Synthesis of N-alkyl-N-hydroxyguanidines: A comparative study using different protecting group strategies

Article abstract of DOI:10.1055/s-0030-1260100

Representing a prime example of the interesting structures found in natural products, the N-alkyl-N-hydroxyguanidine moiety has so far only found limited attention in synthetic organic chemistry. Our studies on Streptomyces-produced muraymycin nucleoside

Full text of DOI:10.1055/s-0030-1260100

FEATURE ARTICLE
2357
Synthesis of N-Alkyl-N-hydroxyguanidines: A Comparative Study Using
Different Protecting Group Strategies
Synthesis of
N
-
Alk
l
yl-
N
-h
i
ydroxy
v
guanidines er Ries, Anne Ochmann, Christian Ducho*
Georg-August-University Göttingen, Department of Chemistry, Institute of Organic and Biomolecular Chemistry, Tammannstr. 2,
37077 Göttingen, Germany
Fax +49(551)399660; E-mail: cducho@gwdg.de
Received 20 May 2011; revised 6 June 2011
Dedicated to Professor Wittko Francke on the occasion of his 70th birthday
The N-alkyl-N-hydroxyguanidine functionality is a part of
five different types of natural products. However, this
functional group should not be mistaken for the N-alkyl-
N¢-hydroxy moiety, which is formed as an NO-releasing
Abstract: Representing a prime example of the interesting struc-
tures found in natural products, the N-alkyl-N-hydroxyguanidine
moiety has so far only found limited attention in synthetic organic
chemistry. Our studies on Streptomyces-produced muraymycin nu-
cleoside antibiotics have led to a systematic investigation of the intermediate (Nw-hydroxy-L-arginine) in the biosynthesis
of nitrogen monoxide.4 One group of naturally occurring
N-alkyl-N-hydroxyguanidines is constituted by the
synthetic access to the N-hydroxyguanidine-functionalized lipid
structure of biologically potent A-series muraymycins. A protect-
ing-group-free approach has been proven to be superior to a strategy
Actinomyces-produced octacosamicin antibiotics
1
employing highly reactive urethane-protected guanidinylation re-
agents, particularly due to the unexpected instability of fully pro-
tected N-alkyl-N-hydroxyguanidines. The protecting-group-free
method has been used for the synthesis of other N-alkyl-N-hydroxy-
guanidine derivatives, namely the antibiotic N5-hydroxy-L-arginine.
These synthetic results will enable the elucidation of the properties
and biological significance of the N-alkyl-N-hydroxyguanidine
functionality.
(Figure 1).5 Furthermore, the non-proteinogenic amino
acid N5-hydroxy-L-arginine (2) was isolated from a Bacil-
lus sp. and displays moderate antibacterial and antifungal
activity.6a It was demonstrated that 2 inhibited the growth
of E. coli, but that this effect could be easily reversed by
the addition of antagonists of 2 such as cyanocobalamin.6b
The structurally more complex antibiotic miharamycin A
(3) formally is a derivative of 2, but the non-N-hydroxy-
lated congener miharamycin B has also been isolated.7
The siderophore asterobactin represents another antibiotic
bearing 2 as a substructure.8
Key words: natural products, antibiotics, guanidines, guanidinyla-
tion, protecting groups
Introduction
However, the probably most interesting class of N-alkyl-
N-hydroxyguanidine natural products is represented by
the muraymycin nucleoside antibiotics of the A-series
(Figure 1). The term ‘nucleoside antibiotics’ usually re-
fers to a class of natural products bearing often unique nu-
cleoside-derived core structures. They inhibit the bacterial
membrane protein translocase I (MraY), a key enzyme in
the intracellular part of peptidoglycan biosynthesis.9 This
clinically yet unexploited mode of action makes them at-
tractive lead structures for the development of novel anti-
bacterial drugs to overcome the emerging problem of
bacterial resistances to established antibiotics.10 Muray-
mycins are a subclass of nucleoside antibiotics isolated
from a Streptomyces sp.11 The 19 structurally related
members of the muraymycin family share a core structure
comprised of a nucleosyl amino acid unit, which is linked
to a peptide moiety by an alkyl spacer. The main structural
differences are constituted by the presence or absence of
an aminoribose unit (R2 in Figure 1) at the 5¢-hydroxy
group of the nucleoside portion as well as the lipidation
pattern of the 3-hydroxy-L-leucine residue. While the
aminoribose is apparently not essential for the antibacteri-
al activity of muraymycins, the lipid moiety has a crucial
influence on the biological potencies of the antibiotics.11
Muraymycin A1 (4a) was found to be the congener with
the highest bioactivity, which apparently results from the
presence of a C13-lipid unit which is w-functionalized with
Natural products are an apparently unlimited source of in-
triguing chemical structures. Some representative exam-
ples are given by the sponge-produced palau’amine,1a
which was recently synthesized,1b or taxol/paclitaxel,
which was isolated from the bark of the pacific yew tree
Taxus brevifolia2a and is now marketed as an anticancer
drug.2b
The structural complexity of many natural products often
results from oligocyclic systems and the dense presence of
functional groups as well as stereogenic centers. Howev-
er, some natural product structures also display function-
alities which are either rarely used in synthetic chemistry
or are known generally to have limited stability. One strik-
ing example of the latter case can be found in the structure
of the antimalarial endoperoxide artemisinin (‘qinghao-
su’),3 which was isolated from the plant Artemisia annua.
An example of a functional group with rare synthetic pre-
cedent is given by the N-alkyl-N-hydroxyguanidine moi-
ety discussed here.
SYNTHESIS 2011, No. 15, pp 2357–2368
Advanced online publication: 08.07.2011
DOI: 10.1055/s-0030-1260100; Art ID: E53111SS
© Georg Thieme Verlag Stuttgart · New York
x
x.
x
x
.
2
0
1
1

2358
O. Ries et al.
FEATURE ARTICLE
an N-hydroxyguanidine group. Most other muraymycins The investigation of such hypotheses would require the
of the A-series also display this structural feature, with synthesis of N-alkyl-N-hydroxyguanidine derivatives, to
muraymycin A2 (4b) having a shorter lipid chain and mu- be followed by detailed chemical and biological studies.
raymycins A4 and A5 differing from 4a only in the ami- In the course of our ongoing research on muraymycin nu-
noribose unit. The notable exception is muraymycin A3, cleoside antibiotics, we surprisingly realized that the syn-
which is not N-hydroxylated at the guanidine moiety.
thetic methodology for the preparation of terminal N-
alkyl-N-hydroxyguanidines is extremely limited. The
usually applied strategy is the guanidinylation12 of N-
alkylhydroxylamine precursors. Simple N-methyl- and N-
butyl-N-hydroxyguanidines were obtained by guanidiny-
lation of the respective hydroxylamines with S-methyl
isothiouronium sulfate in aqueous solution.13 In the first
total synthesis of N5-hydroxy-L-arginine (2), a similar
procedure was used for the guanidinylation step14a as it
was the case for the synthesis of racemic 2.14b All of these
transformations were performed in aqueous media and
are, therefore, obviously of limited use for the synthesis of
more lipophilic N-alkyl-N-hydroxyguanidines. A nucleo-
side-derived N-aryl-N-hydroxyguanidine was synthesized
by guanidinylation of the N-arylhydroxylamine with 3,5-
dimethylpyrazole-1-carboximidamide nitrate in N,N-di-
methylformamide.15 However, it remains unclear if these
conditions are of more general use for the synthesis of oth-
er N-hydroxyguanidines.
The presence of the N-alkyl-N-hydroxyguanidine motif in
different natural products suggests that there might be a
common mechanism or biological effect mediated by this
functional group. In the case of muraymycins, the most
obvious interpretation of the biological data would be to
postulate that the functionalized lipid unit could somehow
assist in the cell penetration process of the antibiotics as
MraY is an intracellular target. The N-hydroxyguanidine
group might interact with the phosphate head groups of
cell membrane lipids, thus inducing a perturbation of the
membrane structure, which would initiate the membrane
penetration process. On the other hand, it cannot be ruled
out that the N-hydroxyguanidine moiety can act as a metal
chelator and therefore contributes to the inhibition of the
magnesium(II)-dependent MraY enzyme. This specula-
tion is further supported by the fact that the aforemen-
tioned
N-hydroxyguanidine-containing
antibiotic
asterobactin acts as a siderophore.8
Biographical Sketches
Oliver Ries was born in completed his Diploma de- tory continuing his synthetic
1983 and studied chemistry gree with a thesis on the studies on muraymycins
at Georg-August-University synthesis of muraymycin with a particular focus on
Göttingen, Germany, and antibiotics, supervised by the lipid structures of the bi-
Newcastle University, New- Christian Ducho. Since ologically most potent A-
castle upon Tyne, UK, from 2009, he has been a Ph.D. series muraymycins.
2003 to 2008. In 2008, he student in the Ducho labora-
Anne Ochmann was born sis on the synthesis of N- this degree in 2011. She is
in 1986 and started her alkyl-N-hydroxyguanidines,
currently working on a
chemistry studies at Georg- supervised by Christian thesis on oligonucleotide
August-University Göttin- Ducho. She subsequently chemistry as a Master candi-
gen, Germany, in 2006. In joined the chemistry Master date in the Ducho group.
2009, she obtained her course at the same universi-
Bachelor degree with a the- ty and is expected to obtain
Christian Ducho was born moved to the University of tant Professor (‘Juniorpro-
in 1976 and studied chemis- Oxford, UK, for postdoctor- fessor’)
position.
His
try at the University of al research (biosynthesis of current research interests
Hamburg, Germany, from naturally occurring antibiot- are synthetic and biological
1996 to 2001. In 2005, he ics) in the group of Prof. studies on nucleoside antibi-
completed his Ph.D. in or- Christopher J. Schofield otics, the biosynthesis of nu-
ganic chemistry under the from 2005 to 2007. In Au- cleoside antibiotics, and the
supervision of Prof. Chris gust 2007, he commenced synthesis and properties of
Meier at the University of his independent research at modified oligonucleotides.
Hamburg with a thesis on Georg-August-University
pronucleotides. He then Göttingen holding an Assis-
Synthesis 2011, No. 15, 2357–2368 © Thieme Stuttgart · New York

FEATURE ARTICLE
Synthesis of N-Alkyl-N-hydroxyguanidines
2359
H2N
NH
OH
O
H2N
N
N
HO
NH
R
O
OH OH
HO
OH
N
OH
H2N
H
O
OH
O
2
octacosamicin A 1a: R = H
octacosamicin B 1b: R = Me
O
N
OH
NH2
HN
NH
H2N
N
O
OR1
HO
O
O
O
O
NH
COOH
H
H
N
OR2
O
N
NH
N
H
5'
HO
HO
HO
N
H
N
O
O
O
N
N
OH OH
HN
HO
muraymycin
core structure
HN
N
H
NH2
miharamycin A 3
H2N
O
NH
O
muraymycin A1 4a: R1
muraymycin A2 4b: R1
=
=
R2
=
N
NH2
11
OH
OH OMe
H2N
=
O
NH
O
R2
N
NH2
9
OH
OH OMe
Figure 1 Structures of some naturally occurring N-alkyl-N-hydroxyguanidines 1–4
It was therefore decided to choose analogue 5 of the lipid and 6c with reagents 8), and a strategy using urethane pro-
structure of the biologically potent muraymycin A1 (4a) tection of the guanidinylation reagent only (precursor 6a
as the first target structure of this study (Scheme 1). In with reagents 8). The use of protecting groups was expect-
analogy to previous syntheses of N-alkyl-N-hydroxy- ed to have three major advantages: firstly, it should enable
guanidines (vide supra), the guanidinylation of hydroxyl- a guanidinylation reaction in nonaqueous organic media,
amine precursor 6 was chosen to obtain the target which is particularly important with respect to the lipophi-
compound. Different protecting group strategies were en- licity of the target molecule. Secondly, the urethane-
visaged: a protecting-group-free approach using reagent 7 protected triflyl guanidines 8 originally introduced by
with precursor 6a, an approach involving a fully protected Goodman16 are known to be highly reactive guanidinyla-
N-alkyl-N-hydroxyguanidine intermediate (precursors 6b tion reagents. Thirdly, protected N-alkyl-N-hydroxy-
guanidines might be useful intermediates in multistep
O
NH
total syntheses of compounds such as the natural products
1–4. It was planned to identify the most suitable guanidi-
nylation strategy for the synthesis of 5 and then apply it to
the preparation of other N-alkyl-N-hydroxyguanidines,
particularly the antibiotic N5-hydroxy-L-arginine (2).
MeO
N
NH2
OH
5
NH
⋅HCl
N
H2N
N
Results and Discussion
7
O
(with 6a)
Hydroxylamine precursors 6 for the guanidinylation reac-
tion were synthesized first (Scheme 2). Starting from
commercially available erucic acid (9), a sequence con-
sisting of epoxidation, epoxide opening, and esterification
provided dihydroxybehenic acid methyl ester 10 in 99%
yield; periodate cleavage then gave aldehyde 11 in 92%
yield. A similar synthetic route had been reported previ-
ously for the preparation of a fatty acid derived insect
pheromone.17 This synthesis of 11 was superior to a short-
MeO
NH
OR1
+
or
11
NTf
6a: R1 = H
6b: R1 = Bn
6c: R1 = All
R2HN
NHR2
8a: R2 = Cbz
8b: R2 = Alloc
Scheme 1 Strategies for the synthesis of target structure 5, an ana-
logue of the lipid moiety of the biologically potent nucleoside antibio-
tic muraymycin A1 (4a); Alloc = allyloxycarbonyl
Synthesis 2011, No. 15, 2357–2368 © Thieme Stuttgart · New York

2360
O. Ries et al.
FEATURE ARTICLE
er approach employing ozonolysis of erucic acid methyl
ester.18 In our hands, the ozonolysis reaction did not result
in quantitative conversion, thus requiring tedious separa-
tion of the starting material and reaction products. Alde-
hyde 11 was then converted into oximes 12a–c as
mixtures of E- and Z-isomers in yields of 81–88%. Subse-
quent reduction of 12a–c with sodium cyanoborohydride
furnished hydroxylamines 6a–c as guanidinylation pre-
cursors in yields of 76–87%.
NH
⋅HCl
N
H2N
N
7
O
NH
O
Et3N, DMF
45%
MeO
N
NH2
MeO
NH
OH
11
11
OH
5
6a
Scheme 3 Protecting-group-free synthesis of target structure 5
The protecting-group-free guanidinylation approach was
tested first (Scheme 3). Treatment of O-unprotected hy-
droxylamine 6a with the commercially available pyra-
zolide 7 gave the desired target compound 5. The addition
of triethylamine was necessary in order to liberate the re-
active guanidinylation reagent from the hydrochloride 7.
The reaction proceeded smoothly, and the NMR spectra
of the crude product indicated quantitative conversion.
However, the purification of 5 was tedious due to the am-
phiphilic nature of the compound and the presence of both
remaining guanidinylation reagent and triethylammonium
chloride as impurities with similar polarities. Column
chromatography finally provided pure product 5, but the
isolated yield was limited to 45%. Interestingly, treatment
of O-protected precursors 6b and 6c with 7 did not lead to
any detectable transformation into the respective O-pro-
tected N-hydroxyguanidines.
their deprotection at a later stage, was investigated. In or-
der to screen potential conditions for the conversion of
guanidinylation precursors 6b and 6c with Goodman’s
guanidine triflates 8, the model reaction of O-benzyl-N-
butylhydroxylamine (13) with the Cbz-protected triflate
8a to provide the fully protected N-butyl-N-hydroxy-
guanidine 14 was investigated first (Table 1). As
Goodman has always performed guanidinylations with
triflates of type 8 in the presence of base,16 different base
and solvent combinations were tested while the amount of
8a was initially limited to one equivalent (entries 1–9).
Table 1 Reaction of O-Benzyl-N-butylhydroxylamine (13) with
Guanidinylation Reagent 8a
NTf
CbzHN
NHCbz
8a
NCbz
base, solvent
1) H2O2, HCO2H
2) KOH, H2O
3) MeOH, SOCl2
N
NHCbz
NH
O
O
HO
OH
OBn
OBn
14
13
HO
MeO
99%
11
7
11
7
9
10
Entry 8a
Basea
Solvent Temp
(°C)
Time Conv.b
H2N-OR⋅HCl
EtOH, pyridine
3 Å MS
(equiv)
(h)
144
24
24
24
24
24
24
24
24
24
24
24
(%)
NaIO4, CH2Cl2
MeOH, H2O
O
O
1
2
1.0
Et3N
Et3N
TMG
TMG
NaH
CH2Cl2 r.t.
–
MeO
H
92%
11
1.0
THF
THF
DMF
THF
THF
THF
THF
THF
THF
THF
THF
r.t.
r.t.
r.t.
66
29
20
35
15
53
53
66
>98
68
>98
11
3
1.0
OR
H
NaBH3CN
AcCl, MeOH
O
O
N
4
1.0
MeO
NH
OR
MeO
11
11
5
1.0
0
6a: R = H
6b: R = Bn
6c: R = All
76%
87%
77%
12a: R = H
12b: R = Bn
12c: R = All
81%
88%
81%
6
1.0
LiHMDS
LDA
BuLi
–
–78
–78
–78
r.t.
7
1.0
Scheme 2 Synthesis of guanidinylation precursors 6a–c
8
1.0
Due to the purification problems encountered during the
isolation of 5, the replacement of triethylamine with other
bases was examined. A variety of different bases was test-
ed, including metal carbonates and phosphates as well as
organic bases such as polymer-supported DBU, but the
purification of 5 was not significantly facilitated.
9
1.0
10
11
12
1.5
Et3N
TMG
TMG
r.t.
2.0
r.t.
3.0
r.t.
a 1.0 Equiv.
Unprotected hydroxylamine or N-hydroxyguanidine moi-
eties are anticipated to be unsuitable intermediates for the
total synthesis of complex naturally occurring N-alkyl-N-
hydroxyguanidines, e.g. muraymycin nucleoside antibiot-
ics of the A-series. Thus, the feasibility of using protected
N-alkyl-N-hydroxyguanidines in the synthesis, including
b Conversion of 13 into 14 according to the crude 1H NMR spectra.
Goodman had originally reported performing the guanid-
inylation of amines with 8a using triethylamine as the
base in dichloromethane.16a Surprisingly, applying these
Synthesis 2011, No. 15, 2357–2368 © Thieme Stuttgart · New York

FEATURE ARTICLE
Synthesis of N-Alkyl-N-hydroxyguanidines
2361
standard conditions to the aforementioned synthesis of 14
gave no conversion. The choice of solvent played an es-
sential role in this case. When dichloromethane was re-
placed with tetrahydrofuran, 66% conversion of 13 into
14 was observed (Table 1, entries 1 and 2). The use of
N,N,N¢,N¢-tetramethylguanidine (TMG) instead of trieth-
ylamine led to lower conversions both in tetrahydrofuran
and N,N-dimethylformamide (entries 3 and 4). Stronger
bases such as butyllithium were also investigated, but pro-
vided moderate conversions at best (entries 5 to 8). How-
ever, the most surprising finding was that, in contrast to
Goodman’s original report, the presence of a base was
completely unnecessary (entry 9 vs. entry 2). This was
also true for the standard guanidinylation of simple
amines (data not shown), and it was thus concluded that a
base can be generally omitted when guanidine triflates of
type 8 are used for guanidinylations. Quantitative conver-
sions of 13 into 14 were observed when the amount of
guanidinylation reagent was increased. This was the case
for using 1.5 equivalents of 8a in the presence of triethyl-
amine (entry 10). Even when an inferior base such as
N,N,N¢,N¢-tetramethylguanidine was added, raising the
amount of 8a (up to 3 equiv) enabled the completion of the
transformation (entries 11 and 12).
Table 2 Reaction of the O-Benzyl-Protected Hydroxylamine 6b
with Guanidinylation Reagent 8aa
NTf
CbzHN
NHCbz
O
8a
different conditions
O
NCbz
N NHCbz
MeO
NH
MeO
11
11
OBn
OBn
6b
15
Entry
8a (equiv)
2.0
Baseb
Et3N
Et3N
Et3N
–
Time (h)
Conv.c (%)
68
1
2
3
4
24
24
72
72
5.0
>98
2.0
>98
2.0
>98
a All reactions were carried out in THF at r.t.
b 1.0 Equiv.
c Conversion of 6b into 15 according to the crude 1H NMR spectra.
of the guanidinylation reagent 8a (2.0 equiv instead of
1.5), only 68% conversion could be obtained (Table 2, en-
try 1). However, the conversion could be brought to com-
pletion by either raising the amount of 8a up to
5.0 equivalents (entry 2) or by prolonging the reaction
time while using the initially applied amount of 8a (2.0
equiv, entry 3). It was thus concluded that the guanidiny-
lation of N-alkylhydroxylamines may require individual
optimization with respect to different substrates. As ob-
served before, the presence of triethylamine as a base was
not essential for the transformation (entry 4).
The formation of the fully protected N-butyl-N-hydroxy-
guanidine 14 resulting from the transformation of 13
with 8a could be proven from NMR and MS data of the
reaction product. Several attempts were made to purify
product 14. Astonishingly, the compound proved to be
highly sensitive towards both acidic and basic conditions.
Chromatography on silica gel or aluminum oxide led to
decomposition in most cases. No defined decomposition
products could be identified though, thus suggesting that
the degradation of fully protected N-hydroxyguanidines
appears to be of unspecific nature. It may be speculated
that the complete derivatization of the N-hydroxyguani-
dine moiety leads to labilization of the N–O bond, and that
a spontaneous N–O bond rupture induces further decom-
position reactions.
In comparison to the synthesis of the O-benzyl-N-Cbz-
protected derivative 15, achieving quantitative conver-
sions for the corresponding reaction of 6c with triflate 8b
towards the O-allyl-N-Alloc congener 16 was more chal-
lenging (Table 3). In this case, the presence of base had a
detrimental effect on the transformation (cf. entries 1 and
2). Attempts were, therefore, made to drive the conversion
to completion under base-free conditions either by raising
the amount of 8b or by prolonging the reaction time.
When only the equivalents of guanidinylation reagent 8b
were increased, no significant improvement in the conver-
sion was observed (entries 3 to 5). Just extending the re-
action period gave nearly identical results as before (entry
7 vs. entry 2), but the combination of prolonged reaction
time and excess 8b led to quantitative conversion (entry
9). It should be noted that the detrimental effect of base on
the reactions was lower when two equivalents of 8b were
employed (entries 8 and 9 vs. entries 1, 2 and 6, 7). As
only one equivalent of triethylamine was used, it is thus
concluded that the base might lead to partial decomposi-
tion of the guanidinylation reagent. This might hint to-
wards a more pronounced sensitivity of reagent 8b in
comparison to the Cbz-protected congener 8a. Such a pro-
posal is in good agreement with the observation that the
best results were achieved with freshly prepared 8b.
The unexpected sensitivity of fully protected N-hydroxy-
guanidines represents a major drawback for total synthe-
ses as any strategy involving the deprotection of this
functional group at a later stage in the synthetic sequence
is impossible. However, as a point of principle, it would
be interesting to investigate a potential sequence consist-
ing of guanidinylation (with triflates of type 8) and imme-
diate subsequent deprotection. Although the atom
economy of such a protocol would be disadvantageous,
some of the purification problems encountered in the pro-
tecting-group-free approach (vide supra) might be over-
come. It was thus decided to transfer the results obtained
for the guanidinylation of 13 to the corresponding trans-
formation of substrates 6b and 6c, before the deprotection
step is studied.
The reaction of precursor 6b with triflate 8a furnishing 15
was investigated first (Table 2). Interestingly, when the
optimized conditions for the synthesis of 14 (Table 1, en-
try 10) were applied with an even slightly higher amount
Synthesis 2011, No. 15, 2357–2368 © Thieme Stuttgart · New York

2362
O. Ries et al.
FEATURE ARTICLE
ment using 10% palladium on charcoal in ethyl acetate,
only two out of these six potential intermediates were de-
tected in significant amounts: (i) the O-benzyl-N-Cbz-
protected derivative 18 and (ii) the N-Cbz-protected
guanidine 19 (Scheme 4). The composition of the reaction
mixture at different time points was determined by HPLC-
MS. Although the figures obtained do not reflect the exact
composition of the analyte, as this would have required te-
dious and time-consuming calibration procedures for each
compound, they provide a solid estimation of the progres-
sion of the transformation. In the case of 10% palladium
on charcoal in ethyl acetate, the starting material 15 rapid-
ly lost one Cbz group to furnish intermediate 18, while
only minor amounts of 19 were found. However, the slow
formation of 5 coincided with concomitant appearance of
byproduct 17, thus indicating that reduction of the N-hy-
droxyguanidine probably also occurred by N–O bond
cleavage at a later stage of the deprotection sequence.
Consequently, prolonged reaction times caused 17 to be-
come the major product of the reaction, which demon-
strated that it can also be formed by N–O bond rupture
from 5 (Scheme 4).
Table 3 Reaction of the O-Allyl-Protected Hydroxylamine 6c with
Guanidinylation Reagent 8ba
NTf
AllocHN
NHAlloc
O
8b
different conditions
O
NAlloc
N NHAlloc
MeO
NH
MeO
11
11
OAll
OAll
6c
16
Entry
8b (equiv)
1.0
Baseb
Time (h)
24
Conv.c (%)
1
2
3
4
5
6
7
8
9
Et3N
41
82
1.0
–
24
1.5
–
24
81
2.0
–
24
78
5.0
–
24
81
1.0
Et3N
–
72
35
1.0
72
75
2.0
Et3N
–
72
74
Several combinations of heterogeneous catalysts and sol-
vents were screened this way in order to determine both
the optimal reaction times (i.e., when the lowest relative
amount of byproduct 17 was formed) and the ratio of 5 to
17 at this specific time point (Table 4). As anticipated, a
complex interplay of satisfying reactivity and product se-
lectivity was observed. With 10% palladium on charcoal,
the amounts of 17 formed were unacceptably high both in
methanol and ethyl acetate as solvents (entries 1 and 2).
When the load of palladium was reduced to 5%, the selec-
tivity improved slightly, but the conversion was limited in
methanol and failed completely in ethyl acetate (entries 3
and 4). Hence, the support material of the palladium cata-
lyst was changed, and best results were achieved with 5%
palladium on aluminum oxide. In methanol as solvent, the
deprotection reaction proceeded rapidly with only 9% un-
desired byproduct 17 being formed (entry 5). The reactiv-
ity of the catalyst dropped drastically in ethyl acetate
though, accompanied by nearly inverted selectivity (entry
6). Palladium on barium sulfate furnished significant
amounts of 17, while palladium on calcium carbonate
gave no product formation (entries 7 to 10). In summary,
it was proven that selective deprotection of 15 largely
2.0
72
>98
a All reactions were carried out in THF at r.t.
b 1.0 Equiv.
c Conversion of 6c into 16 according to the crude 1H NMR spectra.
The main challenge for the deprotection of fully protected
N-alkyl-N-hydroxyguanidines 15 and 16 was to avoid the
undesired rupture of the N–O bond resulting from reduc-
tive deprotection conditions. The hydrogenolytic depro-
tection of the O-benzyl-N-Cbz-derivative 15 was
investigated first.
As anticipated, the outcome of the deprotection reaction
was highly dependent on the choice of catalyst, solvent,
and reaction time. Hence, the time course for hydrogena-
tion experiments with different catalyst and solvent com-
binations was investigated by HPLC-MS. Two final
products could result from the hydrogenation of 15: the
desired target compound 5 or the undesired N-alkylguani-
dine 17 resulting from reductive N–O bond cleavage
(Scheme 4). In addition, six different intermediates with
only partially cleaved protecting groups en route to 5 or 17
were possible. In a representative hydrogenation experi-
Scheme 4 Representative experiment for the hydrogenolytic deprotection of fully protected N-alkyl-N-hydroxyguanidine 15
Synthesis 2011, No. 15, 2357–2368 © Thieme Stuttgart · New York

FEATURE ARTICLE
Synthesis of N-Alkyl-N-hydroxyguanidines
2363
smoothly, but a pronounced amount (15%) of byproduct
17 was obtained (Table 5, entry 1). However, in the pres-
ence of dimedone, byproduct formation could be signifi-
cantly reduced to 5% (entry 2).
Table 4 Hydrogenolytic Deprotection of Fully Protected N-Alkyl-
N-hydroxyguanidine 15
H2 (1 bar)
catalyst, solvent
O
NCbz
O
NH
MeO
N
NHCbz
MeO
N
X
NH2
11
11
Based on the optimization studies for the deprotection
step, the elucidated deprotection conditions were applied
in a two-step synthesis of 5 from hydroxylamines 6b and
6c on a preparative scale. With respect to the pronounced
instability of fully protected synthetic intermediates 15
and 16, the transformation was carried out in a one-pot
manner. However, although the synthesis of 5 could be
performed this way, the purification of the target com-
pound was not simplified in spite of the absence of a polar
guanidinylation reagent such as 7 (vide supra). The ob-
tained yields were consequently lower than for the pro-
tecting-group-free approach with 34% (from 6b) and 28%
(from 6c) of 5 being isolated after repeated chromato-
graphic purification (Scheme 5). The material obtained
this way was proven to be identical to the compound fur-
nished by the protecting-group-free synthesis by means of
rigorous NMR analysis.
OBn
15
5: X = OH
17: X = H
Entry Catalysta
Solvent
Time
(h)
Ratiob
15/(5 + 17) 5/17
Ratiob
1
2
10% Pd/C
10% Pd/C
5% Pd/C
5% Pd/C
Pd/Al2O3
Pd/Al2O3
Pd/BaSO4
Pd/BaSO4
Pd/CaCO3
Pd/CaCO3
MeOH
24
10
2
16:84
55:45
58:42
78:22
EtOAc
MeOH
EtOAc
MeOH
EtOAc
MeOH
EtOAc
MeOH
EtOAc
traces 15c
46:54
3
d
d
4
–
–
–
5
0.75
0.5
10
0.5
–
traces 15c
93:7
91:9
6
26:74
19:81
78:22
7
traces 15c
98:2
8
1) 8a, THF
d
d
9
–
–
2) H2 (1 bar)
5% Pd/Al2O3
MeOH
d
d
O
O
NH
10
–
–
–
a 5% Pd except where indicated.
MeO
MeO
NH
MeO
MeO
N
NH2
34%
11
11
(one-pot)
b Ratios derived from HPLC-MS analysis of the crude reaction mix-
tures.
OBn
OH
6b
5
c Only traces of remaining starting material 15.
d No formation of either product 5 or 17 detected.
1) 8b, THF
2) [Pd(PPh3)4]
dimedone, THF
O
O
NH
NH
N
NH2
28%
(one-pot)
avoiding unwanted N–O bond rupture was feasible when
the reaction conditions were carefully optimized.
11
11
OAll
OH
6c
5
The selective deprotection of O-allyl-N-Alloc-derivative
16 was then studied. The unwanted reduction of the N–O
bond was also a hurdle for this transformation as the pal-
ladium-catalyzed deprotection of allyl moieties requires a
potentially reducing scavenger as additive. The reaction
therefore needed careful fine tuning, which was based on
HPLC-MS optimization studies as described for 15
(Table 5).
Scheme 5 Synthesis of target structure 5 via a one-pot guanidinyla-
tion–deprotection sequence
Overall, the feasibility of using highly reactive urethane-
protected guanidinylation reagents such as 8a and 8b for
the synthesis of N-alkyl-N-hydroxyguanidines has been
demonstrated. However, isolated yields were not better
than for the protecting-group-free approach, and the unex-
pected instability of fully protected N-alkyl-N-hydroxy-
guanidines limits the use of the two-step strategy for total
syntheses. Consequently, we focused our attention on the
application of the protecting-group-free strategy to the
synthesis of other N-alkyl-N-hydroxyguanidines.
With tetrakis(triphenylphosphine)palladium(0) as catalyst
and phenylsilane as additive, the deprotection proceeded
Table 5 Palladium-Catalyzed Deprotection of Fully Protected N-
Alkyl-N-hydroxyguanidine 16
Pd catalyst
additive, solvent
O
NAlloc
O
NH
The scope of the guanidinylation with pyrazole derivative
7 was tested with further hydroxylamines bearing other
functionalities in closer proximity to the guanidinylation
site. Substrates 20a and 20b bearing phosphonate and sul-
fone groups, respectively, were efficiently converted into
the corresponding N-alkyl-N-hydroxyguanidines 21a,b
with quantitative conversions (Table 6, entries 1 and 2).
Similarly, N-benzylhydroxylamine 20c could be trans-
formed into N-benzyl-N-hydroxyguanidine 21c (entry 3).
MeO
N
NHAlloc
MeO
N
X
NH2
11
11
OAll
16
5: X = OH
17: X = H
Entry Catalyst
Additive
PhSiH3
dimedone THF
Solvent Time (h) Ratioa 5/17
1
2
[Pd(PPh3)4]
[Pd(PPh3)4]
CH2Cl2 16
18
85:15
95:5
a Ratios derived from HPLC-MS analysis of the crude reaction mix-
tures.
Synthesis 2011, No. 15, 2357–2368 © Thieme Stuttgart · New York

2364
O. Ries et al.
FEATURE ARTICLE
Me
Me
Table 6 Guanidinylation of Other Hydroxylamines
O
NH
HO
O
N
O
R
7, Et3N, DMF
R
NH
OH
N
NH2
2 steps
OH
Ot-Bu
Ot-Bu
20a–c
21a–c
BocHN
Boc2N
O
O
Entry
Substrate Product
R
Conv.a (%)
>98
22
23
EtO
H
O
1
2
3
20a
20b
20c
21a
21b
21c
EtO
P
H2N-OR⋅HCl
EtOH, pyridine
3 Å MS
O
DIBAL-H, Et2O
87%
Me
O
>98
>98
S
Ot-Bu
91%
O
Boc2N
O
24
H
N
H
N
OH
O
HO
a Conversion of 20 into 21 according to the crude 1H NMR spectra.
NaBH3CN
AcOH
With respect to the apparent broad applicability of the
protecting-group-free guanidinylation reaction, it was
then envisaged that this method could be used for an effi-
cient and concise total synthesis of the antibiotic N5-hy-
droxy-L-arginine (2). It was planned to obtain the required
hydroxylamine precursor from a protected glutamate
semialdehyde derivative.
Ot-Bu
Ot-Bu
53%
Boc2N
Boc2N
O
25
26
H2N
NH
⋅ x HCl
N
HO
7, Et3N, DMF
Synthetic approaches towards glutamate semialdehyde19
include the selective reduction of glutamate derivatives20
as well as the deprotection of hydrazone21 or dioxolane
precursors.22 Commercially available N-Boc-L-glutamic
acid tert-butyl ester 22 was converted into bis-N-Boc-pro-
tected Weinreb amide 23 according to an established two-
step protocol (Scheme 6).20e Selective reduction of 23
with diisobutylaluminum hydride then furnished protect-
ed glutamate semialdehyde derivative 24 in 87% yield.
The presence of two Boc protecting groups was essential
for the subsequent synthetic steps in order to prevent the
formation of the cyclic hemiaminal or imino forms of the
glutamate semialdehyde.19,20e,23 Aldehyde 24 was trans-
formed into oxime 25, and a mixture of E- and Z-isomers
was obtained in 91% yield. The reduction of the oxime to
the hydroxylamine was hindered by the required acidic re-
action conditions, which could lead to unwanted cleavage
of the protecting groups. However, this problem was cir-
cumvented by using acetic acid as the solvent, but due to
moderate reactivity under these conditions, the yield of
hydroxylamine 26 was limited to 53%. Guanidinylation of
26 with 7 using the aforementioned standard protocol pro-
vided 27, apparently as a mixture of the free base and the
hydrochloride. This mixture was deprotected by treatment
with hydrochloric acid, leading to the isolation of 2 (in its
dihydrochloride form) in 67% yield over two steps from
guanidinylation precursor 26 (Scheme 6).
3 M HCl
2 ⋅ 2 HCl
Ot-Bu
67%
(over 2 steps)
Boc2N
O
27
Scheme 6 Concise total synthesis of N5-hydroxy-L-arginine (2)
using the protecting-group-free guanidinylation method
guanidines via the guanidinylation of suitable hydroxyl-
amine precursors. Exemplified by the synthesis of target
structure 5, a mimic of the lipid moiety of some muraymy-
cin antibiotics, a protecting-group-free one-step proce-
dure could be established for the efficient transformation
of N-alkylhydroxylamines into N-alkyl-N-hydroxy-
guanidines. An alternative two-step one-pot protocol in-
volving highly reactive urethane-protected guanidine
triflates as guanidinylation reagents and fully protected N-
alkyl-N-hydroxyguanidine intermediates was also devel-
oped. However, due to the unexpected instability of the
intermediates and the lability of the N–O bond, careful op-
timization was required to make this procedure practically
applicable. The superior protecting-group-free approach
was applied to an efficient and concise total synthesis of
the antibiotic N5-hydroxy-L-arginine (2). Based on this
synthetic methodology, the stage is now set for the eluci-
dation of the detailed biological role of the N-hydroxy-
guanidine moiety as well as the total synthesis of
structurally complex N-alkyl-N-hydroxyguanidines, e.g.
highly potent muraymycin nucleoside antibiotics of the
A-series.
Conclusion
In summary, we report the first comparative study on the
synthesis of biologically relevant N-alkyl-N-hydroxy-
Synthesis 2011, No. 15, 2357–2368 © Thieme Stuttgart · New York

FEATURE ARTICLE
Synthesis of N-Alkyl-N-hydroxyguanidines
2365
The solvent was removed under reduced pressure, and the resultant
crude product was purified by column chromatography.
Chemicals were purchased from Sigma-Aldrich, Acros, Alfa Aesar,
ABCR, VWR, and TRC. Compounds 8a, 8b, and 23 were prepared
according to established procedures.16,20e The synthesis of com-
pound 24 was performed in analogy to a procedure for the reduction
of the analogous methyl ester.24 The syntheses of compounds 13
(carried out differently than described previously)25 and 20b are de-
scribed in the Supporting Information. Reactions involving oxygen-
and/or moisture-sensitive reagents were carried out under an atmo-
sphere of argon using anhydrous solvents. Anhydrous solvents were
obtained in the following manner: Et2O and THF were dried (Na/
benzophenone) and distilled, MeOH and EtOH were dried (activat-
ed molecular sieves 3 Å) and degassed, DMF and EtOAc were dried
(activated molecular sieves 4 Å) and degassed, CH2Cl2 was dried
(CaCl2) and distilled, and pyridine was dried (CaH2) and distilled.
All other solvents were of technical quality and distilled prior to
their use, and deionized H2O was used throughout. Column chroma-
tography was performed on silica gel 60 (0.040–0.063 mm, 230–
400 mesh ASTM, VWR) except where indicated under flash condi-
tions; PE = petroleum ether 30/75. TLC was performed on alumi-
num plates precoated with silica gel 60 F254 (VWR). Visualization
of the spots was carried out using UV light (254 nm) and/or staining
with heating [H2SO4 staining soln: vanillin (4 g), concd H2SO4
(25 mL), AcOH (80 mL), and MeOH (680 mL); KMnO4 staining
soln: KMnO4 (1 g), K2CO3 (6 g), and 5% aq NaOH (1.5 mL), all
dissolved in H2O (100 mL); ninhydrin staining soln: ninhydrin
(0.3 g), AcOH (3 mL), and BuOH (100 mL)]. HPLC-MS analyses
were performed on a Jasco system equipped with a Rheos 4000
pump, 851-AS Intelligent Sampler, Finnigan Surveyor PDA UV de-
tector, Finnigan Ion-Trap-MS LCQ, and Synergi column
(150 × 2 mm, MAA-RP, 4 mm). The method used was as follows:
0–15 min H2O–MeOH gradient 30–100%, then 15–23 min MeOH
N5-Hydroxy-L-arginine Dihydrochloride (2·2 HCl)
To a soln of 26 (53 mg, 0.13 mmol) and Et3N (50 mL, 0.36 mmol)
in anhyd DMF (0.7 mL), 1H-pyrazole-1-carboxamidine hydrochlo-
ride (7, 27 mg, 0.18 mmol) was added and the mixture was stirred
at r.t. for 19 h. The solvent was removed under reduced pressure,
and the resultant crude product was purified by column chromatog-
raphy (RP silica gel 90 C18, H2O–MeCN gradient 5–100%) to fur-
nish 27. This compound was then dissolved in 3 M aq HCl (1 mL),
and the mixture was stirred under reflux for 4 h. The solvent was re-
moved under reduced pressure to give 2 in its dihydrochloride form
as a colorless foam; yield: 23 mg (67%); mp ~200 °C (dec).
[a]D20 +14.1 (c 0.53, 3 M HCl).
IR (ATR): 2850, 1731, 1498, 1449, 1407, 1367, 1205, 1155, 1122,
816, 760, 518 cm–1.
1H NMR (300 MHz, CD3OD): d = 1.82–2.26 (m, 4 H, 3-H2, 4-H2),
3.77 (t, J = 6.6 Hz, 2 H, 5-H2), 4.21 (t, J = 6.1 Hz, 1 H, 2-H).
13C NMR (126 MHz, CD3OD): d = 24.2, 29.4, 53.1, 55.3, 160.5,
174.5.
MS (ESI+): m/z = 191.1 [M + H]+.
HRMS (ESI+): m/z [M + H]+ calcd for C6H15N4O3: 191.1139;
found: 191.1143.
Methyl 13-(1-Hydroxyguanidino)tridecanoate (5)
From 6a: To a soln of 6a (50 mg, 0.19 mmol) and Et3N (60 mL,
0.43 mmol) in anhyd DMF (0.5 mL), 1H-pyrazole-1-carboxami-
dine hydrochloride (7, 39 mg, 0.27 mmol) was added and the mix-
ture was stirred at r.t. for 41 h. The solvent was removed under
reduced pressure and the residue was co-evaporated with toluene
and CH2Cl2. Repeated purification of the resultant crude product by
column chromatography (CH2Cl2–MeOH, 9:1) gave 5 as a colorless
solid; yield: 26 mg (45%).
gradient back to 30%, flow rate 0.3 mL/min. 1H (300 MHz) and 13
C
NMR spectra (75 MHz and 126 MHz) were recorded on Varian
Unity 300, Mercury 300, and Inova 500 spectrometers. All 13C
NMR spectra are 1H-decoupled. All spectra were recorded at r.t. ex-
cept for samples in DMSO-d6, CD3OD, and D2O (standard 35 °C).
All NMR spectra were referenced internally to solvent reference
From 6b: To a soln of 6b (103 mg, 0.295 mmol) in anhyd THF
(1.5 mL), a soln of 8a (263 mg, 0.572 mmol) in anhyd THF
(1.5 mL) was added and the mixture was stirred at r.t. for 72 h. The
solvent was removed under reduced pressure. The residue was dis-
solved in anhyd MeOH (20 mL) and Pd/Al2O3 (5%, 31 mg,
15 mmol) was added. The mixture was stirred under an atmosphere
of H2 (1 bar, balloon) at r.t. for 45 min and then filtered through a
pad of celite. The filtrate was evaporated under reduced pressure
and the resultant crude product was purified by column chromatog-
raphy (CH2Cl2–MeOH, gradient 10–100%) to give 5 as a colorless
solid; yield: 30 mg (34%).
1
frequencies. Assignment of signals was carried out using H,1H-
COSY, HSQC, and HMBC spectra obtained on the aforementioned
spectrometers. Low resolution ESI mass spectrometry was per-
formed on a Varian MAT 311 A spectrometer operating in positive
ionization mode. HRMS ESI were carried out on a Bruker mi-
croTOF or a Bruker 7 T FTICR APEX IV spectrometer. Melting
points were measured on a Büchi instrument and are not corrected.
Optical rotations were recorded on a Perkin-Elmer 241 or on a Jasco
P-2000 polarimeter with a Na source using a 10-cm cell (concentra-
tions in g/100 mL). IR spectroscopy was performed on a Perkin-
Elmer Spectrum BX spectrophotometer with solids being measured
as KBr pellets and liquids as films between NaCl plates or on a
Jasco FT/IR-4100 spectrophotometer equipped with an ATR unit.
UV/Vis spectroscopy was carried out on a Perkin-Elmer Lambda-2
or on a Jasco V-630 spectrophotometer.
From 6c: To a soln of 6c (102 mg, 0.341 mmol) in anhyd THF
(1.8 mL), a soln of 8b (240 mg, 0.668 mmol) in anhyd THF
(1.8 mL) was added and the mixture was stirred at r.t. for 72 h.
[Pd(PPh3)4] (20 mg, 0.017 mmol), dimedone (234 mg, 1.67 mmol),
and anhyd THF (1.5 mL) were then added. The mixture was stirred
with exclusion of light for 18 h and then diluted with H2O (5 mL).
The aqueous layer was extracted with EtOAc (3 × 5 mL) and the
combined organics were dried (Na2SO4). The solvent was removed
under reduced pressure, and the resultant crude product was puri-
fied by column chromatography (CH2Cl2–MeOH, gradient 10–
100%) to give 5 as a colorless solid; yield: 29 mg (28%); mp 72 °C;
Rf = 0.06 (CH2Cl2–MeOH, 4:1).
Oximes from Aldehyde Precursors; General Procedure A
A soln of the aldehyde and the respective hydroxylamine hydro-
chloride in anhyd EtOH and anhyd pyridine (1:1) was stirred over
activated molecular sieves (3 Å) at r.t. for 48–72 h. The mixture was
filtered and the molecular sieves washed with MeOH. The filtrate
was evaporated under reduced pressure, and the resultant crude
product was purified by column chromatography.
IR (KBr): 3334, 3215, 2924, 2851, 2555, 2448, 2371, 1733, 1595,
1546, 1517 cm–1.
1H NMR (300 MHz, DMSO-d6): d = 1.18–1.33 (m, 16 H, 4-H2–11-
H2), 1.43–1.65 (m, 4 H, 3-H2, 12-H2), 2.28 (t, J = 7.4 Hz, 2 H, 2-
H2), 3.15–3.26 (m, 2 H, 13-H2), 3.57 (s, 3 H, OCH3), 7.51 (br s, 4 H,
NH, OH).
Reduction of Oximes to Hydroxylamines; General Procedure B
To a soln of the oxime and a small amount of methyl orange as in-
dicator in MeOH, solid NaBH3CN and methanolic HCl (1 M, fresh-
ly prepared from AcCl and MeOH) were added in alternating
portions until the mixture became cloudy and pink. The mixture was
then stirred at r.t. for 24 h and subsequently neutralized with Et3N.
Synthesis 2011, No. 15, 2357–2368 © Thieme Stuttgart · New York

2366
O. Ries et al.
FEATURE ARTICLE
13C NMR (126 MHz, DMSO-d6): d = 24.3, 25.6, 25.7, 28.3, 28.6,
MS (ESI+): m/z = 322.2 [M + Na]+.
HRMS (ESI+): m/z [M + H]+ calcd for C17H34NO3: 300.2533;
28.6, 28.8, 28.8, 28.9, 33.2, 50.7, 51.0, 157.6, 173.3.
MS (ESI+): m/z = 302.3 [M + H]+.
found: 300.2533.
HRMS (ESI+): m/z [M + H]+ calcd for C15H32N3O3: 302.2438;
found: 302.2440.
13,14-Dihydroxybehenic Acid Methyl Ester (10)
To a suspension of erucic acid (50.0 g, 0.148 mol) in HCO2H
(500 mL), a soln of H2O2 (30% in H2O, 130 mL, 1.27 mol) in
HCO2H (500 mL) was added dropwise at 0 °C. The mixture was
stirred at r.t. for 44 h and the solvent then evaporated under reduced
pressure. The residue was dissolved in 1 M KOH (500 mL) and
stirred under reflux for 1 h. The soln was acidified with concd HCl
and the aqueous phase extracted with EtOAc (3 × 500 mL). The
combined organics were evaporated under reduced pressure to pro-
vide crude 13,14-dihydroxybehenic acid, which was dissolved in
MeOH (1.4 L). At 0 °C, SOCl2 (33 mL, 0.45 mmol) was added
dropwise, and the mixture was stirred under reflux for 4 h. After
cooling to r.t., the solvent was removed under reduced pressure to
give 10 as a colorless solid; yield: 56.8 g (99%); mp 71 °C;
Rf = 0.60 (PE–EtOAc, 1:1).
Methyl 13-(Hydroxyamino)tridecanoate (6a)
The synthesis was carried out according to general procedure B
with oxime 12a (1.06 g, 4.12 mmol), NaBH3CN (390 mg,
6.21 mmol), and MeOH (25 mL). The crude product was purified
by column chromatography (EtOAc) to give 6a as a colorless solid;
yield: 813 mg (76%); mp 72 °C; Rf = 0.09 (EtOAc).
IR (KBr): 2914, 2848, 1742, 1471, 1435, 1225, 1199, 1165, 882,
717 cm–1.
1H NMR (300 MHz, CD3OD): d = 1.21–1.43 (m, 16 H, 4-H2–11-
H2), 1.45–1.68 (m, 4 H, 3-H2, 12-H2), 2.31 (t, J = 7.4 Hz, 2 H, 2-
H2), 2.84 (t, J = 7.4 Hz, 2 H, 13-H2), 3.65 (s, 3 H, OCH3).
13C NMR (126 MHz, CD3OD): d = 26.1, 27.9, 28.3, 30.2, 30.4,
30.6, 30.7, 30.7, 30.7, 30.7, 34.8, 51.9, 54.8, 175.8.
IR (KBr): 3342, 3263, 2912, 2846, 1740, 1468, 1198, 1168, 1142,
1115, 720, 653 cm–1.
1H NMR (300 MHz, DMSO-d6): d = 0.86 (t, J = 6.7 Hz, 3 H, 22-
H3), 1.10–1.60 (m, 34 H, 3-H2–12-H2, 15-H2–21-H2), 2.28 (t,
J = 7.4 Hz, 2 H, 2-H2), 3.15–3.26 (m, 2 H, 13-H, 14-H), 3.58 (s,
3 H, OCH3), 4.05 (d, J = 5.5 Hz, 2 H, OH).
13C NMR (76 MHz, DMSO-d6): d = 13.8, 22.0, 28.3, 28.5, 28.6,
28.8, 28.9, 28.9, 28.9, 29.1, 29.1, 31.2, 24.3, 25.5, 32.3, 33.2, 51.0,
73.0, 173.2.
MS (ESI+): m/z = 260.2 [M + H]+.
HRMS (ESI+): m/z [M + H]+ calcd for C14H30NO3: 260.2222;
found: 260.2220.
Methyl 13-(Benzyloxyamino)tridecanoate (6b)
The synthesis was carried out according to general procedure B
with oxime 12b (1.21 g, 3.49 mmol), NaBH3CN (330 mg,
5.30 mmol), and MeOH (11 mL). The crude product was purified
by column chromatography (PE–EtOAc, 10:1) to give 6b as a col-
orless oil; yield: 1.07 g (87%); Rf = 0.38 (PE–EtOAc, 6:1).
MS (ESI+): m/z = 409.4 [M + Na]+.
HRMS (ESI+): m/z [M + H]+ calcd for C23H47O4: 387.3469;
IR (film): 2926, 2854, 1741, 1437, 1363, 1171, 1126, 1055, 736,
699 cm–1.
found: 387.3468.
1H NMR (300 MHz, DMSO-d6): d = 1.13–1.33 (m, 16 H, 4-H2–11-
H2), 1.40 (tt, J = 6.9, 6.9 Hz, 2 H, 12-H2), 1.51 (tt, J = 7.3, 7.3 Hz,
2 H, 3-H2), 2.27 (t, J = 7.3 Hz, 2 H, 2-H2), 2.76 (td, J = 6.9, 6.4 Hz,
2 H, 13-H2), 3.57 (s, 3 H, OCH3), 4.59 (s, 2 H, CH2-Ph), 6.47 (t,
J = 6.4 Hz, 1 H, NH), 7.21–7.42 (m, 5 H, Ph-H).
13C NMR (126 MHz, DMSO-d6): d = 24.3, 26.6, 26.7, 28.4, 28.6,
28.8, 28.9, 28.9, 33.2, 51.0, 51.2, 75.0, 127.2, 127.9, 128.0, 138.5,
173.2.
Methyl 13-Oxotridecanoate (11)
To a soln of 10 (3.00 g, 7.79 mmol) in CH2Cl2 (19 mL), MeCN
(19 mL), and H2O (3.8 mL), NaIO4 (2.50 g, 11.7 mmol) was added,
and the mixture was stirred at r.t. for 18 h. The solvent was removed
under reduced pressure, and the residue was dissolved in EtOAc and
H2O (60 mL each). The aqueous layer was extracted with EtOAc
(1 × 60 mL). The combined organics were washed with H2O
(1 × 60 mL) and brine (1 × 60 mL) and dried (Na2SO4). The solvent
was removed under reduced pressure, and the resultant crude prod-
uct was purified by column chromatography (PE–EtOAc, 10:1) to
give 11 as a colorless oil; yield: 1.75 g (92%); Rf = 0.27 (PE–
EtOAc, 10:1).
MS (ESI+): m/z = 350.3 [M + H]+.
HRMS (ESI+): m/z [M + H]+ calcd for C21H36NO3: 350.2690;
found: 350.2691.
IR (film): 2924, 2853, 2715, 1736, 1462, 1436, 1363, 1195,
1170 cm–1.
UV/Vis (MeOH): lmax (log e) = 251 (1.93), 257 nm (1.93).
1H NMR (300 MHz, C6D6): d = 0.98–1.26 (m, 14 H, 4-H2–10-H2),
1.32 (tt, J = 7.3, 7.3 Hz, 2 H, 11-H2), 1.57 (tt, J = 7.3, 7.3 Hz, 2 H,
3-H2), 1.83 (td, J = 7.3, 1.6 Hz, 2 H, 12-H2), 2.13 (t, J = 7.3 Hz,
2 H, 2-H2), 3.37 (s, 3 H, OCH3), 9.34 (t, J = 1.6 Hz, 1 H, CHO).
13C NMR (75 MHz, C6D6): d = 22.2, 25.3, 29.4, 29.4, 29.6, 29.7,
29.7, 29.8, 29.8, 34.1, 43.8, 50.9, 173.4, 200.7.
Methyl 13-[(Allyloxy)amino)]tridecanoate (6c)
The synthesis was carried out according to general procedure B
with oxime 12c (3.46 g, 11.6 mmol), NaBH3CN (1.10 g,
17.5 mmol), and MeOH (10 mL). The crude product was purified
by column chromatography (PE–EtOAc, 5:1) to give 6c as a color-
less oil; yield: 2.69 g (77%); Rf = 0.35 (PE–EtOAc, 5:1).
IR (film): 2923, 2852, 1739, 1463, 1436, 1247, 1196, 1170, 994,
922 cm–1.
MS (ESI+): m/z = 243.2 [M + H]+, 265.2 [M + Na]+.
HRMS (ESI+): m/z [M + Na]+ calcd for C14H26NaO3: 265.1774;
1H NMR (300 MHz, C6D6): d = 1.11–1.33 (m, 16 H, 4-H2–11-H2),
1.36–1.49 (m, 2 H, 12-H2), 1.50–1.64 (m, 2 H, 3-H2), 2.12 (t,
J = 7.2 Hz, 2 H, 2-H2), 2.82 (t, J = 7.2 Hz, 2 H, 13-H2), 3.35 (s, 3 H,
OCH3), 4.22 (ddd, J = 5.7, 1.4, 1.4 Hz, 2 H, 1¢-H2), 5.06 (ddt,
J = 10.4, 2.8, 1.4 Hz, 1 H, 3¢-Ha), 5.22 (br s, 1 H, OH), 5.23 (ddt,
J = 17.3, 2.8, 1.4 Hz, 1 H, 3¢-Hb), 5.97 (ddt, J = 17.3, 10.4, 5.7 Hz,
1 H, 2¢-H).
found: 265.1775.
Methyl 13-(Hydroxyimino)tridecanoate (12a)
The synthesis was carried out according to general procedure A
with aldehyde 11 (1.37 g, 5.65 mmol), NH2OH·HCl (600 mg,
8.63 mmol), EtOH (13 mL), and pyridine (13 mL) with a reaction
time of 72 h. The crude product was purified by column chromatog-
raphy (PE–EtOAc, 6:1 → 3:1) to give 12a as a colorless solid as
13C NMR (75 MHz, C6D6): d = 25.3, 27.6, 27.8, 29.4, 29.6, 29.8,
30.0, 30.0, 34.1, 50.9, 52.5, 75.2, 116.6, 135.6, 173.3.
Synthesis 2011, No. 15, 2357–2368 © Thieme Stuttgart · New York

FEATURE ARTICLE
Synthesis of N-Alkyl-N-hydroxyguanidines
2367
mixture of E/Z-isomers; yield: 1.18 g (81%); ratio E/Z 69:31;
Rf = 0.26, 0.34 (PE–EtOAc, 6:1).
tert-Butyl (S)-2-[Bis(tert-butyloxycarbonyl)amino]-5-(hydroxy-
imino)pentanoate (25)
The synthesis was carried out according to general procedure A
with aldehyde 24 (803 mg, 2.07 mmol), NH2OH·HCl (290 mg,
4.17 mmol), EtOH (4 mL), and pyridine (4 mL) with a reaction time
of 48 h. The crude product was purified by column chromatography
(PE–EtOAc, 2:1) to give 25 as a colorless oil as mixture of E/Z-iso-
mers; yield: 759 mg (91%); ratio E/Z 50:50; Rf = 0.28, 0.36 (PE–
EtOAc, 2:1).
1H NMR (300 MHz, CDCl3): d = 1.42 [s, 18 H, 2 × OC(CH3)3],
1.48 [s, 36 H, 2 × Boc-OC(CH3)3], 1.91–2.13 (m, 2 H, 2 × 3-Ha),
2.16–2.34 (m, 4 H, 2 × 3-Hb, 1 × 4-H2), 2.35–2.54 (m, 2 H, 1 × 4-
H2), 4.67–4.79 (m, 2 H, 2 × 2-H), 6.71 (t, J = 5.3 Hz, 1 H, 1 × 5-H),
7.40 (t, J = 5.5 Hz, 1 H, 1 × 5-H), 7.55 (br s, 1 H, 1 × OH), 7.94 (br
s, 1 H, 1 × OH).
1H NMR (300 MHz, CD3OD): d = 1.22–1.40 (m, 28 H, 2 × 4-H2–
10-H2), 1.42–1.53 (m, 4 H, 2 × 11-H2), 1.54–1.68 (m, 4 H, 2 × 3-
H2), 2.15 (td, J = 7.5, 6.2 Hz, 2 H, 1 × 12-H2), 2.31 (t, J = 7.4 Hz,
4 H, 2 × 2-H2), 2.34 (td, J = 7.5, 5.5 Hz, 2 H, 1 × 12-H2), 3.65 (s,
6 H, 2 × OCH3), 6.64 (t, J = 5.5 Hz, 1 H, 1 × 13-H), 7.35 (t,
J = 6.2 Hz, 1 H, 1 × 13-H).
13C NMR (75 MHz, CD3OD): d = 25.9, 26.0, 27.2, 30.2, 30.4, 30.4,
30.5, 30.5, 30.6, 30.6, 34.8, 52.0, 152.5, 153.1, 176.0.
MS (ESI+): m/z = 280.2 [M + Na]+.
HRMS (ESI+): m/z [M + Na]+ calcd for C14H27NNaO3: 280.1883;
found: 280.1893.
13C NMR (75 MHz, CDCl3): d = 22.1, 25.9, 26.4, 26.6, 27.9, 28.0,
Methyl 13-(Benzyloxyimino)tridecanoate (12b)
The synthesis was carried out according to general procedure A
with aldehyde 11 (1.00 g, 4.14 mmol), O-benzylhydroxylamine hy-
drochloride (980 mg, 6.14 mmol), EtOH (7 mL), and pyridine
(7 mL) with a reaction time of 48 h. The crude product was purified
by column chromatography (PE–EtOAc, 9:1) to give 12b as a col-
orless liquid as mixture of E/Z-isomers; yield: 1.26 g (88%); ratio
E/Z 57:43; Rf = 0.47, 0.49 (PE–EtOAc, 6:1).
1H NMR (300 MHz, DMSO-d6): d = 1.10–1.33 (m, 28 H, 2 × 4-H2–
10-H2), 1.34–1.60 (m, 8 H, 2 × 3-H2, 2 × 11-H2), 2.11 (td, J = 7.3,
6.2 Hz, 2 H, 1 × 12-H2), 2.23–2.32 (m, 6 H, 2 × 2-H2, 1 × 12-H2),
3.58 (s, 6 H, 2 × OCH3), 4.98 (s, 2 H, 1 × CH2-Ph), 5.04 (s, 2 H,
1 × CH2-Ph), 6.75 (t, J = 5.5 Hz, 1 H, 1 × 13-H), 7.23–7.42 (m,
10 H, 2 × Ph-H), 7.46 (t, J = 6.2 Hz, 1 H, 1 × 13-H).
13C NMR (75 MHz, DMSO-d6): d = 24.3, 25.2, 25.4, 25.8, 28.3,
28.3, 28.5, 28.6, 28.6, 28.7, 28.7, 28.8, 33.2, 51.0, 74.4, 74.6, 127.4,
127.4, 127.6, 127.8, 128.1, 128.1, 137.9, 138.1, 151.3, 152.0, 173.2.
MS (ESI+): m/z = 348.3 [M + H]+, 370.2 [M + Na]+.
HRMS (ESI+): m/z [M + H]+ calcd for C21H34NO3: 348.2533;
58.2, 58.5, 81.4, 83.0, 151.0, 151.6, 152.3, 169.4.
MS (ESI+): m/z = 425.2 [M + Na]+.
HRMS (ESI+): m/z [M + Na]+ calcd for C19H34N2NaO7: 425.2258;
found: 425.2257.
tert-Butyl (S)-2-[Bis(tert-butyloxycarbonyl)amino]-5-(hy-
droxyamino)pentanoate (26)
To a soln of oxime 25 (506 mg, 1.26 mmol) in AcOH (5 mL),
NaBH3CN (162 mg, 2.58 mmol) was added and the mixture was
stirred at r.t. for 2 h. It was then diluted with EtOAc (50 mL) and
washed with sat. NaHCO3 soln (2 × 50 mL), H2O (1 × 50 mL), and
brine (1 × 50 mL). The organic layer was dried (Na2SO4). The sol-
vent was removed under reduced pressure, and the resultant crude
product was purified by column chromatography (CH2Cl2–MeOH,
9:1) to give 26 as a colorless oil; yield: 270 mg (53%); Rf = 0.47
(CH2Cl2–MeOH, 9:1).
[a]D20 –17.8 (c 0.96, CHCl3).
IR (ATR): 1735, 1698, 1365, 1251, 1231, 1147, 1122, 1035, 845,
794 cm–1.
found: 348.2532.
1H NMR (300 MHz, CDCl3): d = 1.39 [s, 9 H, OC(CH3)3], 1.45 [s,
18 H, Boc-OC(CH3)3], 1.32–1.62 (m, 2 H, 4-H2), 1.70–2.15 (m,
2 H, 3-H2), 2.90 (t, J = 7.0 Hz, 2 H, 5-H2), 4.71 (dd, J = 9.5, 5.2 Hz,
1 H, 2-H), 5.56 (br s, 2 H, NH, OH).
13C NMR (126 MHz, CDCl3): d = 23.5, 26.8, 27.9, 27.9, 53.4, 58.7,
81.1, 82.7, 152.5, 169.7.
MS (ESI+): m/z = 405.3 [M + H]+, 427.3 [M + Na]+.
HRMS (ESI+): m/z [M + H]+ calcd for C19H37N2O7: 405.2595;
Methyl 13-(Allyloxyimino)tridecanoate (12c)
The synthesis was carried out according to general procedure A
with aldehyde 11 (3.50 g, 14.4 mmol), O-allylhydroxylamine hy-
drochloride (1.58 g, 14.4 mmol), EtOH (22 mL), and pyridine
(22 mL) with a reaction time of 48 h. The crude product was puri-
fied by column chromatography (PE–EtOAc, 9:1) to give 12c as a
colorless oil as mixture of E/Z-isomers; yield: 3.48 g (81%); ratio
E/Z 65:35; Rf = 0.32, 0.38 (PE–EtOAc, 9:1).
1H NMR (300 MHz, C6D6): d = 1.02–1.35 (m, 32 H, 2 × 4-H2–11-
H2), 1.48–1.68 (m, 4 H, 2 × 3-H2), 1.97 (td, J = 7.4, 6.1 Hz, 2 H,
1 × 12-H2), 2.12 (t, J = 7.4 Hz, 4 H, 2 × 2-H2), 2.27 (td, J = 7.4,
6.1 Hz, 2 H, 1 × 12-H2), 3.36 (s, 6 H, 2 × OCH3), 4.60 (ddd, J = 5.6,
1.4, 1.4 Hz, 2 H, 1 × 1¢-H2), 4.64 (ddd, J = 5.6, 1.4, 1.4 Hz, 2 H,
1 × 1¢-H2), 5.06 (ddt, J = 10.4, 1.8, 1.4 Hz, 1 H, 2 × 3¢-Ha), 5.23
(ddt, J = 17.4, 1.8, 1.4 Hz, 1 H, 2 × 3¢-Hb), 6.00 (ddt, J = 17.4, 10.4,
5.6 Hz, 1 H, 1 × 2¢-H), 6.01 (ddt, J = 17.4, 10.4, 5.6 Hz, 1 H, 1 × 2¢-
H), 6.52 (t, J = 6.1 Hz, 1 H, 1 × 13-H), 7.34 (t, J = 6.1 Hz, 1 H,
1 × 13-H).
found: 405.2589.
Optimization Studies for the Guanidinylation of Hydroxyl-
amines 6b,c and 13
Guanidinylation reactions described in Tables 1 to 3 were carried
out as follows. A soln of the hydroxylamine derivative and the re-
spective base in the anhyd solvent was stirred at the listed tempera-
ture for 15 min (Et3N, TMG) or 1 h (all other bases). This step was
skipped if no base was used. A soln of the guanidinylation reagent
in the anhyd solvent was then added and stirring was continued for
the listed time. After the addition of sat. NH4Cl soln (1.5 mL), the
solvent was removed under reduced pressure. The residue was sus-
pended in EtOAc (1.5 mL) and filtered through a short pad of celite,
which was then washed with EtOAc. The filtrate was dried
(Na2SO4) and the solvent was removed under reduced pressure. The
resultant crude product was analyzed by 1H NMR.
13C NMR (75 MHz, C6D6): d = 25.3, 26.0, 26.5, 26.9, 29.3, 29.4,
29.6, 29.7, 29.8, 29.8, 29.9, 34.1, 50.9, 74.6, 74.9, 116.9, 135.2,
150.4, 151.5, 173.3.
MS (ESI+): m/z = 320.2 [M + Na]+.
HRMS (ESI+): m/z [M + Na]+ calcd for C17H31NNaO3: 320.2202;
found: 320.2199.
Optimization Studies for the Hydrogenolytic Deprotection of 15
Deprotection reactions described in Table 4 and Scheme 4 were
carried out as follows. To a soln of crude 15 (obtained from a
Synthesis 2011, No. 15, 2357–2368 © Thieme Stuttgart · New York

2368
O. Ries et al.
FEATURE ARTICLE
guanidinylation reaction carried out under optimized conditions,
vide supra) in the respective solvent, a small amount of the catalyst
was added. The mixture was stirred under an atmosphere of H2
(1 bar, balloon) at r.t. for 24 h. Samples (0.3 mL each) were taken
after 0, 0.5, 1, 2, 4, 6, 8, 10, and 24 h, respectively, filtered through
a Spartan syringe filter and evaporated. The resultant residues were
analyzed by HPLC-MS.
(6) (a) Maehr, H.; Blount, J. F.; Pruess, D. L.; Yarmchuk, L.;
Kellett, M. J. Antibiot. 1973, 26, 284. (b) Perlman, D.;
Vlietinck, A. J.; Matthews, H. W.; Lo, F. F. J. Antibiot. 1974,
27, 826.
(7) Seto, H.; Koyama, M.; Ogino, H.; Tsuruoka, T. Tetrahedron
Lett. 1983, 24, 1805.
(8) Nemoto, A.; Hoshino, Y.; Yazawa, K.; Ando, A.; Mikami,
Y. J. Antibiot. 2002, 55, 593.
Optimization Studies for the Palladium-Catalyzed Deprotec-
tion of 16
(9) (a) Kimura, K.-I.; Bugg, T. D. H. Nat. Prod. Rep. 2003, 20,
252. (b) Winn, M.; Goss, R. J. M.; Kimura, K.-I.; Bugg, T.
D. H. Nat. Prod. Rep. 2010, 27, 279.
Deprotection reactions described in Table 5 were carried out as fol-
lows. To a soln of crude 16 (obtained from a guanidinylation reac-
tion carried out under optimized conditions, vide supra) and the
additive in the respective anhyd solvent, [Pd(PPh3)4] (1–2 mol%)
was added. The mixture was stirred at r.t. for the listed time. The
solvent was removed under reduced pressure and the resultant resi-
due was analyzed by HPLC-MS.
(10) (a) Taubes, G. Science 2008, 321, 356. (b) Walsh, C. Nat.
Rev. Microbiol. 2003, 1, 65. (c) Dini, C. Curr. Top. Med.
Chem. 2005, 5, 1221. (d) Bugg, T. D. H.; Lloyd, A. J.;
Roper, D. I. Infect. Disorders Drug Targets 2006, 6, 85.
(11) McDonald, L. A.; Barbieri, L. R.; Carter, G. T.; Lenoy, E.;
Lotvin, J.; Petersen, P. J.; Siegel, M. M.; Singh, G.;
Williamson, R. T. J. Am. Chem. Soc. 2002, 124, 10260.
(12) Jones, J. H. J. Peptide Sci. 2002, 8, 285.
(13) Heesing, A.; Peppmöller, R. Z. Naturforsch., B: Chem. Sci.
1967, 22, 820.
Guanidinylation of Hydroxylamines 20a–c
Guanidinylation reactions described in Table 6 were carried out as
follows. To a soln of the hydroxylamine derivative and Et3N
(2.3 equiv) in anhyd DMF (2.6 mL/mmol hydroxylamine), 1H-
pyrazole-1-carboxamidine hydrochloride (7, 1.4 equiv) was added
and the mixture was stirred at r.t. for 40 h. The solvent was removed
under reduced pressure and the residue was co-evaporated with tol-
uene and CH2Cl2. The resultant crude product was analyzed by 1H
NMR.
(14) (a) Widmer, J.; Keller-Schierlein, W. Helv. Chim. Acta
1974, 57, 657. (b) Maehr, H.; Leach, M. J. Org. Chem.
1974, 39, 1166.
(15) Dutta, S. P.; Bernacki, R. J.; Bloch, A.; Chheda, G. B.
Nucleosides Nucleotides 1990, 9, 151.
(16) (a) Feichtinger, K.; Sings, H. L.; Baker, T. J.; Matthews, K.;
Goodman, M. J. Org. Chem. 1998, 63, 8432. (b) Zapf, C.
W.; Goodman, M. J. Org. Chem. 2003, 68, 10092.
(17) Mithran, S.; Subbaraman, A. S. Molecules 1999, 4, 159.
(18) Adlof, R. O.; Miller, W. R.; Emken, E. A. J. Labelled
Compd. Radiopharm. 1978, 15, 625.
Supporting Information for this article is available online at
http://www.thieme-connect.com/ejournals/toc/synthesis.
(19) Mezl, V. A.; Knox, W. E. Anal. Biochem. 1976, 74, 430.
(20) (a) Burkhart, F.; Hoffmann, M.; Kessler, H. Angew. Chem.
Int. Ed. 1997, 36, 1191; Angew. Chem. 1997, 109, 1240.
(b) Padron, J. M.; Kokotos, G.; Martin, T.; Markidis, T.;
Gibbons, W. A.; Martin, V. S. Tetrahedron: Asymmetry
1998, 9, 3381. (c) Sutherland, A.; Caplan, J. F.; Vederas, J.
C. Chem. Commun. 1999, 555. (d) Luesch, H.; Hoffmann,
D.; Hevel, J. M.; Becker, J. E.; Golakoti, T.; Moore, R. E.
J. Org. Chem. 2003, 68, 83. (e) Sorensen, J. L.; Sleeman, M.
C.; Schofield, C. J. Chem. Commun. 2005, 1155.
(21) Farrant, R. D.; Walker, V.; Mills, G. A.; Mellor, J. M.;
Langley, G. J. J. Biol. Chem. 2001, 276, 15107.
(22) Ducho, C.; Hamed, R. B.; Batchelar, E. T.; Sorensen, J. L.;
Odell, B.; Schofield, C. J. Org. Biomol. Chem. 2009, 7,
2770.
Acknowledgment
We thank the Deutsche Forschungsgemeinschaft (DFG, SFB 803
’Functionality controlled by organization in and between mem-
branes’) and the Fonds der Chemischen Industrie (FCI, Sachkosten-
zuschuss) for financial support.
References
(1) (a) Kinnel, A. B.; Gehrken, H. P.; Scheuer, P. J. J. Am.
Chem. Soc. 1993, 115, 3376. (b) Seiple, I. B.; Su, S.; Young,
I. S.; Lewis, C. A.; Yamaguchi, J.; Baran, P. S. Angew.
Chem. Int. Ed. 2010, 49, 1095; Angew. Chem. 2010, 122,
1113.
(2) (a) Wani, M. C.; Taylor, H. L.; Wall, M. E.; Coggon, P.;
McPhail, A. T. J. Am. Chem. Soc. 1971, 93, 2325. (b) Wall,
M. E.; Wani, M. C. Cancer Res. 1995, 55, 753.
(3) Schmid, G.; Hofheinz, W. J. Am. Chem. Soc. 1983, 105, 624.
(4) Stuehr, D. J.; Kwon, N. S.; Nathan, C. F.; Griffith, O. W.;
Feldman, P. L.; Wiseman, J. J. Biol. Chem. 1991, 266, 6259.
(5) (a) Dobashi, K.; Matsuda, N.; Hamada, M.; Naganawa, H.;
Takita, T.; Takeuchi, T. J. Antibiot. 1988, 41, 1525.
(b) Dobashi, K.; Naganawa, H.; Takahashi, Y.; Takita, T.;
Takeuchi, T. J. Antibiot. 1988, 41, 1533.
(23) Batchelar, E. T.; Hamed, R. B.; Ducho, C.; Claridge, T. D.
W.; Edelmann, M. J.; Kessler, B.; Schofield, C. J. Angew.
Chem. Int. Ed. 2008, 47, 9322; Angew. Chem. 2008, 120,
9462.
(24) Wang, W.; Xiong, C.; Yang, J.; Hruby, V. J. Synthesis 2002,
94.
(25) Gebhardt, P.; Crumbliss, A. L.; Miller, M. J.; Möllmann, U.
Biometals 2008, 21, 41.
Synthesis 2011, No. 15, 2357–2368 © Thieme Stuttgart · New York