Synthesis of Deuterium-Labelled 3-Hydroxy- L -arginine: Comparative Studies on Different Protecting-Group Strategies

Article abstract of DOI:10.1002/ejoc.201501109

The β-hydroxy-α-amino acid 3-hydroxy-L-arginine is an important intermediate in biosynthetic pathways involving 2-oxoglutarate (2-OG) dependent Fe^II oxygenases. It also plays an essential role in post-translational modifications of ribosomal proteins. For profound studies on arginine-hydroxylating enzymes, synthetic reference compounds are of utmost importance. However, the synthesis of β-hydroxy-α-amino acids is often not trivial. In particular, the preparation of a densely functionalized compound such as 3-hydroxy-L-arginine poses a challenge in terms of a protecting-group strategy. We have previously established a concise synthesis of 3-hydroxy-L-arginine that is suitable for the preparation of both 3-epimers of this β-hydroxy-α-amino acid. However, we have since observed a remarkable isomerization reaction upon removal of the hydrogenolytically cleavable protecting groups, thus limiting the reliability of our previously described route. In this paper, we report a comparative study of protecting-group strategies for this synthetic route, including a newly developed, more robust pathway to the target amino acid. The versatility of this improved route is demonstrated by the synthesis of the deuterium-labelled derivatives (3R)- and (3S)-3-hydroxy-[5-²H]-L-arginine. These new isotope-labelled arginine derivatives represent important tools for the elucidation of biosynthetic pathways, such as the formation of the arginine-derived amino acid epicapreomycidine in the biosynthesis of muraymycin nucleoside antibiotics.

Full text of DOI:10.1002/ejoc.201501109

DOI: 10.1002/ejoc.201501109
Full Paper
Amino Acids
Synthesis of Deuterium-Labelled 3-Hydroxy- -arginine:
L
Comparative Studies on Different Protecting-Group Strategies
Anke Lemke[a] and Christian Ducho*[a]
Abstract: The β-hydroxy-α-amino acid 3-hydroxy-L-arginine is remarkable isomerization reaction upon removal of the hydro-
an important intermediate in biosynthetic pathways involving genolytically cleavable protecting groups, thus limiting the reli-
2-oxoglutarate (2-OG) dependent FeII oxygenases. It also plays ability of our previously described route. In this paper, we re-
an essential role in post-translational modifications of ribosomal port a comparative study of protecting-group strategies for this
proteins. For profound studies on arginine-hydroxylating en- synthetic route, including a newly developed, more robust
zymes, synthetic reference compounds are of utmost impor- pathway to the target amino acid. The versatility of this im-
tance. However, the synthesis of β-hydroxy-α-amino acids is of- proved route is demonstrated by the synthesis of the deute-
ten not trivial. In particular, the preparation of a densely func- rium-labelled derivatives (3R)- and (3S)-3-hydroxy-[5-2H]-
tionalized compound such as 3-hydroxy- -arginine poses a chal- arginine. These new isotope-labelled arginine derivatives repre-
lenge in terms of a protecting-group strategy. We have previ- sent important tools for the elucidation of biosynthetic path-
ously established a concise synthesis of 3-hydroxy- -arginine ways, such as the formation of the arginine-derived amino acid
L-
L
L
that is suitable for the preparation of both 3-epimers of this epicapreomycidine in the biosynthesis of muraymycin nucleo-
β-hydroxy-α-amino acid. However, we have since observed a side antibiotics.
Introduction
A).[6] Hydroxylated amino acid (S)-2 is then converted into
(2S,3R)-capreomycidine (“capreomycidine”) (3) by the pyridoxal-
phosphate (PLP) dependent enzyme VioD, with formal inversion
at the β-stereocentre.[2b,2c]
Nonproteinogenic amino acids with densely functionalized
structures are often important constituents of biologically ac-
tive natural products.[1] For instance, oxygenase-catalysed
hydroxylations of arginine or arginine residues within peptides
are interesting transformations that give remarkable structural
motifs (e.g., the biosynthesis of the cyclic arginine derivative
capreomycidine, Scheme 1, A).[2] Such enzyme-mediated oxid-
ations could, for example, also be used as powerful tools for the
preparation of complex natural products. However, this would
require a profound knowledge of the particular overall biosyn-
thetic pathway, and also of the enzymatic mechanisms in-
volved.
The aforementioned cyclic arginine derivative, capreomyc-
idine, is one of the unusual structural building blocks that make
up the structures of the tuberactinomycin peptide antibiotics.[3]
The capreomycins[4] and viomycin,[5] for example, are clinically
used as anti-tuberculosis drugs. As part of the viomycin biosyn-
The 3-epimer of 3, i.e., (2S,3S)-capreomycidine (“epicapreo-
mycidine”; 4), can be found as a component of Streptomyces-
produced muraymycins, a subclass of nucleoside antibiotics
that target bacterial cell-wall biosynthesis.[7] The biosynthesis of
4 has not been elucidated yet, but it is likely that it proceeds
in an analogous way, with either (3S)-3-hydroxy-L-arginine [(S)-
2] or (3R)-3-hydroxy- -arginine [(R)-2] as a direct precursor.
L
Based on homology studies, it has been speculated that Mur15
and Mur16 might be the enzymes encoded by the muraymycin
biosynthetic gene cluster that are responsible for epicapreo-
mycidine formation (Scheme 1, A).[8] It cannot be ruled out,
though, that epicapreomycidine (4) is formed by enzyme-medi-
ated epimerization of capreomycidine (3) at the C-3 position, as
there is precedent for unusual epimerization reactions in other
biosynthetic pathways in bacteria.[9,10]
thesis in Streptomyces vinaceus,
into (3S)-3-hydroxy-
L
-arginine (1) is transformed
3-Hydroxy-L-arginine motifs also occur in the biosynthesis of
clavulanic acid, and as post-translational modifications of pro-
teins. In the biosynthesis of the medically very important β-
L
-arginine [(S)-2] as an intermediate.[2] VioC,
a non-heme 2-oxoglutarate (2-OG) dependent FeII oxygenase,
catalyses this stereoselective hydroxylation reaction (Scheme 1,
lactamase inhibitor clavulanic acid, 3-hydroxy- -arginine (2)
L
does not act as a direct precursor though. Instead, the β-lactam
deoxyguanidino-proclavaminate 5 is stereoselectively hydroxyl-
ated at the C-3 position to give intermediate 6, which is then
transformed into clavulanic acid (7) over several further steps
(Scheme 1, B). This hydroxylation reaction is mediated by an-
other non-heme 2-OG-dependent FeII oxygenase, clavaminate
synthase (CAS).[9,10d,11] One striking example, among others,[12]
[a] Department of Pharmacy, Pharmaceutical and Medicinal Chemistry,
Saarland University,
Campus C2 3, 66123 Saarbrücken, Germany
E-mail: christian.ducho@uni-saarland.de
www.ducholab.de
Supporting information and ORCID(s) from the author(s) for this article
are available on the WWW under http://dx.doi.org/10.1002/
ejoc.201501109.
Eur. J. Org. Chem. 2016, 87–98
87
© 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Scheme 1. A: Biosynthetic formation of capreomycidine 3 (in viomycin biosynthesis), and proposed biosynthesis of epicapreomycidine 4 (in muraymycin
biosynthesis), both involving 3-hydroxy-L-arginine 2 as intermediate. B: Biosynthetic formation of clavulanic acid 7 via 3-hydroxylated arginine derivative 6 as
intermediate. C: Oxygenase-catalysed 3-hydroxylation of an arginine residue of the ribosomal protein Rpl16 (50S subunit of the E. coli ribosome).
of arginine hydroxylation as a post-translational modification is gant, it is limited in terms of its potential to incorporate isotope
the C-3 hydroxylation of one arginine residue of the ribosomal labels in a reasonable number of steps. Consequently, it has
protein Rpl16 (50S subunit of the E. coli ribosome) catalysed by only found application in the synthesis of 5,5-dideuterated 3-
the FeII 2-OG oxygenase YcfD (reaction of 8 to 9, Scheme 1, hydroxy- -arginine.[21] However, a derivative selectively deuter-
L
C).[13]
Synthetic reference compounds for β-hydroxy-α-amino acids,
particularly isotope-labelled congeners, can help to identify
ated at the C-3 (β) position would be very useful in order to
identify a potential epimerization reaction.
We have previously described a concise synthesis of 3-
hydroxylating enzymes, to elucidate biosynthetic routes (includ- hydroxy-
L-arginine that gives access to both 3-epimers [i.e., (S)-
ing stereochemical configurations), to fully characterize 2 and (R)-2]. This route has the potential to introduce isotope
hydroxylated moieties in biomolecules, and to understand the labels at several positions, including C-3.[20] Target structures
biological significance of the modified residues. The preparation (S)-2 and (R)-2 were used as authentic standards to identify the
of β-hydroxy-α-amino acids is not trivial, though, and there are stereochemistry of the hydroxylated arginine residue
several reported strategies for their asymmetric synthesis. One (Scheme 1, C).[13] For the synthesis of deuterium-labelled deriv-
route that proceeds with high enantioselectivity and diastereo- atives of 3-hydroxy-
-arginine, we envisaged that this estab-
9
L
selectivity, and that has consequently found widespread appli- lished route could be used. For the purposes of biosynthetic
cation in the total synthesis of natural products, is the asymmet- feeding studies, we had planned to complete the synthesis of
ric hydrogenation of α-amido-β-keto esters.[14] However, syn- C-5-deuterated 3-hydroxy-
L-arginine first. However, we encoun-
theses of the required α-amido-β-keto ester precursors usually tered unexpected challenges in a hydrogenation step of this
involve multiple steps, and the chiral transition-metal catalysts route (vide infra), and this motivated us to carry out a compara-
used for the hydrogenation reaction are often expensive, and tive study of different protecting-group strategies. We focussed
not always commercially available. Furthermore, the hydrogena- our attention in particular on protecting groups for the 3-
tion reaction usually has to be conducted at high pressure in hydroxy and the guanidine moieties. In this work, three differ-
an autoclave. Other approaches such as Sharpless amino- ent protecting-group approaches towards 3-hydroxy-
L-arginine
hydroxylation,[15] dihydroxylation,[16] or asymmetric aldol reac- derivatives will be discussed (3-OH/guanidine): (i) O-Bn/tris-N-
tions[17] are effective methods, but are often specifically de- Cbz; (ii) O-TBDPS/tris-N-Cbz, and (iii) O-TBDPS/tris-N-Alloc, with
signed for only a limited number of transformations. Hence, strategy (i) representing the initial route[20] that proved to be
variations in the conditions or substrates often result in poor problematic (Cbz = benzyloxycarbonyl; TBDPS = tert-butyldi-
selectivity. For the synthesis of 3-hydroxy-L-leucine, another im- phenylsilyl; Alloc = allyloxycarbonyl). These detailed studies en-
portant component of muraymycin antibiotics, a diastereo- abled the development of a more robust synthetic route to 3-
selective Grignard addition to a serine-derived aldehyde was hydroxy-
used as the key step.[18]
L-arginine and its isotope-labelled derivatives.
Overall, there is no generally applicable route for the synthe-
sis of β-hydroxy-α-amino acids. In particular, densely functional-
Results and Discussion
ized structures, such as selectively isotope-labelled 3-hydroxy-
-arginine derivatives, require an individually designed strategy For all three of the aforementioned protecting-group strategies,
for an efficient preparation. To date, there are only two reported homoallylic alcohol 10 (Scheme 2) was an essential key inter-
syntheses of 3-hydroxy-
-arginine.[19,20] One route involves a mediate. As reported by Zabriskie and coworkers,[22] Grignard
1,3-dipolar cycloaddition of a vinylglycine derivative and a addition to the (R)-configured Garner aldehyde (which can be
nitrone as the key step.[19] Although this approach is very ele- readily prepared from -serine[23]) created the C-1′ stereocentre,
L
L
D
Eur. J. Org. Chem. 2016, 87–98
www.eurjoc.org
88
© 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
and thus gave a mixture of both epimers of 10, which was,
After protection of the hydroxy group, the preparation of
however, difficult to separate.[20] When a suitable 3-hydroxy the differently protected 3-hydroxy-
L-arginine derivatives was
protecting group was introduced, the resulting diastereomeric achieved as shown in Scheme 3. Homoallylic alcohols (R)-11b,
mixture of products 11 could be separated by column chroma- (S)-11b, and (R)-11c were hydrolysed under acidic conditions to
tography. To study the different protecting-group strategies, it give amino alcohols (R)-12b, (S)-12b, and (R)-12c, respectively,
was necessary to synthesize differently protected homoallylic in yields of 87–98 %. When TBDMS-protected homoallyllic alco-
alcohols 11a–11c. The separation of the diastereomers was hols 11a were used in this step, unwanted concomitant desilyl-
most efficient for the tert-butyldimethylsilyl (TBDMS)-protected ation occurred (vide supra).[20] Oxidation of (R)-12b, (S)-12b,
derivative (i.e., 11a), which gave 58 % of (R)-11a, and 22 % of and (R)-12c to the corresponding amino acids using TEMPO
(S)-11a (Scheme 2).
(2,2,6,6-tetramethyl-1-piperidinyloxy) and (bisacetoxy)iodo-
benzene (BAIB),[24] and subsequent esterification under Eschen-
moser conditions[25] gave tert-butyl esters (R)-13b, (S)-13b, and
(R)-13c, respectively, in yields of 54–62 % over two steps. The
introduction of the deuterium label at the C-5 position was
accomplished by ozonolysis of (R)-13b, (S)-13b, and (R)-13c, fol-
lowed by reduction with sodium borodeuteride. The resulting
5-deuterated alcohols [i.e., (R)-14b, (S)-14b, and (R)-14c] were
obtained in yields of 71–89 %. The multiplet signals for C-5 in
the 13C NMR spectra indicated that these products were iso-
lated as mixtures of C-5 epimers, implying that the reduction
step occurred without pronounced diastereoselectivity. It is also
noteworthy that the primary products of the ozonolysis reac-
tion were derivatives of glutamate semialdehyde, which have a
strong preference for the cyclic hemiaminal form.[26] This might
have contributed to the reaction requiring a large excess of
Scheme 2. Synthesis of differently protected homoallylic alcohols 11a–11c as
key intermediates for the three different protecting-group strategies. Boc =
tert-butoxycarbonyl; TBAF = tetrabutylammonium fluoride; TBAI = tetrabutyl-
ammonium iodide.
For the O-Bn/tris-N-Cbz strategy, desilylation of (R)-11a and
(S)-11a gave the diastereomerically pure homoallylic alcohols
[i.e., (R)-10 and (S)-10], which were then benzylated to give (R)-
11b and (S)-11b in yields of 51 and 20 %, respectively, over
three steps from 10. This exchange of the alcohol protecting
group was necessary as chromatographic separation of Bn-pro-
tected diastereomers (R)-11b and (S)-11b was not feasible. At-
tempts to separate the corresponding Bn-protected derivatives
at a later stage of the synthesis were similarly unsuccessful. An
attempted approach in which the TBDMS protecting group was
retained throughout the synthetic route led to the silyl ether
being concomitantly hydrolysed in the subsequent acidic
acetonide cleavage (vide infra).[20] Consequently, for the other
strategies, the tert-butyldiphenylsilyl (TBDPS) ether was chosen
as the hydroxyl protecting group because it is pronouncedly
more stable under acidic conditions. Chromatographic separa-
tion of TBDPS-protected derivatives (R)-11c and (S)-11c (which
were readily obtained by silylation of 10) gave isolated yields
of 46 and 4 %, respectively (Scheme 2). On a larger scale, the
pure (S) isomer [i.e., (S)-11c] could no longer be isolated. The
low yields of pure (S)-11c resulted from the challenging
chromatographic separation of the TBDPS-protected diastere-
omers. Therefore, the strategy used for Bn-protected derivatives
(R)-11b and (S)-11b, i.e., separation of TBDMS-protected deriva-
tives 11a, desilylation, and 3-O-protection, might also be a
good alternative for the synthesis of both of the TBDPS-pro-
tected diastereomers [i.e., (R)-11c and (S)-11c]. Due to the small
amount of the TBDPS-protected (S)-configured epimer [i.e., (S)-
11c] available, subsequent steps were studied only for the (R)-
configured diastereomer [i.e., (R)-11c]. The stereochemistry of the
homoallylic alcohol derivatives was assigned as reported previ-
ously.[20,22]
Scheme 3. Synthesis of differently protected 3-hydroxy-L-arginine derivatives
17b–17d. BAIB = (bisacetoxyiodo)benzene, DIAD = diisopropyl azodicarbox-
ylate.
Eur. J. Org. Chem. 2016, 87–98
www.eurjoc.org
89
© 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
sodium borodeuteride to achieve complete conversion. Particu- man's catalyst that had previously been used coincidentally had
larly on a larger reaction scale, it was essential to ensure a bal- the “right” composition for the reductive deprotection reaction,
anced ratio between the solvent volume and the amount of thus leading to the desired product in a nonreproducible way.
sodium borodeuteride, i.e., a suitable concentration of sodium Hence, we decided to screen different reaction conditions using
borodeuteride in deuterated methanol. Otherwise, reductive Pearlman's catalyst for a more robust protocol. This included
side-reactions were observed. Deuterated alcohols (R)-14b, (S)- reactions with increased hydrogen pressure (Table 1, entry 3),
14b, and (R)-14c were then converted with tris-Cbz-guanidine higher temperatures and more effective hydrogen introduction
15 using Goodman's guanidinylation method under Mitsunobu (entry 4), different additives (entry 5), solvents (entry 6), and
conditions[27] to give protected guanidines (R)-17b, (S)-17b, and hydrogen sources (entry 7). However, we found that the activity
(R)-17c, respectively, in yields of 74–81 %. Using tris-Alloc-guan- of Pearlman's catalyst was either too low for complete Bn de-
idine 16[28] instead of tris-Cbz-guanidine 15 under analogous protection, or too high, resulting in the formation of side-pro-
conditions, TBDPS- and Alloc-protected derivative 17d was ob- ducts. Other catalysts[29] such as Pd/C (Table 1, entry 8), Pd
tained in 90 % yield (Scheme 3).
The deprotection strategy for the Cbz and Bn protecting
groups had already been established for the synthesis of both
black (entry 9), Raney-Ni (entry 10), or Zn (entry 11) resulted in
worse conversions than that obtained with Pearlman's catalyst.
A fully Cbz-deprotected compound had been isolated before
diastereomers of non-deuterated 3-hydroxy- -arginines (R)-2 without any obvious isomerization (i.e., no double set of NMR
L
and (S)-2.[20] Consequently, these reaction steps were not ex- signals was observed), thus indicating that it was primarily the
pected to represent major hurdles. However, using Pearlman's 3-O-Bn protecting group that was responsible for the aforemen-
catalyst in the hydrogenation reaction of (R)-17b and (S)-17b tioned problems with the hydrogenolytic deprotection step.
and hydrochloric acid in the subsequent acidic hydrolysis, a Therefore, we planned to use a route involving 3-O-TBDPS pro-
double set of signals was observed in the 13C NMR spectra of tection instead. Initial attempts to deprotect (R)-17c were car-
both products. This finding suggested that an isomerization ried out using TBAF trihydrate in THF, with a reaction time of
process had taken place, most probably during the hydrogenol- 23 h (Table 2, entry 1). Under these conditions, the removal of
ysis reaction, as indicated by the NMR spectra of the hydrogen- one Cbz group and elimination was observed, and compound
ation product. Experiments with the previously synthesized, 22 was formed with a yield of 35 %. Using ammonium fluoride
non-deuterated compound (S)-18 and its Cbz- and Bn-depro- as the deprotecting agent, only the removal of Cbz groups
tected derivative proved that this assumption was correct: nei- could be observed (Table 2, entry 2). Further deprotection at-
ther the presence of the stereogenic deuterium label nor the tempts with TBAF trihydrate (Table 2, entry 3) and with a TBAF
acidic hydrolysis was responsible for the double set of signals. solution (1.0
M in THF; Table 2, entry 4) with shortened reaction
The NMR spectra indicated, though, that not only isomerization, times showed that elimination increased when the reaction
but also the formation of other side-products occurred in the time was extended. Under buffered conditions, such as with the
hydrogenation reaction (Table 1, entries 1 and 2). Due to the addition of acetic acid (Table 2, entry 5)[30] or phosphate buffer
high polarity of the compounds, purification of the resulting (Table 2, entry 6), partial elimination was also observed. How-
mixtures was challenging, and therefore only qualitative results ever, when acetic acid was used, a migration (N,O-shift) of a
are reported in Table 1. It was only possible to isolate different Cbz group was observed, thus giving by-product 25. Urethane-
Cbz-deprotected species. We propose that the batch of Pearl- protected guanidines are known to cause problems due to such
Table 1. Different conditions for the hydrogenolytic deprotection of (R)-17b, (S)-17b, and (S)-18.
Entry
Compd. Catalyst
Conditions[a]
Product(s)
1
2
3
4
5
6
7
8
(R)-17b
(S)-18
(R)-17b
(S)-18
Pd(OH)2/C (20 mol-%) H2, MeOH, room temp., 24 h
Pd(OH)2/C (40 mol-%) H2, MeOH, room temp., 48 h
Pd(OH)2/C (20 mol-%) H2, MeOH, 3 bar, room temp., 12 h
Pd(OH)2/C (20 mol-%) H2,[c] MeOH, 60 °C, 7.5 h
no complete conversion,[b] OBn species
complete conversion,[b] side-products
no complete conversion,[b] mixture of deprotected species
complete conversion,[b] mixture of isomers
complete conversion,[b] side-products
(S)-18
Pd(OH)2/C (60 mol-%) H2, MeOH, AcOH, room temp., 6 d
Pd(OH)2/C (20 mol-%) H2, EtOAc, room temp., 3 d
(R)-17b
(S)-17b
(S)-18
(R)-17b
(S)-17b
(R)-17b
complete conversion,[b] mixture of isomers
Pd(OH)2/C (40 mol-%) 1,4-cyclohexadiene, MeOH, room temp., 4 d no complete conversion,[b] mixture of deprotected species
Pd/C (60 mol-%)
Pd black (40 mol-%)
Raney-Ni
H2, MeOH, room temp., 4 d
H2, MeOH, room temp., 40 h
H2, EtOH, room temp., 3 d
no complete conversion,[b] mixture of deprotected species
no complete conversion,[b] mixture of deprotected species
starting material
9
10
11
Zn (60 mol-%)
NH4COOH, MeOH, room temp., 22 h
no complete conversion,[b] OBn species
[a] If not indicated otherwise, a hydrogen pressure of 1 bar was used. [b] Based on the remaining aromatic signals in the 1H NMR spectrum. [c] Hydrogen
was bubbled through the solution.
Eur. J. Org. Chem. 2016, 87–98
www.eurjoc.org
90
© 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Table 2. Attempted TBDPS-deprotection of (R)-17c under different conditions.
Entry
Reagent
Conditions
Product(s) (Yield)
1
2
3
4
5
6
7
8
TBAF·3H2O (2.5 equiv.)
NH4F (10 equiv.)
THF, room temp., 23 h
MeOH, room temp., 6 d
THF, room temp., 4.5 h
22 (35 %)
23,[a] 24[a]
TBAF·3H2O (2.5 equiv.)
TBAF (1.1 equiv.) in dry THF
TBAF (1.1 equiv.) in dry THF
TBAF, buffered solution (3.3
HF·Py
(R)-17c (43 %), 22 (23 %)
22 (23 %)
THF, room temp., 3 h
THF, AcOH, room temp., 3 h
THF, KH2PO4, room temp., 10 d
THF, 0 °C to room temp., 4 h
THF, NEt3, room temp., 11 d
(R)-17c (15 %), 22,[a] 23 (29 %), 25 (17 %)
22 (29 %), 23 (20 %)
M)
22 (23 %), 24 (37 %)
HF·NEt3
(R)-17c,[a] 25 (16 %)
[a] Obtained as a mixture, or as impure material.
rearrangement reactions,[27] particularly under basic conditions.
A migration of the Cbz group from N5 to the hydroxyl group
would also explain the observation of elimination product 22.
Elimination reactions of β-carbonate-substituted carboxyl deriv-
atives under basic conditions are used, for example, for the syn-
thesis of didehydro amino acids.[31] Deprotection attempts us-
ing hydrogen fluoride pyridine complex (Table 2, entry 7), or
hydrogen fluoride triethylamine complex (Table 2, entry 8)[32]
also failed. We therefore concluded that it did not seem to be
feasible to carry out the deprotection of the TBDPS group be-
fore the hydrogenolysis, most probably due to migration of the
N5-Cbz group. Hence, we planned to remove the Cbz groups
first.
Scheme 4. Hydrogenolysis of (R)-17c with subsequent acidic hydrolysis.
For this Cbz deprotection, an initial attempt using
Pearlman's catalyst for the hydrogenolysis of (R)-17c unexpect-
edly led to problems similar to those seen in the previous
As acidic hydrolysis of pure (R)-26 had given the desired pure
hydrogenation reactions (see Table 1). However, further hydro- diastereomer (R)-27 without isomerization, a new strategy
genation conditions were tested, as N-Cbz groups should be avoiding hydrogenolysis was developed. Alloc-protected deriva-
less difficult to remove than O-benzyl groups (vide supra). tive (R)-17d was deprotected by allylic substitution, i.e., with
When palladium on charcoal was used as the catalyst and EtOH/ tetrakis(triphenylphosphine)palladium(0) as catalyst and dime-
EtOAc (1:1) as the solvent,[33] pure diastereomer (R)-26 was iso- done in THF.[33] As some catalyst residues could not be removed
lated in 9 % yield after normal-phase chromatography by extractive workup nor by column chromatography, Alloc-
(Scheme 4). The 13C NMR spectrum of the crude hydrogenation deprotected compound (R)-26 could only be obtained as im-
product showed though that isomerization or the formation of pure material. However, subsequent acidic hydrolysis of (R)-26
side-products had taken place. Deprotection of the TBDPS with hydrochloric acid (6 ) gave pure diastereomer (R)-27. The
M
group using TBAF trihydrate resulted in purification problems limited solubility of (R)-26 in aqueous hydrochloric acid, how-
due to the polarity of the putative product. Alternatively, acidic ever, resulted in moderate yields of up to 40 %. In this context,
hydrolysis conducted in hydrochloric acid (6 M) led to a global the global deprotection strategy using hydrochloric acid could
deprotection by cleavage of the Boc group, the tert-butyl ester, be further improved by ultrasonic treatment, which was repeat-
and the TBDPS ether. Using these conditions, pure diastereomer edly performed over 24 h. Diastereomerically pure fully depro-
(R)-27 was obtained in 43 % yield for the final step (Scheme 4). tected 3-hydroxy-[5-2H]-
L-arginine was thus obtained in an ex-
Eur. J. Org. Chem. 2016, 87–98
www.eurjoc.org
91
© 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
cellent overall yield of 90 % over both deprotection steps
(Scheme 5).
Scheme 5. Two-step deprotection of Alloc-protected derivative (R)-17d.
By using this new TBDPS/Alloc protecting-group strategy,
palladium-catalysed hydrogenolysis, which had led to isomeri-
zation and/or the formation of side-products (vide supra), could
be avoided. Nevertheless, we wanted to elucidate the possible
structures of the side-products of the hydrogenation reaction.
The high polarity of all of the compounds present in the crude
product mixture, as well as their limited amounts, made separa-
tion of the mixture for NMR analysis impossible. Therefore, only
ESI-MS-based structural assignment of the putative side-pro-
ducts of the hydrogenolytic deprotection was feasible. The
crude product obtained from the Pd/C-catalysed hydrogenation
of TBDPS-protected compound (R)-17c in a solvent mixture of
EtOH and EtOAc (see Scheme 4) showed two significant peaks
in the ESI-HRMS spectrum, one at m/z = 586.3524 (100 %), and
one at m/z = 330.2238 (61 %). The most intense peak at m/z =
586.3524 would be consistent with the desired Cbz-depro-
tected product (i.e., 26). The other significant peak at m/z =
330.2238 would be consistent with two structural isomers: elim-
ination product 28, and capreomycidine derivative 29, formed
by a putative ring-closure reaction (see Figure 1). As a reliable
interpretation of the NMR spectra was not possible, it could not
be ruled out that an isomerization of 26 had already taken
place, resulting in a diasteromeric mixture. After acidic hydroly-
sis, ESI-MS revealed signals that would be consistent with com-
pounds 30, 31, and 27 (Figure 1). It should be noted that we
Figure 1. Putative side-products of the hydrogenolysis reaction and the sub-
sequent acidic hydrolysis (see Scheme 4), based on ESI-HRMS analysis.
pairs of enantiomers, not taking into account the mono-deuter-
ated stereogenic centre at C-5).
Considering the metal-binding properties of guanidine,[35]
coordination of this moiety to the palladium catalyst during the
hydrogenation reaction may be assumed. This would keep the
3-hydroxy-L-arginine derivative in close proximity to the cata-
lyst. As such a coordination would compete with the hydrogen-
loading ability of the catalyst, the reactivity might be decreased.
The palladium catalyst, which would be blocked for the activa-
tion of hydrogen in such a scenario, might then just “snatch”
the 2-H and 3-H of the coordinated 3-hydroxy-L-arginine deriva-
tive. Subsequent nonselective reattachment of the hydrogen
atoms at C-2 and C-3 would then lead to a loss of stereochemi-
cal information, and isomerization would occur. For the forma-
tion of a putative elimination product 28, or the corresponding
ring-closed derivative [i.e., 29 (formed from 28)], radical mecha-
nisms might be conceivable, leading to a very unselective reac-
tion outcome. The ability of guanidine to coordinate metals is
very pH-dependent. Consequently, we speculated that suffi-
ciently acidic reaction conditions could prevent the formation
of side-products and the isomerization. By adding trifluoro-
acetic acid (TFA) to the hydrogenation mixture of (R)-17c, we
were able to isolate pure diastereomer (R)-26 in quantitative
yield (Scheme 6).
have previously reported
a domino-guanidinylation–aza-
Michael-addition for the ring closure of didehydro amino acid
precursors to give capreomycidine derivatives, thus providing a
biomimetic, but nonstereocontrolled access to (epi)capreo-
mycidine.[34] Hence, the formation of cyclization products 29
and 31 from elimination by-products 28 and 30, respectively,
1
Analysis of the H and 13C NMR spectra revealed the pres-
appears to be likely.
ence of only small amounts of impurities, and no indications of
1
A comparison of the H NMR spectra of pure diastereomers isomerization or of the formation of the aforementioned side-
(R)-27 and (S)-27 with the spectra of the product mixtures con- products were observed. Acidic hydrolysis of pure diastereomer
taining the proposed by-products (i.e., 30 and 31, together with (R)-26 gave (3R)-3-hydroxy-
-[5-2H]arginine (R)-27 in a yield of
L
27) showed that, besides an isomeric mixture of 27, a signifi- 77 % over two steps from (R)-17c. The hydrogenation of the Bn-
cant amount of more than one side-product was obtained. As protected congeners [i.e., (R)-17b and (S)-17b] was also revised
the 13C NMR spectra did not contain any signals that would be in the light of these results. In the reactions carried out previ-
consistent with the typical shifts of double bonds, we propose ously, Pearlman's catalyst was considered to be essential for the
that product 29 was formed during the hydrogenation reaction, effective removal of the Bn protecting group. However, after
resulting in up to four possible stereoisomers (diastereomeric the addition of TFA, Pearlman's catalyst proved to be too unre-
Eur. J. Org. Chem. 2016, 87–98
www.eurjoc.org
92
© 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
cation problems, the elucidation of the structures of the side-
products was nontrivial. In this context, changing the strategy
to the TBDPS-based route proved to be advantageous as the
more hydrophobic derivative (R)-26 could be easily isolated
from the crude mixture by normal-phase chromatography. This
achievement led to an improved understanding of the hydro-
genation reaction, and it supported our identification of the
putative side-products. Overall, the detailed investigations of
different protecting-group strategies finally led to an optimized
synthetic strategy. The TBDPS/Alloc route now offers a robust
synthesis of both 3-epimers of 3-hydroxy-L-arginine and its iso-
tope-labelled derivatives. We have shown that the preparation
of C-5-deuterated target structures is feasible. Future work will
include the synthesis of the C-3-deuterated derivatives, which
will be accessible from our optimized route, using a deuterated
analogue of Garner's aldehyde as starting material. The result-
Scheme 6. Hydrogenolysis of (R)-17b, (S)-17b, and (R)-17c with the addition
ing C-3-deuterated 3-epimers of 3-hydroxy-L-arginine will shed
of TFA, and subsequent global hydrolytic deprotection.
light on a possible epimerization reaction during the biosyn-
thetic formation of epicapreomycidine within muraymycin bio-
synthesis. It has been proposed that the enzymes Mur15 and
Mur16 are involved in the biosynthesis of epicapreomycidine,
and in vitro studies of these enzymes will be possible using
active for a complete Cbz- and Bn-deprotection of compound
(S)-17b. This change in reactivity under acidic conditions had
already been observed upon the addition of acetic acid (see
Table 1, entry 5). Surprisingly, when Pd/C was used in EtOH and
EtOAc with the addition of TFA, the desired diastereomer was
formed with no side-product formation or isomerization ob-
served. Accordingly, ca. 40 mol-% of the catalyst was required
to remove the Bn group. For most of the hydrogenolysis reac-
synthetic 3-hydroxy-
geners) as substrates. Therefore, the synthesis of C-3-deuterated
3-hydroxy- -arginine will also be carried out in our laboratories.
L-arginine epimers (and their labelled con-
L
1
tions, minor aromatic impurities were observed in the H NMR
Experimental Section
spectra of the crude products. This observation indicated that
complete removal of the Bn group could not be achieved. How-
ever, pure (R)-19 and (S)-19 were isolated by reverse-phase
chromatography after hydrogenation. Subsequent acidic
hydrolysis gave the pure target compounds [i.e., (R)-27 and (S)-
27] in good yields of ca. 80 % over two steps from (R)-17b and
(S)-17b, respectively (Scheme 6).
General Remarks: Chemicals were purchased from standard sup-
pliers, and were used without further purification. Starting material
10 and compounds (R)-11a, (S)-11a, (R)-11b, (S)-11b, (R)-12b, (S)-
12b, (R)-13b, (S)-13b, and 15 were synthesized as previously re-
ported.[20,22,23b,27] Ozone was generated using a Fischer ozone gen-
erator model 502 or model 500. Reactions involving oxygen- and/
or moisture-sensitive reagents were carried out under an atmos-
phere of argon using anhydrous solvents. The glass equipment used
for these reactions was dried by heating before use. Anhydrous
solvents were obtained in the following manner: THF was dried
with sodium/benzophenone and distilled, CH2Cl2 was dried with
CaCl2 and distilled, and pyridine was dried with CaH2 and distilled.
DMF and MeOH were purchased as absolute grade, and were de-
gassed and stored over molecular sieves (4 Å and 3 Å, respectively).
The following solvents, used for reactions without inert conditions,
extractions, and chromatography, were purchased as technical
grade, and were distilled before use: CH2Cl2, Et2O, EtOAc, isohexane,
and petroleum ether (35–70). All other solvents were p.a. grade,
and distilled water was used throughout. For RP (reverse-phase)
column chromatography, bidistilled ultra-pure water (Milli-Q) was
Conclusions
In summary, pure C-5-deuterated 3-epimeric 3-hydroxy-L-argin-
ine derivatives (R)-27 and (S)-27 were both synthesized in a
sufficient amount for them to be used in biosynthetic studies.
The synthesis of (R)-27 was achieved by three alternative pro-
tecting-group strategies starting from protected homoallylic al-
cohols 11b and 11c, respectively. Comparing the different pro-
tection strategies from these homoallylic alcohols towards pro-
tected amino acids (R)-17b, (S)-17b, (R)-17c, and (R)-17d over
five subsequent steps, the yields obtained were quite similar
(31 % and 30 % by the Bn/Cbz route, 21 % by the TBDPS/Cbz used. Column chromatography was carried out on silica gel 60
(0.040–0.063 mm, 230–400 mesh ASTM, VWR) under flash condi-
tions. RP column chromatography was carried out on RP silica gel
90 C18 (0.040–0.063 nm, Fluka). TLC was carried out on aluminium
plates precoated with silica gel 60 F254 (VWR). Spots were visualized
using UV light (254 nm) where appropriate, and/or staining under
heating [KMnO4 staining solution: KMnO4 (1 g), K2CO3 (6 g), and
NaOH. (aq., 5 % w/v; 1.5 mL), all dissolved in H2O (100 mL); nin-
hydrin staining solution: ninhydrin (0.3 g), AcOH (3 mL), all dissolved
route, and 27 % by the TBDPS/Alloc route). It is reasonable to
assume that the proposed substrate–catalyst coordination dur-
ing the Pd-catalysed hydrogenolysis of Cbz-protected deriva-
tives (R)-17b, (S)-17b, and (R)-17c was probably responsible for
the observed isomerization reaction, and that it probably also
promoted an elimination–cyclization sequence, thus leading to
by-product formation. During our investigation of the different
deprotection strategies, another major issue was the polarity of
in 1-butanol (100 mL)]. 1H NMR spectra (300 and 500 MHz) and 13
NMR spectra (75, 76, and 126 MHz) were recorded with Varian
well as the polarity of the observed side-products. Due to purifi- UNITY 300, MERCURY 300, INOVA 500, and INOVA 600, and Bruker
C
the partially deprotected 3-hydroxy-L-arginine derivatives, as
Eur. J. Org. Chem. 2016, 87–98
www.eurjoc.org
93
© 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
AVANCE-500 and AVANCE-300 spectrometers. 13C NMR spectra were
2.9 Hz, 1 H, 1′-H), 4.73–4.76 (m, 1 H, 4′a-H), 4.76–4.80 (m, 1 H, 4′b-
1H-decoupled. All spectra were recorded at room temperature ex- H), 5.46–5.56 (m, 1 H, 3′-H), 7.37–7.49 (m, 6 H, Ph-H), 7.64–7.70 (m,
cept for samples in [D6]DMSO and D2O (standard 35 °C) and where
otherwise indicated. Spectra were referenced internally to solvent
reference frequencies. Chemical shifts (δ) are quoted in ppm. Cou-
pling constants (J) are reported in Hz to the nearest 0.1 Hz. The
4 H, Ph-H) ppm. 13C NMR (126 MHz, [D6]DMSO, 100 °C): δ = 19.4
[SiC(CH3)3], 26.5 [C(CH3)2], 26.8 [C(CH3)2], 27.5 [SiC(CH3)3], 28.6 (tBu-
CH3), 39.8 (C-2′), 60.6 (C-4), 63.2 (C-5), 72.7 (C-1′), 79.9 (tBu-C), 94.4
(C-2), 117.5 (C-4′), 128.0 (Ph-CH), 128.0 (Ph-CH), 130.1 (Ph-CH), 130.1
(Ph-CH), 134.2 (C-3′), 134.9 (Ph-C), 135.0 (Ph-C), 135.8 (Ph-C), 135.9
1
assignment of signals was carried out using H,1H-COSY and HSQC
spectra obtained on the spectrometers mentioned above. Low-reso- (Ph-C), 153.0 (Boc-C=O) ppm. IR (ATR): ν = 2935, 1697, 1389, 1365,
˜
lution ESI mass spectrometry was carried out with a Finnigan ion-
trap mass spectrometer LCQ. High-resolution (HRMS) ESI mass spec-
trometry was carried out with a Bruker microTOF spectrometer, a
Bruker 7 T FTICR APEX IV, or a Waters QTOF Synapt 2G spectrometer.
Melting points (m.p.) were measured with a Büchi instrument. Opti-
cal rotations were recorded with a Perkin–Elmer 241 polarimeter or
a Jasco P-2000 with a Na source using a 10 cm cell. Infrared spectro-
scopy (IR) was carried out either with a Bruker Vector 22 spectrome-
ter, with liquids and oils being measured as films on NaCl plates
and solids as KBr pellets, or with spectrometers equipped with an
ATR unit (Jasco FTIR 4100 with Pike GladiATR™, or Bruker Vertex 70
1255, 1174, 1092, 1075, 1011, 854, 824, 743, 703, 616 cm–1. UV
(MeCN): λmax (log ε) = 219 (4.63), 266 (3.22) nm. HRMS (ESI): calcd.
for C30H43NNaO4Si [M + Na]+ 532.2854; found 532.2865. TLC (iso-
hexane/EtOAc, 12:1): Rf = 0.22.
N-Boc-(2R,3R)-2-amino-3-(tert-butyldiphenylsilyloxy)hex-5-ene-
1,3-diol [(R)-12c]: A solution of (R)-11c (8.44 g, 16.6 mmol) in acetic
acid (125 mL) and H2O (25 mL) was stirred at room temp. for 4 d.
The solvent was evaporated under reduced pressure, and the result-
ing crude product was purified by column chromatography (isohex-
ane/EtOAc, 3:1) to give (R)-12c (6.77 g, 87 %) as a colourless oil.
[α]D20 = –22.1 (c = 0.96, CHCl3). 1H NMR (500 MHz, C6D6): δ = 1.07 [s,
9 H, SiC(CH3)3], 1.44 (s, 9 H, tBu-CH3), 2.15–2.24 (m, 1 H, 4a-H), 2.39–
2.48 (m, 1 H, 4b-H), 3.47–3.61 (m, 1 H, 1a-H), 3.61–3.71 (m, 1 H, 1b-
H), 3.94–4.02 (m, 1 H, 2-H), 4.05–4.12 (m, 1 H, 3-H), 4.83–4.92 (m, 2
H, 6-H), 5.12 (d, J = 8.7 Hz, 1 H, NH), 5.52–5.66 (m, 1 H, 5-H), 7.12–
7.23 (m, 6 H, Ph-H), 7.67–7.76 (m, 4 H, Ph-H) ppm. 13C NMR
(126 MHz, C6D6): δ = 19.3 [SiC(CH3)3], 26.9 [SiC(CH3)3], 28.1 (tBu-
CH3), 38.8 (C-4), 54.4 (C-2), 63.3 (C-1), 72.3 (C-3), 78.9 (tBu-C), 117.9
(C-6), 127.9 (Ph-CH), 128.0 (Ph-CH), 129.6 (Ph-CH), 130.0 (Ph-CH),
133.4 (Ph-C), 133.4 (C-5), 133.5 (Ph-C), 135.9 (Ph-CH), 135.9 (Ph-CH),
with a Platinum ATR crystal). Wavenumbers (ν) are quoted in cm–1
.
˜
UV spectroscopy was carried out with a Perkin–Elmer Lambda-2, a
Jasco V-630, or an Agilent Cary 50 spectrophotometer. Wavelengths
of maximum absorption (λmax) are reported in nm, with the corre-
sponding logarithmic molar extinction coefficient (log ε) given in
parentheses.
N-Boc-(4R,1′R)-4-[1′-(tert-Butyldiphenylsilyloxy)-3′-butenyl]-2,2-
dimethyl-1,3-oxazolidine [(R)-11c] and N-Boc-(4R,1′S)-4-[1′-
(tert-Butyldiphenylsilyloxy)-3′-butenyl]-2,2-dimethyl-1,3-ox-
azolidine [(S)-11c]: Imidazole (2.18 g, 32.0 mmol) and tert-butyldi-
phenylsilyl chloride (6.40 mL, 24.7 mmol) were added to a solution
of homoallylic alcohol 10 (2.17 g, 8.00 mmol) in dry DMF (7 mL).
156.1 (Boc-C=O) ppm. IR (ATR): ν = 2935, 1697, 1493, 1476, 1429,
˜
1365, 1249, 1168, 1109, 1087, 1051, 1011, 918, 824, 743, 703 cm–1
.
UV (MeCN): λmax (log ε) = 219 (4.23), 265 (2.83) nm. HRMS (ESI):
The mixture was stirred at room temp. for 3 d. Water (30 mL) was calcd. for C27H39NNaO4Si [M + Na]+ 492.2541; found 492.2545. TLC
added, and the aqueous layer was extracted with Et2O (3 × 70 mL).
The combined organic extracts were washed with saturated aque-
ous NaHCO3 solution (3 × 70 mL) and brine (70 mL), and dried with
Na2SO4. The solvent was evaporated under reduced pressure, and
the resulting crude product was purified by column chromatogra-
phy (isohexane/Et2O, 30:1) to give (R)-11c (1.88 g, 46 %) as a colour-
less oil, and (S)-11c (163 mg, 4 %) as a colourless oil. Furthermore,
a diastereomeric mixture of (R)-11c and (S)-11c (ratio 5:4, 1.17 g,
29 %) was isolated.
(isohexane/EtOAc, 2:1): Rf = 0.54.
tert-Butyl N-Boc-(2S,3R)-2-amino-3-(tert-butyldiphenylsilyloxy)-
hex-5-enoate [(R)-13c]: A mixture of (R)-12c (6.56 g, 14.0 mmol),
BAIB (9.90 g, 30.7 mmol), and TEMPO (655 mg, 4.19 mmol) in MeCN
(40 mL) and water (40 mL) was stirred at room temp. for 3 h. The
organic layer was washed with HCl (1
M aq.; 50 mL), and the aque-
ous layer was extracted with EtOAc (3 × 50 mL). The combined
organic extracts were dried with Na2SO4, and the solvent was evap-
orated under reduced pressure. The resulting crude product was
Data for (R)-11c: [α]D20 = +1.1 (c = 1.3, CHCl3). 1H NMR (500 MHz, purified by column chromatography (CH2Cl2/MeOH, 95:5) to give
[D6]DMSO, 100 °C): δ = 1.07 [s, 9 H, SiC(CH3)3], 1.31 (s, 9 H, tBu-CH3), the acid (6.31 g, 93 %) as colourless crystals.
1.38 [s, 3 H, C(CH3)2], 1.52 [s, 3 H, C(CH3)2], 2.20 (ddd, J = 14.7, 7.1,
The subsequent esterification was carried out as described previ-
1.3 Hz, 1 H, 2′a-H), 2.29–2.36 (m, 1 H, 2′b-H), 3.90–3.96 (m, 2 H, 4-
ously for compound (R)-13b[20] with the acid (6.18 g, 12.8 mmol),
H, 5a-H), 4.21 (dd, J = 8.6, 1.8 Hz, 1 H, 5b-H), 4.41–4.47 (m, 1 H, 1′-
tBuOH (16.0 mL, 12.0 g, 16.0 mol), N,N-dimethylformamide dineo-
H), 4.82–4.90 (m, 2 H, 4′-H), 5.57–5.68 (m, 1 H, 3′-H), 7.39–7.49 (m,
pentyl acetal (11.0 mL, 9.10 g, 39.0 mmol), and toluene (100 mL).
6 H, Ph-H), 7.61–7.68 (m, 4 H, Ph-H) ppm. 13C NMR (126 MHz,
The resulting crude product was purified by column chromatogra-
[D6]DMSO, 100 °C): δ = 19.4 [SiC(CH3)3], 23.6 [C(CH3)2], 26.5 [C(CH3)2],
phy (isohexane/EtOAc, 15:1) to give (R)-13c (3.99 g, 58 %) as a col-
27.4 [SiC(CH3)3], 28.5 (tBu-CH3), 36.2 (C-2′), 60.2 (C-4), 63.5 (C-5), 72.5
1
ourless oil. [α]D20 = –20.2 (c = 0.84, CHCl3). H NMR (500 MHz, C6D6):
(C-1′), 79.9 (tBu-C), 94.7 (C-2), 116.8 (C-4′), 127.9 (Ph-CH), 128.0 (Ph-
δ = 1.10 [s, 9 H, SiC(CH3)3], 1.30 (s, 9 H, tBu-CH3), 1.44 (s, 9 H, tBu-
CH), 130.1 (Ph-CH), 130.2 (Ph-CH), 134.2 (Ph-C), 134.3 (Ph-C), 135.9
CH3), 2.19–2.27 (m, 1 H, 4a-H), 2.47–2.58 (m, 1 H, 4b-H), 4.49 (ddd,
(Ph-CH), 136.0 (Ph-CH), 136.0 (C-3′), 152.3 (Boc-C=O) ppm. IR (ATR):
J = 10.2, 4.1, 1.7 Hz, 1 H, 3-H), 4.72 (dd, J = 9.9, 1.7 Hz, 1 H, 2-H),
ν = 2932, 1699, 1391, 1374, 1364, 1262, 1174, 1106, 1070, 914, 853,
˜
4.79–4.95 (m, 2 H, 6-H), 5.47–5.58 (m, 1 H, 5-H), 5.63 (d, J = 9.9 Hz,
826, 741, 700, 612 cm–1. UV (MeCN): λmax (log ε) = 219 (4.40), 266
1 H, NH), 7.13–7.20 (m, 6 H, Ph-H), 7.65–7.80 (m, 4 H, Ph-H) ppm.
(3.00) nm. HRMS (ESI): calcd. for C30H43NNaO4Si [M + Na]+ 532.2854;
13C NMR (126 MHz, C6D6): δ = 19.2 [SiC(CH3)3], 26.8 [SiC(CH3)3], 27.7
found 532.2864. TLC (isohexane/EtOAc, 12:1): Rf = 0.26.
(tBu-CH3), 28.1 (tBu-CH3), 39.0 (C-4), 56.8 (C-2), 74.6 (C-3), 79.1 (tBu-
Data for (S)-11c: [α]D20 = +38.0 (c = 1.1, CHCl3). 1H NMR (500 MHz, C), 81.1 (tBu-C), 118.6 (C-6), 127.5 (Ph-CH), 127.6 (Ph-CH), 129.7 (Ph-
[D6]DMSO, 100 °C): δ = 1.08 [s, 9 H, SiC(CH3)3], 1.41 (s, 9 H, tBu-CH3), CH), 129.8 (Ph-CH), 133.0 (C-5), 133.0 (Ph-C), 134.2 (Ph-C), 135.9 (Ph-
1.45 [s, 3 H, C(CH3)2], 1.53 [s, 3 H, C(CH3)2], 2.00–2.06 (m, 2 H, 2′-H), CH), 136.0 (Ph-CH), 156.0 (Boc-C=O), 170.6 (ester-C=O) ppm. IR
3.91 (ddd, J = 7.1, 3.8, 2.9 Hz, 1 H, 4-H), 3.97 (dd, J = 8.4, 7.1 Hz, 1
H, 5a-H), 4.17 (dd, J = 8.4, 3.8 Hz, 1 H, 5b-H), 4.43 (ddd, J = 7.8, 4.8,
(ATR): ν = 2975, 2929, 1720, 1493, 1429, 1365, 1325, 1151, 1109,
˜
1087, 906, 824, 743, 703, 616 cm–1. UV (MeCN): λmax (log ε) = 219
Eur. J. Org. Chem. 2016, 87–98
www.eurjoc.org
94
© 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
(4.68), 266 (3.30) nm. MS (ESI): m/z = 562.3 [M + Na]+. HRMS (ESI): (R)-13c (1.78 g, 3.30 mmol), MeOH (60 mL), CH2Cl2 (6.9 mL), pyridine
calcd. for C31H45NNaO5Si [M + Na]+ 562.2959; found 562.2958. TLC (1.10 mL, 1.10 g, 14.0 mmol), dimethyl sulfide (4.00 mL, 3.40 g,
(isohexane/EtOAc, 12:1): Rf = 0.31.
55.0 mmol), sodium borodeuteride (2.56 g, 61.2 mmol), and CD3OD
(40 mL). The mixture was stirred for 49 h at room temp. until TLC
indicated a nearly complete conversion. Column chromatography
(petroleum ether/EtOAc, 3:1→2:1) gave (R)-14c (1.27 g, 71 %) as a
colourless oil. [α]D20 = +1.7 (c = 0.66, CHCl3). 1H NMR (500 MHz, C6D6):
δ = 1.12 [s, 9 H, SiC(CH3)3], 1.31 (s, 9 H, tBu-CH3), 1.40 (s, 9 H, tBu-
CH3), 1.66–1.74 (m, 1 H, 4a-H), 1.87–1.96 (m, 1 H, 4b-H), 3.21 (dd,
J = 5.9, 5.9 Hz, 0.5 H, 5-H), 3.35 (dd, J = 5.4, 5.4 Hz, 0.5 H, 5-H), 4.73
(ddd, J = 9.5, 4.4, 2.0 Hz, 1 H, 3-H), 4.86 (dd, J = 9.5, 2.0 Hz, 1 H, 2-
H), 5.75 (d, J = 9.5 Hz, 1 H, NH), 7.17–7.22 (m, 6 H, Ph-H), 7.71–7.80
(m, 4 H, Ph-H) ppm. 13C NMR (126 MHz, C6D6): δ = 19.3 [SiC(CH3)3],
26.9 [SiC(CH3)3], 27.7 (tBu-CH3), 28.0 (tBu-CH3), 37.0 (C-4), 58.1 (C-2),
58.2–58.9 (m, C-5), 73.3 (C-3), 79.4 (tBu-C), 81.3 (tBu-C), 127.6 (Ph-
CH), 127.8 (Ph-CH), 129.7 (Ph-CH), 129.8 (Ph-CH), 133.3 (Ph-C), 134.5
(Ph-C), 135.9 (Ph-CH), 135.9 (Ph-CH), 156.4 (Boc-C=O), 170.3 (ester-
tert-Butyl N-Boc-(2S,3R)-2-amino-3-(benzyloxy)-[5-2H]-5-
hydroxyvalerate [(R)-14b]: A solution of (R)-13b (1.87 g,
4.78 mmol) in MeOH (60 mL), CH2Cl2 (6.9 mL), and pyridine
(1.50 mL, 1.50 g, 19.0 mmol) was cooled to –78 °C. Ozone was
bubbled through the solution at –78 °C for 30 min. After this time,
dimethyl sulfide (4.00 mL, 3.40 g, 55.0 mmol) was added, and the
reaction mixture was allowed to warm to room temp. overnight.
The solvent was then evaporated under reduced pressure.
The resulting crude aldehyde was immediately dissolved in CD3OD
(50 mL) under an argon atmosphere. Sodium borodeuteride (2.40 g,
57.3 mmol) was added portionwise to this solution at 0 °C, and the
reaction mixture was stirred at room temp. for 22 h, while being
monitored by TLC (petroleum ether/EtOAc, 2:1). The reaction was
quenched by the addition of satd. NH4Cl solution (10 mL), and the
aqueous layer was extracted with EtOAc (4 × 50 mL). The combined
organic extracts were washed with water (2 × 50 mL), and dried
with Na2SO4, and the solvent was evaporated under reduced pres-
sure. The resulting crude product was purified by column chroma-
tography (petroleum ether/EtOAc, 3:1→2:1) to give (R)-14b (1.68 g,
89 %) as a colourless oil. [α]D20 = +34.2 (c = 1.1, CHCl3). 1H NMR
(301 MHz, C6D6): δ = 1.30 (s, 9 H, tBu-CH3), 1.38 (s, 9 H, tBu-CH3),
1.62–1.89 (m, 2 H, 4-H), 2.12 (br. s, 1 H, OH), 3.48–3.59 (m, 1 H, 5-
H), 4.22–4.31 (m, 1 H, 3-H), 4.44 (s, 2 H, CH2-Ph), 4.70 (dd, J = 9.5,
2.1 Hz, 1 H, 2-H), 5.55 (d, J = 9.5 Hz, 1 H, NH), 7.02–7.30 (m, 5 H,
Ph-H) ppm. 13C NMR (126 MHz, C6D6): δ = 27.9 (tBu-CH3), 28.3 (tBu-
CH3), 34.1 (C-4), 57.6 (C-2), 58.7–59.2 (m, C-5), 72.6 (CH2-Ph), 78.2
(C-3), 79.5 (tBu-C), 81.6 (tBu-C), 127.8 (Ph-CH), 127.9 (Ph-CH), 128.5
(Ph-CH), 138.7 (Ph-C), 156.6 (Boc-C=O), 170.6 (ester-C=O) ppm. IR
C=O) ppm. IR (ATR): ν = 2935, 1720, 1493, 1372, 1249, 1151, 1104,
˜
1087, 1064, 941, 824, 790, 743, 703, 610 cm–1. UV (MeCN): λmax
(log ε) = 219 (3.67), 266 (2.42) nm. HRMS (ESI): calcd. for
C30H44DNNaO6Si [M + Na]+ 567.2971; found 567.2974. TLC (isohex-
ane/EtOAc, 1:1): Rf = 0.35.
N,N′,N′′-Tris-Alloc-guanidine (16): Benzyltriethylammonium chlor-
ide (107 mg, 0.470 mmol) and CH2Cl2 (40 mL) were added to a
solution of guanidine hydrochloride (2.01 g, 21.0 mmol) in NaOH
(6
M aq.; 14 mL). The mixture was stirred at 0 °C for 15 min, then
allyloxycarbonyl chloride (13.5 mL, 15.3 g, 127 mmol) was added,
and the reaction mixture was stirred for an additional 8 h at 0 °C.
The mixture was then partitioned between water (40 mL) and
CH2Cl2 (40 mL), and the aqueous layer was extracted with CH2Cl2
(2 × 40 mL). The combined organic extracts were dried with Na2SO4,
and the solvent was evaporated under reduced pressure. The result-
ing crude product was purified by column chromatography (CH2Cl2/
Et2O, 98:2→97:3→95:5→90:10) to give 16 (3.05 g, 47 %) as a colour-
(ATR): ν = 2975, 1714, 1499, 1453, 1372, 1342, 1255, 1151, 1064,
˜
1028, 848, 737, 703 cm–1. UV (MeCN): λmax (log ε) = 205 (3.92), 257
(2.38) nm. MS (ESI): m/z = 419.2 [M + Na]+. HRMS (ESI): calcd. for
C21H32DNNaO6 [M + Na]+ 419.2263; found 419.2260. TLC (petroleum
ether/EtOAc, 1:1): Rf = 0.36.
1
less oil. H NMR (500 MHz, CDCl3): δ = 4.65 (dt, J = 5.9, 1.4 Hz, 6 H,
1-H), 5.26 (ddt, J = 10.3, 1.4, 1.4 Hz, 3 H, 3a-H), 5.35 (ddt, J = 17.2,
1.4, 1.4 Hz, 3 H, 3b-H), 5.92 (ddt, J = 17.2, 10.3, 5.9 Hz, 3 H, 2-H),
10.99 (br. s, 2 H, NH) ppm. 13C NMR (126 MHz, CDCl3): δ = 67.3 (C-
1), 119.3 (C-3), 131.3 (C-2), 147.9 (guanidine-C), 152.2 (C=O) ppm. IR
tert-Butyl N-Boc-(2S,3S)-2-amino-3-(benzyloxy)-[5-2H]-5-
hydroxyvalerate [(S)-14b]: Isomer (S)-14b was prepared in the
same way as compound (R)-14b, using (S)-13b (1.64 g, 4.19 mmol),
MeOH (60 mL), CH2Cl2 (6.9 mL), pyridine (1.30 mL, 1.30 g,
16.0 mmol), dimethyl sulfide (4.00 mL, 3.40 g, 55.0 mmol), sodium
borodeuteride (3.22 g, 76.9 mmol), and CD3OD (31 mL). The mixture
was stirred for 64 h at room temp. until TLC indicated a nearly
complete conversion. Column chromatography (petroleum ether/
EtOAc, 3:1→2:1) gave (S)-14b (1.40 g, 84 %) as a colourless oil.
(ATR): ν = 3249, 1784, 1731, 1639, 1546, 1446, 1418, 1365, 1278,
˜
1214, 1174, 1132, 1057, 994, 935, 807, 767 cm–1. UV (MeCN): λmax
(log ε) = 205 (4.48), 238 (4.50) nm. MS (ESI): m/z = 312.1 [M + H]+.
HRMS (ESI): calcd. for C13H18N3O6 [M + H]+ 312.1190; found
312.1198. TLC (CH2Cl2/MeOH, 20:1): Rf = 0.40.
N2-Boc-N5,N7,N7′-tris-Cbz-(3R)-3-(Benzyloxy)- -[5-2H]arginine
L
tert-Butyl Ester [(R)-17b]: Tris-Cbz-guanidine 15 (11.2 g,
24.3 mmol) and triphenylphosphine (3.20 g, 12.2 mmol) were added
to a solution of (R)-14b (3.22 g, 8.13 mmol) in THF (100 mL) at room
temp. DIAD (2.40 mL, 2.46 g, 12.2 mmol) was then added dropwise
at 0 °C at such a rate that the reaction mixture was completely
colourless before the addition of the next drop. The reaction mix-
ture was warmed to room temp., and was stirred overnight at room
temp. The reaction was quenched by the addition of water (40 mL),
and the solvent was evaporated under reduced pressure. The result-
ing crude product was purified by column chromatography (petro-
leum ether/EtOAc, 5:1→4:1→3:1) to give (R)-17b (5.05 g, 74 %) as
a colourless viscous foam. [α]D20 = +8.6 (c = 0.78, CHCl3). 1H NMR
(300 MHz, C6D6): δ = 1.29 (s, 9 H, tBu-CH3), 1.40 (s, 9 H, tBu-CH3),
1.95 (ddd, J = 13.8, 6.8, 6.8 Hz, 1 H, 4a-H), 2.14 (ddd, J = 13.8, 6.8,
6.8 Hz, 1 H, 4b-H), 3.92 (dd, J = 6.8, 6.8 Hz, 0.5 H, 5-H), 4.06 (dd, J =
6.8, 6.8 Hz, 0.5 H, 5-H), 4.21 (ddd, J = 6.8, 6.8, 1.9 Hz, 1 H, 3-H), 4.44
(d, J = 11.3 Hz, 1 H, CH2-Ph), 4.52 (d, 1 H, J = 11.3 Hz, CH2-Ph), 4.60–
5.33 (m, 4 H, Cbz-CH2), 4.74 (dd, J = 9.6, 1.9 Hz, 1 H, 2-H), 4.91 (s, 2
1
[α]D20 = +1.9 (c = 0.89, CHCl3). H NMR (301 MHz, C6D6): δ = 1.27 (s,
9 H, tBu-CH3), 1.38 (s, 9 H, tBu-CH3), 1.61–1.84 (m, 2 H, 4-H), 1.98
(br. s, 1 H, OH), 3.53–3.63 (m, 1 H, 5-H), 3.96–4.05 (m, 1 H, 3-H), 4.32
(d, J = 11.4 Hz, 1 H, CH2-Ph), 4.61 (d, J = 11.4 Hz, 1 H, CH2-Ph), 4.84
(dd, J = 7.9, 2.8 Hz, 1 H, 2-H), 5.58 (d, J = 7.9 Hz, 1 H, NH), 7.02–7.31
(m, 5 H, Ph-H) ppm. 13C NMR (76 MHz, C6D6): δ = 27.9 (tBu-CH3),
28.3 (tBu-CH3), 33.9 (C-4), 56.5 (C-2), 59.7–59.9 (m, C-5), 72.1 (CH2-
Ph), 78.6 (C-3), 79.6 (tBu-C), 81.9 (tBu-C), 127.8 (Ph-CH), 128.0 (Ph-
CH), 128.5 (Ph-CH), 138.8 (Ph-C), 155.8 (Boc-C=O), 169.6 (ester-C=O)
ppm. IR (ATR): ν = 3435, 2978, 1708, 1498, 1366, 1249, 1150, 1096,
˜
1056, 1027, 737, 697 cm–1. UV (MeCN): λmax (log ε) = 205 (3.97), 258
(2.45) nm. MS (ESI): m/z = 419.2 [M + Na]+. HRMS (ESI): calcd. for
C21H32DNNaO6 [M + Na]+ 419.2263; found 419.2261. TLC (petroleum
ether/EtOAc, 1:1): Rf = 0.43.
tert-Butyl N-Boc-(2S,3R)-2-amino-3-(tert-butyldiphenylsilyloxy)-
[5-2H]-5-hydroxyvalerate [(R)-14c]: O-TBDPS-protected derivative
(R)-14c was prepared in the same way as compound (R)-14b, using
Eur. J. Org. Chem. 2016, 87–98
www.eurjoc.org
95
© 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
H, Cbz-CH2), 5.43 (d, J = 9.6 Hz, 1 H, NHBoc), 6.94–7.36 (m, 20 H,
Ph-H), 11.42 (s, 1 H, NHCbz) ppm. 13C NMR (126 MHz, C6D6): δ =
N2-Boc-N5,N7,N7′-tris-Alloc-(3R)-3-(tert-Butyldiphenylsilyloxy)-
L-
[5-2H]arginine tert-Butyl Ester [(R)-17d]: Compound (R)-14c
27.9 (tBu-CH3), 28.4 (tBu-CH3), 30.5 (C-4), 44.2–45.2 (m, C-5), 57.4 (C- (785 mg, 1.44 mmol) was dissolved in THF (18 mL). A solution of tris-
2), 68.1 (Cbz-CH2), 68.9 (Cbz-CH2), 72.2 (CH2-Ph), 77.7 (C-3), 79.3 Alloc-guanidine 16 (1.35 g, 4.34 mmol) in THF (12 mL) was added
(tBu-C), 81.4 (tBu-C), 127.6 (Ph-CH), 127.8 (Ph-CH), 127.9 (Ph-CH),
128.3 (Ph-CH), 128.5 (Ph-CH), 128.6 (Ph-CH), 128.6 (Ph-CH), 128.7
(Ph-CH), 128.8 (Ph-CH), 135.4 (Ph-C), 138.7 (Ph-C), 152.3 (guanidine-
C), 154.7 (Cbz-C=O), 156.3 (Boc-C=O), 170.5 (ester-C=O) ppm. IR
at room temp., followed by triphenylphosphine (570 mg,
2.17 mmol). DIAD (425 μL, 436 mg, 2.16 mmol) was then added
dropwise at 0 °C at such a rate that the reaction mixture was com-
pletely colourless before the addition of the next drop. The reaction
(ATR): ν = 2981, 1767, 1720, 1656, 1610, 1499, 1453, 1372, 1208, mixture was warmed to room temp., and stirred overnight at room
˜
1191, 1151, 1051, 1028, 737, 697 cm–1. UV (MeCN): λmax (log ε) = temp. The reaction was quenched by the addition of water (20 mL),
205 (4.64), 245 (4.01) nm. MS (ESI): m/z = 862.4 [M + Na]+. HRMS
(ESI): calcd. for C46H53DN4NaO11 [M + Na]+ 862.3744; found
862.3759. TLC (petroleum ether/EtOAc, 3:1): Rf = 0.36.
and the solvent was evaporated under reduced pressure. The result-
ing crude product was purified by column chromatography (isohex-
ane/EtOAc, 6:1) to give (R)-17d (1.08 g, 90 %) as a colourless viscous
1
oil. [α]D20 = –7.3 (c = 0.88, CHCl3). H NMR (500 MHz, C6D6): δ = 1.14
N2-Boc-N5,N7,N7′-tris-Cbz-(3S)-3-(Benzyloxy)- -[5-2H]arginine
L
[s, 9 H, SiC(CH3)3], 1.37 (s, 9 H, tBu-CH3), 1.42 (s, 9 H, tBu-CH3), 2.01–
2.11 (m, 1 H, 4a-H), 2.11–2.22 (m, 1 H, 4b-H), 3.66 (dd, J = 5.7, 5.7 Hz,
0.5 H, 5-H), 4.00 (dd, J = 7.4, 7.4 Hz, 0.5 H, 5-H), 4.22–4.73 (m, 7 H,
3-H, Alloc-1-H), 4.87–5.27 (m, 7 H, 2-H, Alloc-3-H), 5.58–5.89 (m, 4
H, Alloc-2-H, NHBoc), 7.19–7.34 (m, 6 H, Ph-H), 7.73–7.86 (m, 4 H,
Ph-H), 11.17 (br. s, 1 H, NHAlloc) ppm. 13C NMR (126 MHz, C6D6):
δ = 19.3 [SiC(CH3)3], 27.0 [SiC(CH3)3], 27.8 (tBu-CH3), 28.1 (tBu-CH3),
33.0 (C-4), 43.7–44.4 (m, C-5), 57.4 (C-2), 66.4 (Alloc-C-1), 67.5 (Alloc-
C-1), 72.7 (C-3), 79.0 (tBu-C), 81.2 (tBu-C), 127.5 (Ph-CH), 127.6 (Ph-
CH), 129.6 (Ph-CH), 131.4 (Alloc-C-2), 132.2 (Alloc-C-2), 133.5 (Ph-C),
134.2 (Ph-C), 136.0 (Ph-CH), 136.0 (Ph-CH), 151.0 (guanidine-C),
154.1 (Alloc-C=O), 156.0 (Boc-C=O), 170.3 (ester-C=O) ppm. IR (ATR):
tert-Butyl Ester [(S)-17b]: Isomer (S)-17b was prepared in the same
way as compound (R)-17b, using (S)-14b (1.24 g, 3.13 mmol), tris-
Cbz-guanidine 15 (4.33 g, 9.38 mmol), triphenylphosphine (1.23 g,
4.69 mmol), DIAD (930 μL, 955 mg, 4.72 mmol), and THF (35 mL).
Column chromatography (petroleum ether/EtOAc, 5:1→4:1→3:1)
gave (S)-17b (2.11 g, 80 %) as a colourless viscous foam. [α]D20 = +3.4
(c = 0.92, CHCl3). 1H NMR (300 MHz, C6D6): δ = 1.27 (s, 9 H, tBu-
CH3), 1.33 (s, 9 H, tBu-CH3), 1.75–1.97 (m, 2 H, 4-H), 3.88–3.99 (m, 2
H, 3-H, 5-H), 4.32 (d, J = 11.0 Hz, 1 H, CH2-Ph), 4.67 (d, 1 H, J =
11.0 Hz, CH2-Ph), 4.71–5.24 (m, 4 H, Cbz-CH2), 4.82 (s, 2 H, Cbz-CH2),
4.89 (dd, J = 8.5, 3.3 Hz, 1 H, 2-H), 5.33 (d, J = 8.5 Hz, 1 H, NHBoc),
6.94–7.40 (m, 20 H, Ph-H), 11.34 (s, 1 H, NHCbz) ppm. 13C NMR
(126 MHz, C6D6): δ = 27.9 (tBu-CH3), 28.3 (tBu-CH3), 29.7 (C-4), 44.7–
45.5 (m, C-5), 56.0 (C-2), 68.1 (Cbz-CH2), 68.9 (Cbz-CH2), 71.6 (CH2-
Ph), 77.7 (C-3), 79.3 (tBu-CH3), 82.0 (tBu-CH3), 127.5 (Ph-CH), 127.8
(Ph-CH), 127.9 (Ph-CH), 128.3 (Ph-CH), 128.5 (Ph-CH), 128.5 (Ph-CH),
128.6 (Ph-CH), 128.7 (Ph-CH), 135.4 (Ph-C), 139.0 (Ph-C), 151.7
(guanidine-C), 154.8 (Cbz-C=O), 155.7 (Boc-C=O), 169.7 (ester-C=O)
ν = 2929, 1748, 1731, 1708, 1673, 1603, 1522, 1365, 1208, 1151,
˜
1064, 935, 767, 703, 616 cm–1. UV (MeCN): λmax (log ε) = 219 (4.76),
243 (4.37) nm. HRMS (ESI): calcd. for C43H59DN4NaO11Si [M + Na]+
860.3983; found 860.3990. TLC (isohexane/EtOAc, 3:1): Rf = 0.38.
N2-Boc-(3R)-3-hydroxy- -[5-2H]arginine tert-Butyl Ester [(R)-19]:
L
Compound (R)-17b (1.05 g, 1.25 mmol) was dissolved in EtOAc
ppm. IR (ATR): ν = 2981, 1767, 1714, 1650, 1610, 1499, 1453, 1372,
(20 mL) and EtOH (20 mL). TFA (180 μL, 266 mg, 2.33 mmol) and
˜
1219, 1197, 1151, 1051, 1028, 737, 697 cm–1. UV (MeCN): λmax Pd/C (10 %; 530 mg, 0.498 mmol Pd) were added, and the reaction
(log ε) = 206 (4.60), 235 (4.09) nm. MS (ESI): m/z = 862.4 [M + Na]+.
HRMS (ESI): calcd. for C46H53DN4NaO11 [M + Na]+ 862.3744; found
862.3754. TLC (petroleum ether/EtOAc, 3:1): Rf = 0.28.
mixture was stirred under an atmosphere of hydrogen (1 bar) at
room temp. for 22 h. The mixture was then filtered through Celite.
The Celite was washed thoroughly with EtOAc and EtOH, and the
solvent of the filtrate was evaporated under reduced pressure. The
resulting crude product was purified by column chromatography
(RP silica gel 90 C18, H2O/MeCN, 4:1→2:1) to give (R)-19 (429 mg,
98 %) as a colourless solid, m.p. 56 °C. [α]D20 = +1.1 (c = 1.1, MeOH).
1H NMR (500 MHz, CD3OD): δ = 1.48 (s, 9 H, tBu-CH3), 1.51 (s, 9 H,
tBu-CH3), 1.70–1.83 (m, 2 H, 4-H), 3.28–3.36 (m, 1 H, 5-H), 4.06–4.17
(m, 2 H, 2-H, 3-H) ppm. 13C NMR (126 MHz, CD3OD): δ = 26.9 (tBu-
CH3), 27.3 (tBu-CH3), 32.8 (C-4), 37.4–38.2 (m, C-5), 58.8 (C-2), 68.7
(C-3), 79.6 (tBu-C), 81.9 (tBu-C), 156.9 (guanidine-C), 157.4 (Boc-C=
N2-Boc-N5,N7,N7′-tris-Cbz-(3R)-3-(tert-Butyldiphenylsilyloxy)-
L-
[5-2H]arginine tert-Butyl Ester [(R)-17c]: O-TBDPS-protected de-
rivative (R)-17c was prepared in the same way as compound (R)-
17b, using (R)-14c (1.27 g, 2.33 mmol), tris-Cbz-guanidine 15
(3.23 g, 7.00 mmol), triphenylphosphine (917 mg, 3.49 mmol), DIAD
(690 μL, 709 mg, 3.51 mmol), and THF (40 mL). Column chromatog-
raphy (petroleum ether/EtOAc, 5:1→4:1→3:1) gave (R)-17c (1.86 g,
81 %) as a colourless oil. [α]D20 = –10.4 (c = 0.99, CHCl3). 1H NMR
(500 MHz, C6D6): δ = 1.12 [s, 9 H, SiC(CH3)3], 1.34 (s, 9 H, tBu-CH3),
1.42 (s, 9 H, tBu-CH3), 1.98–2.09 (m, 1 H, 4a-H), 2.10–2.21 (m, 1 H,
4b-H), 3.61 (dd, J = 6.9, 6.9 Hz, 0.5 H, 5-H), 3.98 (dd, J = 7.6, 7.6 Hz,
0.5 H, 5-H), 4.55–5.43 (m, 7 H, 2-H, Cbz-CH2), 4.64–4.71 (m, 1 H, 3-
H), 5.64 (d, J = 9.6 Hz, 1 H, NHBoc), 6.50–7.46 (m, 21 H, Ph-H), 7.74–
7.83 (m, 4 H, Ph-H), 11.20 (br. s, 1 H, NHCbz) ppm. 13C NMR
(126 MHz, C6D6): δ = 19.3 [SiC(CH3)3], 27.0 [SiC(CH3)3], 27.8 (tBu-
CH3), 28.1 (tBu-CH3), 33.0 (C-4), 43.4–44.5 (m, C-5), 57.4 (C-2), 67.6
(Cbz-CH2), 68.6 (Cbz-CH2), 72.7 (C-3), 79.1 (tBu-C), 81.2 (tBu-C), 127.6
O), 170.1 (ester-C=O) ppm. IR (ATR): ν = 3353, 1667, 1621, 1510,
˜
1372, 1255, 1202, 1179, 1138, 1064, 843, 801, 726 cm–1. HRMS (ESI):
calcd. for C15H30DN4O5 [M + H]+ 348.2357; found 348.2355. TLC (iP-
rOH/H2O/AcOH, 5:2:1 as satd. NaCl solution): Rf = 0.67–0.76. TLC
(CH2Cl2/MeOH, 5:1): Rf = 0.00–0.12.
N2-Boc-(3S)-3-hydroxy- -[5-2H]arginine tert-Butyl Ester [(S)-19]:
L
Isomer (S)-19 was prepared in the same way as compound (R)-19,
using (S)-17b (2.00 g, 2.38 mmol), TFA (330 μL, 488 mg, 4.28 mmol),
(Ph-CH), 127.8 (Ph-CH), 128.2 (Ph-CH), 128.3 (Ph-CH), 128.3 (Ph-CH), Pd/C (10 %; 1.01 g, 0.952 mmol Pd), EtOAc (40 mL), and EtOH
128.4 (Ph-CH), 128.4 (Ph-CH), 129.6 (Ph-CH), 129.7 (Ph-CH), 133.5
(Ph-C), 134.0 (Ph-C), 135.1 (Ph-C), 135.9 (Ph-CH), 136.0 (Ph-CH),
150.8 (guanidine-C), 154.4 (Cbz-C=O), 156.1 (Boc-C=O), 170.3 (ester-
(40 mL). After 23 h, additional Pd/C (10 %; 240 mg, 0.226 mmol Pd)
was added, and the reaction mixture was stirred for a total of 60 h.
Column chromatography (RP silica gel 90 C18, H2O/MeCN, 4:1→2:1)
C=O) ppm. IR (ATR): ν = 2935, 1767, 1714, 1661, 1610, 1499, 1372, gave (S)-19 (838 mg, 100 % yield: 827 mg) as a slightly impure
˜
1
1214, 1191, 1151, 1109, 1051, 1028, 743, 697 cm–1. UV (MeCN): λmax
(log ε) = 210 (5.28), 244 (4.62) nm. HRMS (ESI): calcd. for
C55H65DN4NaO11Si [M + Na]+ 1010.4452; found 1010.4476. TLC (iso-
hexane/EtOAc, 3:1): Rf = 0.53.
colourless solid, m.p. 56 °C. [α]D20 = –10.0 (c = 1.2, MeOH). H NMR
(500 MHz, CD3OD): δ = 1.47 (s, 9 H, tBu-CH3), 1.51 (s, 9 H, tBu-CH3),
1.74–1.88 (m, 2 H, 4-H), 3.30–3.36 (m, 1 H, 5-H), 3.91 (ddd, J = 9.1,
4.8, 4.8 Hz, 1 H, 3-H), 4.09 (d, J = 4.8 Hz, 1 H, 2-H) ppm. 13C NMR
Eur. J. Org. Chem. 2016, 87–98
www.eurjoc.org
96
© 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
(126 MHz, CD3OD): δ = 26.9 (tBu-CH3), 27.3 (tBu-CH3), 31.9 (C-4),
37.4–38.3 (m, C-5), 59.3 (C-2), 69.3 (C-3), 79.5 (tBu-C), 81.8 (tBu-C),
156.5 (guanidine-C), 157.4 (Boc-C=O), 169.8 (ester-C=O) ppm. IR
calcd. for C6H14DN4O3 [M + H]+ 192.1207; found 192.1200. TLC (iP-
rOH/H2O/AcOH, 5:2:1 as satd. NaCl solution): Rf = 0.15.
Procedure 2: A solution of (R)-19 (273 mg, 0.786 mmol) in HCl (6
M
(ATR): ν = 3353, 1667, 1627, 1510, 1372, 1255, 1197, 1179, 1138,
˜
aq.; 8 mL) was stirred at room temp. for 3 h. The solvent was then
evaporated under reduced pressure. The resulting crude product
was purified by column chromatography (RP silica gel 90 C18, H2O)
to give (R)-27 [164 mg, 79 %, 77 % over two steps from (R)-17b] as
a colourless-to-yellowish hygroscopic solid.
1064, 1017, 843, 801, 720 cm–1. HRMS (ESI): calcd. for C15H30DN4O5
[M + H]+ 348.2357; found 348.2355. TLC (iPrOH/H2O/AcOH, 5:2:1 as
satd. NaCl solution): Rf = 0.63–0.75. TLC (CH2Cl2/MeOH, 5:1): Rf =
0.00–0.20.
L
N2-Boc-(3R)-3-(tert-butyldiphenylsilyloxy)- -[5-2H]arginine tert-
(3S)-3-Hydroxy-L
-[5-2H]arginine Dihydrochloride [(S)-27]: Isomer
(S)-27 was prepared in the same way as compound (R)-27 by Proce-
dure 2 using (S)-19 (739 mg slightly impure material, vide supra)
Butyl Ester [(R)-26]
Procedure 1: Dimedone (572 mg, 4.08 mmol) and tetrakis(triphen-
ylphosphine)-palladium (47 mg, 41 μmol) were added to a solution
of (R)-17d (683 mg, 0.815 mmol) in THF (15 mL), and the resulting
mixture was stirred in the dark at room temp. overnight. The reac-
tion mixture was then partitioned between Et2O (30 mL) and water
and HCl (6
M aq.; 20 mL). Column chromatography (RP silica gel 90
C
18, H2O) gave (S)-27 [464 mg, 83 % over two steps from (S)-17b]
as a colourless-to-yellowish hygroscopic solid. [α]D20 = +0.9 (c = 2.0,
H2O). 1H NMR (500 MHz, D2O): δ = 1.84–1.98 (m, 2 H, 4-H), 3.28–
3.37 (m, 1 H, 5-H), 4.13–4.23 (m, 2 H, 2-H, 3-H) ppm. 13C NMR
(126 MHz, D2O): δ = 31.0 (C-4), 37.5–38.0 (m, C-5), 57.7 (C-2), 67.0
(50 mL), and the organic layer was washed with HCl (1
M aq.; 3 ×
20 mL). The combined aqueous layers were extracted with Et2O (4 ×
100 mL). The combined organic extracts were dried with Na2SO4,
and the solvent was evaporated under reduced pressure. The result-
ing crude product was purified by column chromatography (CH2Cl2/
MeOH, 8:1→5:1) to give (R)-26 (586 mg, 100 % yield: 477 mg) as an
impure yellowish foam, m.p. 90 °C. [α]D20 = –24.9 (c = 1.1, MeOH). 1H
NMR (500 MHz, CD3OD): δ = 1.05 [s, 9 H, SiC(CH3)3], 1.45 (s, 9 H,
tBu-CH3), 1.51 (s, 9 H, tBu-CH3), 1.63–1.72 (m, 1 H, 4a-H), 1.74–1.85
(m, 1 H, 4b-H), 2.89 (dd, J = 7.8, 7.8 Hz, 0.5 H, 5-H), 3.07 (dd, J = 6.9,
6.9 Hz, 0.5 H, 5-H), 4.21–4.27 (m, 1 H, 2-H), 4.29–4.36 (m, 1 H, 3-H),
7.41–7.54 (m, 6 H, Ph-H), 7.69–7.76 (m, 4 H, Ph-H) ppm. 13C NMR
(126 MHz, CD3OD): δ = 18.3 [SiC(CH3)3], 26.1 [SiC(CH3)3], 27.0 (tBu-
CH3), 27.3 (tBu-CH3), 33.2 (C-4), 36.8–37.4 (m, C-5), 57.2 (C-2), 71.9
(C-3), 80.0 (tBu-C), 82.3 (tBu-C), 127.4 (Ph-CH), 127.6 (Ph-CH), 129.8
(Ph-CH), 129.9 (Ph-CH), 132.5 (Ph-C), 133.6 (Ph-C), 136.6 (Ph-CH),
156.9 (guanidine-C), 157.0 (Boc-C=O), 169.9 (ester-C=O) ppm. IR
(C-3), 156.9 (guanidine-C), 169.4 (C=O) ppm. IR (ATR): ν = 3330,
˜
3179, 2899, 1731, 1650, 1616, 1499, 1429, 1214, 1161, 1057 cm–1
.
HRMS (ESI): calcd. for C6H14DN4O3 [M + H]+ 192.1207; found
192.1205. TLC (iPrOH/H2O/AcOH, 5:2:1 as satd. NaCl solution): Rf =
0.12.
Supporting Information (see footnote on the first page of this
1
article): H and 13C NMR spectra of the synthesized compounds.
Acknowledgments
The authors thank the Fonds der Chemischen Industrie (FCI)
and the State of Lower Saxony [Lichtenberg doctoral fellowship
(CaSuS program) to A. L.] for financial support.
(ATR): ν = 2935, 1703, 1673, 1505, 1476, 1372, 1330, 1255, 1151,
˜
Keywords: Synthesis design · Natural products · Amino
acids · Protecting groups · Deuterium · Isotopic labeling
1109, 1087, 824, 743, 703, 610 cm–1. UV (MeCN): λmax (log ε) = 219
(4.35), 266 (3.18) nm. HRMS (ESI): calcd. for C31H48DN4O5Si [M + H]+
586.3530; found 586.3524. TLC (CH2Cl2/MeOH, 5:1): Rf = 0.30.
[1] a) D. Schwarzer, R. Finking, M. A. Marahiel, Nat. Prod. Rep. 2003, 20, 275–
287; b) C. T. Walsh, R. V. O'Brien, C. Khosla, Angew. Chem. Int. Ed. 2013,
52, 7098–7124; Angew. Chem. 2013, 125, 7238–7265.
[2] a) X. Yin, T. M. Zabriskie, ChemBioChem 2004, 5, 1274–1277; b) X. Yin,
K. L. McPhail, K.-j. Kim, T. M. Zabriskie, ChemBioChem 2004, 5, 1278–
1281; c) J. Ju, S. G. Ozanick, B. Shen, M. G. Thomas, ChemBioChem 2004,
5, 1281–1285.
[3] a) A. Nagata, T. Ando, R. Izumi, H. Sakakibara, T. Take, J. Antibiot. 1968,
21, 681–687; b) H. Yoshioka, T. Aoki, H. Goko, K. Nakatsu, T. Noda, H.
Sakakibara, T. Take, A. Nagata, J. Abe, T. Wakamiya, T. Shiba, T. Kaneko,
Tetrahedron Lett. 1971, 12, 2043–2046.
[4] E. B. Herr, Antimicrob. Agents Chemother. 1962, 201–202.
[5] B. W. Bycroft, L. R. Croft, A. W. Johnson, T. Webb, J. Antibiot. 1969, 22,
133–134.
[6] V. Helmetag, S. A. Samel, M. G. Thomas, M. A. Marahiel, L.-O. Essen, FEBS
J. 2009, 276, 3669–3682.
[7] L. A. McDonald, L. R. Barbieri, G. T. Carter, E. Lenoy, J. Lotvin, P. J. Petersen,
M. M. Siegel, G. Singh, R. T. Williamson, J. Am. Chem. Soc. 2002, 124,
10260–10261.
[8] L. Cheng, W. Chen, L. Zhai, D. Xu, T. Huang, S. Lin, X. Zhou, Z. Deng, Mol.
BioSyst. 2011, 7, 920–927.
[9] R. B. Hamed, J. R. Gomez-Castellanos, L. Henry, C. Ducho, M. A. McDon-
ough, C. J. Schofield, Nat. Prod. Rep. 2013, 30, 21–107.
Procedure 2: Compound (R)-17c (213 mg, 0.215 mmol) was dis-
solved in EtOAc (4 mL) and EtOH (4 mL). TFA (30 μL, 44 mg,
0.39 mmol) and Pd/C (10 %; 50 mg, 47 μmol Pd) were added, and
the reaction mixture was stirred under an atmosphere of hydrogen
(1 bar) at room temp. for 80 min. The mixture was then filtered
through a syringe filter, and the syringe filter was thoroughly
washed with EtOH. The solvent of the filtrate was evaporated under
reduced pressure to give (R)-26 (141 mg, 100 % yield: 126 mg) as
an impure colourless foam.
(3R)-3-Hydroxy-
L
-[5-2H]arginine Dihydrochloride [(R)-27]
Procedure 1: A suspension of (R)-26 [298 mg impure material from
Procedure 1 for the synthesis of (R)-26, vide supra] in HCl (6 aq.;
M
8 mL) was stirred at room temp. Every 2 h for the first 10 h, the
reaction mixture was put into an ultrasonic bath for 15 min. After
24 h, the solvent was evaporated under reduced pressure. The re-
sulting crude product was purified by column chromatography (RP
silica gel 90 C18, H2O) to give (R)-27 [105 mg, 90 % over two steps
from (R)-17d] as a colourless-to-yellowish hygroscopic solid. [α]D20
=
+23.5 (c = 2.2, H2O). 1H NMR (500 MHz, D2O): δ = 1.82 (ddd, J =
14.1, 10.5, 6.1 Hz, 1 H, 4a-H), 1.97 (ddd, J = 14.1, 7.7, 3.1 Hz, 1 H,
4b-H), 3.31–3.39 (m, 1 H, 5-H), 4.04 (d, J = 4.1 Hz, 1 H, 2-H), 4.28
(ddd, J = 10.5, 4.1, 3.1 Hz, 1 H, 3-H) ppm. 13C NMR (126 MHz, D2O):
δ = 32.0 (C-4), 37.4–37.8 (m, C-5), 57.9 (C-2), 66.5 (C-3), 156.9 (guan
[10] a) R. Li, A. Stapon, J. T. Blanchfield, C. A. Townsend, J. Am. Chem. Soc.
2000, 122, 9296–9297; b) A. Stapon, R. Li, C. A. Townsend, J. Am. Chem.
Soc. 2003, 125, 15746–15747; c) M. Topf, G. M. Sandala, D. M. Smith, C. J.
Schofield, C. J. Easton, L. Radom, J. Am. Chem. Soc. 2004, 126, 9932–
9933; d) N. J. Kershaw, M. E. C. Caines, M. C. Sleeman, C. J. Schofield,
Chem. Commun. 2005, 4251–4263; e) R. B. Hamed, E. T. Batchelar, J. Me-
idine-C), 170.3 (C=O) ppm. IR (ATR): ν = 3341, 3173, 2918, 1737,
˜
1661, 1621, 1516, 1429, 1278, 1226, 1081, 1040 cm–1. HRMS (ESI):
Eur. J. Org. Chem. 2016, 87–98
www.eurjoc.org
97
© 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
cinovic, T. D. W. Claridge, C. J. Schofield, ChemBioChem 2009, 10, 246–
250.
[19]
J. Wityak, S. J. Gould, S. J. Hein, D. A. Keszler, J. Org. Chem. 1987, 52,
2179–2183.
[11] a) J. E. Baldwin, M. D. Lloyd, B. Wha-Son, C. J. Schofield, S. W. Elson, K. H.
Baggaley, N. H. Nicholson, J. Chem. Soc., Chem. Commun. 1993, 500–502;
b) M. D. Lloyd, K. D. Merritt, V. Lee, T. J. Sewell, B. Wha-Son, J. E. Baldwin,
C. J. Schofield, S. W. Elson, K. H. Baggaley, N. H. Nicholson, Tetrahedron
1999, 55, 10201–10220; c) R. W. Busby, C. A. Townsend, Bio-
org. Med. Chem. 1996, 4, 1059–1064.
[12] a) D. J. Slade, V. Subramanian, J. Fuhrmann, P. R. Thompson, Biopolymers
2014, 101, 133–143; b) V. V. Papov, T. V. Diamond, K. Biemann, J. H. Waite,
J. Biol. Chem. 1995, 270, 20183–20192; c) P. Hänzelmann, H. Dobbek, L.
Gremer, R. Huber, O. Meyer, J. Mol. Biol. 2000, 301, 1221–1235; d) J. Car-
roll, S. Ding, I. M. Fearnley, J. E. Walker, J. Biol. Chem. 2013, 288, 24799–
24808.
[13] W. Ge, A. Wolf, T. Feng, C.-h. Ho, R. Sekirnik, A. Zayer, N. Granatino, M. E.
Cockman, C. Loenarz, N. D. Loik, A. P. Hardy, T. D. W. Claridge, R. B.
Hamed, R. Chowdhury, L. Gong, C. V. Robinson, D. C. Trudgian, M. Jiang,
M. M. Mackeen, J. S. McCullagh, Y. Gordiyenko, A. Thalhammer, A. Yama-
moto, M. Yang, P. Liu-Yi, Z. Zhang, M. Schmidt-Zachmann, B. M. Kessler,
P. J. Ratcliffe, G. M. Preston, M. L. Coleman, C. J. Schofield, Nat. Chem.
Biol. 2012, 8, 960–962.
[14] a) J.-P. Genet, Pure Appl. Chem. 1996, 68, 593–596; b) J.-P. Genet, Pure
Appl. Chem. 2002, 74, 77–83; c) V. Ratovelomanana-Vidal, J.-P. Genet, Can.
J. Chem. 2000, 78, 846–851; d) J.-P. Genet, Acc. Chem. Res. 2003, 36, 908–
918; e) R. Noyori, T. Ikeda, T. Ohkuma, M. Widhalm, M. Kitamura, H. Tak-
aya, S. Akutagawa, N. Sayo, T. Saito, J. Am. Chem. Soc. 1989, 111, 9134–
9135; f) J. Ohtaka, A. Hamajima, T. Nemoto, Y. Hamada, Chem. Pharm.
Bull. 2013, 61, 245–250.
[15] a) J. A. Bodkin, M. D. McLeod, J. Chem. Soc. Perkin Trans. 1 2002, 2733–
2746; b) S. Mekki, S. Bellahouel, N. Vanthuyne, M. Rolland, A. Derdour, J.
Martinez, M. Vignes, V. Rolland, Tetrahedron: Asymmetry 2012, 23, 94–99.
[16] H. C. Kolb, M. S. VanNieuwenhze, K. B. Sharpless, Chem. Rev. 1994, 94,
2483–2547.
[17] a) M. Horikawa, J. Busch-Petersen, E. J. Corey, Tetrahedron Lett. 1999, 40,
3843–3846; b) N. Yoshikawa, M. Shibasaki, Tetrahedron 2002, 58, 8289–
8298; c) J. Wehbe, T. Kassem, V. Rolland, M. Rolland, M. Tabcheh, M.-L.
Roumestant, J. Martinez, Org. Biomol. Chem. 2003, 1, 1938–1942; d) J.
Kobayashi, M. Nakamura, Y. Mori, Y. Yamashita, S. Kobayashi, J. Am. Chem.
Soc. 2004, 126, 9192–9193.
[20]
[21]
[22]
A. Lemke, M. Büschleb, C. Ducho, Tetrahedron 2010, 66, 208–214.
S. J. Gould, J. Lee, J. Wityak, Bioorg. Chem. 1991, 19, 333–350.
M. D. Jackson, S. J. Gould, T. M. Zabriskie, J. Org. Chem. 2002, 67, 2934–
2941.
[23]
[24]
[25]
[26]
a) P. Garner, J. M. Park, J. Org. Chem. 1987, 52, 2361–2364; b) A. Dondoni,
D. Perrone, Org. Synth. 2000, 77, 64–70.
A. De Mico, R. Margarita, L. Parlanti, A. Vescovi, G. Piancatelli, J. Org.
Chem. 1997, 62, 6974–6977.
H. Brechbühler, H. Büchi, E. Hatz, J. Schreiber, A. Eschen-
moser, Helv. Chim. Acta 1965, 48, 1746–1771.
a) J. L. Sorensen, M. C. Sleeman, C. J. Schofield, Chem. Commun. 2005,
1155–1157; b) C. Ducho, R. B. Hamed, E. T. Batchelar, J. L. Sorensen, B.
Odell, C. J. Schofield, Org. Biomol. Chem. 2009, 7, 2770–2779; c) E. T.
Batchelar, R. B. Hamed, C. Ducho, T. D. W. Claridge, M. J. Edelmann, B.
Kessler, C. J. Schofield, Angew. Chem. Int. Ed. 2008, 47, 9322–9325; Angew.
Chem. 2008, 120, 9462–9465.
[27]
K. Feichtinger, H. L. Sings, T. J. Baker, K. Matthews, M. Goodman, J. Org.
Chem. 1998, 63, 8432–8439.
[28] C. W. Zapf, M. Goodman, J. Org. Chem. 2003, 68, 10092–10097.
[29] a) D. C. Gowda, K. Abiraj, Lett. Pept. Sci. 2002, 9, 153–165; b) G. R. Sriniv-
asa, S. N. N. Babu, C. Lakshmi, D. C. Gowda, Synth. Commun. 2004, 34,
1831–1837.
[30] S. Hirano, S. Ichikawa, A. Matsuda, Bioorg. Med. Chem. 2008, 16, 5123–
5133.
[31] R. Ramapanicker, R. Mishra, S. Chandrasekaran, J. Pept. Sci. 2010, 16, 123–
125.
[32] T. D. Nelson, R. D. Crouch, Synthesis 1996, 1031–1069.
[33] D. Fishlock, J. G. Guillemette, G. A. Lajoie, J. Org. Chem. 2002, 67, 2352–
2354.
[34] M. Büschleb, M. Granitzka, D. Stalke, C. Ducho, Amino Acids 2012, 43,
2313–2328.
[35] a) M. P. Coles, Dalton Trans. 2006, 985–1001; b) J. Shao, W. Chen, M. A.
Giulianotti, R. A. Houghten, Y. Yu, Org. Lett. 2012, 14, 5452–5455.
[18] a) T. Laïb, J. Chastanet, J. Zhu, J. Org. Chem. 1998, 63, 1709–1713; b) O.
Ries, M. Büschleb, M. Granitzka, D. Stalke, C. Ducho, Beilstein J. Org. Chem.
2014, 10, 1135–1142.
Received: August 26, 2015
Published Online: November 24, 2015
Eur. J. Org. Chem. 2016, 87–98
www.eurjoc.org
98
© 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim