way round).

Full text of DOI:10.1002/chem.200700084

DOI: 10.1002/chem.200700084
Highly Stereoselective Peptide Modifications through Pd-Catalyzed Allylic
Alkylations ofChelated Peptide Enolates
Jan Deska and Uli Kazmaier*[a]
Dedicated to Professor Lutz Tietze on the occasion of his 65th birthday
Abstract: Deprotonation of peptides in
the presence of zinc chloride gives rise
to highly reactive nucleophiles that can
be subjected to palladium-catalyzed al-
lylic alkylation reactions. Excellent dia-
stereoselectivities are obtained that are
nearly independent of the allylic sub-
strate used. By using this protocol,
highly functionalized side chains can
also be incorporated in excellent yields
and selectivities. The stereochemical
outcome of the reaction is exclusively
controlled by the peptide chain as long
as terminal p-allyl–palladium com-
plexes are involved. Probably, there is
a threefold coordination, at least, of
the deprotonated peptide chain to the
chelating zinc ion. In such metal pep-
tide complexes, one face of the gener-
ated enolate is shielded by the side
chain of the adjacent amino acid, thus
directing the electrophilic attack onto
the opposite face. This behavior ex-
plains why an S amino acid always gen-
erates an R amino acid (and the other
way round).
Keywords: allylic alkylation · asym-
metric synthesis · chelates · palladi-
um · peptide modification · pep-
tides
Introduction
generate libraries of structurally related derivatives for
structure–activity studies.[5] An alternative to the classical
approach, the stepwise coupling of commercially available
or separately synthesized amino acids, is the incorporation
of a (unusual) side chain into a given peptide, a so-called
backbone modification. This route has the advantage that a
wide range of derivatives can readily be prepared from a
single peptide without repetition of the peptide synthesis for
each modification. In principle, various types of intermedi-
Small peptides, linear as well as cyclic, are found as metabo-
lites in a wide range of microorganisms, sponges, and fungi.
Many of these peptides show highly interesting biological
properties[1] and are therefore suitable candidates as lead
structures for the synthesis of drugs.[2] On the basis of their
non-ribosomal peptide synthesis (NRPS),[3] these “lower or-
ganisms” can incorporate unusual amino acids with “exotic”
side chains and N-methylated or d-amino acids into their
metabolites.[4] In addition, cyclic structures have been found.
In most cases, the unusual amino acids participate signifi-
cantly in binding towards an enzyme or receptor and are
therefore (at least in part) responsible for the biological ac-
tivity.
ꢀ
ates can be involved in the C C coupling step, such as gly-
cine cations,[6] radicals,[7] or anions.[8] The major problem of
this efficient process is control of the stereochemical out-
ꢀ
come of the C C bond formation, especially in the modifi-
cation of linear peptides. For cyclic peptides, the selectivities
are generally better, because one face of the reaction inter-
mediate can be shielded by the peptide ring.[9] The most
spectacular example reported so far was the modification of
cyclosporine, as reported by Seebach et al.[10] In this case, a
single sarcosine subunit was alkylated stereoselectively. Re-
cently, Maruoka and co-workers[11] reported the stereoselec-
tive functionalization of peptide Schiff bases[12] by chiral
phase-transfer catalysis. The fine tuning of the catalyst al-
lowed the introduction of various substituents at the N-ter-
minus of the peptide chain with remarkable selectivity.
Our group has been involved in peptide modifications for
some time,[13] and our aim is to use the chiral information of
For synthetic and medicinal chemists, this behavior leads
to the challenge of developing synthetic strategies towards
these unusual peptides that are as flexible as possible to
[a] Dipl.-Chem. J. Deska, Prof. Dr. U. Kazmaier
Institut für Organische Chemie
Universität des Saarlandes
Im Stadtwald, Geb. C4.2, 66041 Saarbrücken (Germany)
Fax : (+49)681-302-2409
E-mail: u.kazmaier@mx.uni-saarland.de
Supporting information for this article is available on the WWW
under http://www.chemeurj.org/ or from the author.
6204
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2007, 13, 6204 – 6211

FULL PAPER
the peptide chain to control the stereochemical outcome of
the modification step by taking advantage of metal–peptide
complexes.[8d] In the case of peptide Claisen rearrangements,
good results were obtained with tosyl-protected peptides,
such as 1.[14] Although the yields and selectivities varied de-
pending on the chelating metal salt (MXn) used, the
S,R diastereomer was always formed preferentially
(Scheme 1). This behavior can be explained by a multifold
Results and Discussion
Herein, we report the transfer of this protocol from amino
acids to peptides and its application to highly stereoselective
peptide modifications. We began our investigations with the
protected leucine dipeptides 6 and methallyl carbonate 7
under the reaction conditions shown (Table 1). The tert-
Table 1. Allylic alkylations of leucine dipeptides.
Entry Peptide Protecting Equiv
Product Yield d.r. (S,S)/(S,R)
A
group
LHMDS
[%]
1
2
3
4
5
6
6a
6b
6c
9a
9a
9a
Boc
Z
Ts
TFA
TFA
TFA
4
4
4
4
3.5
3.2
8a
8b
8c
10a
10a
10a
67
67
40
63
74
63
30:70
37:64
14:86
13:87
16:84
17:83
Scheme 1. Stereoselective modifications of peptides. LDA=lithium diiso-
propylamide, M=metal, Ts=tosyl.
butyl ester was chosen to avoid side reactions, such as trans-
esterifications with the ethylate moiety liberated from 7
(which is a serious problem if the corresponding methyl
esters are used). Zinc chloride was superior to other metal
salts with respect to a clean conversion, which is in good
agreement with observations made from other alkylations of
chelated enolates. LHMDS was preferred as the base (in-
stead of LDA) to avoid epimerization of the newly formed
stereogenic center. To obtain complete conversion of the
peptide, we used an excess of 7. Interestingly, the results ob-
tained were comparable to those of the Claisen rearrange-
ment and the alkylation reactions. The carbamate-protected
peptides 6a and 6b gave the best yields (Table 1, entries 1
and 2), whereas the tosyl derivative 6c showed the highest
selectivity (Table 1, entry 3). Based on our positive experi-
ences with TFA protecting groups in the allylic alkylation of
glycine enolates, we subjected the TFA-protected peptide
9a to our reaction conditions, and indeed, the yield could be
increased relatively to 6c, whereas the selectivity was un-
changed. Regarding the amount of base used, 3.5 equiva-
lents of LHMDS seemed to be a good compromise with re-
spect to yield and selectivity (Table 1, entry 5).
coordination of the peptide chain to the metal ion,[8d] and
the shielding of one face of the enolate by the side chain of
the neighboring amino acid. The highly ordered transition
state of the Claisen rearrangement might also favor the
transfer of chirality.
Very recently, we showed that tosyl-protected peptides 3
also undergo diastereoselective modifications in an intermo-
lecular fashion, for example, through alkylation or aldol re-
actions.[15] Unfortunately, the tosyl protecting group is diffi-
cult to remove, which results in a serious limitation of this
protocol and the need to find other protecting groups that
give good yields and selectivities.
Independent of this consideration, we were able to devel-
op new protocols towards unsaturated amino acids through
Pd- or Rh-catalyzed allylic alkylations[16,17] of chelated gly-
cine ester enolates 5. These enolates are much more reactive
relative to the generally used stabilized enolates and under-
go allylation already at ꢀ788C.[18] With these enolates, there-
fore, isomerization processes (e.g., of the double bond) can
be suppressed completely, a precondition for highly selective
reactions (Scheme 2).[19]
Under these optimized conditions, we investigated the al-
lylation of a range of different TFA-protected dipeptides.
Unfortunately, these peptides can not be directly obtained
by simple peptide coupling of TFA-protected amino acids
because of their configurational lability. In some cases,
nearly complete racemization was observed,[20] probably
through the formation of azlactone intermediates.[21] There-
fore, the required dipeptides were prepared from the corre-
sponding Z-protected derivatives, for example, the synthesis
Scheme 2. Isomerization-free allylation of chelated glycine ester 5.
TFA=trifluoroacetyl.
Chem. Eur. J. 2007, 13, 6204 – 6211
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
6205

U. Kazmaier and J. Deska
of 9a (Scheme 3). For the introduction of the TFA group,
methyl trifluoroacetate proved superior to the generally
used trifluoroacetic anhydride, which showed side reactions
tion of the allylation product, the peptides 10 were subjected
to catalytic hydrogenation (Scheme 4), and the HPLC chro-
matograms were compared with those of the corresponding
Scheme 3. Preparation of TFA-protected dipeptides. Z=BnOCO.
with several substrates. But with the methyl ester, the re-
quired peptides were obtained in 71–99% yield.
Scheme 4. Determination of configuration.
The results obtained with the different TFA-protected
peptides 9 are summarized in Table 2. As expected, the se-
lectivity increased with increasing steric demand of the in-
peptides 11 containing (S)-leucine, as illustrated with pep-
tide 10e. We found that the reference S,S peptides corre-
spond to the minor isomer, so
obviously the S,R stereoisomer
is formed preferentially, which
Table 2. Allylic alkylations of various dipeptides 9.
is in good agreement with re-
sults obtained earlier, for exam-
ple, in the peptide Claisen rear-
rangements.[14,22]
In case of methallyl carbon-
ate 7, the use of an excess of al-
Entry
Peptide
R
Reaction
conditions
Product
Yield
[%]
d.r.
(S,S)/(S,R)
lylating agent caused no prob-
lems as the reagent is inexpen-
sive. But bearing in mind that
this protocol might be applied
to drug synthesis, in which
more complicated and syntheti-
cally valuable side chains could
G
E
1
2
3
4
5
6
9a
9b
9c
9d
9a
9a
iBu
Me
CH2OMOM
CH2CH2SMe
iBu
ꢀ788C to rt, overnight
ꢀ788C to rt, overnight
ꢀ788C to rt, overnight
ꢀ788C to rt, overnight
ꢀ788C, 17 h
10a
10b
10c
10d
10a
10a
74
52
61
16:84
29:71
24:76
11:89
16:84
16:84
30–70
50
iBu
ꢀ78!ꢀ208C, 4.5 h,
2 mol% [{allylPdCl}2]
ꢀ78!ꢀ208C, 4.5 h,
2 mol% [{allylPdCl}2]
ꢀ788C to ꢀ208C, 4.5 h,
2 mol% [{allylPdCl}2]
81
be incorporated,
a
twofold
7
8
9e
9 f
Bn
10e
10 f
82
59
11:89
9:91
excess is not really attractive,
especially if the substrate
cannot be recovered. Therefore,
we were interested in whether
it is possible to change the ratio
tBu
ducing amino acid. The alanine peptide 9b gave the worst
result, whereas the tert-leucine derivative 9 f gave the best,
although in this case the yield was lower. Surprisingly, with
some substrates, such as 9d, the yields varied dramatically
and it was difficult to get reproducible results (Table 2,
entry 4). Therefore we had a closer look at the reaction con-
ditions with our model peptide 9a. We found, that the reac-
tion proceeded under very mild conditions, and 10a was ob-
tained in 50% yield if the temperature was kept at ꢀ788C
overnight (Table 2, entry 7). But in general, the yield was
higher if the reaction mixture was warmed to ꢀ208C over-
night in the presence of 4 mol% Pd0 (Table 2, entries 8–10).
Interestingly, the selectivity was nearly unchanged under
the different reaction conditions. The selectivity was deter-
mined by HPLC using the chiral column Reprosil Chiral
NR. In all the cases investigated so far, the minor diastereo-
mer was eluted first. To determine the absolute configura-
and use the peptide in excess to get a clean conversion of
the allylating reagent (Table 3; the yields are calculated rela-
tive to the allyl substrate).
Surprisingly, the yield could be increased with an excess
of enolate, and even more interestingly, the selectivity also
increased. For reactions with valuable allylic substrates, the
optimal conditions seem to be a 1.5-fold excess of the pep-
tide enolate (Table 3, entry 4). As long as only proteinogenic
amino acids are incorporated, such an excess can be tolerat-
ed, especially because the excess peptide can be recovered
relatively easily by chromatography.[22]
Therefore, the allylation of our TFA-protected peptides 9
was reinvestigated again under these “optimized” conditions
(Table 4). Indeed, both the yield and the selectivity in-
creased in most cases. Even with 9b, the peptide with the
“smallest” amino acid alanine, the selectivity could be in-
creased to 83% (Table 4, entry 1). All other peptides gave
6206
www.chemeurj.org
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2007, 13, 6204 – 6211

Pd-Catalyzed Alkylations of Peptides
Table 3. Inversion of the reagent excess.
FULL PAPER
Entry
Equiv
9a
Yield
[%]
d.r.
(S,S)/(S,R)
A
1
2
3
4
5
0.5
1.0
1.2
1.5
2.0
81[a]
60
16:84
12:88
10:90
8:92
77[b]
90[b]
91[b]
10:90
[a] Calculated relative to the peptide enolate. [b] Calculated relative to
the allyl carbonate.
Table 4. Allylic alkylations of various dipeptides 9 revisited.
Scheme 5. Introduction of several unfunctionalized side chains. Reaction
conditions: allylic substrate (0.7 equiv), [{allylPdCl}2] (1.4 mol%), PPh3
(6.3 mol%), LHMDS (3.5 equiv), ZnCl2 (1.2 equiv), THF, ꢀ78!ꢀ558C.
AA=amino acid, E=electrophile.
Entry
Peptide
R
Product
Yield
[%]
d.r.
(S,S)/(S,R)
U
1
2
3
4
5
6
7
9b
9c
9d
9e
9 f
9g
9h
Me
10b
10c
10d
10e
10 f
10g
10h
85
72
89
92
73
93
98
17:83
25:75
10:90
7:93
7:93
8:92
CH2OMOM
CH2CH2SMe
Bn
tBu
p-MeOBn
CH2OTBDPS
The formation of the branched product is by far less dra-
matic if allylic substrates with more sterically demanding
substituents are used (Scheme 6). For example, with carbon-
ate 19, the linear product 20 was the major product and
formed with excellent diastereo- and regioselectivity. Using
the terpenoid substrate 22, no reaction was observed on
8:92
selectivities of ꢁ90%, except the MOM-protected serine
peptide 9c (Table 4, entry 2), which gave the worst result.
But this problem could be readily solved by introduction of
the more sterically demanding, and therefore stronger di-
recting, TBDPS protecting group (9h; Table 4, entry 7).
To assess out the scope and limitations of this protocol,
we varied the allylating reagents (Scheme 5). The simple
allyl carbonate 12 seems to be the worst-case scenario. Inde-
pendent of the peptide used, the diastereomers of 13 are
only formed in a ratio of 2:1–3:1. But the selectivities were
much better with all other substrates. For example, the cin-
namyl derivative 14 gave the allylation products 15 in quan-
titative yield and excellent diastereoselectivities (ds). Al-
though an unsymmetrical p-allyl complex was formed as an
intermediate, only attack of the peptide enolate at the ter-
minal position of the allyl complex was observed, thus
giving rise to the product with the linear side chain as the
sole regioisomer. The situation with crotyl carbonate 16 is
slightly different: As expected, the regioisomeric branched
product 18 was formed in a comparable yield; however, a
second set of diastereomers was formed, which makes sub-
strates of this type less attractive. But this problem is not a
result of the peptide allylation but of the allylic alkylation
itself.
Scheme 6. Regioselective introduction of several aliphatic side chains.
Reaction conditions: a) allylic substrate (0.7 equiv), [{allylPdCl}2]
(1.4 mol%), PPh3 (6.3 mol%), LHMDS (3.5 equiv), ZnCl2 (1.2 equiv),
THF, ꢀ78!ꢀ558C; b) 22 (0.5 equiv), [{allylPdCl}2] (1 mol%), PPh3
(4.5 mol%), LHMDS (3.5 equiv), ZnCl2 (1.2 equiv), THF, ꢀ78!558C.
Chem. Eur. J. 2007, 13, 6204 – 6211
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
6207

U. Kazmaier and J. Deska
warming the reaction mixture to room temperature. In this
special case, heating to 558C was required to get consump-
tion of the carbonate, and allylation product 23 was ob-
tained as the only regioisomer in good yield. This case im-
pressively demonstrates the extraordinary thermal stability
of the chelated peptide enolates. Substrate 24, readily pre-
pared from citronellal, opens up access to another terpene-
type side chain in both excellent yield and selectivity.
For drug syntheses, it might be also interesting to incorpo-
rate side chains bearing functional groups. A few examples
of peptides with oxygenated side chains are shown in
Scheme 7. The oxygenated allylic substrates seem to be
the peptide Claisen rearrangement.[13,14] In such complexes,
one face of the enolate should be shielded by the side chain
of the adjacent amino acids. It is known from X-ray analysis
of metal–peptide complexes that the phenyl rings of aromat-
ic side chains can especially interact with the chelating
metal center,[23] which would explain why these peptides
show the highest selectivities. If this assumption is correct,
one might expect that the selectivity would fall if the phe-
nylalanine is replaced by a more flexible b amino acid. In
addition, alkylation of the TFA–amide functionality should
suppress the proposed threefold coordination completely,
thus resulting in a loss of stereoselectivity. To prove this pre-
diction, the required substrates were prepared according to
Scheme 8. For the preparation of a prolonged phenylalanine
Scheme 8. Preparation of modified dipeptides. NMM=N-methylmorpho-
line.
derivative 35, a Z-protected phenylalanine was subjected to
an Arndt–Eistert homologation. Decomposition of the inter-
mediate diazoketone was accomplished directly in the pres-
ence of tert-butyl glycinate, thus giving rise to the Z-protect-
ed b-dipeptide 34.[24] Replacing the protecting group with
TFA was carried out analogously to the dipeptide prepared
previously. TFA-protected peptides can be readily N-allylat-
ed under neutral conditions by using palladium catalysis.[25]
Using this protocol, peptide 36 was obtained directly from
9e.
Scheme 7. Regioselective introduction of several functionalized side
chains. Reaction conditions: a) allylic substrate (0.7 equiv), [{allylPdCl}2]
(1.4 mol%), PPh3 (6.3 mol%), LHMDS (3.5 equiv), ZnCl2 (1.2 equiv),
THF, ꢀ78!ꢀ558C; b) the same as (a), but warming to 08C was re-
quired. TBDMS=tert-butyldimethylsilyl, TBDPS=tert-butyldiphenylsilyl
ether.
even more suitable than the unfunctionalized side chains. In
all the examples investigated so far, the selectivities were
95% or higher and the yields were very high. These findings
allowed the direct incorporation of sugar-type side chains
into a given peptide. Substrates 30 and 32 could readily be
obtained by addition of a vinyl Grignard reagent to the cor-
responding aldehyde. In this case, the configuration of the
leaving group had no influence as terminal p-allyl com-
plexes isomerize readily and the stereochemical outcome of
the allylation reaction is solely controlled by the peptide.
To explain the stereochemical outcome of the reaction,
one might assume a threefold coordination of the peptide
chain towards the chelating zinc ion, as proposed earlier for
And indeed, allylation of these two peptides gave exactly
the expected results (Scheme 9). With the prolonged peptide
35, the selectivities dropped dramatically, and the N-alkylat-
ed derivative gave a 1:1 diastereomeric mixture of 39 in
quantitative yield. These results support the theory of multi-
fold coordination of the peptide chain.
Conclusion
In conclusion, we have shown that chelated peptide enolates
are excellent nucleophiles for stereoselective palladium-cat-
alyzed allylic alkylations. All the results indicate that a
6208
www.chemeurj.org
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2007, 13, 6204 – 6211

Pd-Catalyzed Alkylations of Peptides
FULL PAPER
showed complete conversion (20–90 min). After addition of diethyl ether
(2 mLmmolꢀ1), the reaction mixture was filtered through celite and con-
centrated in vacuo. The free amine was redissolved in methanol
(2 mLmmolꢀ1), triethylamine (2 equiv) and methyl trifluoroacetate
(2 equiv) were added at 08C and the reaction mixture was stirred over-
night at room temperature. The solvent was evaporated and the residue
was partitioned between diethyl ether and 1n HCl. The organic layer
was dried over Na2SO4 and concentrated in vacuo. The title compound
was purified by column chromatography or recrystallization.
tert-Butyl (N-trifluoroacetyl-(S)-leucyl)glycinate (9a): According to the
general procedure for peptide coupling, starting from Z-(S)-leucine, 9a
was obtained after recrystallization (hexane, dichloromethane) as a color-
less solid in a yield of 8.06 g (23.7 mmol, 90%). M.p. 113–1158C; [a]D20
=
ꢀ25.6 (c=1.0, CHCl3); 1H NMR (500 MHz, CDCl3): d=7.50 (d, J=
8.5 Hz, 1H; NH), 6.71 (t, J=5.1 Hz, 1H; NH), 4.58 (dt, J=8.5, 5.7 Hz,
1H; NCH), 3.84 (dd, J=18.0, 5.4 Hz, 1H; NCH2), 3.86 (dd, J=18.3,
5.1 Hz, 1H; NCH2), 1.66 (m, 3H; CH
(CH3)3), 0.92 (d, J=3.5 Hz, 3H; CH(CH3)2), 0.91 ppm (d, J=3.8 Hz, 3H;
CH
(CH3)2); 13C NMR (125 MHz, CDCl3): d=168.4, 170.7 (CON, COO),
157.0 (q, J=37.2 Hz; CF3CO), 115.7 (q, J=286 Hz; CF3), 82.7 (C(CH3)3),
52.0 (NCH), 42.1 (NCH2), 41.5 (CHCH2CH), 28.0 (C(CH3)3), 24.7 (CH-
(CH3)2), 22.7 (CH(CH3)2), 21.9 ppm (CH(CH3)2); HRMS (CI) calcd for
N
A
ACHTREUNG
AHCTREUNG
G
G
Scheme 9. Allylic alkylations of modified dipeptides. Reaction condi-
tions: allylic substrate (0.7 equiv), [{allylPdCl}2] (1.4 mol%), PPh3
(6.3 mol%), LHMDS (3.5 equiv), ZnCl2 (1.2 equiv), THF, ꢀ78!ꢀ558C.
(S)-tert-Butyl (N-trifluoroacetyl-(S)-phenylalanyl)glycinate (9e): Accord-
ing to the general procedure for peptide coupling, starting from Z-(S)-
phenylalanine, 9e was obtained after recrystallization (hexane, dichloro-
methane) as a colorless solid in a yield of 1.45 g (3.87 mmol, 95%).
M.p. 1348C; [a]2D0 =+19.5 (c=1.0, CHCl3); 1H NMR (500 MHz, CDCl3):
d=7.38 (d, J=7.5 Hz, 1H; NH), 7.17–7.30 (m, 5H; C6H5), 6.18 (t, J=
4.5 Hz, 1H; NH), 4.71 (dt, J=7.5, 6.5 Hz, 1H; NCH), 3.88 (dd, J=18.2,
5.1 Hz, 1H; NCH2), 3.81 (dd, J=18.2, 5.0 Hz, 1H; NCH2), 3.13 (dd, J=
13.8, 6.4 Hz, 1H; C6H5CH2), 3.08 (dd, J=13.7, 7.5 Hz, 1H; C6H5CH2),
threefold coordination of the peptide chain is the reason for
the high selectivities obtained. Even highly functionalized
and complex side chains can be introduced in excellent
yields and selectivities. Applications of this protocol towards
natural product synthesis and investigations of chiral p-allyl
complexes are currently in progress.
1.43 ppm (s, 9H; C
(CON, COO), 156.7 (q, J=37.2 Hz; CF3CO), 127.5, 128.8, 129.2, 135.3
(C6H5), 115.6 (q, J=286 Hz; CF3), 82.8 (C(CH3)3), 54.7 (NCH), 42.0
(NCH2), 38.5 (C6H5CH2), 28.0 ppm (C(CH3)3); HRMS (CI) calcd for
A
AHCTREUNG
Experimental Section
AHCTREUNG
C17H21F3N2O4 [M]+: 374.1453; found: 374.1464; elemental analysis calcd
(%) for C17H21F3N2O4 (374.39): C 54.54, H 5.65, N7.48; found: C 54.53,
H 5.58, N7.46.
General remarks: All the reactions were carried out in oven-dried glass-
ware (1008C) under argon. All the solvents were dried before use: THF
was distilled from LiAlH4 and CH2Cl2 from CaH2 and stored over molec-
ular sieves. The products were purified by flash chromatography on silica
gel (32–63 mm). Mixtures of ethyl acetate and hexanes were generally
used as the eluents. Analysis by TLC was carried out on commercially
precoated Polygram SIL-G/UV 254 plates (Machery—Nagel, Dueren).
Visualization was accomplished with UV light, KMnO4 solution, or
iodine. Melting points were determined on a Büchi melting-point appara-
General procedure for allylic alkylations of peptides: n-Butyllithium
(1.6m in hexanes, 0.82 mL, 1.31 mmol) was added dropwise to a solution
of hexamethyldisilazane (233 mg, 1.44 mmol) in dry THF (2 mL) at
ꢀ788C. The cooling bath was removed, and the solution was allowed to
warm up to room temperature. In
a second flask, ZnCl2 (57 mg,
0.42 mmol) was carefully dried in vacuo with a heat gun. After the mix-
ture was cooled to room temperature, the corresponding dipeptide
(0.375 mmol) was added dissolved in dry THF (2 mL). The freshly pre-
pared LHMDS solution was cooled to ꢀ788C and the ZnCl2/peptide so-
lution was slowly added by syringe. In a third flask, [{allylPdCl}2] (1.8 mg,
5.0 mmol) and PPh3 (5.9 mg, 22.5 mmol) were dissolved in dry THF
(0.5 mL), allylic carbonate (0.25 mmol) was added, and the catalyst/car-
bonate solution was transferred to the cold solution of the zinc enolate
by syringe. The excess dry ice was removed from the cooling bath, and
the reaction mixture was allowed to warm up. After diluting with diethyl
ether, the mixture was quenched by the addition of 1n HCl (sat. NH4Cl
was added if substrates with acid labile groups were used). The aqueous
layer was extracted twice with diethyl ether and the combined organic
layers were dried over Na2SO4. The solvent was evaporated and the resi-
due was purified by column chromatography (silica gel, hexane/ethyl ace-
tate).
1
tus and are uncorrected. H and 13C NMR spectroscopic analysis was per-
formed on Bruker AC-500 or Bruker DRX-500 spectrometers. Selected
signals for the minor diastereomers are extracted from the spectra of the
diastereomeric mixture. The values for diastereomeric excess were deter-
mined on by analytical HPLC using a Trentec Reprosil-100 Chiral-NR
8 mm-column and a Shimadzu UV detector. Optical rotations were mea-
sured on a Perkin–Elmer polarimeter PE 241. Chemical ionization (CI)
mass-spectrometric analysis was performed on a Finnigan MAT 95 ma-
chine. Elemental analyses were carried out at the Department of Chemis-
try, University of Saarland (Germany).
General procedure for the synthesis of TFA-protected peptides: N-meth-
ylmorpholine (1.05 equiv) and isobutyl chloroformate (1.0 equiv) were
added to a solution of the Z-protected amino acid (1.0 equiv) in dry THF
(2 mLmmolꢀ1) at ꢀ208C, and reaction mixture was stirred for 5 min fol-
lowed by addition of tert-butyl glycinate (1.0 equiv). The reaction mixture
was diluted with diethyl ether after stirring at room temperature over-
night, washed with 1n HCl, sat. NaHCO3 solution, and brine. The organ-
ic layer was dried over Na2SO4 and the solvent was evaporated in vacuo.
The crude Z dipeptide was dissolved in methanol (1 mLmmolꢀ1), ammo-
nium formate (2–3 equiv) and palladium on charcoal were added at room
temperature and the reaction mixture was stirred until analysis by TLC
tert-Butyl
(N-trifluoroacetyl-(S)-leucyl)-(R)-4,5-didehydroleucinate
(10a): Following the general procedure, peptide 9a (128 mg, 0.375 mmol)
and carbonate 7 (36 mg, 0.25 mmol) were subjected to the peptide allyla-
tion. After column chromatography (silica gel, hexane/ethyl acetate 92:8)
peptide 10a (88 mg, 0.223 mmol; 89%) was obtained as a colorless oil.
[a]2D0 =ꢀ30.7 (c=1.0, CHCl3, 92% ds). Major diastereomer: 1H N MR
Chem. Eur. J. 2007, 13, 6204 – 6211
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
6209

U. Kazmaier and J. Deska
(500 MHz, CDCl3): d=7.42 (d, J=7.1 Hz, 1H; NH), 6.49 (d, J=7.3 Hz,
1H; NH), 4.81 (s, 1H; C=CH2), 4.72 (s, 1H; C=CH2), 4.50–4.56 (m, 2H;
NCH), 2.51 (dd, J=14.0, 5.6 Hz, 1H; H2C=CCH2), 2.36 (dd, J=14.0,
8.3 Hz, 1H; H2C=CCH2), 1.72 (s, 3H; CCH3), 1.51–1.68 (m, 3H; CH-
13.97 min,
t
R(S,R) =19.62 min; HRMS (CI) calcd for C21H30F3N2O4
[M+H]+: 431.2158; found: 431.2177; elemental analysis calcd (%) for
C21H29F3N2O4 (430.47): C 58.59, H 6.79, N6.51; found: C 59.05, H 6.76, N
6.26.
A
G
tert-Butyl (N-trifluoroacetyl-(S)-phenylalanyl)-(S)-leucinate ((S,S)-11e):
(S)-Cbz-Phenylalanine (90 mg, 0.30 mmol) and (S)-tert-butyl leucinate
hydrotosylate (108 mg, 0.30 mmol) were treated according to the general
procedure for peptide coupling. Column chromatography (silica gel,
hexane/ethyl acetate 85:15) yielded 11e (108 mg, 0.25 mmol, 83%) as a
colorless solid. M.p. 95–968C; [a]2D0 =+7.5 (c=1.0; CHCl3); 1H N MR
(500 MHz, CDCl3): d=7.55 (d, J=7.6 Hz, 1H; NH), 7.12–7.24 (m, 5H;
C6H5), 6.28 (d, J=8.0 Hz, 1H; NH), 4.67 (dt, J=7.4, 6.7 Hz, 1H; NCH),
4.35 (dt, J=8.2, 5.5 Hz, 1H; NCH), 3.07 (dd, J=13.8, 6.8 Hz, 1H;
C6H5CH2), 3.03 (dd, J=13.8, 6.4 Hz, 1H; C6H5CH2), 1.37–1.54 (m, 3H;
ACHTREUNG
AHCTREUNG
(CH
G
U
(selected signals): 1H NMR (500 MHz, CDCl3): d=6.42 (d, J=7.4 Hz,
1H; NH), 4.80 (s, 1H; C=CH2), 4.70 (s, 1H; C=CH2), 1.70 (s, 3H;
CCH3), 1.43 ppm (s, 9H; C
140.2 (C=CH2), 82.5 (C(CH3)3), 41.2 (CHCH2CH), 40.6 (H2C=CCH2),
27.7 ppm (C(CH3)3); HPLC (Reprosil 100 Chiral-NR 8 mm, hexane/
iPrOH 98:2, 1.0 mLminꢀ1
219 nm): tR(S,S) =5.79 min, tR(S,R) =7.55 min;
E
AHCTREUNG
ACHTREUNG
CH
A
ACHTREUNG
,
CH
A
ACHTREUNG
HRMS (CI) calcd for C18H30F3N2O4 [M+H]+: 395.2113; found: 395.2119;
elemental analysis calcd (%) for C18H29F3N2O4 (394.44): C 54.81, H 7.41,
N7.10; found: C 54.97, H 7.26, N6.94.
(125 MHz, CDCl3): d=168.8, 171.3 (CON, COO), 156.7 (q, J=37.8 Hz,
CF3CON), 127.4, 129.3, 129.7, 135.2 (C6H5), 115.6 (q, J=287 Hz, CF3),
82.2 (C
ACHTREUNG
(C6H5CH2), 27.9 (C
A
N
ACHTREUNG
tert-Butyl
(N-trifluoroacetyl-(S)-phenylalanyl)-(R)-4,5-didehydroleuci-
(CH(CH3)2); HPLC (Reprosil 100 Chiral-NR 8 mm, hexane/iPrOH
A
nate (10e): Following the general procedure, peptide 9e (140 mg,
0.375 mmol) and carbonate 7 (36 mg, 0.25 mmol) were subjected to the
peptide allylation. After column chromatography (silica gel, hexane/ethyl
acetate 92:8) peptide 10e (99 mg, 0.231 mmol; 92%) was obtained as a
colorless solid. M.p. 112–1148C; [a]2D0 =+3.8 (c=1.0, CHCl3, 93% ds).
Major diastereomer: 1H NMR (500 MHz, CDCl3): d=7.51 (d, J=7.7 Hz,
1H; NH), 7.11–7.25 (m, 5H; C6H5), 6.33 (d, J=7.7 Hz, 1H; NH), 4.71 (s,
1H; C=CH2), 4.64 (dt, J=7.7, 7.0 Hz, 1H; NCH), 4.56 (s, 1H; C=CH2),
4.41 (dt, J=7.7, 6.3 Hz, 1H; NCH), 3.08 (dd, J=13.8, 6.3 Hz, 1H;
C6H5CH2), 3.03 (dd, J=13.7, 7.5 Hz, 1H; C6H5CH2), 2.28 (dd, J=14.1,
6.3 Hz, 1H; H2C=CCH2), 2.20 (dd, J=14.0, 7.8 Hz, 1H; H2C=CCH2),
99.5:0.5, 1.5 mLminꢀ1, 206 nm): tR(S,S) =13.94 min.
Financial support from the Deutsche Forschungsgemeinschaft (Ka880/6)
and the Fonds der Chemischen Industrie is gratefully acknowledged.
1.63 (s, 3H; CCH3), 1.35 ppm (s, 9H; C
(CH3)3); 13C NMR (125 MHz,
A
[1] a) D. J. Faulkner, Nat. Prod. Rep. 1998, 15, 113–158; b) A. Schaller,
in Studies in Natural Products Chemistry, Vol. 25 (Ed.: A. U.
Rahman), Elsevier Science, Amsterdam, Netherlands, 2001,
pp. 367–411.
[2] F. von Nussbaum, M. Brands, B. Hinzen, S. Weigand, D. Häbich,
Angew. Chem. 2006, 118, 5194–5254; Angew. Chem. Int. Ed. 2006,
45, 5072–5129.
[3] a) M. A. Marahiel, T. Stachelhaus, H. D. Mootz, Chem. Rev. 1997,
97, 2651–2673; b) H. v. Dçhren, U. Keller, J. Vater, R. Zocher,
Chem. Rev. 1997, 97, 2675–2705.
[4] a) N. Fusetani, S. Matsunaga, Chem. Rev. 1993, 93, 1793–1806;
b) D. R. W. Hogson, J. M. Sanderson, Chem. Soc. Rev. 2004, 33,
422–430.
CDCl3): d=168.8, 170.5 (CON, COO), 156.6 (q, J=37.2 Hz; CF3CON),
140.3 (C=CH2), 127.4, 128.8, 129.3, 135.3, (C6H5), 115.8 (q, J=236 Hz;
CF3), 114.6 (C=CH2), 82.6 (C
ACHTREUNG
(H2C=CCH2), 38.6, (C6H5CH2), 27.9 (C
AHCTREUNG
diastereomer (selected signals): 1H NMR (500 MHz, CDCl3): d=6.15 (d,
J=7.4 Hz, 1H; NH), 4.53 (s, 1H; C=CH2), 3.06 (d, J=6.3 Hz, 1H;
C6H5CH2), 2.39 (dd, J=14.0, 6.1 Hz, 1H; H2C=CCH2), 1.62 (s, 3H;
CCH3), 1.39 (s, 9H; C
(C=CH2), 82.4 (C(CH3)3), 40.5 (H2C=CCH2), 28.0 ppm (s, C
(CH3)3); 13C NMR (125 MHz, CDCl3): d=140.3
(CH3)3);
A
ACHTREUNG
HPLC (Reprosil 100 Chiral-NR 8 mm, hexane/iPrOH 99.5:0.5,
1.5 mLminꢀ1, 209 nm): tR(S,S) =16.65 min, tR(S,R) =22.07 min; HRMS (CI)
calcd for C21H28F3N2O4 [M+H]+: 429.2001; found: 429.2041; elemental
analysis calcd (%) for C21H27F3N2O4 (428.46): C 58.87, H 6.35, N6.54;
found: C 59.02, H 6.30, N6.54.
[5] a) A. N. Eberle, Chimia 1991, 45, 145–159; b) V. De Filippis, D.
Quarzago, A. Vindigni, E. Di Cera, A. Fontana, Biochemistry 1998,
37, 13507–13515.
Determination ofthe absolute conifguration : For the determination of
the absolute configuration of the allylation products, the dehydroleucine
derivatives were hydrogenated to the corresponding (S,S/R)-AA-Leu di-
peptides, which could be compared with their S,S analogues (e.g., phenyl-
alanine–leucine dipeptide 11e).
[6] a) P. Renaud, D. Seebach, Angew. Chem. 1986, 98, 836–837; Angew.
Chem. Int. Ed. Engl. 1986, 25, 843–844; b) P. Renaud, D. Seebach,
Helv. Chim. Acta 1986, 69, 1704–1710; c) D. Seebach, R. Charczuk,
C. Gerber, P. Renaud, H. Berner, H. Schneider, Helv. Chim. Acta
1989, 72, 401–425; d) C. J. Easton, I. M. Scharfbillig, E. W. Tan, Tet-
rahedron Lett. 1988, 29, 1565–1568; e) G. Apitz, W. Steglich, Tetra-
hedron Lett. 1991, 32, 3163–3166; f) W. Steglich, M. Jäger, S. Jaroch,
P. Zistler, Pure Appl. Chem. 1994, 66, 2167–2170.
[7] a) C. J. Easton, Chem. Rev. 1997, 97, 53–82; b) M. Ricci, L. Madar-
iaga, T. Skrydstrup, Angew. Chem. 2000, 112, 248–252; Angew.
Chem. Int. Ed. 2000, 2039, 242–246; c) M. Ricci, P. Blakskjaer, T.
Skrydstrup, J. Am. Chem. Soc. 2000, 122, 12413–12421.
[8] Reviews: a) D. Seebach, Angew. Chem. 1988, 100, 1685–1715;
Angew. Chem. Int. Ed. Engl. 1988, 27, 1624–1654; b) D. Seebach,
Aldrichimica Acta 1992, 25, 59–66; c) D. Seebach, A. K. Beck, S. A.
, in Modern Synthetic Methods, Vol. 7 (Eds.: B. Ernst, C. Leumann),
Helvetica Chimica Acta, Basel, 1995, pp. 1–179; d) K. Severin, R.
Bergs, W. Beck, Angew. Chem. 1998, 110, 1722–1743; Angew.
Chem. Int. Ed. 1998, 37, 1634–1654.
tert-Butyl (N-trifluoroacetyl-(S)-phenylalanyl)-(R)-leucinate ((S,R)-11e):
Pd/C (20 mg) was added to peptide 10e (43 mg, 0.10 mmol) in methanol
(2 mL). The reaction mixture was stirred in an atmosphere of H2 for 30
min. After filtration over silica, 11e (43 mg, 0.10 mmol, 100%) was ob-
tained as a colorless solid. M.p. 101–1028C; [a]2D0 =+3.3 (c=1.0, CHCl3,
93% ds). Major diastereomer: 1H NMR (500 MHz, CDCl3): d=7.55 (d,
J=7.6 Hz, 1H; NH), 7.12–7.24 (m, 5H; C6H5), 6.28 (d, J=8.0 Hz, 1H;
NH), 4.67 (dt, J=7.4, 6.7 Hz, 1H; NCH), 4.35 (dt, J=8.2, 5.5 Hz, 1H;
NCH), 3.07 (dd, J=13.8, 6.8 Hz, 1H; C6H5CH2), 3.03 (dd, J=13.8,
6.4 Hz, 1H; C6H5CH2), 1.37–1.54 (m, 3H; CH
(s, 9H; C(CH3)3), 0.83 (d, J=6.0 Hz, 3H; CH
6.0 Hz, 3H; CH
(CH3)2); 13C NMR (125 MHz, CDCl3): d=168.8, 171.3
(CON, COO), 156.7 (q, J=37.8 Hz; CF3CON), 127.4, 129.3, 129.7, 135.2
(C6H5), 115.6 (q, J=287 Hz; CF3), 82.2 (C(CH3)3), 54.6 (NCH), 51.8
(NCH), 41.6 (CHCH2CH), 38.3, (C6H5CH2), 27.9 (C(CH3)3), 24.8 (CH-
(CH3)2), 22.5 (CH(CH3)2), 22.0 ppm (CH(CH3)2); HPLC (Reprosil 100
Chiral-NR 8 mm, hexane/iPrOH 99.5/0.5, 1.5 mLminꢀ1, 206 nm): tR(S,S)
ACHTREUNG
A
ACHTREUNG
ACHTREUNG
AHCTREUNG
AHCTREUNG
A
G
A
[9] S. A. Miller, S. L. Griffiths, D. Seebach, Helv. Chim. Acta 1993, 76,
563–595.
=
6210
www.chemeurj.org
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2007, 13, 6204 – 6211

Pd-Catalyzed Alkylations of Peptides
FULL PAPER
[10] D. Seebach, A. K. Beck, H. G. Bossler, C. Gerber, S. Y. Ko, C. W.
Murtiashaw, R. Naef, S.-I. Shoda, A. Thaler, M. Krieger, R. Wenger,
Helv. Chim. Acta 1993, 76, 1564–1590.
[11] T. Ooi, E. Tayama, K. Maruoka, Angew. Chem. 2003, 115, 599–602;
Angew. Chem. Int. Ed. 2003, 4, 579.
[12] a) M. J. OꢁDonnell, C. Zhou, W. L. Scott, J. Am. Chem. Soc. 1996,
118, 6070–6071; b) M. J. OꢁDonnell, Aldrichimica Acta 2001, 34, 3–
15.
[13] a) U. Kazmaier, J. Org. Chem. 1994, 59, 6667–6670; b) U. Kazmaier,
S. Maier, Chem. Commun. 1998, 2535–2536; c) U. Kazmaier, S.
Maier, Org. Lett. 1999, 1, 1763–1766; d) U. Kazmaier, S. Maier,
F. L. Zumpe, Synlett 2000, 1523–1535; e) S. Maier, U. Kazmaier,
Eur. J. Org. Chem. 2000, 1241–1251; f) U. Kazmaier, C. Hebach, A.
Watzke, S. Maier, H. Mues, V. Huch, Org. Biomol. Chem. 2005, 3,
136–145.
[14] U. Kazmaier, S. Maier, J. Org. Chem. 1999, 64, 4574–4575.
[15] U. Kazmaier, J. Deska, A. Watzke, Angew. Chem. 2006, 118, 4973–
4976; Angew. Chem. Int. Ed. 2006, 45, 4855–4858.
[16] a) T. D. Weiß, G. Helmchen, U. Kazmaier, Chem. Commun. 2002,
1270–1271; b) M. Bauer, U. Kazmaier, Recent Res. Dev. Org. Chem.
2005, 9, 49–69; c) U. Kazmaier, T. Lindner, Angew. Chem. 2005,
117, 3368–3371; Angew. Chem. Int. Ed. 2005, 44, 3303–3306; d) T.
Lindner, U. Kazmaier, Adv. Synth. Catal. 2005, 347, 1687–1695.
[17] U. Kazmaier, D. Stolz, Angew. Chem. 2006, 118, 3143–3146; Angew.
Chem. Int. Ed. 2006, 45, 3072–3075.
[18] U. Kazmaier, F. L. Zumpe, Angew. Chem. 1999, 111, 1572–1574;
Angew. Chem. Int. Ed. 1999, 38, 1468–1470.
[19] a) U. Kazmaier, F. L. Zumpe, Angew. Chem. 2000, 112, 805–807;
Angew. Chem. Int. Ed. 2000, 39, 802–804; b) U. Kazmaier, F. L.
Zumpe, Eur. J. Org. Chem. 2001, 4067–4076; c) U. Kazmaier, M.
Pohlman, Synlett 2004, 623–626.
[20] F. Weygand, A. Prox, L. Schmidhammer, W. Kçnig, Angew. Chem.
1963, 75, 282–287; Angew. Chem. Int. Ed. Engl. 1963, 2, 183–188.
[21] F. Weygand, W. Steglich, H. Tanner, Liebigs Ann. Chem. 1962, 658,
128–150.
[22] No epimerization of the stereogenic center was observed, as deter-
mined by HPLC, which is in good agreement with the results ob-
tained in the peptide Claisen rearrangement;[13a] reference samples
were obtained by using TFA-protected amino acids in the peptide
coupling step, which occurred with significant racemization.
[23] H. Masuda, A. Odani, O. Yamauchi, Inorg. Chem. 1989, 28, 624–
625.
[24] J. Podlech, D. Seebach, Angew. Chem. 1995, 107, 507–509; Angew.
Chem. Int. Ed. Engl. 1995, 34, 471–472.
[25] a) F. L. Zumpe, U. Kazmaier, Synlett 1998, 1199–1200; b) F. L.
Zumpe, U. Kazmaier, Synthesis 1999, 10, 1785–1791.
Received: January 19, 2007
Published online: May 7, 2007
Chem. Eur. J. 2007, 13, 6204 – 6211
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
6211