Convenient synthesis of the antibiotic linezolid via an oxazolidine-2,4-dione intermediate derived from the chiral building block isoserine

Article abstract of DOI:10.1002/ejoc.201402888

We describe a new synthesis of the 5-(aminomethyl)oxazolidin-3-one core of linezolid in enantiomerically pure form. The expedient cyclization of the α-hydroxy amide derived from isoserine and 3-fluoro-4-morpholinoaniline to give the corresponding (aminomethyl)oxazolidine-2,4-dione, followed by its mild selective reduction at the C(4)-position, gave linezolid in almost quantitative overall yield. The 1,3-oxazolidin-2-one core of linezolid was obtained from isoserine in just three steps and with almost quantitative overall yield; the key features of the protocol are the expedient formation of the intermediate oxazolidine-2,4-dione, and its regioselective reduction at the 4-position.

Full text of DOI:10.1002/ejoc.201402888

FULL PAPER
DOI: 10.1002/ejoc.201402888
Convenient Synthesis of the Antibiotic Linezolid via an Oxazolidine-2,4-dione
Intermediate Derived from the Chiral Building Block Isoserine
Arianna Greco,[a] Rossella De Marco,[a] Sara Tani,[a] Daria Giacomini,[a] Paola Galletti,[a]
Alessandra Tolomelli,[a] Eusebio Juaristi,[b] and Luca Gentilucci*[a]
Dedicated to C.I.N.M.P.I.S. on the occasion of its 20th anniversary
Keywords: Medicinal chemistry / Asymmetric synthesis / Antibiotics / Amino acids / Nitrogen heterocycles
We describe a new synthesis of the 5-(aminomethyl)oxazol-
idin-3-one core of linezolid in enantiomerically pure form.
The expedient cyclization of the α-hydroxy amide derived
from isoserine and 3-fluoro-4-morpholinoaniline to give the
corresponding (aminomethyl)oxazolidine-2,4-dione, fol-
lowed by its mild selective reduction at the C(4)-position,
gave linezolid in almost quantitative overall yield.
Gram-positive pathogens, particularly those that have been
the cause of resistance development.[2,3]
Introduction
The development of bacterial resistance to currently
available antibacterial agents is a growing global health
problem.[1] Indeed, infections caused by multidrug-resistant
pathogens are responsible for significant morbidity and
mortality in both hospital and community settings.[2] One
possible approach to face this problem is to design new
classes of antibacterial agents that employ new mechanisms
of action. Such agents should exhibit a lack of cross-resis-
tance with existing antimicrobial drugs. However, from the
early 1970s to 1999 the innovative antibiotic pipeline dried
up. All newly launched antibiotics were analogues of exi-
sting drugs except for mupirocin, an antibiotic active
against Gram-positive bacteria launched in 1985. Since
2000, the situation has improved, with five more new classes
of antibiotics approved and launched: linezolid (LZD, 1),[3]
daptomycin, retapamulin, fidaxomicin, and bedaquiline.[2]
Discovered in the 1990s and approved in the U.S. by the
FDA in 2000, LZD (Figure 1) is currently the only commer-
cially available antibiotic in the oxazolidinone class. Com-
pounds containing the oxazolidinone moiety are interesting
because their spectrum of activity covers the important
Figure 1. The syntheses of LZD (1) utilize chiral building blocks.
LZD selectively inhibits bacterial protein synthesis by
binding the peptidyltransferase center on the bacterial ribo-
some, thus preventing the bacterial translation process.[4]
This action mechanism is unique among protein synthesis
inhibitors and explains why LZD retains antibacterial ac-
tivity against Gram-positive organisms that are resistant to
members of other classes.
When LZD came onto the market, it was claimed that
there would be no cross-resistance to this antibiotic and
that resistance would be rare and difficult for the bacteria
to develop.[5] Nevertheless, with the increased wide use of
LZD and abuse to some degree in clinics in recent years,
some clinically resistant strains to LZD have been found
worldwide. These include S. aureus and Enterococcus sp.,[6]
and although the LZD-resistant strains (LinR) appear rela-
tively infrequently in the clinic, the infections caused by
them were found to be life-threatening.
[a] Dept. of Chemistry “G, Ciamician”, University of Bologna,
Via Selmi 2, 40126-Bologna, Italy
E-mail: luca.gentilucci@unibo.it
rossella.demarco2@unibo.it
http://www.ciam.unibo.it/gentilucci
[b] Dept. of Chemistry, Centro de Investigacion y de Estudios
Avanzados del IPN,
Av. Instituto Politécnico Nacional 2508, 07360 México D.F.,
México
Consequently, several researchers have recently focused
on modifying the oxazolidinone structure to seek analogues
with an extended spectrum of antibacterial activity covering
E-mail: ejuarist@cinvestav.mx
Supporting information for this article is available on the
WWW under http://dx.doi.org/10.1002/ejoc.201402888.
7614
© 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2014, 7614–7620

Synthesis of the Antibiotic Linezolid
Gram-negative organisms, activity against LinR strains,
These heterocycles have been described and utilized only
and an improved safety profile. At present, only four oxaz- occasionally in organic and medicinal chemistry.[17] Our re-
olidinones have entered clinical development.[7] As a conse- sult prompted us to design a plausible retrosynthetic path-
quence, the identification of new synthetic protocols to ob- way for the synthesis of optically pure LZD (1) via the in-
tain the 5-(aminomethyl)oxazolidin-3-one core in enantio- termediate 5-(aminomethyl)oxazolidine-2,4-dione, taking
merically pure form can be regarded of great interest for advantage of the very convenient commercially available
the preparation of LZD and analogues.
chiral reagent (S)-isoserine (Scheme 2). The rationale for
The key step in LZD synthesis is the formation of the S- the other steps of the proposed strategy is discussed in the
configured 5-substituted oxazolidinone A-ring. Most proto- following sections.
cols described in patents or publications made use of chiral
The approach depicted in Scheme 2 could be exploited
oxiranes[3,7,8] such as epichlorohydrin[9] (Figure 1) or of in the preparation of a number of unprecedented LZD de-
other building blocks such as glyceraldehyde[10] (Figure 1), rivatives bearing diverse substituents on the A-ring and/or
whereas very few papers have reported approaches based on the methylacetamide tail (see Figure 1). Isoser can also
on asymmetric protocols.[11] The optimized large-scale syn- be purchased in the R configuration, and several enantio-
thesis of LZD involves a total of nine steps, with the oxazol- merically pure Isoser derivatives are commercially avail-
idinone ring being built by adding enantiomerically pure able.[18] Furthermore, many nonracemic substituted isoser-
(S)-3-chloropropane-1,2-diol to an appropriate Cbz-pro- ines can be obtained from chiral starting materials or by
tected aniline (Scheme 1).[3,8]
enzymatic resolution, by the use of chiral auxiliaries,[19] or
by asymmetric catalysis, such as through catalytic enantio-
selective Henry reactions.[20] These synthetic alternatives
will be developed in due course.
According to the retrosynthetic plan, the first step corre-
sponded to the coupling of 3-fluoro-4-morpholin-4-ylphen-
ylamine (2), prepared as reported in the literature,[3] and
Isoser. Initially, we optimized the conditions for the cou-
pling of Isoser bearing different protecting groups at the
N terminus. The reaction between 2 and N-Boc-IsoserOH
(Scheme 3) was attempted according to a protocol for the
coupling of anilines with chiral α-hydroxy acids described
in the literature. This strategy appeared to be a suitable
process for the synthesis of the enantiomerically pure α-
hydroxy amide 3a.[21] Arylamine 2 was treated with SOCl2/
IMD (imidazole) to give the corresponding N-sulfinylaryl-
amine, and treatment of this with N-Boc-IsoserOH was per-
formed in the presence of 1,2,4-triazole. Despite several
modifications designed to optimize the procedure (see the
Supporting Information), in our hands the reaction af-
forded the expected amide 3a only in moderate yields of up
to 40% (Table 1, Entry 1).
Consequently, reactions between 2 and N-Boc-IsoserOH
were conducted under various conditions (Table 1) and in
different solvents (only the best results are reported in
Table 1). All runs were monitored by tlc and terminated
after the disappearance of the reagents, or in any case after
12 h. The reaction between 2 and the mixed anhydride ob-
tained from N-Boc-IsoserOH and ethyl chloroformate/TEA
gave only traces of 3a after 12 h (Table 1, Entry 2). The use
of EDC and HOBt as coupling activating agents in the
presence of DIPEA (or TEA, not shown) at room temp.
afforded 3a only in low yield (15%, Entry 3). The use of
Scheme 1. Synthesis of LZD (1) on a process scale.
The large majority of LZD analogues reported in the
literature display modifications at the B or C ring and/or
in the C(5)-side chain.[2,3,7] In contrast, replacement of the
oxazolidinone ring A with other heterocycles has been de-
scribed only a very few times.[12]
In this context, here we report an expedient synthesis of
the oxazolidinone core from the very convenient chiral rea-
gent isoserine, taking advantage of a synthetic strategy that
is also suitable for furnishing new LZD analogues bearing
diverse substituents on the A-ring.
Results and Discussion
In the last few years we have been interested in the rapid DBU (1,8-diazabicyclo[5.4.0]undec-7-ene) gave a 40% yield
and simple formation of 1,3-oxazolidin-2-one scaffolds for (Entry 4), whereas NMM (N-methylmorpholine) gave a
the design of β-turn-like peptidomimetics[13] or as chiral more satisfactory 60% yield (Entry 5).
auxiliaries for the stereoselective synthesis of unusual amino
On the other hand, the use of the activating agents
acids.[14] Very recently, we observed that treatment of pept- HBTU/HOBt/NMM or HATU/HOBt/NMM at room
ide sequences containing the α-hydroxy-β2-amino acid[15] temp. for 12 h led to very good yields of 96% and 94%
isoserine (Isoser) with carbonate gave rise to 5-(amino- (Entries 6 and 7 in Table 1), respectively. These two reac-
methyl)oxazolidine-2,4-dione rings.[16]
tions were then accelerated by heating for 10 min at 80 °C
Eur. J. Org. Chem. 2014, 7614–7620
© 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
7615

L. Gentilucci et al.
FULL PAPER
Scheme 2. Retrosynthetic approach to LZD (1) via an oxazolidine-2,4-dione intermediate, readily obtained in turn from the chiral building
block isoserine.
Scheme 3. Coupling of 2 and Isoser.
Table 1. Coupling reactions of isoserine derivatives and 2 to afford amides 3a–d.
Entry
1
P
Coupling agents
Temperature [°C]
Solvent
3
a
Yield [%][a]
40[b]
Boc
1. SOCl2/IMD
2. triazole
EtOCOCl/TEA
1. –30
CH2Cl2/DMF
2. room temp.
–20 to room temp.
0 to room temp.
0 to room temp.
0 to room temp.
0 to room temp.
0 to room temp.
80 (MW)
80 (MW)
0 to room temp.
80 (MW)
0 to room temp.
80 (MW)
80 (MW)
[b,c]
2
3
4
5
6
7
8
9
10
11
12
13
14
Boc
Boc
Boc
Boc
Boc
Boc
Boc
Boc
Ac
CH2Cl2/DMF
CH2Cl2/DMF
CH2Cl2/DMF
CH2Cl2/DMF
CH2Cl2/DMF
CH2Cl2/DMF
DMF
DMF
CH2Cl2/DMF
DMF
–
a
a
a
a
a
a
a
b
b
b
c
–
EDC/HOBt/DIPEA
EDC/HOBt/DBU
EDC/HOBt/NMM
HBTU/HOBt/NMM
HATU/HOBt/NMM
HBTU/HOBt/NMM
HATU/HOBt/NMM
HBTU/HOBt/NMM
HATU/HOBt/NMM
EDC/HOBt/NMM
HATU/HOBt/NMM
HATU/HOBt/NMM
15[b]
40[b]
60[b]
96[b]
94[b]
98[d]
98[d]
97
Ac
98[d]
20
Ac
CH2Cl2/DMF
DMF
DMF
Ts[e]
Ns[e]
97[d]
96[d]
d
[a] Determined after isolation by flash chromatography over silica gel. [b] Reaction time 12 h. [c] Traces. [d] Reaction time 10 min. [e] Not
further examined.
by MW irradiation, giving 3a in quantitative yields after tained a low yield (20%, Entry 12). This observation con-
isolation by flash chromatography over silica gel (Entry 8 firms the efficacy of HATU/HOBt/NMM as coupling rea-
with HBTU and Entry 9 with HATU).
gents and MW activation.
Notably, Isoser was incorporated without any need for
We observed that N-Ts- and N-Ns-IsoserOH (N-nosyl-
protection of the OH function. Under the reaction condi- IsoserOH) behaved similarly to N-Boc- and N-Ac-Iso-
tions reported above,[13] the acylation of the OH function serOH in the coupling with 2 under the same reaction con-
and thus the formation of isoserine-isoserine side products ditions as used in Entry 9 (Table 1), giving the correspond-
1
was not observed, as confirmed by the H NMR and RP- ing α-hydroxy arylamides 3c and 3d in almost quantitative
HPLC ESI MS analyses of the crude reaction mixtures.
yields (Entries 13 and 14). The arylsulfonyl protecting
The availability of N-Ac-IsoserOH offered the opportu- groups might be of some interest for future developments
nity to synthesize a direct precursor of LZD (1) already of LZD analogues. Thanks to its simpler cleavage, the Boc
equipped with the desired methylacetamide tail (Scheme 2). protecting group represents the most convenient option.[22]
Therefore, we treated 2 and N-Ac-IsoserOH under the best- Nevertheless, the literature reports several methods for the
yielding conditions determined for the preparation of 3a mild cleavage of arylsulfonyl groups.[23]
either at room temp. (Entry 6 in Table 1) or with MW acti-
The next synthetic step of the strategy depicted in
vation (Entry 9 in Table 1). In particular, the use of HBTU/ Scheme 2 consisted of the cyclization of the Isoser aryl-
HOBt/NMM at room temp. gave 3b in 97% yield after 12 h amides 3 to the five-membered heterocycles 4 (Scheme 4).
(Entry 10), whereas HATU/HOBt/NMM at 80 °C under As anticipated, we observed that oligopeptides containing
MW irradiation gave the same product quantitatively in Isoser underwent cyclization to 5-(aminomethyl)oxazolid-
10 min (Entry 11). As a reference control, we repeated the ine-2,4-diones on treatment with 1.1 equiv. of DSC (N,NЈ-
coupling with EDC/HOBt/NMM (see Entry 5) and ob- disuccinimidyl carbonate) and a catalytic amount of
7616
www.eurjoc.org
© 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2014, 7614–7620

Synthesis of the Antibiotic Linezolid
Scheme 4. Cyclization of 3a and 3b to 3-aryl-5-(aminomethyl)-1,3-oxazolidine-2,4-diones 4a and 4b.
DIPEA in DMF at room temp. for 1 h.[16] Under these con-
ditions, 3a and 3b gave the corresponding 4a and 4b in very
good yields of 96% (Entry 1 in Table 2) and 92% (Entry 2
in Table 2), respectively, after purification by flash
chromatography over silica gel.
Table 2. Cyclization of 3a and 3b to 4a and 4b: reagents and yields.
Entry
3
Carbonate
4
Yield [%][a]
1
2
3
4
5
6
a
b
a
a
a
a
DSC
DSC
Boc2O
triphosgene
a
b
a
a
a
a
96[b]
92[b]
15[c]
35[c]
30[c]
45[c]
Scheme 5. Reduction of the carbonyl bond at C(4) of the 1,3-ox-
azolidine-2,4-dione: synthesis of LZD (1).
CDI
Table 3. Mild reductions of 4a and 4b at C(4) with several reducing
agents.
pNO2-phenyl chloroformate
[a] Determined after isolation by flash chromatography over silica
gel. [b] Reaction time 1 h. [c] Reaction time 12 h.
Entry
4
Reducing agent
Temp. [°C] Solvent
Prod. Yield [%][a]
1
2
b
b
NaBH4
NaBH4
0
0
H2O/THF
MeOH/
THF
5
5
50[b]
95[b]
Other carbonates or dicarbonates converted 3a into the
1,3-oxazolidine-2,4-dione 4a in lower yields: 15% with
Boc2O (Table 2, Entry 3),[24] 35% with triphosgene
(Entry 4),[24] 30% with CDI (1,1Ј-carbonyldiimidazole)[25]
(Entry 5), or 45% with p-nitrophenyl chloroformate[26] (En-
try 6). All reactions were conducted at room temp. for 12 h;
increasing the equivalents of the reagents, time, and tem-
perature, in different solvents, had little effect on the reac-
tions.
With the Boc-aminomethyl compound 4a and the acet-
amidomethyl compound 4b to hand, the last step to LZD
(1) consisted of the reduction of the carbonyl group at C(4)
(Scheme 5 and Table 3). This step raised some concerns
with regard to the risk of simultaneous reduction of the
acetamide segment. The reduction of amides to amines un-
der mild conditions is regarded as a highly valuable but
challenging transformation; many of the reported pro-
[b,c]
3
4
5
6
7
b
b
b
b
b
LiAlH4
0
THF
–
–
–
–
–
–
[d]
H2/Pd-C
r.t.
MeOH
iPrOH
(HOCH2)2
THF
–
–
[d]
HCOO–NH4+/Pd-C 60
[d]
NH2NH2
100
r.t. to 40
–
[d]
Zn(OAc)2/
(EtO)3SiH
BH3/OEt2
BH3·DMS
BH3·DMS/MS
BH3·OEt2
–
[d]
8
9
10
11
12
13
b
b
b
a
a
a
0 to r.t.
0
0
0 to r.t.
0
0
THF
THF
THF
THF
THF
THF
–
1
1
–
7
7
–
55[e]
88[b]
[d]
–
BH3·DMS
BH3·DMS/MS
64[b,f]
97[b]
[a] Determined after isolation by flash chromatography over silica
gel. [b] Reaction time 2 h. [c] Traces of 1, with 6 as the major by-
product. [d] Reaction time 24 h; 4 was recovered almost quantita-
tively. [e] Compounds 5 and 6 were the major byproducts. [f] Com-
pound 8 was the major byproduct.
cedures utilize costly or harmful heavy metal catalysts, high try 2), whereas other solvents gave inferior results (not
pressure, and/or high temperatures.[27] The convenience of shown). The trans relationship was confirmed by the very
1
the strategy depicted in Scheme 2 was substantiated by the small coupling constant J(H4,H5) in the H NMR spectra;
observations reported in the literature that in a 1,3-oxazol- trans-4,5-disubstituted 1,3-oxazolidin-2-ones are charac-
idine-2,4-dione, the N-carbamate moiety increases the reac- terized by coupling constants much lower than those of cis
tivity of the internal amide bond towards reduction by stereoisomers.[24] Interestingly, the RP-HPLC and NMR
NaBH4.[17] This precedent motivated us to attempt conver- analyses ruled out an equilibrium between anomers at C(4).
sion of 4b into LZD (1) by identification of a suitable re-
ducing agent (Table 3).
Not unexpectedly, treatment with LiAlH4 at 0 °C re-
sulted in the reduction of both the carbonyl at C(4) and the
Indeed, the 5-methylacetamide 4b was regioselectively re- N-Ac function, giving a small amount of 1, together with
duced with 1.5 equiv. of NaBH4 in H2O/THF at 0 °C, giv- the 5-ethylaminomethyl compound 6 (Scheme 5) as the
ing exclusively the 5-aminomethyl-4-hydroxy compound 5 major byproduct (Table 3, Entry 3). Catalytic hydrogen-
(Scheme 5) with trans relative stereochemistry (Table 3, En- ation (Entries 4 and 5),[27] Wolff–Kishner reduction (En-
try 1). This LZD derivative might be successfully utilized try 6),[28] Zn(OAc)2-catalyzed reduction with (EtO)3SiH (En-
for the introduction of substituents at C(4) in the oxazol- try 7),[27d] or use of 2 m borane in diethyl ether (Entry 8)[27]
idinone ring. Yields improved to 95% in MeOH/THF (En- gave only traces of the desired product after 24 h.
Eur. J. Org. Chem. 2014, 7614–7620
© 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
7617

L. Gentilucci et al.
FULL PAPER
literature (see also the Supporting Information). Flash chromatog-
raphy was performed with silica gel (230–400 mesh) and mixtures
of distilled solvents. Compounds’ purities were assessed by analyti-
cal RP-HPLC and elemental analysis. Analytical RP-HPLC was
performed with an Agilent 1100 series apparatus and a RP column
Phenomenex mod. Gemini 3μ C18 110A 100ϫ3.0 mm (P/No 00D-
4439-Y0); column description: stationary phase octadecyl carbon
chain-bonded silica (C18) with TMS endcapping, fully porous or-
gano-silica solid support, particle size 3 μm, pore size 110 Å, length
100 mm, internal diameter 3 mm; DAD 210 nm; mobile phase:
On the other hand, use of borane dimethyl sulfide com-
plex in anhydrous THF at 0 °C gave 1, which was easily
isolated in 55% yield by flash chromatography (Entry 9 in
Scheme 5). Analysis of the crude reaction mixture by RP-
HPLC ESI revealed the presence of significant amounts of
byproducts 5 and 6 (not isolated). When the reaction was
repeated in the presence of activated (dried) 4 Å molecular
sieves, which effectively prevented the formation of 4-
hydroxy compound 5, product 1 was obtained in the very
satisfactory yield of 88% (Entry 10), whereas 6 was present from H2O/CH3CN (9:1) to H2O/CH3CN (2:8) in 20 min at a flow
rate of 1.0 mLmin–1, followed by 10 min at the same composition.
Epimerization of intermediates and products was ruled out by chi-
ral HPLC analysis, performed with an Agilent 1200 series appara-
in traces. Use of an increased reaction temperature resulted
in larger amounts of N-ethylamino 6.
Finally, we examined the reduction of 4a under the con-
tus and a CHIRALPAK IC column (P/No 83325); column descrip-
ditions described in Entries 8–10 in Scheme 5 and Table 3.
tion: chiral stationary phase cellulose tris(3,5-dichlorophenylcarb-
Use of borane in diethyl ether was again almost ineffective
amate) immobilized on silica, particle size 5 μm, length 250 mm,
(Entry 11), whereas use of borane dimethyl sulfide complex
internal diameter 4.6 mm, DAD 210/254 nm; mobile phase: n-hex-
gave the 1,3-oxazolidin-4-one 7[29] (64%, Entry 12) and
traces of 4-hydroxy-1,3-oxazolidin-4-one 8 (Scheme 5, not
isolated). In contrast, in the presence of molecular sieves,
ane/propan-2-ol (1:1), at 0.8 mLmin–1. The synthetic procedures
involving MW irradiation were performed with a microwave oven
(MicroSYNTH Microwave Labstation for Synthesis) equipped
use of borane dimethyl sulfide complex afforded 7 quantita- with a built-in ATC-FO advanced fiber-optic automatic tempera-
ture control. 1H NMR spectra were recorded with a Varian Gemini
apparatus at 400 MHz in 5 mm tubes, with use of 0.01 m peptide
tively (Entry 13). Subsequent treatment with TFA followed
by acetic anhydride/TEA afforded
(Scheme 5).
1 in 97% yield
at room temperature. 13C NMR spectra were recorded at 100 MHz.
Chemical shifts are reported as δ values relative to residual CHCl3
δH ( 7.26 ppm), DMSO δH (2.50 ppm), and CDCl3 δC (77.16 ppm)
as internal standards. The unambiguous assignment of 1H NMR
resonances was performed with 2D gCOSY.
The enantiomeric purities of the reported compounds
and intermediates were confirmed by HPLC analysis on a
chiral stationary phase. Analyses of compounds 1, 3a, 3b,
4a, 4b, 5, and 7 performed on a CHIRALPAK IC column
were compared with analyses of the corresponding race-
mates obtained from rac-isoserine (see the Supporting In-
formation). In the cases of compounds 1 (LZD) and 7 the
spectroscopic characterization and specific optical rotations
nicely matched the data reported in the literature (1,[3,11a]
7;[29] see Exp. Sect.).
N-Protected Isoserine-Arylamides 3: 3-Fluoro-4-morpholinoaniline
(2, 1.0 mmol) was added at 0 °C to a stirred solution of the appro-
priate N-protected isoserine (1.05 mmol) in CH2Cl2/DMF (9:1,
5 mL). After 15 min, HOBt (1.1 mmol) and NMM (2.0 mmol) were
added, followed after 30 min by either HBTU or HATU
(1.1 mmol). The mixture was stirred at room temp. for 12 h. Alter-
natively, the reaction was performed under MW irradiation for
10 min, with an initial irradiation power of 150 W and monitoring
of the internal reaction temperature at 80 °C. Then the reaction
mixture was concentrated at reduced pressure, and the residue was
diluted with AcOEt (30 mL) and washed with HCl (0.1 m, 5 mL)
and a saturated solution of NaHCO3 (5 mL). The organic layer
was dried with Na2SO4 and the solvent was evaporated at reduced
Conclusions
We have demonstrated the feasibility of a new approach
to the synthesis of LZD, based on the use of the chiral
starting material N-acetylisoserine, which was efficiently pressure, to afford the appropriate compound 3 as a crude residue,
which was purified (yields: see Table 1) by flash chromatography
over silica gel (eluent: cyclohexane/EtOAc 40:60).
coupled to 3-fluoro-4-morpholinoaniline. Cyclization of the
resulting α-hydroxy β-acetamido arylamide provided a 5-
(acetamidomethyl)-1,3-oxazolidine-2,4-dione intermediate,
which was reduced regioselectively to give LZD in very
good yield under mild reaction conditions. The use of N-
1
Compound 3a: [α]2D0 = –98.6 (c = 0.5, CHCl3). H NMR (CDCl3):
δ = 1.44 (s, 9 H, tBu), 3.12 (t, J = 4.4 Hz, 4 H, morphH3,5), 3.59
(ddd, J = 2.2, 5.8, 14.8 Hz, 1 H, CH2N), 3.69 (ddd, J = 2.2, 5.8,
Boc-protected isoserine further improved the efficiency of 14.8 Hz, 1 H, CH2N), 3.92 (t, J = 4.4 Hz, 4 H, morphH2,6), 4.30
(m, 1 H, CH2CH), 5.19 (br. t, 1 H, BocNH), 5.72 (br. d, 1 H, OH),
6.93 (t, J = 8.8 Hz, 1 H, ArH2), 7.16 (m, 1 H, ArH5), 7.58 (dd, J
= 2.4, 14 Hz, 1 H, ArH6), 8.85 (s, 1 H, CONH) ppm. 13C NMR
(CDCl3): δ = 28.2, 44.8, 51.0, 66.9, 71.8, 81.1, 108.6, 108.8, 115.5
118.7, 118.8, 132.3, 132.4, 136.6, 136.7, 154.2, 156.6, 170.1 ppm.
MS (ESI) calcd. 384.2 [M + H]+; found 384.2. C18H26FN3O5
(383.42): calcd. C 56.39, H 6.84, F 4.96, N 10.96; found C 56.67,
H 6.89, F 5.02, N 10.90.
the reduction, requiring only an extra deprotection/acetyl-
ation step to afford LZD. Further work directed towards
the synthesis of optically pure substituted α-hydroxy β-
amino acids of β2-, β3-, β2,3-, and β3,3-types, suitable precur-
sors of a diverse range of LZD analogues substituted in
ring A, is currently ongoing.
1
Compound 3b: [α]2D0 = –77.1 (c = 0.7, CHCl3). H NMR (CDCl3):
Experimental Section
δ = 2.04 (s, 3 H, Ac), 3.07 (t, J = 4.4 Hz, 4 H, morphH3,5), 3.68
(ddd, J = 2.2, 5.8, 14.8 Hz, 1 H, CH2N), 3.82 (ddd, J = 2.2, 5.8,
14.8 Hz, 1 H, CH2N), 3.89 (t, J = 4.4 Hz, 4 H, morphH2,6), 4.32
(m, 1 H, CH2CH), 6.07 (br. s, 1 H, OH), 6.45 (br. t, 1 H, AcNH),
General Methods: Standard chemicals, including (S)- and rac-iso-
serine, were purchased from commercial sources and used without
further purification. Compound 2 was prepared as reported in the
7618
www.eurjoc.org
© 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2014, 7614–7620

Synthesis of the Antibiotic Linezolid
6.95 (t, J = 8.6 Hz, 1 H, ArH2), 7.15 (m, 1 H, ArH5), 7.55 (m, 1 inert atmosphere at 0 °C. After having been stirred for 7 h, the reac-
H, ArH6), 8.89 (s, 1 H, CONH) ppm. 13C NMR (CDCl3): δ = 22.7, tion mixture was quenched with HCl (0.1 m, 10 mL). The mixture
44.8, 51.1, 66.9, 74.2, 108.6, 108.8, 115.5, 119.0, 127.0, 128.3, 154.2,
was concentrated at reduced pressure to half of the initial volume,
and extracted three times with EtOAc (3ϫ 15 mL). The combined
170.1, 174.7 ppm. MS (ESI) calcd. 326.2 [M + H]+; found 326.3.
C15H20FN3O4 (325.34): calcd. C 55.38, H 6.20, F 5.84, N 12.92; organic layers were dried with Na2SO4, filtered, and concentrated
found C 55.71, H 6.25, F 5.87, N 13.00.
at reduced pressure. The products, 1 or 7 (yields: see Table 3), were
purified by flash chromatography over silica gel (eluent: cyclohex-
ane/EtOAc, 75:25).
5-(Aminomethyl)oxazolidine-2,4-diones 4: DSC (1.1 mmol) was
added at room temp. under inert atmosphere to a stirred solution
of the appropriate compound 3 (1.0 mmol) in CH2Cl2/DMF (3:1,
4 mL), followed by DIPEA (0.1 mmol). After 1 h, the solvent was
distilled at reduced pressure, the residue was diluted with HCl (1 m,
5 mL), and the aqueous phase was extracted three times with Ac-
OEt (3ϫ 15 mL). The combined organic layers were dried with
Na2SO4, filtered, and concentrated at reduced pressure. The oily
residue was purified by flash chromatography over silica gel (eluent:
cyclohexane/EtOAc 60:40) to give the appropriate product 4
(yields: see Table 2).
LZD (1):[3,11a] [α]2D0 = –9.0 (c = 0.5, CHCl3), ref.[11a] [α]2D3 = –9.8 (c
1
= 2.5, CHCl3). H NMR (CDCl3): δ = 2.04 (s, 3 H, Ac), 3.11 (t, J
= 4.2 Hz, 4 H, morphH3,5), 3.62 (ddd, J = 3.2, 6.0, 14.8 Hz, 1 H,
CH2N), 3.71 (ddd, J = 3.2, 6.0, 14.8 Hz, 1 H, CH2N), 3.76 (dd, J
= 7.2, 9.0 Hz, 1 H, OxdH4), 3.91 (t, J = 4.4 Hz, 4 H, morphH2,6),
4.04 (dd, J = 7.2, 9.0 Hz, 1 H, OxdH4), 4.78 (m, 1 H, OxdH5), 5.99
(br. t, J = 6.0 Hz, 1 H, AcNH), 7.05–7.10 (m, 2 H, ArH1,6), 7.49
(dd, J = 2.8, 15.2 Hz, 1 H, ArH5) ppm. 13C NMR (CDCl3): δ =
23.1, 42.1, 47.7, 51.3, 66.5, 71.9, 107.5, 107.8, 113.9, 119.8, 132.8,
133.0, 136.4, 136.5, 154.2, 154.4, 156.8, 171.1 ppm. MS (ESI) calcd.
338.2 [M + H]+; found 338.2. C16H20FN3O4 (337.35): calcd. C
56.97, H 5.98, F 5.63, N 12.46; found C 57.23, H 6.05, F 5.61, N
12.68.
Compound 4a: [α]2D0 = –32.4 (c = 0.5, CHCl3). H NMR (CDCl3):
1
δ = 1.45 (s, 9 H, tBu), 3.13 (t, J = 4.4 Hz, 4 H, morphH3,5), 3.75–
3.85 (m, 2 H, CH2N), 3.88 (t, J = 4.4 Hz, 4 H, MorphH2,6), 4.90
(t, J = 6.0 Hz, 1 H, CH2CH), 5.00 (br. t, 1 H, BocNH), 7.01 (t, J
= 8.4 Hz, 1 H, ArH2), 7.16–7.19 (m, 2 H, ArH5,6) ppm. 13C NMR
(CDCl3): δ = 28.2, 40.4, 40.7, 50.5, 66.7, 82.2, 113.8, 118.6, 118.7,
121.7, 124.3, 124.4, 140.4, 140.6, 153.6, 153.7, 156.1, 169.8,
172.9 ppm. MS (ESI) calcd. 410.2 [M + H]+; found 410.1.
C19H24FN3O6 (409.41): calcd. C 55.74, H 5.91, F 4.64, N 10.26;
found C 56.01, H 6.05, F 4.68, N 10.32.
Compound 7:[29] [α]2D0 = –34.4 (c = 0.3, CH3CN), ref. [α]2D5 = –36 (c
1
= 0.71, CH3CN). H NMR (CDCl3): δ = 1.41 (s, 9 H, tBu), 3.06
(t, J = 4.8 Hz, 4 H, morphH3,5), 3.46–3.54 (m, 2 H, CH2N), 3.81
(dd, J = 6.6, 9.0 Hz, 1 H, OxdH4), 3.87 (t, J = 4.8 Hz, 4 H,
morphH2,6), 4.00 (dd, J = 6.6, 9.0 Hz, 1 H, OxdH4), 4.74 (m, 1 H,
OxdH5), 5.02 (br. t, 1 H, BocNH), 6.94 (t, J = 9.0 Hz, 1 H, ArH1),
7.09 (ddd, J = 1.2, 2.8, 8.8 Hz, 1 H, ArH6), 7.44 (dd, J = 2.8,
14.2 Hz, 1 H, ArH5) ppm. 13C NMR (CDCl3): δ = 28.2, 47.5, 51.0,
66.9, 72.0, 80.3, 107.4, 107.6, 113.9, 118.9, 119.0, 133.2, 133.3,
136.4, 154.3, 156.2, 156.7 ppm. MS (ESI) calcd. 396.2 [M + H]+;
found 396.2. C19H26FN3O5 (395.43): calcd. C 57.71, H 6.63, F 4.80,
N 10.63; found C 57.98, H 6.62, F 4.91, N 10.89.
Compound 4b: [α]2D0 = –27.3 (c = 0.6, CHCl3). H NMR (CDCl3):
1
δ = 2.06 (s, 3 H, Ac), 3.13 (t, J = 4.8 Hz, 4 H, morphH3,5), 3.83–
3.94 (m, 6 H, CH2N, morphH2,6), 5.05 (t, J = 5.4 Hz, 1 H,
CH2CH), 5.93 (br. t, 1 H, AcNH), 7.01 (t, J = 8.8 Hz, 1 H, ArH2),
7.15–7.19 (m, 2 H, ArH6,5) ppm. 13C NMR (DMSO): δ = 23.0,
39.4, 51.0, 66.8, 79.1, 115.2, 115.5, 119.7, 124.0, 125.4, 125.5, 140.8,
153.5, 154.9, 171.1, 171.2 ppm. MS (ESI) calcd. 352.1 [M + H]+;
found 352.1. C16H18FN3O5 (351.33): calcd. C 54.70, H 5.16, F 5.41,
N 11.96; found C 55.02, H 5.19, F 5.45, N 12.04.
Synthesis of 1 via 7: Compound 7 (0.40 g, 1.0 mmol) was treated
with TFA/CH2Cl2 (1:3, 5 mL) with stirring at room temp. After
15 min, the solution was evaporated at reduced pressure, and the
residue was subjected to the same treatment. The crude residue was
suspended in Et2O (20 mL), and the precipitate was collected by
centrifuge. A stirred suspension of the crude TFA salt was sus-
pended in EtOAc (5 mL) and treated with Ac2O (0.14 mL,
1.5 mmol) and TEA (0.28 mL, 2.0 mmol) at room temp. After 3 h,
the mixture was diluted with EtOAc (40 mL), and washed with HCl
(0.1 m, 5 mL) and brine (5 mL). The organic layer was dried with
Na2SO4, filtered, and concentrated at reduced pressure. The prod-
uct 1 was isolated (0.33 g, 97%) by flash chromatography over silica
gel (eluent: cyclohexane/EtOAc, 75:25).
5-Acetamidomethyl-4-hydroxyoxazolidin-2-one (5): Fresh NaBH4
(0.057 g, 1.5 mmol) was added at 0 °C under inert atmosphere to a
stirred solution of 4b (0.35 g, 1.0 mmol) in MeOH/THF (1:1,
5 mL). The reaction was quenched after 24 h by addition of acet-
one. The mixture was concentrated at reduced pressure, and the
residue was diluted with water (5 mL). The mixture was extracted
twice with CH2Cl2 and once with EtOAc. The organic layers were
collected and dried with Na2SO4. The solution was filtered, and
solvent was evaporated at reduced pressure. The oily 5 was isolated
(0.33 g, 95%) by flash chromatography over silica gel (eluent: cyclo-
hexane/EtOAc 50:50). [α]2D0 = +1.8 (c = 0.5, CHCl3). 1H NMR
(CDCl3): δ = 1.98 (s, 3 H, Ac), 3.07 (dd, J = 3.0, 5.8 Hz, 4 H,
morphH3,5), 3.55 (ddd, J = 5.0, 9.0, 14.4 Hz, 1 H, CH2N), 3.73
(ddd, J = 5.0, 9.0, 14.4 Hz, 1 H, CH2N), 3.87 (dd, J = 3.0, 5.8 Hz,
4 H, morphH2,6), 4.47 (br. t, 1 H, CH2CH), 5.02 (br. s, 1 H, OH),
5.60 (s, 1 H, CHOH), 6.36 (t, J = 6.0 Hz, 1 H, AcNH), 6.93 (t, J
= 9.2 Hz, 1 H, ArH1), 7.22 (dd, J = 2.0, 10.4 Hz, 1 H, ArH6), 7.31
(dd, J = 2.2, 13.8 Hz, 1 H, ArH5) ppm. 13C NMR (CDCl3): δ =
22.9, 40.7, 50.9, 66.8, 80.9, 83.1, 111.0, 111.3, 118.3, 119.0, 125.5,
154.0, 154.8, 156.5, 171.9 ppm. MS (ESI) calcd. 354.1 [M + H]+;
found 354.0. C16H20FN3O5 (353.35): calcd. C 54.39, H 5.71, F 5.38,
N 11.89; found C 54.67, H 5.75, F 5.36, N 11.99.
Supporting Information (see footnote on the first page of this arti-
1
cle): Synthesis of 2, 3a; chiral HPLC analyses; H and 13C NMR
spectra.
Acknowledgments
The authors thank the Ministero dell’Università e della Ricerca
(MIUR) (PRIN 2010), and the Fondazione per la Ricerca sulla
Fibrosi Cistica, Verona (FFC#11/2011) for financial support, and
Dr. S. Rupiani for collaboration.
[1] World Health Organization, 2014, Antimicrobial resistance:
Global Report on Surveillance. http://www.who.int/drugresist-
ance/documents/surveil-lancereport/en/.
[2] a) J. M. Rolain, R. Canton, G. Cornaglia, Clin. Microbiol. In-
fect. 2012, 18, 615–616; b) M. S. Butler, M. A. Blaskovich,
5-(Aminomethyl)oxazolidin-2-ones 1 and 7: Dry molecular sieves
(4 Å, 100% w/w) were added under inert atmosphere to a solution
of the appropriate compound 4 (1.0 mmol) in freshly distilled THF
(10 mL). BH3·DMS (2.0 mmol) was added to the suspension under
Eur. J. Org. Chem. 2014, 7614–7620
© 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
7619

L. Gentilucci et al.
FULL PAPER
M. A. Cooper, J. Antibiot. 2013, 66, 571–591; c) J. Rybak, Anti-
microb. Agents Chemother. 1998, 42, 721–724.
M. R. Barbachyn, Angew. Chem. Int. Ed. 2003, 42, 2010–2023;
Angew. Chem. 2003, 115, 2056.
a) S. M. Swaney, H. Aoki, M. C. Ganoza, D. L. Shinabarger,
Antimicrob. Agents Chemother. 1998, 42, 3251–3255; b) D. N.
Wilson, F. Schluenzen, J. M. Harms, A. L. Starosta, S. R. Con-
nell, P. Fucini, Proc. Natl. Acad. Sci. USA 2008, 105, 13339–
13344; c) K. L. Leach, S. J. Brickner, Ann. N. Y. Acad. Sci.
2011, 1222, 49–54.
105–110; c) S. Kano, T. Yokomatsu, H. Nemoto, S. Shibuya, J.
Am. Chem. Soc. 1986, 108, 6746–6748.
[18] Selected examples of commercially available isoserines: 3-phen-
ylisoserine, isothreonine, 3-amino-2-hydroxy-2-methylpropa-
noic acid, 3-amino-2-hydroxy-4-methylpentanoic acid, 3-
amino-2-hydroxy-4-phenylbutyric acid, 3-amino-2-hydroxy-
5-methylhexanoic acid, 3-amino-2,4-dihydroxybutanoic acid,
3-amino-2-hydroxy-3-(4-hydroxyphenyl)propanoic acid, 3-
amino-3-(3-fluorophenyl)-2-hydroxypropionic acid.
[19] For some recent examples, see: a) J. E. Semple, T. D. Owens,
K. Nguyen, O. E. Levy, Org. Lett. 2000, 2, 2769–2772; b) Y.
Aoyagi, R. P. Jain, R. M. Williams, J. Am. Chem. Soc. 2001,
123, 3472–3477; c) F. Fringuelli, F. Pizzo, M. Rucci, L. Vac-
caro, J. Org. Chem. 2003, 68, 7041–7045; d) F. Gassa, A. Con-
tini, G. Fontana, S. Pellegrino, M. L. Gelmi, J. Org. Chem.
2010, 75, 7099–7106; e) Y. Jiang, X. Chen, Y. Zheng, Z. Xue,
C. Shu, W. Yuan, X. Zhang, Angew. Chem. Int. Ed. 2011, 50,
7304–7307; f) F. Rodriguez, F. Corzana, A. Avenoza, J. H. Bu-
sto, J. M. Peregrina, M. D. M. Zurbano, Curr. Top. Med. Chem.
2014, 14, 1225–1234.
[3]
[4]
[5]
[6]
M. Fines, R. Leclercq, J. Antimicrob. Chemother. 2000, 45, 797–
802.
a) K. S. Long, B. Vester, Antimicrob. Agents Chemother. 2012,
56, 603–612; b) B. Gu, T. Kelesidis, S. Tsiodras, J. Hindler,
R. M. Humphries, J. Antimicrob. Chemother. 2013, 68, 4–11.
a) M. F. Gordeev, Z. Y. Yuan, J. Med. Chem. 2014, 57, 4487–
4497; b) A. R. Renslo, G. W. Luehr, M. F. Gordeev, Bioorg.
Med. Chem. 2006, 14, 4227–4240; c) K. Michalska, I. Karpiuk,
M. Król, S. Tyski, Bioorg. Med. Chem. 2013, 21, 577–591.
B. A. Pearlman, W. R. Perrault, M. R. Barbachyn, P. R. Man-
ninen, D. S. Toops, D. J. Houser, T. J. Fleck, Upjohn,
US5837870, 1998 [Chem. Abstr. 1998, 130, 25061].
[7]
[20] C. Christensen, K. Juhl, R. G. Hazell, K. A. Jørgensen, J. Org.
Chem. 2002, 67, 4875–4881.
[8]
[9]
[21] R. Chidambaram, J. Zhu, K. Penmetsa, D. Kronenthala, J.
Kant, Tetrahedron Lett. 2000, 41, 6017–6020.
G. Madhusudhan, G. Om Reddy, T. Rajesh, J. Ramanatham,
P. K. Dubey, Tetrahedron Lett. 2008, 49, 3060–3062.
[22] A. Isidro-Llobet, M. Alvarez, F. Albericio, Chem. Rev. 2009,
109, 2455–2504.
[10] G. Y. Xu, Y. Zhou, M. C. Xu, Chin. Chem. Lett. 2006, 17, 302–
[23] a) T. Fukuyama, C. K. Jow, M. Cheung, Tetrahedron Lett.
1995, 36, 6373–6374; b) G. Sabitha, B. V. S. Reddy, S. Abra-
ham, J. S. Yadav, Tetrahedron Lett. 1999, 40, 1569–1570; c) T.
Ankner, G. Hilmersson, Org. Lett. 2009, 11, 503–506.
[24] G. Zappia, E. Gacs-Baitz, G. Delle Monache, D. Misiti, L. Ne-
vola, B. Botta, Curr. Org. Synth. 2007, 4, 81–135.
304.
[11] a) L. Song, X. Chen, S. Zhang, H. Zhang, P. Li, G. Luo, W.
Liu, W. Duan, W. Wang, Org. Lett. 2008, 10, 5489–5492; b) A.
Palumbo Piccionello, P. Pierro, A. Accardo, S. Buscemia, A.
Pace, RSC Adv. 2013, 3, 24946–24951.
[12] A. Palumbo Piccionello, R. Musumeci, C. Cocuzza, C. G. For-
tuna, A. Guarcello, P. Pierro, A. Pace, Eur. J. Med. Chem. 2012,
50, 441–448.
[25] S. Cutugno, G. Martelli, L. Negro, D. Savoia, Eur. J. Org.
Chem. 2001, 3, 517–522.
[26] J. B. Blanco-Canosa, P. E. Dawson, Angew. Chem. Int. Ed.
2008, 120, 6957–6961; Angew. Chem. 2008, 120, 7064.
[27] a) A. M. Smit, R. Whyman, Chem. Rev. 2014, 114, 5477–5510;
b) D. L. Dodds, D. J. Cole-Hamilton, in: Sustainable catalysis:
challenges and practices for the pharmaceutical and fine chemical
industries [Eds.: P. J. Dunn, K. K. (Mimi) Hii, M. J. Krische],
Wiley, 2013, p. 1–36; c) M. Stein, B. Breit, Angew. Chem. Int.
Ed. 2013, 52, 2231–2234; d) S. Das, D. Addis, S. Zhou, K.
Junge, M. Beller, J. Am. Chem. Soc. 2010, 132, 1770–1771.
[28] Z. Luo, M. Naguib, Tetrahedron Lett. 2012, 53, 3316–3318.
[29] W. R. Perrault, B. A. Pearlman, D. B. Godrej, A. Jeganathan,
K. Yamagata, J. J. Chen, C. V. Lu, P. M. Herrinton, R. C. Gad-
wood, L. Chan, M. A. Lyster, M. T. Maloney, J. A. Moeslein,
M. L. Greene, M. R. Barbachyn, Org. Process Res. Dev. 2003,
7, 533–546.
[13] a) L. Gentilucci, A. Tolomelli, R. De Marco, C. Tomasini, S.
Feddersen, Eur. J. Org. Chem. 2011, 4925–4930; b) R.
De Marco, A. Tolomelli, M. Campitiello, P. Rubini, L. Gentil-
ucci, Org. Biomol. Chem. 2012, 10, 2307–2317; c) R. De Marco,
A. Greco, S. Rupiani, A. Tolomelli, C. Tomasini, S. Pieraccini,
L. Gentilucci, Org. Biomol. Chem. 2013, 11, 4273–4420.
[14] R. De Marco, L. Cavina, A. Greco, L. Gentilucci, Amino Acids
2014, 0000–0000. DOI: 10.1007/s00726-014-1839-3.
[15] a) G. Lelais, D. Seebach, Biopolymers 2004, 76, 206–243; b)
Enantioselective Synthesis of β-Amino Acids, 2nd ed. (Eds.: E.
Juaristi, V. Soloshonok), Wiley-VCH, New York, 2005.
[16] A. Greco, S. Tani, R. De Marco, L. Gentilucci, Chem. Eur. J.,
20, 0000–0000. DOI: 10.1002/chem.201402519.
[17] a) Ö. Demir-Ordu, I. Dog˘an, Chirality 2010, 22, 641–654; b)
G. Chen, C. Chunling Fu, S. Ma, Org. Biomol. Chem. 2011, 9,
Received: July 8, 2014
Published Online: October 28, 2014
7620
www.eurjoc.org
© 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2014, 7614–7620