Exercises in pyrrolidine chemistry: Gram scale synthesis of a Pro-Pro dipeptide mimetic with a polyproline type II helix conformation

Article abstract of DOI:10.1002/chem.201101704

A practical and scalable synthesis of a Fmoc-protected tricyclic dipeptide mimetic (6), that is, a 1,4-diaza-tricyclo-[8.3.0^{3, 7}]-tridec-8-ene derivative resembling a rigidified di-L-proline in a polyproline type II (PPII) helix conformation, was developed. The strategy is based on a Ru-catalyzed ring-closing metathesis of a dipeptide (4) prepared by PyBOP coupling of cis-5-vinylproline tert-butylester (2) and trans-N-Boc-3-vinylproline (rac-3) followed by chromatographic diastereomer separation. Building block 2 was prepared from L-proline in six steps via electrochemical C5-methoxylation, cyanation and conversion of the nitrile into a vinyl substituent. Building block rac-3 was prepared in five steps exploiting a Cu-catalyzed 1,4-addition of vinyl-MgBr to a 2,3-dehydroproline derivative in the key step. In the course of the investigation subtle dependencies of protecting groups on the reactivity of the 2,3- and 2,5-disubstituted pyrrolidine derivatives were observed. The configuration and conformational preference of several intermediates were determined by X-ray crystallography. The developed synthesis allows the preparation of substantial amounts of 6, which will be used in the search for new small molecules for the modulation of protein-protein interactions involving prolin-rich motifs (PRDs).

Full text of DOI:10.1002/chem.201101704

FULL PAPER
DOI: 10.1002/chem.201101704
Exercises in Pyrrolidine Chemistry: Gram Scale Synthesis of a Pro–Pro
Dipeptide Mimetic with a Polyproline Type II Helix Conformation
[
a]
[a]
[a]
[b]
Cꢀdric Reuter, Peter Huy, Jçrg-Martin Neudçrfl, Ronald Kꢁhne, and
[
a]
Hans-Gꢁnther Schmalz*
Abstract: A practical and scalable syn-
thesis of a Fmoc-protected tricyclic di-
peptide mimetic (6), that is, a 1,4-
block 2 was prepared from l-proline in
six steps via electrochemical C5-me-
thoxylation, cyanation and conversion
of the nitrile into a vinyl substituent.
Building block rac-3 was prepared in
five steps exploiting a Cu-catalyzed
1,4-addition of vinyl-MgBr to a 2,3-de-
hydroproline derivative in the key step.
In the course of the investigation
subtle dependencies of protecting
groups on the reactivity of the 2,3- and
2,5-disubstituted pyrrolidine derivatives
were observed. The configuration and
conformational preference of several
intermediates were determined by X-
ray crystallography. The developed syn-
thesis allows the preparation of sub-
stantial amounts of 6, which will be
used in the search for new small mole-
cules for the modulation of protein–
protein interactions involving prolin-
rich motifs (PRDs).
3,7
diaza-tricyclo-[8.3.0 ]-tridec-8-ene de-
rivative resembling a rigidified di-l-
proline in a polyproline type II (PPII)
helix conformation, was developed.
The strategy is based on a Ru-cata-
lyzed ring-closing metathesis of a di-
peptide (4) prepared by PyBOP cou-
pling of cis-5-vinylproline tert-butyles-
ter (2) and trans-N-Boc-3-vinylproline
Keywords: conformation analysis ·
electrochemistry · olefin metathesis ·
peptidomimetics · proline
(
rac-3) followed by chromatographic
diastereomer separation. Building
Introduction
teractions represent attractive targets for pharmaceutical in-
tervention.
The selective modulation of relevant protein–protein inter-
actions by means of small synthetic molecules represents an
important challenge for chemical biology and medicinal
A common property of all PRMs is that they preferential-
ly adopt a left-handed polyproline type II helix (PPII) con-
formation. By viewing along the helix axis, these units ex-
hibit a shape resembling a triangular prism (Figure 1).
PPII helices are characterized by a helical pitch of 9.3 ꢀ,
[1]
[9a]
chemistry. In nature, small peptides usually take over this
function. Thus, the search for suitable synthetic peptidomi-
[2]
metics targeting such interactions has nowadays led to in-
three residues per turn, F and Y angles centered around
[3]
[10]
tensive research activities in many laboratories. A class of
prominent protein–protein interactions, which are well un-
derstood from a structural biology point of view, are the in-
teractions of so-called proline-rich motif-recognizing do-
mains (PRDs) with proteins containing proline-rich motifs
À758 and 1458, respectively, and a complete lack of main-
chain hydrogen bonding.
[4]
(
PRMs). These are widely utilized by nature and are in-
volved in several relevant processes such as tyrosine kinase
[
5]
[6]
receptor signaling, endocytosis, cytoskeletal rearrange-
[7]
[8]
[9]
ments, transcription, and splicing. Accordingly, such in-
+
+
[
a] Dipl.-Chem. C. Reuter, Dr. P. Huy, Dr. J.-M. Neudçrfl,
Prof. Dr. H.-G. Schmalz
Figure 1. Ideal PPII helix as seen from the side (left) or along the helix
axis (right).
Department of Chemistry, Institute of Organic Chemistry
University of Cologne
Greinstrasse 4, 50939 Kçln (Germany)
E-mail: schmalz@uni-koeln.de
By examination of molecular models of a Pro–Pro unit of
a PPII helix we envisioned to lock the system in this confor-
mation by simply introducing a vinylidene bridge between
the two proline rings in a stereo-defined manner. The re-
sulting tricyclic compound X would represent a conforma-
tionally restricted di-proline analogue almost perfectly mim-
icking the desired PPII structure (Figure 2).
[
b] Dr. R. Kꢁhne
Leibniz-Institut fꢁr Molekulare Pharmakologie (FMP)
Campus BerlinBuch
Robert-Rçssle-Str. 10, 13125 Berlin (Germany)
[11]
+
[
] These authors contributed equally to this work.
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.201101704.
Chem. Eur. J. 2011, 17, 12037 – 12044
ꢂ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
12037

Figure 2. Diproline (Pro–Pro) in a PPII helix conformation (left) and the
designed conformationally restricted analogue X (right). The backbone
angles of the calculated structure of X (F=À788 and Y=1468) corre-
spond to an ideal PPII helix.
Scheme 1. First-generation synthesis of Fmoc-X (6) according to Zaminer
[12]
et al.:
5 mol%), CH
Cl, KHCO , MeCN, 08C, RT, 15 h.
a) PyBop, DIPEA, MeCN, 08C, RT, 15 h; b) Grubbs II
(
2
2
Cl , reflux, 24 h; c) TFA, CH Cl , 08C, RT, 1 h; d) Fmoc-
2
2
3
We recently communicated a synthesis of X (as its Fmoc
derivative) and demonstrated the feasibility of the structural
design by successfully incorporating X as a PPII Pro–Pro
substitute into the core-motif of a SH3 domain-binding pep-
tide (Figure 3). This way it was demonstrated for the first
time that, in contrast to previous reports, it is indeed possi-
ble to replace proline units of a PRM without loosing the af-
finity to the PRD.
Results and Discussion
Synthetic concept: Our synthetic plan for 2,5-disubstituted
pyrrolidine building blocks of type is depicted in
[12]
7
Scheme 2. It is based on the idea to generate the vinyl
group from a cyano substituent. The required (N-protected)
precursor of type 8 could eventually be prepared by (cis-se-
[14]
lective)
cyanation of a 5-methoxy-proline derivative (9)
which in turn should be accessible from l-proline (10)
[15]
through diprotection and electrochemical oxidation.
Figure 3. SH3-binding peptide ligand bearing the scaffold X in the core-
motif.
Scheme 2. Retrosynthetic analysis of building blocks of type 7 (related to
).
[12]
2
In our first-generation synthesis of the scaffold Fmoc-X
Scheme 1) we had exploited a ring-closing metathesis reac-
(
tion (RCM) to form the seven-membered ring. The required
dipeptide intermediate 4 was assembled by coupling of the
two vinylproline derivatives 2 and 3. Actually, the by far
most demanding part of the whole synthesis was the stereo-
selective preparation of the two building blocks (2 and 3),
which was accomplished in nine and eleven steps, respec-
Our plan for the synthesis of the second required building
block, that is, of type 11, is outlined in Scheme 3. As a key
consideration we envisioned introducing the vinyl group
I
through Cu -catalyzed 1,4-addition to a N-protected dehy-
[16]
droproline ester of type 12. This type of intermediate in
turn could be prepared from l-proline by protection and ox-
[13]
tively, starting from l-pyroglutamic acid.
[12]
[17]
After preliminary investigations had revealed promising
biological properties of PRM-related peptides containing
the diproline analogue X, we felt challenged to develop a
more efficient and scalable access to the required vinylpro-
line building blocks. We here disclose the results of a study,
which indeed has resulted in a greatly improved synthesis of
Fmoc-X on a gram scale.
idation. Our concept included the possibility to employ a
building block of type 11 eventually as a racemic mixture
because the emerging diastereoisomers formed after cou-
Scheme 3. Retrosynthetic analysis of building blocks of type 11 (related
to 3).
12038
www.chemeurj.org
ꢂ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 12037 – 12044

Pyrrolidine Chemistry
FULL PAPER
pling with the second (enantiopure) building block of type 7
could possibly be separated by means of chromatography.
and we therefore decided to probe the electrochemistry-
based strategy (in analogy to Scheme 4) for the direct syn-
thesis of building block 2. Noteworthy, related anodic oxida-
tions have only been described for methyl, ethyl and benzyl
esters while, to the best of our knowledge, tert-butyl esters
Synthesis of 5-vinylproline derivatives of type 7: We started
our synthesis by treating l-proline (>99% ee) with an
excess (3 equiv) of methyl chloroformate in methanol in the
[15,22]
have never been employed.
presence of NEt (3.5 equiv) to obtain the diprotected deriv-
The required substrate, that is, the N-Boc-proline-tert-bu-
tylester (19) was prepared from l-proline using our protocol
which exploits the solubilization of proline as a triethylam-
monium salt to allow the N-Boc protection and the subse-
quent DMAP-promoted esterification being performed in a
3
ative 13 in 88% yield. The crude (NMR pure) product was
[15]
then subjected to anodic oxidation. The reaction was per-
formed in a 0.05m solution of NBu BF in MeOH using
4
4
graphite plate electrodes (48ꢃ28 mm, 600 mA, 5 mm, 08C)
to give 14 as a 1:1 mixture of diastereomers in virtually
quantitative yield (Scheme 4). Noteworthy, this remarkably
[23]
one-pot fashion (Scheme 5).
Best results were obtained
using an excess (2.7 equiv) of Boc O in tert-BuOH as a sol-
2
vent. After heating the wet reaction mixture (to destroy
[23b]
excess reagent)
tive yield.
the product 19 was obtained in quantita-
Scheme 5. One-pot synthesis of N-Boc-proline-tert-butyl ester (19) and its
unexpected conversion to 20 under electrochemical conditions: a) Boc
2.7 equiv), NEt (0.3 equiv), 4 h, RT, then DMAP (0.3 equiv), tert-
2
O
(
3
À
BuOH, 15 h, RT, then 508C 1 h; b) À2e (300 mA, 13 mm), MeOH,
Scheme 4. Synthesis of 18 as a mixture of cis/trans isomers: a) ClCO
NEt NBF , 08C;
, MeOH, RT, 15 h; b) À2e (600 mA, 5 mm), MeOH, Bu
c) TMSCN, SnCl , CH Cl H/H O, 75 8C,
, À408C, 1.5 h; d) Ra/Ni, HCO
h; e) Nysted reagent, BF ·Et O, THF, 08C, RT, 2 h; f) TMSI, CHCl
reflux, 3 h.
2
Me,
Bu NBF , 08C.
4
4
3
4
4
4
2
2
2
2
5
3
2
3
,
When we then subjected 19 to similar electrochemical
conditions as used before for the oxidation of 13 to 14
(
graphite plate electrodes, 48ꢃ28 mm, 300 mA, 13 mm) we
efficient transformation could easily be carried out on an
were much surprised to isolate the N-Moc-protected dimeth-
yl glutamate 20 as the only product in 88% yield after pu-
rification (Scheme 5). Obviously, an overoxidation at C5
had taken place followed by opening of the pyrolidine ring
8
0 mmol scale in an open 100 mL beaker within 10 h
À1
(
2.80 Fmol ). Without purification, the crude N,O-acetal 14
GC purity ꢀ95%) was reacted with TMSCN in the pres-
(
[24]
ence of SnCl in CH Cl to afford the cyanide 15 in 86%
and transesterification.
Nevertheless, by using a lower current and a larger dis-
4
2
2
yield as a mixture of diastereomers (1:1). Noteworthy, no
significant excess of one diastereomer was observed in this
[25]
tance between the electrodes (48ꢃ28 mm, 260 mA, 5 cm)
[14]
case. The planned conversion of the nitrile function to an
aldehyde was then achieved through reduction/hydrolysis
we succeeded in obtaining the desired methoxylated product
21 in good yields. The crude product was contaminated only
with traces of the starting material 19 and could directly be
used for the subsequent cyanation step (Scheme 6).
using commercial Raney nickel (50% slurry in H O) in
2
8
8% aqueous formic acid to give the aldehyde 16 in 63%
[18]
yield (as a 1:1 mixture of diastereomers). While our at-
tempts to convert 16 into the vinylproline derivative 17
through Wittig olefination (using H C=PPh under different
2
3
[19]
conditions)
only resulted in mediocre yields of about
4
0%, the desired transformation was achieved with Nysted
[20]
reagent in 86% yield. After cleavage of the methyl carba-
[21]
mate (N-deprotection) with TMSI in CHCl3 the building
block 18 was obtained as a 1:1 mixture of cis- and trans-dia-
stereomers, which could not be separated by column chro-
matography.
[12]
In the course of our previous work we had found that
the diastereomers of the tert-butyl ester 2 (see Scheme 1)
are readily separable by chromatography. Thus, in principle
we could have converted cis/trans-18 into the corresponding
tert-butyl ester (2) followed by diastereomer separation.
However, this would have required three additional steps
À
Scheme 6. Improved synthesis 2: a) Bu
60 mA, 5 cm; b) TMSCN, TMSOTf, CH Cl
Py/AcOH/H O 2:1:1, 808C, 72 h; d) Ph PMe Br , KHMDS, THF, RT,
2.5 h; e) TMSOTf, CH Cl , 08C, 5 min.
4
NBF
4
,
À2e , MeOH, 08C,
2
2
2
, À408C, 1.5 h; c) Ra/Ni, H
2
,
+
À
2
3
2
2
Chem. Eur. J. 2011, 17, 12037 – 12044
ꢂ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
12039

H.-G. Schmalz et al.
Noteworthy, the expected cyanation could not be ach-
ieved in this case through Wittig olefination with Ph P=
3
[19b]
ieved under the conditions (TMSCN, SnCl ) used before for
CH2.
This way, the vinylproline 24 was obtained in 88%
4
the cyanation of 14 (Scheme 4) due to decomposition. How-
yield, still as cis/trans mixture (1.7:1). After N-deprotection
[12]
ever, when we replaced SnCl (as a rather aggressive Lewis
(Boc cleavage) with TMSOTf the diastereomers could be
separated to give 49% of the desired cis-building block 2
along with 33% of its trans-epimer (epi-2). The relative
(and absolute) configuration of both products were proven
by X-ray crystal structure analysis of the hydro-triflate of 2
and the hydro-chloride of epi-2 (Figure 5).
4
acid) by TMSOTf the cyanated product 22 was obtained as
[26]
a cis/trans mixture (2.9:1) in 81% overall yield from 19.
After separation by column chromatography the relative
configuration of both diastereomers could be unambiguous-
ly determined by X-ray crystal structure analysis. Interest-
ingly, both cis-22 and trans-22 take a conformation in which
the cyano substituent adopts
Figure 4). This may reflect a stereoelectronic preference re-
sulting from interaction of the ring nitrogen lone pair with
a pseudo-axial position
(
[27]
the s*-orbital of the adjacent CÀCN bond.
Figure 4. Top: Structure of cis-22 and trans-22 in the crystalline state;
bottom: stereodrawings of cis-22 and trans-22 emphasizing the envelope
conformation of the pyrrolidine rings with the nitrile moieties in a
Figure 5. Structure of the hydro-triflate salt of 2 (top) and of the hydro-
chloride salt of epi-2 (bottom) in the crystalline state.
“
pseudo-axial” position.
Synthesis of 3-vinylproline derivatives of type 11: Having
successfully elaborated a reliable gram-scale synthesis of 2,
the next goal was to develop a short and operationally con-
venient protocol for the synthesis of a trans-3-vinylproline
building block of type 11. Following the concept outlined in
Having succeeded in preparing nitrile 22 in a straightfor-
ward fashion (Scheme 6) the next challenge was to convert
the cyano function into an aldehyde group. Preliminary ex-
periments had shown that the tert-butyl esters will neither
survive the Raney-nickel reduction in 88% aqueous formic
Scheme 3, which is based on a Cu-catalyzed 1,4-addition to
[18]
[16]
acid (also acting as a hydrogen source) nor allow a che-
a 2,3-dehydroproline derivative,
we first focused on the
[23b]
moselective DIBAH reduction
of the nitrile function.
synthesis of the racemic building block rac-29.
Thus, we probed the Raney-nickel reduction in a slightly
basic solvent system (pyridine/acetic acid/water 2:1:1)
The synthesis of rac-29 (Scheme 7) started with the
[28]
[29]
esterification of l-proline (with SOCl2 in MeOH)
fol-
using hydrogen gas as a stoichiometric reductant. First ex-
periments using the separated diastereomers of 22 revealed
that the less hindered trans-isomer reacts much faster and
that the transformation is not stereospecific, as the resulting
aldehyde 23 epimerizes under the basic conditions. There-
fore, on a preparative scale (Scheme 6), the nitrile 22 was
employed as a cis/trans mixture to afford (after a reaction
time of 72 h) the aldehyde 23 in 51% isolated yield (cis/
trans 1.7:1) besides small amounts (8%) of remaining cis-22,
lowed by a one-pot oxidation/protection sequence devel-
[16a]
oped in our lab
based on a two step protocol of Hꢄusler
[17]
et al. In the first step 25 is oxidized with NCS (instead of
[30]
tert-BuOCl)
to the corresponding N-chloro amine from
which the imine 26 is formed by HCl elimination. Without
isolation, 26 is then treated with methyl chloroformate in
the presence of pyridine to give the dehydroproline deriva-
tive 27 in 90% overall yield. The key 1,4-addition was then
achieved in good yield by very slow addition of a THF solu-
tion of 27 at À408C to a stirred mixture prepared from vinyl
separated off by column chromatography. While the BF -
3
[20]
promoted methylenation of 23 with Nysted reagent gave
the N-Boc-protected vinylproline 24 in varying yields (24 to
magnesium bromide (15 wt% in THF) and CuBr·SMe
2
[31]
(0.3 equiv) in THF. After quenching the reaction with a
concentrated aqueous NH /NH Cl solution, the vinylproline
67%), the desired transformation could be efficiently ach-
3
4
12040
www.chemeurj.org
ꢂ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 12037 – 12044

Pyrrolidine Chemistry
FULL PAPER
Scheme 7. Synthesis of rac-29: a) SOCl
2
, MeOH, RT, 1 d; b) NCS, NEt
3
,
CH
30 mol%), vinylMgCl, THF, À308C; e) KOH, H
reflux.
2
Cl
2
, RT, 3 h; c) ClCO
2
Me, pyridine, À208C, RT, 15 h; d) CuBr·SMe
2
(
2
O/tBuOH 1:8, 3 h,
derivative rac-28 was isolated as a 8.6:1 (trans/cis) mixture
of diastereomers. Hydrolysis of the ester function with KOH
Scheme 8. Attempted synthesis of 6: a) PyBOP, DIPEA, MeCN, 08C,
RT, overnight; b) Grubbsꢅ II, CH Cl , reflux, 24 h; c) TMSI (2 equiv),
2 2
in H O/tBuOH then gave the N-methoxycarbonyl- (Moc)
CHCl , reflux, 3 h; d) Fmoc-Cl, NaHCO , H O/THF 2.5:1, 08C, RT, 15 h;
2
3
3
2
protected building block rac-29 in a diastereomeric ratio of
3 2
e) Alloc-Cl, NaHCO , H O/THF 2.5:1, RT, 15 h.
12:1 (trans/cis). The increase of the diastereomeric purity
can be attributed to a slower saponification rate of the steri-
[
32]
cally more demanding cis-isomer. The relative configura-
tion of the main diastereomer (rac-29) was initially assigned
by NOE experiments
ly, we did not succeed in separating the diastereomers at
any stage of the sequence shown in Scheme 8.
[16a]
and confirmed by comparison of
An attempt to convert the mixture (31/dia-31) into the
Fmoc-protected target compound (6/dia-6) under similar
conditions used for the preparation of 32/dia-32 failed
(Scheme 8). This failure was somewhat surprising because
we had probed this type of protecting group manipulation
successfully in the conversion of 33 to the N-Fmoc-protected
proline derivative 34 (Scheme 9).
the NMR data with those of rac-3, the configuration of
which was proven by X-ray crystal structure analysis (see
below).
Synthesis of the tricyclic diproline mimetics: With the build-
ing blocks 2 and rac-29 (instead of 3; compare Scheme 1) in
our hands, we considered it worthwhile trying to conclude
the synthesis of the target structures according to the
proven strategy. The hope was that we could separate the
diastereomeric coupling products and exchange the Moc
protecting group by an Fmoc group at a later stage of the
[12]
synthesis. As expected, the PyBOP-mediated coupling of
with rac-29 (d.r. 12:1) afforded a (ca. 1:1) mixture of the
Scheme 9. Conversion of 33 to 34: a) TMSI, CHCl
3
, reflux, 3 h, then
2
Fmoc-Cl, NaHCO , H O/THF 2.5:1, 08C, RT, overnight.
3
2
diastereomeric dipeptides 30 and dia-30 (Scheme 8). While
the yield was only 50% it is noteworthy that no impurities
resulting from the cis-diastereomer of rac-29 could be de-
tected. This “self-purification” certainly results from the ki-
netic torpidness of the cis-epimers (see above). Unfortunate-
ly, the diastereomers (30/dia-30) could not be separated. We
To circumvent the difficulties associated with the diaste-
reomer separation and the final Fmoc protection (required
for solid phase peptide synthesis) we decided to return to
the Boc-protected building block rac-3. The experience
[12,33]
[12]
thus decided to perform the RCM
first and to separate
gained in the course of our previous study (Scheme 1)
the diastereomers at a later stage. Treatment of the divinyl-
suggested that the N-Boc-/CO -tert-butyl combination of
2
substrate (30/dia-30 1:1) with 5 mol% of the Grubbs II cata-
protecting groups should be most promising with respect to
the diastereomer separation problem. Also, the installation
of the Fmoc-protecting group could be achieved using the
existing protocol (5 to 6).
A first attempt to prepare the methyl ester of rac-3 direct-
ly through Cu-catalyzed 1,4-addition was hampered by the
[34]
lyst in CH Cl at 708C under microwave conditions
af-
2
2
forded a 6:1 mixture of the tricyclic products 31 and dia-31
in 46% yield. The pronounced resolution of diastereomers
associated to this transformation reflects a higher tendency
of formation of the product 31 with the desired relative con-
figuration. After cleaving off both protecting groups with
low reactivity of the imine intermediate 26 towards Boc O,
2
[21]
TMSI in refluxing CHCl3 the resulting crude product mix-
ture was directly treated with Alloc-Cl in the presence of
which is much less reactive as compared to the related chlor-
[35]
oformates. As the Boc analogue of 27 was thus not acces-
sible, we decided to take the 1,4-addition product rac-28
(d.r. 8.6:1) and convert it into rac-3 by means of “global de-
aqueous NaHCO to give the N-Alloc-protected diproline
3
analogue (32/dia-32) in 72% yield (Scheme 8). Unfortunate-
Chem. Eur. J. 2011, 17, 12037 – 12044
ꢂ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
12041

H.-G. Schmalz et al.
protection” followed by N-re-protection with Boc O. The
2
elaborated sequence (Scheme 10) starts with the TMSI-
mediated selective cleavage of the methyl carbamate moiety
Scheme 10. Conversion of rac-28 into rac-3: a) TMSI (2 equiv), CH
reflux, 3 h, b) LiOH, H O/MeOH/THF 1:1:3, RT, 15 h, c) AcOH, Boc
d.
2
Cl
2
,
2
2
O,
3
of rac-28. The ester function of the resulting hydroiodide
rac-35 is then hydrolyzed by treatment with LiOH in a
Scheme 11. Synthesis of 6: a) PyBOP, DIPEA, MeCN, 08C, RT, 15 h, dia-
stereomer separation by flash chromatography; b) Grubbs II (10 mol%),
CH
2
Cl
2
,
reflux, 24 h; c) TFA, CH
2 2
Cl , 08C, RT, 1 h; d) Fmoc-Cl,
MeOH/H O/THF solvent mixture, and the carboxylate in-
2
NaHCO
3
, H O/THF (2:1), 08C, RT, 15 h.
2
termediate rac-36 is directly further converted to rac-3 by
treatment with Boc O. Remarkably, the success of this one-
2
pot protocol for ester cleavage and Boc-protection depends
on a seemingly small detail: It proved to be essential to add
some AcOH to the solution of rac-36 prior to the addition
crystal X-ray structure analysis. Indeed, the crystal structure
of 5 (Figure 7) clearly shows the predicted PPII helix-type
conformation (F=À52.48, Y=172.18).
of Boc O because the formation of a LiOAc-buffered solu-
2
tion avoids the hydrolysis of Boc O. Under the optimized
2
conditions the N-Boc-protected vinylproline rac-3 was ob-
tained in 83% yield (d.r. 15:1) after recrystallization.
A nearly diastereomerically
pure sample of rac-3 (d.r.ꢀ
97:1) was obtained by very slow
recrystallization allowing us to
prove the relative (trans) con-
figuration by X-ray crystallog-
raphy (Figure 6).
Having thus elaborated a re-
liable and efficient access to the
Boc-protected building block
rac-3 (Scheme 7 and 10) the
way was paved to tackle the
final stages of the projected
synthesis. Thus, PyBOP-mediat-
ed peptide coupling of rac-3
with the second (enantiomeri-
cally pure) building block 2 af-
forded two diastereomers (4
and dia-4), which fortunately
Figure 7. Structure of 5 in the crystalline state.
Conclusion
Figure 6. Structure of rac-3 in
the crystalline state.
Starting from l-proline, we have elaborated new and effi-
cient routes to the vinylproline building blocks 2 and rac-3,
which were then fused and converted into the conformation-
ally locked diproline mimetic 6 through peptide coupling,
ring-closing metathesis and protecting group adjustment. In
comparison to our original approach towards 6 the new
synthesis is significantly shorter, more convenient and gives
access to 6 on a multi-gram scale. Of course, an obvious pos-
sibility for further (future) improvement would result from
employing the building block 3 in non-racemic form. This
might be achieved either through resolution of rac-3 (or rac-
28) or by performing the Cu-catalyzed 1,4-addition in an
enantioselective manner.
could be easily separated by flash chromatography
Scheme 11). The desired diastereomer 4, obtained in 42%
(
yield, was then subjected to RCM (using 10 mol% of
Grubbs II catalyst) to give the tricyclic scaffold 5 in 91%
yield. Finally, deprotection of both the tert-butyl ester and
the N-Boc group with TFA and subsequent Fmoc re-protec-
tion of the amino function afforded the Pro–Pro mimetic 6
in 86% yield. We were pleased to succeed in crystallizing a
sample of the pre-target compound 5. This allowed us to un-
equivocally confirm not only the configurational assign-
ments but also the conformational details by means of single
[12]
12042
www.chemeurj.org
ꢂ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 12037 – 12044

Pyrrolidine Chemistry
FULL PAPER
Nevertheless, the sequence developed has already allowed
us to prepare substantial amounts of 6 required for the syn-
thesis of peptide-analogue ligands specifically interfering
with protein-protein interactions based on the recognition of
proline-rich motifs in a PPII helix conformation. The results
of these ongoing biological studies will be published sepa-
rately.
Linding, B. Mayer, M. Nagai, M. Sudol, U. Walter, S. Winder, FEBS
Lett. 2002, 513, 141–144; c) L. Ball, T. Jarchau, H. Oschkinat, U.
Walter, FEBS Lett. 2002, 513, 45–52.
[
6] I. Geisler, J. Chmielewski, Bioorg. Med. Chem. Lett. 2007, 17, 2765–
2
768.
[7] a) M. Dustin, M. Olszowy, A. Holdorf, J. Bromley, N. Desai, P.
Widder, F. Rosenberger, P. van der Merwe, P. Allen, A. Shaw, Cell
1
998, 94, 667–677; b) V. Laurent, T. Loisel, B. Harbeck, A.
Wehman, L. Grçbe, B. Jockusch, J. Wehland, F. Gertler, M. Carlier,
J. Cell. Biochem. 1999, 72-75, 1245–1258.
8] A. Goldstrohm, T. Albrecht, C. SuÇꢇ, M. Bedford, M. Garcia-
Blanco, Mol. Cell. Biol. 2001, 21, 7617–7628.
Finally, we would like to mention that the investigation
disclosed here has also opened insights into new structural
and reactivity aspects of substituted pyrrolidines. For in-
[
[36]
stance, various X-ray crystal structure analyses
have re-
[9] a) B. Laggerbauer, S. Liu, E. Makarov, H.-P. Vornlocher, O. Makar-
ova, D. Ingelfinger, T. Achsel, R. Lꢁhrmann, RNA 2005, 11, 598–
vealed interesting conformational preferences of the pyrroli-
dine intermediates. Moreover, the identification of subtle
dependencies of the chemical reactivity of the compounds
on the protecting groups used may allow a more precise
planning of syntheses in the future. Thus, we are optimistic
that the developed protocols will find application in the pro-
spective synthesis of biologically relevant pyrrolidine deriva-
tives, including conformationally restricted peptidomimetics.
In this context, the use of vinyl-substituted proline deriva-
tives represents a valuable addition to the molecular con-
struction kit based on the combination of peptide coupling
6
08; b) M. Kofler, M. Schuemann, C. Merz, D. Kosslick, A.
Schlundt, A. Tannert, M. Schaefer, R. Luhrmann, E. Krause, C.
Freund, Mol. Cell. Proteomics 2009, 8, 2461–2473.
[
10] a) C. M. Deber, F. A. Bovey, J. P. Carver, E. R. Blout, J. Am. Chem.
Soc. 1970, 92, 6191–6198; b) N. Helbecque, M. H. Loucheux-Lefeb-
vre, Int. J. Peptide Protein Res. 1982, 19, 94–101; c) H. Okabayashi,
T. Isemura, S. Sakakibara, Biopolymers 1968, 6, 323–330; d) R. K.
Dukor, T. A. Kiederling, Biopolymers 1991, 31, 1747–1761; e) R. K.
Dukor, T. A. Kiederling, V. Gut, Int. J. Peptide Protein Res. 1991, 38,
1
98–203; f) P. M. Cowan, S. McGavin, Nature 1955, 176, 501–503.
[11] R. Kꢁhne, H. Oschkinat, C. Brockmann, H.-G. Schmalz, GER, WO
008/040332A1, 2008.
2
[37]
[12] J. Zaminer, C. Brockmann, P. Huy, R. Opitz, C. Reuter, M. Beyer-
mann, C. Freund, M. Mꢁller, H. Oschkinat, R. Kꢁhne, H.-G.
Schmalz, Angew. Chem. 2010, 122, 7265–7269; Angew. Chem. Int.
Ed. 2010, 49, 7111–7115.
and ring closing metathesis.
[
13] Compare also: a) C. Herdeis, H. P. Hubmann, Tetrahedron: Asym-
metry 1992, 3, 1213–1221; b) C. Herdeis, H. P. Hubmann, H. Latter,
Tetrahedron: Asymmetry 1994, 5, 351–354; c) N. Langlois, R. Z. An-
driamialisoa, Tetrahedron Lett. 1991, 32, 3057–3058; d) C. Acevedo,
E. Kogut, M. A. Lipton, Tetrahedron Lett. 2001, 57, 6353–6359; e) P.
Somfai, H. Ming He, D. Tanner, Tetrahedron Lett. 1991, 32, 283–
Acknowledgements
This work was supported by the Deutsche Forschungsgemeinschaft
DFG FOR 806) and the Fonds der Chemischen Industrie (PhD fellow-
ship to P.H.). The authors also gratefully acknowledge Chemetall GmbH
and Evonik for valuable gifts of chemicals.
(
2
86; f) O. Hara, K. Sugimoto, K. Maikno, Y. Hamada, Synlett 2004,
1
625–1627.
[
[
14] N. Langlois, A. Rojas, Tetrahedron 1993, 49, 77–82.
15] a) T. Shono, H. Hamaguchi, Y. Matsumura, J. Am. Chem. Soc. 1975,
[
[
[
1] For selected reviews, see: a) O. Keskin, A. Gursoy, B. Ma, R. Nussi-
nov, Chem. Rev. 2008, 108, 1225–1244; b) D. Gonzꢆlez-Ruiz, H.
Gohlke, Curr. Med. Chem. 2006, 13, 2607–2625; c) Y. Pommier, J
Cherfils, Trends Pharmacol. Sci. 2005, 26, 138–145; d) A. Reayi, P.
Arya, Curr. Opin. Chem. Biol. 2005, 9, 240–247; e) T. Berg, Angew.
Chem. 2003, 115, 2566–2586; Angew. Chem. Int. Ed. 2003, 42, 2462–
9
7, 4264–4268; b) T. Shono, Tetrahedron 1984, 40, 811–850; c) M.
Skrinjar, L.-G. Wistrand, Tetrahedron 1991, 47, 573–582.
[
16] a) P. Huy, J.-M. Neudçrfl, H.-G. Schmalz, Org. Lett. 2011, 13, 216–
2
19; see also: b) J. Ezquerra, A. Escribano, A. Rubio, M. J. Re-
muiÇꢆn, J. J. Vaquero, Tetrahedron: Asymmetry 1996, 7, 2613–2626;
c) N. Toyooka, M. Okumura, T. Himiyama, A. Nakazawa, H.
Nemoto, Synlett 2003, 55–58.
2
481.
2] For selected recent reviews, see: a) M. J. Pꢇrez de Vega, M. Martꢈn-
Martꢈnez, R. Gonzꢆlez-MuÇiz, Curr. Top. Med. Chem. 2007, 7, 33–
[
17] a) H. Poisel, U. Schmidt, Chem. Ber. 1975, 108, 2547–2553; b) U.
Schmidt, J. Hꢄusler, Liebigs Ann. Chem. 1979, 1881–1889; c) J.
Hꢄusler, Liebigs Ann. Chem. 1981, 1073–1088.
62; b) J. Vagner, H. Qu, V. J. Hruby, Curr. Opin. Chem. Biol. 2008,
12, 292–296; c) A. Grauer, B. Kçnig, Eur. J. Org. Chem. 2009,
5099–5111; d) J. M. Mason, Future Med. Chem. 2010, 2, 1813–1822;
[
[
18] N. Moss, J.-M. Ferland, S. Goulet, I. Guse, E. Malenfant, L. Plamon-
don, R. Plante, R. Dꢇziel, Synthesis 1997, 32–34.
19] a) G. A. Molander, J. A. C. Romero, Tetrahedron 2005, 61, 2631–
see also: e) N. Srinivas, K. Moehle, K. Abou-Hadeed, D. Obrecht,
J. A. Robinson, Org. Biomol. Chem. 2007, 5, 3100–3105.
3] Selected references: a) R. E. Moellering, M. Cornejo, T. N. Davis, C.
Del Bianco, J. C. Aster, S. C. Blacklow, A. L. Kung, D. G. Gilliland,
G. L. Verdine, J. E. Bradner, Nature 2009, 462, 182–190; b) E. Ko, J.
Liu, L. M. Perez, G. Lu, A. Schaefer, K. Burgess, J. Am. Chem. Soc.
2
643; b) S. Scapecchi, R. Matucci, C. Bellucci, M. Buccioni, S. Dei,
L. Guandalini, C. Martelli, D. Manetti, E. Martini, G. Marucci, M.
Nesi, M. N. Romanelli, E. Teodori, F. Gualtieri, J. Med. Chem. 2006,
4
9, 1925–1931; c) T. Sato, K. Tsujimoto, K. Matsubayashi, K. Ishiba-
2
011, 133, 462–477; c) J. H. Lee, Q. Zhang, S. Jo, S. C. Chai, M. Oh,
shi, M. Ikeda, Chem. Pharm. Bull. 1992, 40, 2308–2312; d) M. Ari-
sawa, M. Takahashi, E. Takezawa, T. Yamaguchi, Y. Torisawa, A.
Nishida, M. Nakagawa, Chem. Pharm. Bull. 2000, 48, 1593–1596.
20] a) Nysted reagent (L. N. Nysted, US Patent 3865848, 1975; [Chem.
Abstr. 1975, 83, 10406q]) is commercially available from Aldrich; for
a reliable protocol, see: b) S. Matsubara, M. Sugihara, K. Utimoto,
Synlett 1998, 313–315.
[21] a) M. E. Jung, M. A. Lyster, J. Chem. Soc. Chem. Comm. 1978, 315–
316; b) R. S. Lott, V. S. Chauhan, C. H. Stammer, J. Chem. Soc.
Chem. Commun. 1979, 495–496; c) F. Manfrꢇ, J. M. Kern, J. F. Biell-
mann, J. Org. Chem. 1992, 57, 2060–2065.
W. Im, H. Lu, H.-S. Lim, J. Am. Chem. Soc. 2011, 133, 676–679.
4] a) L. Ball, R. Kꢁhne, J. Schneider-Mergener, H. Oschkinat, Angew.
Chem. 2005, 117, 2912–2930; Angew. Chem. Int. Ed. 2005, 44, 2852–
[
[
2
869; b) Protein–Protein Interactions as New Drug Targets, Hand-
book of Experimental Pharmacology, Vol. 186 (Eds.: E. Klussmann,
J. Scott), Springer, Heidelberg, 2008.
[
5] a) L. Ball, R. Kꢁhne, B. Hoffmann, A. Hꢄfner, P. Schmieder, R.
Volkmer-Engert, M. Hof, M. Wahl, J. Schneider-Mergener, U.
Walter, H. Oschkinat, T. Jarchau, EMBO J. 2000, 19, 4903–4914;
b) R. Aasland, C. Abrams, C. Ampe, L. Ball, M. Bedford, G. Cesar-
eni, M. Gimona, J. Hurley, T. Jarchau, V. Lehto, M. Lemmon, R.
Chem. Eur. J. 2011, 17, 12037 – 12044
ꢂ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
12043

H.-G. Schmalz et al.
[
[
[
22] a) P. G. Kirira, M. Kuriyama, O. Onomura, Chem. Eur. J. 2010, 16,
[29] R. G. Webb, M. W. Haskell, C. H. Stammer, J. Org. Chem. 1969, 34,
576–580; b) S. Guttmann, Helv. Chim. Acta 1961, 44, 721–744.
[30] Examples for NCS oxidations of amines to N-chloroamines:
a) R. H. Furneaux, G. Limberg, P. C. Tyler, V. L. Schramm, Tetrahe-
dron 1997, 53, 2915–2930; b) A. Hꢄberli, C. J. Leumann, Org. Lett.
2001, 3, 489–492.
3
5
970–3982; b) A. G. M. Barrett, D. Pilipauskas, J. Org. Chem. 1991,
6, 2787–2800.
23] a) P. Huy, H.-G. Schmalz, Synthesis 2011, 954–960; b) P. G. M. Wuts,
T. W. Greene, Greeneꢀs Protective Groups in Organic Synthesis, 4th
ed., Wiley-Interscience, New York, 2007.
24] To explain the formation of 20 under the electrochemical conditions,
we suggest the following mechanism: over-oxidation of the primary
product 21 would lead to the formation of a well-stabilized cation A
from which an oxazetidinone intermediate (B) may be formed
under cleavage of the tert-butyl group. The fragmentation of B
[
31] G. Quinkert, H.-G. Schmalz, E. Walzer, S. Gross, T. Kowalczyk-
Przewloka, C. Schierloh, G. Dꢁrner, J. W. Bats, H. Kessler, Liebigs
Ann. Chem. 1988, 283–315.
[
32] a) B. Mauger, F. Irreverre, B. Witkop, J. Am. Chem. Soc. 1966, 88,
2
019–2024; b) R. Sarges, J. R. Tretter, J. Org. Chem. 1974, 39, 1710–
1
716; c) J. Y. L. Chung, J. T. Wasicak, W. A. Arnold, C. S. May,
(either via a zwitterionic species or through direct retro-[2+2] cyclo-
A. M. Nadzan, M. W. Holladay, J. Org. Chem. 1990, 55, 270–275.
33] S. J. Miller, R. H. Grubbs, J. Am. Chem. Soc. 1995, 117, 5855–5856.
34] For reviews on microwave-assisted synthesis, see: a) C. O. Kappe,
Angew. Chem. 2004, 116, 6408–6443; Angew. Chem. Int. Ed. 2004,
addition) would then give rise to an isocyanate C which on metha-
nolysis and neighboring group-assisted transesterification finally
would lead to the isolated product 20.
[
[
4
3, 6250–6284; b) A. de La Hoz, A. Diaz-Ortiz, A. Moreno, Chem.
Soc. Rev. 2005, 34, 164–178; c) Y. Coquerel, J. Rodriguez, Eur. J.
Org. Chem. 2008, 1125–1132.
[
35] K. Hashimoto, H. Shirahama, Tetrahedron Lett. 1991, 32, 2625–
2
628.
36] CCDC-828266 (cis-22), 828265 (trans-22), 828263 (2-HOTf), 828262
epi-2-HCl), 828267 (rac-3), 828264 (5) contain the supplementary
[
(
crystallographic data for this paper. These data can be obtained free
of charge from The Cambridge Crystallographic Data Centre via
www.ccdc.cam.ac.uk/data_request/cif.
[
[
[
25] K. F. McClure, P. Renold, D. S. Kemp, J. Org. Chem. 1995, 60, 454–
4
57.
26] G. Sunilkumar, D. Nagamani, N. P. Argade, K. N. Ganesh, Synthesis
003, 2304–2306.
[
37] a) L. M. Beal, K. D. Moeller, Tetrahedron Lett. 1998, 39, 4639–4642;
b) L. M. Beal, B. Liu, W. Chu, K. D. Moeller, Tetrahedron 2000, 56,
2
27] a) A. J. Kirby, The Anomeric Effect and Related Stereoelectronic Ef-
fects at Oxygen, Springer, Berlin, 1983, pp. 20–23; b) M. Riediker,
W. Graf, Helv. Chim. Acta 1979, 62, 2053–2061.
1
0113–10125; c) J. Einsiedel, H. Lanig, R. Waibel, P. Gmeiner, J.
Org. Chem. 2007, 72, 9102–9113; d) G. Lesma, A. Colombo, A. Sa-
chetti, A. Silvani, Tetrahedron Lett. 2008, 49, 7423–7425.
[
28] a) O. G. Backeberg, B. Staskun, J. Chem. Soc. 1962, 3961–3963;
b) G. Billek, H. Kindl, A. Schimpl, F. P. Schmook, J. Labelled
Compd. 1968, 5, 3–7; c) H. P. Albrecht, D. B. Repke, J. G. Moffatt,
J. Org. Chem. 1973, 38, 1836–1840.
Received: June 3, 2011
Published online: September 7, 2011
12044
www.chemeurj.org
ꢂ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 12037 – 12044