A new synthetic approach to high-purity (15R)-latanoprost

Article abstract of DOI:10.1002/ejoc.200600749

This paper describes a new synthesis of latanoprost (1) that afforded high purity latanoprost in 16.9 % overall yield in eight synthetic steps from sulfone 4. The "a chain" in a derivative of the (-)-"Corey lactone" was elongated first, followed by the attachment of a novel, enantiomerically pure "ω chain" synthon. This ensured the absence of the undesired (15S)-1 diastereomer in the synthesized prostaglandin. The crystalline nature of the novel sulfone 4 facilitated its purification. A variation of the new synthesis of latanoprost is described, where the laboratory-scale synthesis was further adapted to a hundred-gram scale. In the course of the present synthesis, a new prostaglandin sulfone intermediate, 21, which may find application in the synthesis of diverse prostaglandin analogs, was introduced. A practical synthesis of novel, enantiomerically pure "ω chain" synthons 15, 16, and 17 has also been carried out, employing diol 12, which was obtained from derivatives of the D-mannitol chiral pool. Wiley-VCH Verlag GmbH & Co. KGaA, 2007.

Full text of DOI:10.1002/ejoc.200600749

FULL PAPER
DOI: 10.1002/ejoc.200600749
A New Synthetic Approach to High-Purity (15R)-Latanoprost
Jacek G. Martynow,*[a][‡] Julita Józ´wik,[a] Wiesław Szelejewski,[a] Osman Achmatowicz,*[a]
[a]
[a]
[a]
[a]
´
Andrzej Kutner, Krzysztof Wisniewski, Jerzy Winiarski, Oliwia Zegrocka-Stendel,
and Piotr Gołebiewski[a]
˛
Keywords: Prostaglandins / Chiral pool / Alkylation / Nucleophilic substitution / Reductive desulfonylation
This paper describes a new synthesis of latanoprost (1) that
adapted to a hundred-gram scale. In the course of the pres-
afforded high purity latanoprost in 16.9% overall yield in
ent synthesis, a new prostaglandin sulfone intermediate, 21,
eight synthetic steps from sulfone 4. The “α chain” in a deriv- which may find application in the synthesis of diverse prosta-
ative of the (–)-“Corey lactone” was elongated first, followed
by the attachment of a novel, enantiomerically pure “ω
chain” synthon. This ensured the absence of the undesired
(15S)-1 diastereomer in the synthesized prostaglandin. The
crystalline nature of the novel sulfone 4 facilitated its purifi-
cation. A variation of the new synthesis of latanoprost is de-
scribed, where the laboratory-scale synthesis was further
glandin analogs, was introduced. A practical synthesis of
novel, enantiomerically pure “ω chain” synthons 15, 16, and
17 has also been carried out, employing diol 12, which was
obtained from derivatives of the D-mannitol chiral pool.
(© Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim,
Germany, 2007)
Latanoprost continues to hold a key position in the anti-
glaucoma market, although certain newer prostaglandin an-
alogs [e.g. travoprost and bimatoprost (Figure 1)] show
similarly high levels of intraocular pressure reduction[5] in
clinical trials. All[4,6] but one[6j] of the reported synthetic
approaches toward 1 make use of one of known variations
of the Corey method,[2a–2c] in which the side chains are se-
quentially attached in a specific order to a derivative of the
commercially available (–)-“Corey lactone”.[4,6h] Thus, the
Corey strategy, as applied to the synthesis of latanoprost,
comprises a sequence of the following reactions: (a) ad-
dition of an “ω chain” precursor to the (–)-“Corey lactone”
resulting in a 13-en-15-one, (b) reduction of the enone to a
13-en-15-ol, (c) hydrogenation of the 13,14-alkene, (d) ad-
dition of the “α chain”, leading to a cis-5,6-alkene, and (e)
further transformations (the numbering used herein corre-
sponds to the prostaglandin numbering system[2c]).
Introduction
Natural prostaglandins are unsaturated fatty acids de-
rived from arachidonic acid. Many classes of prostaglan-
dins are characterized by very high biological activity, their
KD values being within 10–8–10–12 . Prostaglandin F2α
(PGF2α), for example, is known to be a very potent vaso-
constrictor and oxytoxic agent. The medicinal chemistry of
these natural products has experienced dynamic growth
since their discovery in 1934 by von Euler.[1] The synthesis
of prostaglandins was the subject of considerable scientific
effort, which led to many discoveries and developments that
have now become a part of the standard arsenal of organic
chemistry methodologies.[1c,2] More recently, solid-phase
strategies have been developed for the design and synthesis
of combinatorial libraries of antiviral and other bioactive
compounds based on prostaglandin templates including
PGF analogs.[3] Importantly, research into the pharmaco-
The very high potency of prostaglandins dictates that
logical properties of PGF derivatives has led to the discov- contamination with diastereomeric prostaglandins of un-
ery[4] of latanoprost (1) and structurally related derivatives known activity should be avoided in prostaglandins used as
(Figure 1), which are very effective as anti-glaucoma agents pharmaceutical active ingredients.[1] However, the multiple
and are currently broadly marketed against glaucoma.
chiral centers present in these molecules continue to create
practical challenges, and existing synthetic routes to latano-
prost by the Corey strategy, in spite of the considerable pro-
gress made in this field,[6] are not completely free of the
contaminating diastereomers. Besides the occurrence of the
5,6-trans olefin, particular difficulty has been associated
with maintaining the purity at the C(15) carbinol asymmet-
ric center, which typically is formed by a stereoselective re-
duction. Such reductions require expensive reagents[2a,6g,7]
and often low-temperature conditions. However, in practice
[a] Pharmaceutical Research Institute,
Rydygiera 8 Street, Warsaw, Poland
Fax: +48-22-456-38-38
E-mail: o.achmatowicz@ifarm.waw.pl
[‡] Present address: Department of Chemistry and Chemical
Biology, Indiana University, Purdue University Indianapolis,
Indiana 46202, USA
E-mail: jmartynow@chem.iupui.edu
Supporting information for this article is available on the
WWW under http://www.eurjoc.org or from the author.
Eur. J. Org. Chem. 2007, 689–703
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
689

J. G. Martynow, O. Achmatowicz et al.
FULL PAPER
Figure 1. Structures of the (–)-“Corey lactone”, synthetic PGF2α analogs, and key impurities (15S)-1 and (5E)-1 of latanoprost (1).
these methods do not lead to a completely stereoselective proaches was poor. On the other hand, a “reversed” order
formation of the desired (15R) isomer. The (15S) dia- of side chain attachment was conducted by Schaaf and Co-
stereomer of latanoprost [(15S)-1 in Figure 1] has been de- rey[9a] during their synthesis of PGF1α and more recently
scribed,[4a] and it has been reported to typically occur[6a] by the Procter & Gamble research group,[9b] but the re-
in the product in amounts exceeding 1%. In addition, the duction of a 13-en-15-one could not be avoided. Moreover,
physicochemical properties of 1 and (15S)-1 are very sim- a synthesis of 13,14-dihydro-PGF2α analogs by this route
ilar. Indeed, it is not a straightforward task to discern be- would require a difficult regioselective hydrogenation of the
tween these two compounds by 500 MHz 1H NMR, optical 13,14-alkene in the presence of the 5,6-alkene. Recently, the
rotation, or even by analytical reversed-phase HPLC. 1,4-addition approach[2d,e,f] to prostaglandin synthesis has
Furthermore, the application of preparative scale HPLC to been applied to the synthesis of latanoprost (1), which re-
rid 1 of the (15S) diastereomer is very costly and far from sulted in a series of elegant but laborious transformations[6j]
trivial, while the 5,6-trans isomer of 1 [(5E)-1] is much easier that had to be extensively supported by Simulated Moving
to separate by HPLC.[6a] Other methods for the preparation Bed (SMB) chiral chromatography in order to provide in-
of prostaglandin F analogs[8] from the (–)-Corey lactone termediates of sufficient enantiomeric purity. Another
were reported, in which the introduction of the ω chain al- lengthy synthetic route of this type has been reported re-
ready in the correct stereochemical arrangement preceded cently.[6k] Thus, the Corey method[2a,2b] still promises to
the attachment of the α chain. Thus, the problematic re- provide the most facile and practical access to latanoprost
duction of a 13-en-15-one to the corresponding (15R)-13- (1), yet the difficulty with the (15S)-1 stereoisomer needs to
en-15-ol was avoided. However, the efficiency of these ap- be overcome. Therefore, a synthesis of latanoprost (1) cap-
690
www.eurjoc.org
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2007, 689–703

A New Synthetic Approach to High-Purity (15R)-Latanoprost
FULL PAPER
able of delivering the desired product free of its (15S) dia- C could be synthesized by a Wittig reaction from lactol E,
stereomer is expected to have considerable scientific and which is readily accessible from commercially available de-
practical merit.
rivatives of the (–)-Corey lactone and a suitably protected α
chain precursor D. The 4-methyl-2,6,7-trioxabicyclo[2.2.2]-
octane protective group was selected as the latent carboxyl
group X introduced in D, due to its ease of formation and
stability under the basic conditions[11] required to success-
fully carry out the alkylation of sulfone C.
Results and Discussion
As shown in Scheme 1, we envisioned that latanoprost
(1) could be prepared by the reductive desulfonation of
phenylsulfone A, followed by simple protecting group trans-
formations. Intermediate A could be formed from key pros-
taglandin sulfone C through an SN2 alkylation of the carb-
Synthesis of the Central Building Block, Hemiacetal 5
The commercially available (3aR,4S,5R,6aS)-4-(hy-
anion generated at C(13), α to the phenylsulfonyl group, droxymethyl)-5-(triethylsilyloxy)-hexahydro-cyclopenta[b]fu-
with an electrophilic alkylating agent B possessing the ran-2-one (2) (Ͼ98% de) was selected as the starting mate-
stereochemical arrangement corresponding to the (15R) rial for the synthesis of latanoprost. The high enantiomeric
stereochemistry in the product. Critical to this disconnec- purity of this material was expected to be further enhanced
tion, the enantiomeric purity of the ω chain synthon B upon crystallization of one of the subsequent intermediates.
should be very high (e.g. above 99.8%), thus a priori estab- The transformation of alcohol 2 to sulfide 3 was carried out
lishing very high diastereomeric excess of the (15R) prod- over 40 h in the presence of diisopropyl azodicarboxylate
uct. With a well-chosen protecting group, alkylating agent (DIAD)[12] and required slightly elevated temperatures. The
B should not be capable of epimerization under the reaction oxidation of compound 3 to sulfone 4 was conveniently car-
conditions. We expected that the utility of synthon C would ried out with magnesium monoperoxyphthalate (MMPP) in
not be limited to the synthesis of latanoprost (1), as alkylat- the biphasic H2O/CH2Cl2 solvent system.[13] Use of an ex-
ing synthons other than B could easily be proposed. Ad- cess of the oxidizing agent ensured a fast and clean conver-
ditionally, the utility of the -CH2SO2Ar group in the syn- sion to the sulfone. Contrary to the report on microwave-
thesis of substituted alkenes is well documented.[10] Synthon catalyzed MMPP oxidation of sulfides to sulfoxides under
Scheme 1. Projected synthesis of latanoprost (1).
Eur. J. Org. Chem. 2007, 689–703
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
691

J. G. Martynow, O. Achmatowicz et al.
FULL PAPER
similar conditions,[13b] we have not observed the sulfoxide
during the entire course of the reaction. Sulfone 4 crys-
tallized readily from EtOAc/hexanes or from Et2O/hexanes,
and the crystalline product was completely dia-
stereomerically pure, as indicated by chiral HPLC.
DIBALH reduction of lactone 4 to hemiacetal 5 (Scheme 2)
afforded the product in excellent purity after column
chromatography.
Scheme 3. The Synthesis of triphenylphosphonium iodide 8. Rea-
gents and conditions: a) ref.[14]; b) (i) PPh3, sulfolane + trace
C5H5N, 80 °C, 1 h; (ii) Et2O/EtOAc/CHCl3 + trace C5H5N,
crystallization, 48%, or (ii) 2-propanol + trace C5H5N, crystalli-
zation, 65.6%.
Synthesis of the ω Chain Synthon, 15
The synthon used for the construction of the ω chain of
latanoprost (1), according to the strategy of our synthesis
(Scheme 1), contains a stereogenic center at the carbon cor-
responding to C(15) in the target molecule, with chirality
corresponding to the (R) configuration in the product, and
with a very high diastereomeric excess, preferably above
99.8%. Individual impurity content below 0.1% facilitates
the registration process of active pharmaceutical ingredi-
ents.
Scheme 2. The Synthesis of hemiacetal 5. Reagents and conditions:
a) DIAD, Ph3P, PhSH, THF/CH2Cl2, 15–40 °C, 72 %; b) (i)
MMPP, H2O/CH2Cl2, 95%; (ii) crystallization, 79%; c) DIBALH,
THF, –70 °C, 91.3%.
A number of structural arrangements were possible for
synthon B (Scheme 1) including ones in which the –OR
group and the leaving group (LG) are a part of the same
functionality (e.g. an epoxide, cyclic sulfate, or cyclic sul-
Synthesis of the α Chain Synthon, 8
Alkyl iodide 7, possessing the latent C(1)-carboxyl group fite).[17] In fact, we synthesized these three ω chain precur-
protected in the form of a bicyclic orthoester, was obtained sors (compounds 17, 18, and 19) in enantiomerically pure
from commercially available 5-iodopentanoic acid chloride form as (S) enantiomers, which is illustrated in Scheme 4.
6, according to previously described procedures.[11a,14] However, at the next stage of the synthesis of 1, the best
Compound 7 was then converted into quaternary phospho- results for ω chain introduction were obtained with (S)-1-
nium salt 8 with triphenylphosphane (PPh3) in sulfolane in iodo-4-phenyl-2-(triethylsiloxy)butane (15), which was syn-
the presence of pyridine (Scheme 3). Acetonitrile is typically thesized from the corresponding tosylate 14 (Scheme 4) in
used[15a] as the solvent for Wittig salt formation from alkyl a straightforward manner.
halides and PPh3, but in our hands iodide 7 did not react
α-Hydroxytosylates (such as 13) are best accessed from
with in acetonitrile, even after a prolonged reflux. However, the corresponding nonracemic diols, which are available by
the use of melted sulfolane[15b] as the reaction medium re- asymmetric dihydroxylation of terminal alkenes.[18a–18d] We
sulted in a clean conversion to the desired salt, 8. Interest- carried out the synthesis of the key optically active sub-
ingly, this Wittig salt has not been described in the literature strate, (S)-4-phenylbutane-1,2-diol[18e–18i] 12 in two ways: by
to date, although the corresponding bromide is known.[16] asymmetric synthesis and by a natural, chiral-pool starting
Due to its limited stability in acidic media, the 4-methyl- material. An asymmetric synthesis of (R)-4-phenylbutane-
2,6,7-trioxabicyclo[2.2.2]octane (methyl-OBO) carboxyl 1,2-diol[18d] (ent-12) from 4-phenylbutene, in the presence of
masking group has been used only sporadically in the pros- (DHQD)2PHAL, was reported to give the product with
taglandin field.[2a,2b,15] However, the methyl-OBO group 84% enantiomeric excess (84% ee). To directly obtain the
shows good stability under basic conditions and easily sur- (S)-diol 12 in an analogous reaction, we had to use DHQ-
vives hydrolysis or alcoholysis,[11] which are transformations based chiral catalysts.[18a] Diol 12 was obtained with
occurring in the final steps of the present synthesis of (DHQ)2PYR with a diminished ee of 65%. Exactly the
latanoprost.
same ee was realized when we employed the newer genera-
692
www.eurjoc.org
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2007, 689–703

A New Synthetic Approach to High-Purity (15R)-Latanoprost
FULL PAPER
Scheme 4. The Synthesis of alkyl iodide 15 and other potential ω chain synthons. Reagents and conditions: a) ref.[19a], 53 %; b)
PhCH2P(Ph)3Br, THF, 0 – Ͼ25 °C, 62%; c) H2, 10% Pd/C, 99%; d) PTSA, MeOH, 40 °C, 99%, 99.3% ee; e) (i) nBu2SnO, TsCl, Et3N,
0 – Ͼ20 °C; (ii) Et2O, crystallization, 67.4%, 99.92% ee for the chiral-pool run; f) Et3SiCl, imidazole, DMF, 99.5%; g) NaI, DMF, 89%,
99.92% ee; h) pTsOH, acetone/H2O, 72%; i) KOH, THF/H2O, 81%; j) SOCl2, diisopropylethylamine, CH2Cl2, 96%, 99.92% ee; k) (i)
NaIO4, RuCl3, MeCN/H2O; (ii) Et2O, crystallization, 42%, Ͼ99.99% ee; l) (DHQ)2AQN, K3Fe(CN)6, K2OsO2(OH)4, tBuOH/H2O,
K2CO3, 0 °C, 17 h, 98%, 65.3% ee.
tion catalyst, (DHQ)2AQN.[18b] A much higher enantio- of iodide 15 required heating of compound 14 with sodium
meric purity of diol 12 was achieved when we employed iodide in DMF at about 75 °C for 2.5–3.5 h, which ensured
(R)-2,3-O-isopropylideneglyceraldehyde[19] (9) as the source almost complete conversion while still preventing decompo-
of chirality (Scheme 4). Aldehyde 9 was synthesized in high sition of the product. At higher temperatures, desilylation
enantiomeric purity from 1,2:5,6-di-O-isopropylidene-- became noticeable, while at 60 °C the reaction was not com-
mannitol, which is available commercially or can be pre- plete after several days. Synthon 15 was very easy to purify
pared from -mannitol.[19a] A chain elongation protocol by on a silica gel column, and quantities up to about 300 g of
the Wittig reaction of 9 with the ylide obtained in situ from highly enantiomerically pure 15 were thus prepared
benzyltriphenylphosphonium bromide gave alkenes 10 as a (99.93% ee).
7:3 cis/trans mixture, which was then successfully hydroge-
nated in methanol solution to acetonide 11. Removal of the
acetonide in the next step (11 to 12) required high dilution
in methanol at 40 °C and quenching with triethylamine
The Synthesis of Latanoprost (1)
prior to workup to circumvent the reversibility of the hy-
With the three necessary components 5, 8, and 15 in
drolysis process. Diol 12 was thus prepared in quantities up hand, we continued the synthesis of target compound 1 ac-
to 250 g in 99.3% ee. Selective monotosylation of 12 was cording to the strategy described in Scheme 1. Initial
best conducted with tin-derived catalysts, of which dibu- attempts to obtain 21 were discouraging. The Wittig reac-
tyltin oxide proved to be the most efficient and very easy tion of pure lactol 5 with the ylide generated from com-
to use.[20] Importantly, we discovered that known[21] com- pound 8 was very sluggish. When tBuOK alone was used
pound 13 crystallized readily from Et2O with efficient en- as the base, only 25% of product 20a was obtained after
hancement of enantiomeric purity. This property was criti- stirring for 5 h at room temperature, with 10% recovery of
cal to the synthesis of enantiomerically pure ω chain syn- 5. Longer reaction times resulted in decomposition. Fortu-
thons. The silylation of hydroxytosylate 13 afforded pro- nately, in the course of the optimization of the reaction con-
tected derivative 14 in high yield. The optimized synthesis ditions, we observed that the yields improved by about 20%
Eur. J. Org. Chem. 2007, 689–703
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
693

J. G. Martynow, O. Achmatowicz et al.
FULL PAPER
when crude 5 was used as the substrate. The known oxo- to the moderate reactivity of the iodide 15 and the lability
philic character[22] of many aluminum compounds led us to of the C(11)-O-silyl group in 21. Based on the known be-
speculate that aluminum impurities present in 5 prior to havior of alkyl or aryl sulfones,[23,24] we chose the lithium
chromatography must positively influence the reactivity of hexamethyldisilazane (LHMDS)/THF conditions for the al-
this hemiacetal, perhaps by assisting in the lactol ring open- kylation of the carbanion generated at C(13) of sulfone 21,
ing. Indeed, the addition of a catalytic amount of aluminum which eventually led to the desired substitution product 22,
tri-tert-butoxide resulted in a marked improvement in the albeit in a modest isolated yield of 46%. This alkylation
efficiency for this Wittig reaction (Scheme 5). However, un- reaction required room temperature. However, reaction
der the optimized conditions, we also observed desilylation, times longer than about 4 h did not result in a better yield
which led to 20a (65.5% yield), and a partial triethylsilyl of 22. In spite of stringently maintained basic conditions
group migration, which gave a 35/65 mixture of 20b and throughout the entire process, including chromatography,
20c (15.4% yield). Analogous silyl group migrations have OBO hydrolysis product 23 (13.5% yield) was isolated after
been reported[6] in the prostaglandin field. Alcohols 20a–c the purification procedure. Clearly, the acidity of the silica
were readily resilylated, which afforded the important pros- gel may have been sufficient to cause partial hydrolysis of
taglandin sulfone intermediate, 21. Orthoester 21 was stable 22, even though we attempted to attenuate it with traces of
when stored in pure form at about 0 °C in the presence of pyridine.
a trace of pyridine. After routine chromatography, both
Alternatively, the lithium salts present in the crude reac-
1
HPLC analysis as well as H and 13C NMR spectra indi- tion product may have facilitated the oxabicyclooctane ring
cated the absence of the trans isomer of 21. The 5,6-cis ste- opening. This is additionally corroborated by our observa-
reochemistry of the sulfone was in agreement with the tion on thin layer chromatograms (TLC) of the crude reac-
NMR spectra.
tion product of a spot corresponding to compound 23 (ca.
The ω chain elongation of sulfone 21 to the prostaglan- 5–10% estimated yield; the TLC analyses of compounds
din 22 initially encountered serious difficulties, mainly due 20–24 were performed in solvent systems containing 0.2%
Scheme 5. Laboratory scale synthesis of latanoprost. Reagents and conditions: a) (i) 8, tBuOK, THF, 0 – Ͼ25 °C; (ii) 5, (tBuO)3Al,
0 – Ͼ25 °C, isolate products: 65.5% of 20a plus 15.4% of 20b/20c; b) Et3SiCl, Et3N, imidazole, DMF, 87% from 20a or 67% from
20b/20c; c) 15, LHMDS, THF, 46.2% of 22 plus 13.5% of 23; d) 22, 10% Na/Hg, Na2HPO4, MeOH, 99%; e) PPTS, acetone/H2O, 95 %;
f) (i) LiOH, MeOH/H2O; (ii) NaHSO4, NH4Cl, H2O, 99%; g) iPrI, DBU, acetone, 63.2% after chromatography; chiral HPLC purity
99.83%.
694
www.eurjoc.org
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2007, 689–703

A New Synthetic Approach to High-Purity (15R)-Latanoprost
FULL PAPER
triethylamine). Compounds 22 and 23 were both obtained din 24 in 99% yield. This result was obtained after much
as a 5:2 mixture of two diastereomers, epimeric at the C(13) less successful yet extensive experimentation, which in-
center (the prostaglandin numbering system).[2c] In order to cluded reductions with calcium, sodium, or lithium in li-
gain a better insight into this alkylation reaction and to quid ammonia or n-propylamine and Mg in ethanol/HgCl2
determine the structure of these isomers, a small sample of or Mg/MeOH conditions.[25b] Desired compound 24 was
22 was additionally separated into the components, (13R)- formed in all the dissolving metal reductions, but in each
1
22 and (13S)-22. 500 MHz NMR spectra including H-1H case the product was contaminated with 15–25% of a phen-
COSY, ROESY, and 13C-1H HETCOR were acquired for ylsulfinic acid elimination product, which was not separable
these compounds, which revealed very similar coupling con- from 24.[25a] In our hands, the most successful application
stants to C(13)–H in both isomers, and nOe/ROESY data of the sodium amalgam method of desulfonylation specifi-
were inconclusive. To examine the substitution geometry, cally required that an excess of newly purchased 10% Na/
MacroModel/BatchMin conformational searches for com- Hg reagent was used. Compound 24 showed no signs of
pound 21 were carried out with the MMFF force field, cis-trans isomerisation of the 5,6-olefin by HPLC or NMR
Polak–Ribiere conjugate gradient (PRCG) minimization, spectroscopy. However, during the final steps of the synthe-
and the Monte Carlo method.[25] The structures of the low- sis (24 Ǟ 1) the formation of a small amount (ca. 5–9%)
est energy conformers thus found suggested that the pro-R of the 5,6-trans isomer of 1 could not be avoided. Fortu-
approach of the electrophile should be sterically less encum- nately, this contamination was easily removed from 1 by
bered than a pro-S approach. Therefore, we propose that preparative HPLC on silica gel stationary phases. The tri-
the major isomer of 22 had (13R) configuration, and the ethylsilyl protecting groups in compound 24 were all re-
minor had a (13S) configuration.
moved under the action of pyridinium p-toluenesulfonate
Using the phosphate-buffered conditions of Trost,[26a] the (PPTS) in aqueous acetone, which also led to the intended
epimeric mixture of compounds 22 was subsequently sub- hydrolytic opening of the OBO group, affording pentol 25
jected to reductive desulfonylation to give OBO-prostaglan- in 95% yield. This 2,2-bis(hydroxymethyl)propyl ester was
Scheme 6. Multigram scale synthesis of 1. Reagents and conditions: a) 15 (99.93% ee), LHMDS, THF, 62.8% of 22 plus 16.4% of 23;
b) 22 (81%) + 23 (19%), 10% Na/Hg, Na2HPO4, MeOH, 90%; c) PPTS, acetone/H2O, 60%; d) (i) LiOH, MeOH/H2O; (ii) NaHSO4,
NH4Cl, H2O, 91.5 %; e) iPrI, DBU, acetone, flash chromatography, then preparative HPLC, 65 % from 26, chiral HPLC purity of
1 Ͼ 99.9%.
Eur. J. Org. Chem. 2007, 689–703
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
695

J. G. Martynow, O. Achmatowicz et al.
FULL PAPER
subsequently cleaved with lithium hydroxide and, after acid- sired (15S)-1 diastereomer in the synthesized prostaglandin.
ification with NaHSO4/NH4Cl, latanoprost acid[4a] 26 was The crystalline nature of our starting material (sulfone 4)
isolated in an almost quantitative yield. In the final step of eliminates the need for the troublesome ester protecting
the synthesis, acid 26 was successfully converted to latanop- groups at the C(11)-hydroxy, such as a benzoate or a p-
rost (1), with iPrI/DBU/acetone, conditions commonly used phenylbenzoate, which were previously used by others to
in the prostaglandin field.[2,6] The crude isopropyl ester impart crystallinity and thus facilitate purifications during
(91% purity) was further purified by preparative HPLC on the course of prostaglandin synthesis.[2,4,6] We also describe
silica gel, which afforded latanoprost (1) in 63.2% yield a variation of the new synthesis of latanoprost, where the
from 25 and in 99.83% HPLC purity. Latanoprost was ob- laboratory scale synthesis is further adapted to a multigram
tained from the readily available sulfone 5 in 17.3% overall scale by carrying out a series of transformations, in parallel
yield in seven synthetic steps. This result was unoptimized, and in one pot, with two prostaglandin compounds: (a) a
and further improvements were clearly possible.
compound protected at C(1) with the 4-methyl-OBO group
With this goal in mind, we undertook a multigram-scale and (b) a compound protected with an ester group at the
(“proof of concept”) synthesis of latanoprost (Scheme 6). same position. The absence of (15S)-1 in the crude final
Very gratifyingly, the alkylation of synthon 21, carried out product substantially facilitated the preparative HPLC pu-
on a 104 g scale, gave a noticeably higher yield of the alky- rification of 1 from the (5E) diastereomer. This resulted in
lation products (79.2%) than did the smaller scale experi- pure latanoprost obtained from the prostaglandin precursor
ment discussed above. TLC monitoring indicated that, in 21 in 25% yield.
the present case, the formation of 22 was still slowly pro-
Importantly, in the course of the present synthesis, a new
gressing after 8 h of reaction at room temperature, and prostaglandin sulfone intermediate 21, which may find ap-
therefore, the large-scale reaction mixture was stirred over- plication in the synthesis of diverse prostaglandin analogs,
night. Compound 23 (as a 5:2 mixture of both C(13) epi- was introduced. A practical synthesis of novel, enantiomer-
mers) was isolated in 16.4% yield. Due to the quantity of ically pure ω chain synthons 17, 18, and 19 has also been
material and high enantiomeric purity of the C(15) frag- carried out, employing diol 12, which was obtained from
ment incorporated in 23, we decided that the practical syn- derivatives of the -mannitol chiral pool.
thesis of 1 should be continued without separating com-
pounds differing in the substitution pattern at C(1). Thus,
a mixture of 22 (81%) and 23 (19%) was subjected to the
reductive desulfonylation conditions, as discussed above.
Experimental Section
Desulfonylation of 22 and 23 occurred as expected, af-
fording an 80:20 mixture of compound 24 and the methyl
ester[6a] 27, respectively, in a combined isolated yield of
90%. This mixture was then hydrolyzed with PPTS/acetone/
H2O, which gave an 80:20 mixture of 25 and the known[6c]
latanoprost acid ester 28. The next step led, by efficient hy-
drolysis of the esters (91.5% yield), to the latanoprost acid
26. Selective alkylation of 26 with 2-iodopropane provided
The starting material, (3aR,4S,5R,6aS)-4-(hydroxymethyl)-5-(tri-
ethylsilyloxy)hexahydrocyclopenta[b]furan-2-one (2), is commer-
cially available[28] under the name “(–)-Triethylsilyloxy Corey-diol”:
[α]2D0 = –47.5 (c = 1.00, CHCl3). 1-(4-Iodobutyl)-4-methyl-2,6,7-
trioxabicyclo[2.2.2]octane (7) was obtained as described by Matsa-
mura et al.[14] (R)-2,3-O-Isopropylideneglyceraldehyde (9) was ob-
tained from -mannitol according to the known procedure and was
distilled immediately before use (b.p. 47–51 °C/20 Torr).[19]
a 76.4% yield of latanoprost, in which the (15S)-1 isomer (3aR,4R,5R,6aS)-4-(Phenylthio)methyl-5-(triethylsilyloxy)hexahy-
drocyclopenta[b]furan-2-one (3): To a solution of (3aR,4S,5R,6aS)-
4-(hydroxymethyl)-5-(triethylsilyloxy)-hexahydro-cyclopenta[b]-
furan-2-one (2, 11.45 g, 40 mmol) in CH2Cl2 (40 mL), stirred under
Ar at room temperature, THF (20 mL) and PPh3 (13.1 g, 50 mmol)
were added. After the solids dissolved, the solution was cooled to
15 °C, and benzenethiol (5.51 g, 50 mmol) was added. The mixture
was stirred for 5 min, after which time a solution of DIAD (95%,
10.4 mL, 50 mmol) in THF (20 mL) was slowly added by syringe.
Stirring was continued while the mixture was allowed to gradually
warm to 25 °C. After 15 h, THF (25 mL) was added, and the mix-
could not be detected. However, a few percent of the 5,6-
trans diastereomer of 1 were present in the product. Prepar-
ative HPLC purification of 1 on silica gel gave latanoprost
in better than 99.9% purity and 65% yield from acid 26.[27]
Conclusions
A new synthesis of latanoprost (1) was designed and car-
ried out. High-purity latanoprost was obtained in 15.8% ture was heated at 40 °C for 9 h and then stirred at room tempera-
ture for 16 h. Afterwards, the solvents were evaporated to give an
oil (41 g), which was subjected to flash chromatography on silica
gel (230–400 mesh, 440 g). Elution with hexanes/EtOAc/CH2Cl2
(75:12.5:12.5) gave sulfide 3 (10.84 g, 71.6% yield) as a colorless,
overall yield in eight synthetic steps from sulfone 4 or in
26.5% yield and six steps from prostaglandin intermediate
21. The occurrence of the (15S)-1 diastereomer has been an
“Achilles heel” of most existing syntheses of 1. Therefore,
the present method for latanoprost synthesis targeted a very
high enantiomeric purity at C(15) of the product. The α
chain in a carefully selected, novel derivative of the (–)-Co-
rey lactone was elongated first, followed by the attachment
of a novel, enantiomerically pure ω chain synthon. This en-
viscous oil. [α]2D5 = –31.0 (c = 1.00, CHCl3); H NMR (200 MHz,
1
CDCl3): δ = 0.56 (q, J = 8.0 Hz, 6 H), 0.92 (t, J = 8.0 Hz, 9 H),
2.00 (m, 1 H), 2.08 (m, 1 H), 2.22 (m, 1 H), 2.45–2.85 (m, 4 H),
3.02 (dd, J = 13.2, 5.9 Hz, 1 H), 4.08 (q, J = 5.0 Hz, 1 H), 4.95
(ddd, J = 7.0, 6.8, 2.8 Hz, 1 H), 7.31 (m, 5 H) ppm; 13C NMR
(50 MHz, CDCl3): δ = 4.6 (3 C), 6.7 (3 C), 35.8, 36.1, 40.6, 41.8,
sured the absence of an appreciable quantity of the unde- 54.0, 76.3, 83.1, 126.2, 129.0 (2 C), 129.1 (2 C), 135.4, 176.9 ppm;
696
www.eurjoc.org
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2007, 689–703

A New Synthetic Approach to High-Purity (15R)-Latanoprost
FULL PAPER
EI-MS: m/z (%) = 379 [M + H] (2). C20H30O3SSi (378.6): calcd. C
63.45, H 7.99, S 8.47; found C 63.37, H 8.03, S 8.46.
(ddd, J = 8.8, 8.6, 5.8 Hz, 0.75 H), 3.98 (br. q, J = 5.1 Hz, 0.25 H),
4.60 (m, 1.25 H), 5.48 (ddd, J = 6.4, 4.5, 1.7 Hz, 0.25 H), 5.62 (dd,
J = 3.7, 3.5 Hz, 0.75 H), 7.91 (m, 2 H), 7.61 (m, 3 H) ppm; 13C
NMR (50 MHz, CDCl3): δ = 4.45,4.73, 6.63, 6.73, 40.36, 40.68,
40.99, 42.54, 44.18, 46.43, 48.45, 49.08, 59.04, 59.80, 77.39, 78.46,
79.90, 83.62, 100.42 and 101.22 (hemiacetal), 127.61, 127.80,
129.19, 129.23, 133.54, 133.66, 139.28, 139.58 ppm; C20H32O5SSi
(412.6): calcd. C 58.22, H 7.82, S 7.77; found C 57.98, H 7.78, S
8.05.
(3aR,4R,5R,6aS)-4-(Phenylsulfonyl)methyl-5-(triethylsilyloxy)-hexa-
hydro-cyclopenta[b]furan-2-one (4): (a) Sulfide 3 (9.55 g, 25.2 mmol)
was dissolved in CH2Cl2 (150 mL). The flask was cooled in a H2O
bath at 17 °C. Upon vigorous stirring, a suspension of magnesium
monoperoxyphthalate hexahydrate (MMPP·6H2O, 80 % pure,
89.7 g, ca. 145 mmol) in H2O (230 mL) was added over 2 min. Vig-
orous stirring was continued for 65 min, after which time saturated
aqueous NaHCO3 (350 mL) was added dropwise over 20 min. Stir-
ring was continued for another 15 min. The phases were separated,
and the aqueous phase was extracted with CH2Cl2 (40 mL). The
organic phases were combined and extracted with saturated aque-
ous NaHCO3 (300 mL). The phases were separated, and the aque-
ous phase was re-extracted with CH2Cl2 (40 mL). The organic
phases were combined, extracted with 10 % aqueous Na2S2O3
(500 mL) and saturated aqueous NaHCO3 (300 mL). The organic
layer was dried with MgSO4 (50 g). The drying agent was filtered
and then washed with CH2Cl2 (50 mL). The filtrates were com-
bined, concentrated in vacuo, and dried in vacuo (1 Torr, 30 °C,
30 min), affording compound 4 as a colorless, solidifying oil
(9.81 g, 94.8% yield). [α]2D5 = +24.9 (c = 1.00, CHCl3); 1H NMR
(200 MHz, CDCl3): δ = 0.53 (q, J = 8.0 Hz, 6 H), 0.89 (t, J =
8.0 Hz, 9 H), 1.98 (m, 1 H), 2.15 (m, 1 H), 2.33 (m, 1 H), 2.62–
[4-(4-Methyl-2,6,7-trioxabicyclo[2.2.2]oct-1-yl)butyl]triphenylphos-
phonium Iodide (8): A mixture of iodide 7 (15.8 g, 50.6 mmol), PPh3
(14.6 g, 55.66 mmol), sulfolane (20 mL), and anhydrous pyridine
(0.1 mL) was stirred under Ar at 80 °C for 70 min. After cooling
to 40 °C, CHCl3 containing 0.3% of pyridine (70 mL) was added.
The resulting solution was added dropwise to a mixture of Et2O
(1.5 L) and EtOAc (0.6 L), vigorously stirred under Ar. Stirring
was continued for 30 min, and then the solution was decanted. The
remaining precipitate was dissolved in CHCl3 containing 0.2% pyr-
idine (75 mL). The solution thus obtained was added dropwise to
the mixture of Et2O (1.2 L) and EtOAc (0.5 L), vigorously stirred
under Ar. The stirring was continued for 20 min at 25 °C. The solu-
tion was decanted, and the precipitate was washed with Et2O
(100 mL) and dried (1 Torr, 25 °C, 1.5 h). The crude product
(28.95 g) was dissolved in MeOH containing 0.04 % pyridine
2.98 (m, 4 H), 3.18 (dd, J = 13.9, 4.4 Hz, 1 H), 4.01 (q, J = 5.1 Hz, (35 mL), EtOAc containing 0.04% pyridine (65 mL) was added,
1 H), 4.98 (ddd, J = 7.1, 6.8, 3.5 Hz, 1 H) ppm; 13C NMR
(50 MHz, CDCl3): δ = 4.6 (3 C), 6.7 (3 C), 35.8, 40.4, 41.9, 49.4,
58.5, 76.7, 82.6, 127.7 (2 C), 129.4 (2 C), 133.9, 139.0, 176.6 ppm;
HRMS (ESI): calcd. for C20H31O5SSi [M + H]+ 411.1656; found
411.1676; C20H30O5SSi (410.6): calcd. C 58.50, H 7.36, S 7.81;
found C 58.36, H 7.22, S 8.01.
and the mixture was stored at 4 °C. The crystals were filtered and
dried (1 Torr, 25 °C, 1 h) to give phosphonium iodide 8 (14.01 g,
48.3% yield) as colorless thick prisms. M.p. 130–134 °C; 1H NMR
(200 MHz, CDCl3): δ = 0.78 (s, 3 H), 1.58–1.80 (m, 6 H), 3.45 (m,
2 H), 3.79 (s, 6 H), 7.70–7.88 (m, 15 H) ppm; 13C NMR (50 MHz,
CDCl3): δ = 14.5, 21.7 (d, J = 4.0 Hz), 22.7 (d, J = 50.8 Hz), 24.0
(d, J = 16.8 Hz), 30.2, 34.7, 72.3 (3ϫC), 108.5, 117.8 (3ϫC, d, J
= 86 Hz), 130.5 (6 ϫ C, d, J = 12.4 Hz), 133.5 (6 ϫ C, d, J =
10.0 Hz), 135.1 (3ϫC, d, J = 2.8 Hz) ppm.
(b) In another series of two reactions, compound 2 (211.2 g) gave
sulfone 4 (179.5 g, 69.2% yield as a solidifying oil), as described
above. This sample of 4 was crystallized from EtOAc/hexanes (1:1,
400 mL). The white precipitate thus obtained was then recrystal-
lised from Et2O/hexanes (3:1). This gave crystalline sulfone 4
(141.25 g, 46.7% yield from 2) as white needles; m.p. 75–76 °C,
other analytical data were identical with that described above for
compound 4. Chiral HPLC analysis [Chiralcel OD 10 µm,
250ϫ4.6 mm column, n-hexanes/2-propanol (85:15), and 1.0 mL/
(S)-4-Phenylbutane-1,2-diol (12). (a) By Asymmetric Synthesis: A
mixture of tert-butyl alcohol (450 mL), distilled H2O (450 mL),
(DHQ)2AQN (Aldrich, 95%, 990 mg, 1.1 mmol), K2CO3 (38.7 g,
280 mmol), K3Fe(CN)6 (93.1 g, 280 mmol), and K2OsO2(OH)4
(133 mg, 0.33 mmol) was vigorously stirred at room temperature
for 1.5 h. The mixture was cooled to 0 °C, 4-phenyl-1-butene
min] gave only one peak for the crystallized sample of compound (13.52 mL, 11.90 g, 90.0 mmol) was added, and stirring at 0 °C was
4.
continued for 17 h. Na2S2O5 (130 g) was added in a few portions,
and the mixture was warmed to room temperature over 1 h. The
product was extracted with EtOAc twice (400 mL and 100 mL).
The combined organic phases were dried with anhydrous Na2SO4
(100 g) and concentrated in vacuo. The residue was purified by
flash chromatography on silica gel (350 g). Elution with EtOAc
gave diol 12 (14.70 g, 98% yield, 65.3% ee) as a viscous oil. 1H
NMR (200 MHz, CDCl3): δ = 1.72 (m, 2 H), 2.70 (m, 2 H), 3.31
(br. s, 2 H) 3.42 (dd, J = 11.2, 7.7 Hz, 1 H), 3.61 (dd, J = 11.2,
2.9 Hz, 1 H), 3.69 (m, 1 H), 7.13–7.31 (m, 5 H) ppm; 13C NMR
(50 MHz, CDCl3): δ = 31.8, 34.6, 66.6, 71.5, 125.8, 128.2 (2 C),
128.3 (2 C), 141.5 ppm; chiral HPLC: Chiralcel OD, 10 µm, (250
+ 20) ϫ4.6 mm column, hexanes/2-propanol, 80:20 (v/v), 1.0 mL/
min, Rt 8.47 min (17.2% yield), Rt 11.06 min (81.9% yield), 65.3%
ee. Racemic diol 12, used as a standard for HPLC, was prepared
from 4-phenyl-1-butene by a dihydroxylation reaction as described
above, in which DABCO was used in place of (DHQ)2AQN.
(2R/S,3aR,4R,5R,6aS)-4-[(Phenylsulfonyl)methyl]-5-(triethylosilyl-
oxy)hexahydro-2H-cyclopenta[b]furan-2-ol (5): To a solution of lac-
tone 4 (9.26 g, 22.55 mmol) in THF (120 mL), stirred under Ar at
–75 °C, DIBALH (1.4  toluene solution, 35.4 mL, 49.6 mmol) was
added dropwise over 5 min. Stirring at –75 °C was continued for
2 h, and then MeOH (9.5 mL, 234 mmol) was added slowly (foam-
ing!) with vigorous stirring. The mixture was warmed to –5 °C, and
H2O (130 mL) was added dropwise, followed by aqueous NaHSO4
(2 , 100 mL). Stirring was continued for 5 min, EtOAc (100 mL)
was added, and the phases were separated. The aqueous layer was
extracted with EtOAc (2ϫ80 mL). The organic layers were com-
bined, washed with brine (2ϫ200 mL), dried with Na2SO4 (50 g),
filtered, and concentrated under reduced pressure. The viscous, col-
orless oil thus obtained (10.1 g) was purified by flash chromatog-
raphy on silica gel (200 g). Elution with 42–50% EtOAc/hexanes +
0.02% C5H5N gave hemiacetal 5 (3:1 mixture of epimers) as a col-
orless, viscous oil (8.50 g, 91.3 % yield). 1H NMR (200 MHz,
(b) Chiral-Pool Synthesis. (E/Z,4S)-2,2-Dimethyl-4-styryl-1,3-diox-
CDCl3): δ = 0.52 (q, 6 H), 0.89 (t, 9 H), 1.62 (m, 0.75 H), 1.98– olane (10): To a suspension of benzyltriphenylphosphonium bro-
2.37 (m, 4 H), 2.62 (m, 1.25 H), 2.90–3.04 (m, 1.75 H), 3.23 (dd, J mide (189.4 g, 0.437 mol) in THF (1.2 L), vigorously stirred under
= 14.1, 4.4 Hz, 0.25 H), 3.36 (dd, J = 14.0, 3.1 Hz, 0.75 H), 3.78
Ar at 0 °C, hexyllithium (2.5  solution in hexanes, 170 mL,
Eur. J. Org. Chem. 2007, 689–703
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
697

J. G. Martynow, O. Achmatowicz et al.
FULL PAPER
0.425 mol) was added dropwise while the temperature was main-
tained below 5 °C. The temperature of the mixture was then al-
lowed to increase to 15 °C during 1 h. The mixture was stirred at
15 °C for 30 min, and then it was cooled to 0 °C. A solution of
freshly distilled (R)-2,3-O-isopropylideneglyceraldehyde (9, 52.3 g,
0.402 mol) in THF (150 mL), cooled to Ͻ5 °C, was added drop-
wise. The mixture was stirred at 20 °C for 2 h, and then MeOH
(10 mL) was slowly added. The suspension thus formed was filtered
through Celite (100 g), which was additionally washed with hex-
anes/EtOAc (2:1, 2ϫ200 mL). The filtrates were combined, con-
centrated, and subjected to flash chromatography on silica gel
(500 g). Elution with hexanes/EtOAc (1:1) gave dioxolane 10
(50.9 g, 62% yield) as a colorless, viscous oil. 1H NMR (200 MHz,
CDCl3): δ = 1.39 (s, 2.3 H), 1.43 (s, 0.7 H), 1.47 (br. s, 3 H), 3.68
(m, 1 H), 4.08 (m, 0.3 H), 4.16 (m, 0.7 H), 4.67 (m, 0.3 H), 4.92
(m, 0.7 H), 5.70 (dd, J = 11.6, 9.0 Hz, 0.7 H), 6.16 (dd, J = 15.8,
7.6 Hz, 0.3 H), 6.70 (m, 1 H), 7.23–7.42 (m, 5 H) ppm; 13C NMR
(50 MHz, CDCl3): δ = 25.89, 25.94, 26.75, 26.86, 69.51, 69.70,
72.41, 77.24, 109.37, 109.43, 126.61, 126.71, 127.52, 127.98, 128.29,
128.57, 128.70, 129.20, 133.38, 133.97, 136.13, 136.24 ppm.
J = 4.6 Hz, 1 H), 2.45 (s, 3 H), 2.70 (m, 2 H), 3.85 (m, 2 H), 4.02
(dd, J = 9.5, 2.9 Hz, 1 H), 7.11–7.37 (m, 7 H), 7.79 (ddd, J = 8.4,
2.0, 1.8 Hz, 2 H) ppm; 13C NMR (50 MHz, CDCl3): δ = 21.6, 31.3,
34.1, 68.5, 73.8, 125.9, 127.8 (2 C), 128.2 (2 C), 128.3 (2 C), 129.8
(2 C), 132.4, 140.9, 144.9 ppm; chiral HPLC: Chiralcel OD, 10 µm,
(250+20) ϫ 4.6 mm column, hexanes/2-propanol, 80:20 (v/v),
1.0 mL/min, Rt 13.28 min (99.59 % yield), Rt 15.53 min (0.37%
yield), 99.26% ee; C17H20O4S (272.4): calcd. C 63.73, H 6.29, S
10.01; found C 63.79, H 6.19, S 10.16.
(b) Chiral-Pool Synthesis: Diol 12 (40 g, 99.3% ee) was monotosyl-
ated as described above for the asymmetric dihydroxylation prod-
uct. The usual workup and flash chromatography on silica gel (1
kg, elution with 25% EtOAc/hexanes) gave compound 13 (60 g,
77.8% yield) as a colorless, viscous oil. Crystallization from Et2O
(210 mL) gave p-toluenesulfonate 13 (52 g, 67.4% yield) as colorless
needles. [α]2D0 = +1.0 (c = 1.00, CHCl3); 1H NMR (200 MHz,
CDCl3) was identical with that given above for compound 13, ob-
tained from the asymmetric synthesis; chiral HPLC: Chiralcel OD,
10 µm, (250+20) ϫ 4.6 mm column, hexanes/2-propanol, 80:20
(v/v), 1.0 mL/min, Rt 13.2 min (98.370 % yield), Rt 15.9 min
(0.039% yield), 99.92% ee.
(4S)-2,2-Dimethyl-4-phenethyl-1,3-dioxolane (11): The (E/Z) mix-
ture of dioxolanes (4S)-10 (50.0 g) was dissolved in MeOH (0.5 L),
and the solution was hydrogenated in a Parr reactor over 10% Pd/
C (5.0 g) under a pressure of 10 bar at 30 °C for 24 h. The catalyst
was filtered, washed with MeOH (3ϫ50 mL), the solvent was evap-
orated under reduced pressure, and the residue was dried in vacuo
(S)-4-Phenyl-1-(p-tolylsulfonyloxy)-2-(triethylsilyloxy)butane (14):
(a) By Asymmetric Synthesis: To a solution of p-toluenesulfonate
13 (4.537 g, 14.16 mmol, 99.26% ee) in anhydrous DMF (38 mL),
stirred under Ar at 0 °C, imidazole (1.07 g, 15.7 mmol) and Et3N
(2.0 mL, 1.45 g, 14.3 mmol) were added, followed by chlorotrieth-
ylsilane (2.52 mL, 2.26 g, 15.0 mmol). Stirring at 0 °C was contin-
ued for 1 h and then at room temperature for 20 min. The mixture
was diluted with hexanes (100 mL) and washed with saturated
aqueous NaHCO3 (90 mL). The aqueous layer was extracted with
hexanes (2ϫ40 mL). The organic phases were combined, washed
with saturated aqueous NaHCO3 (100 mL), dried with Na2SO4
(20 g), and concentrated in vacuo. The residue thus obtained (7.0 g)
was purified by flash chromatography on silica gel (185 g). Elution
with 10% EtOAc/hexanes gave triethylsilyl derivative 14 (6.127 g,
99.5% yield, 99.26% ee) as a colorless, viscous oil. [α]2D5 = +4.9 (c
= 1.00, CHCl3); 1H NMR (200 MHz, CDCl3): δ = 0.56 (q, 8.0 Hz,
6 H), 0.91 (t, J = 8.0 Hz, 9 H), 1.74 (m, 2 H), 2.44 (s, 3 H), 2.60
(m, 2 H), 3.90 (br. s, 3 H), 7.09–7.36 (m, 7 H), 7.78 (ddd, J = 8.4,
2.0, 1.8 Hz, 2 H) ppm; 13C NMR (50 MHz, CDCl3): δ = 4.8 (3 C),
6.8 (3 C), 21.6, 31.0, 35.8, 69.4, 72.8, 125.8, 127.8 (2 C), 128.1 (2
C), 128.3 (2 C), 129.7 (2 C), 132.7, 141.4, 144.7 ppm; C23H34O4SSi
(434.7): calcd. C 63.55, H 7.88, S 7.38; found C 63.62, H 7.72, S
7.40.
to give dioxolane 11 (49.9 g, 98.8% yield) as a colorless oil. [α]2D0
=
+3.8 (c = 1.00, CHCl3); 1H NMR (200 MHz, CDCl3): δ = 1.36 (br.
s, 3 H), 1.43 (br. s, 3 H), 1.88 (m, 2 H), 2.70 (m, 2 H), 3.52 (dd, J
= 7.5, 7.0 Hz, 1 H), 4.05 (m, 2 H), 7.14–7.33 (m, 5 H) ppm; 13C
NMR (50 MHz, CDCl3): δ = 25.73, 26.99, 32.02, 35.33, 69.31,
75.35, 108.71, 125.95, 128.35, 125.41, 141.53 ppm.
(S)-4-Phenylbutan-1,2-diol (12): A solution of dioxolane 11 (49.8 g,
0.241 mol) in MeOH (600 mL) was treated with PTSA (0.5 g), and
the mixture was stirred at 40 °C for 4 h. Et3N (2 mL) was added,
the solvent was evaporated, and the residue was subjected to flash
chromatography on silica gel (500 g). Elution with EtOAc afforded
diol 12 (40.0 g, 99% yield, 99.3% ee) as a colorless oil, which solidi-
fied at 4 °C. M.p. 34–36 °C; [α]2D0 = –13.6 (c = 1.00, CHCl3); [α]2D0
1
= –33.1 (c = 1.00, EtOH); H NMR (200 MHz, CDCl3) was iden-
tical with that given for 12 above; chiral HPLC: Chiralcel OD,
10 µm, (250+20) ϫ 4.6 mm column, hexanes/2-propanol, 80:20
(v/v), 1.0 mL/min, Rt 8.5 min (0.33% yield), Rt 11.1 min (98.86%
yield), 99.3% ee. For HPLC calibration, rac-4-phenyl-1,2-butane-
diol was used.
(b) Chiral-Pool Synthesis: Crystalline compound 13 (10.0 g, 99.92%
ee) was silylated with chlorotriethylsilane according to the pro-
cedure described above, to give triethylsilyl derivative 14 (13.43 g,
(S)-4-Phenyl-1-(p-tolylsulfonyloxy)butan-2-ol (13). (a) By Asymmet-
ric Synthesis: To a solution of (S)-4-phenyl-1,2-butanediol (12,
13.58 g, 81.7 mmol, 65 % ee) in anhydrous CH2Cl2 (190 mL),
stirred under Ar, was added Bu2SnO (720 mg, 2.89 mmol). The sus-
pension was stirred for 5 min prior to adding Et3N (11.40 mL,
8.276 g, 81.79 mmol). The mixture was cooled to 0 °C, and p-tolu-
enesulfonyl chloride (16.03 g, 84.08 mmol) was added. The mixture
was stirred at 0 °C for 5 min, gradually warmed to room tempera-
ture over 1.5 h, and then left overnight at 4 °C. The mixture was
then concentrated to 100 mL and purified by flash chromatography
on silica gel (500 g). Elution with 25% EtOAc/hexanes gave nonra-
cemic tosylate 13 (20.1 g, 76.8% yield) as a colorless, viscous oil.
This sample of 13 (19.85 g) was crystallized from Et2O (60 mL),
leading to crystals (10.51 g). This sample was further recrystallised
from Et2O twice, affording compound 13 (4.77 g, 18.45% yield,
99.26% ee) as needles. M.p. 68–69 °C; [α]2D5 = +0.70 (c = 1.00,
1
99% yield, 99.92% ee); H NMR (200 MHz, CDCl3) was identical
to that described above for compound 14.
(S)-4-Phenyl-1-iodo-2-(triethylsilyloxy)butane (15). (a) By Asym-
metric Synthesis: To a solution of tosylate 14 (3.145 g, 7.23 mmol,
99.26% ee) in anhydrous DMF (30 mL), vigorously stirred under
Ar at room temperature, was added sodium iodide (4.60 g,
30.7 mmol), and the mixture was heated and stirred at about 80 °C
for 2 h. After cooling to about 40 °C, saturated aqueous NaHCO3
(70 mL) was added, and the mixture was extracted with hexanes
(70 mL and 2 ϫ 40 mL). The combined hexanes solutions were
washed with saturated aqueous NaHCO3 (70 mL), dried (Na2SO4,
16 g), and concentrated in vacuo to give a yellow oil (2.88 g), which
was subjected to flash chromatography on silica gel (100 g). Elution
with 3–10 % EtOAc/hexanes afforded iodide 15 (2.423 g, 85.8%
yield, 99.2% ee) as a colorless oil {[α]2D0 = –9.1 (c = 1.00, CHCl3);
1
CHCl3); H NMR (200 MHz, CDCl3): δ = 1.73 (m, 2 H), 2.19 (d,
698
www.eurjoc.org
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2007, 689–703

A New Synthetic Approach to High-Purity (15R)-Latanoprost
1H NMR (200 MHz, CDCl3): δ = 0.62 (q, J = 7.6 Hz, 6 H), 0.98
FULL PAPER
was carefully introduced under the surface of the stirred solution
(t, J = 7.6 Hz, 9 H), 1.78 (m, 2 H), 2.65 (m, 2 H), 3.23 (dd, J = over 10 min such that the temperature was maintained at 0–5 °C,
5.1, 0.6 Hz, 2 H), 3.66 (m, 1 H), 7.14–7.33 (m, 5 H) ppm; 13C NMR
and stirring was continued for another 1 h. The mixture was
(50 MHz, CDCl3): δ = 5.05 (3 C), 6.90 (3 C), 13.25, 31.26, 38.61, poured into pH 7.2 phosphate buffer (0.1 , 500 mL) and stirred
71.01, 125.73, 128.16 (2 C), 128.25 (2 C), 141.58 ppm; HRMS at 0 °C. The organic phase was separated, washed with 2% aqueous
(ESI): calcd. for C16H27IONaSi [M + Na]+ 413.0768; found
413.0764} and recovered 14 [317 mg, 10.1 % yield, 1H NMR
(200 MHz, CDCl3) was identical to that cited above for this
compound]}.
NaCl (500 mL), dried with Na2SO4, and concentrated under re-
duced pressure, yielding cyclic sulfite 18 [16.5 g, 96% yield, 2 dia-
1
stereomers (1:1)] as a thick oil. H NMR (200 MHz, CDCl3): δ =
1.83–2.38 (m, 2 H), 2.62–2.97 (m, 2 H), 3.90 (dd, J = 8.1, 6.8 Hz,
0.5 H), 4.24–4.52 (m, 1 H), 4.62 (dd, J = 8.5, 6.4 Hz, 0.5 H), 4.87–
5.00 (m, 0.5 H), 7.16–7.35 (m, 5 H) ppm; 13C NMR (50 MHz,
CDCl3): δ = 31.58, 31.99, 34.09, 35.13, 70.20, 71.50, 79.41, 82.98,
126.45, 126.49, 128.42, 128.46, 128.6, 128.70, 140.06, 140.16 ppm.
(b) Chiral-Pool Synthesis: Tosylate 14, obtained from the chiral-
pool synthesis (13.2 g, 99.92% ee) was treated with NaI in DMF
for 2.5 h, essentially as described above, but the temperature was
very strictly maintained within the 75–80 °C range. Workup and
chromatography as described above gave iodide 15 (10.55 g, 89%
yield, 99.92% ee); 1H NMR (200 MHz, CDCl3) was identical to
that cited above for this compound.
(2R/S,4S)-4-(2-Phenylethyl)-1,3,2-dioxathiolane 2,2-Dioxide (19).
Chiral-Pool Synthesis: Sulfite 18 (7.8 g, 36.7 mmol) was dissolved in
acetonitrile (100 mL) and sodium periodate (11.0 g), RuCl3·3H2O
(81 mg), and H2O (20 mL) were added to the stirred solution. The
reaction mixture was warmed to 40 °C over 10 min, after which
time the oxidation was complete (TLC, hexanes/EtOAc, 2:1). The
mixture was cooled to 20 °C, Et2O (100 mL) and H2O (80 mL)
were added, and after extraction, the phases were separated. The
aqueous phase was extracted with Et2O (2ϫ200 mL). The Et2O
layers were combined, dried with Na2SO4, filtered, and concen-
trated in vacuo to give a dark solid (7.9 g), which was crystallized
from Et2O (20 mL), affording cyclic sulfate 19 (3.50 g, 42% yield,
100% ee) as white crystals. M.p. 50–51 °C; [α]2D0 = –42.9 (c = 1,
(S)-2-Hydroxy-1-iodo-4-phenylbutane (16). Chiral-Pool Synthesis:
Compound 15 (3.52 g, 9.0 mmol, 99.2% ee) was dissolved in ace-
tone (20 mL). H2O (2 mL) was added. This solution was stirred at
20 °C, and pyridinium p-toluenesulfonate (200 mg) was added. Af-
ter 20 h, the mixture was poured into 3 % aqueous NaHCO3
(150 mL), and the mixture was extracted with Et2O/EtOAc (1:1,
100 mL). The phases were separated, the organic phase was dried
with Na2SO4 (15 g), and the drying agent was filtered and washed
with EtOAc (20 mL). The filtrates were combined and concentrated
in vacuo, and the oil thus obtained was purified by flash
chromatography on silica gel (120 g). Elution with 15% EtOAc/
hexanes gave pure fractions by TLC, which were concentrated in
vacuo to 10 mL and allowed to crystallize at 0 °C for 2 h. The
crystals were filtered and dried, giving compound 16 (1.80 g, 72.3%
yield, 99.85 % ee) as white needles. M.p. 76–77 °C; 1H NMR
(200 MHz, CDCl3): δ = 1.87 (m, 2 H), 2.74 (m, 2 H), 3.24 (dd, J
= 10.2, 6.8 Hz, 1 H), 3.38 (dd, J = 10.1, 3.5 Hz, 1 H), 3.52 (m, 1
H), 7.15–7.34 (m, 5 H) ppm; 13C NMR (50 MHz, CDCl3): δ =
16.59, 31.91, 38.14, 70.12, 126.07, 128.42 (2 C), 128.50 (2 C), 141.26
ppm; C10H13IO (276.1): calcd. C 43.50, H 4.75, I 45.96; found C
43.49, H 4.61, I 46.00; chiral HPLC: Chiralcel OD, 10 µm,
(250+20) ϫ 4.6 mm column, hexanes/2-propanol, 85:15 (v/v),
1.0 mL/min, Rt 8.2 min (99.867% yield), Rt 11.1 (0.073% yield),
99.85% ee.
MeOH); IR (KBr): ν = 651, 702, 757, 783, 853 (s), 965 (s), 1013,
˜
1
1038, 1210 (s), 1381 (s), 1602 cm–1; H NMR (200 MHz, CDCl3):
δ = 1.93–2.11 (m, 1 H), 2.19–2.37 (m, 1 H), 2.65–2.94 (m, 2 H),
4.28 (t, J = 8.1 Hz, 1 H), 4.60 (dd, J = 9.0, 6.0 Hz, 1 H), 4.85–4.98
(m, 1 H), 7.16–7.38 (m, 5 H) ppm; 13C NMR (50 MHz, CDCl3): δ
= 30.73, 33.86, 72.68, 81.99, 126.68, 128.31, 128.76, 139.07 ppm;
C10H12O4S (228.3): calcd. C 52.62, H 5.30, S 14.05; found C 52.66,
H 5 . 3 4 , S 1 4 . 0 9 ; c h i ra l H P LC : C h i ra l c el OD, 10 µm,
(250+20) ϫ 4.6 mm column, hexanes/2-propanol, 85:15 (v/v),
1.0 mL/min, Rt 41.0 min (100% yield), the other peak was not de-
tected; chiralcel AD, 10 µm, (250+20)ϫ4.6 mm column, hexanes/
2-propanol, 85:15 (v/v), 1.0 mL/min, Rt 10.5 min (100% yield), the
other peak was not detected, 100% ee.
(1R,3S,4R,5R)-4-[(Z)-6-(4-Methyl-2,6,7-trioxabicyclo[2.2.2]octan-1-
yl)hex-2-enyl]-5-[(phenylsulfonyl)methyl]cyclopentan-1,3-diol (20a)
and a Mixture of (1S,2R,3R,4R)-2-[(Z)-6-(4-Methyl-2,6,7-trioxa-
bicyclo[2.2.2]octan-1-yl)hex-2-enyl]-3-[(phenylsulfonyl)methyl]-4-(tri-
ethylsilyloxy)cyclopentanol (20b) and (1R,2R,3R,4S)-3-[(Z)-6-(4-
Methyl-2,6,7-trioxabicyclo[2.2.2]octan-1-yl)hex-2-enyl]-2-[(phenyl-
sulfonyl)methyl]-4-(triethylsilyloxy)cyclopentanol (20c): To a suspen-
sion of phosphonium iodide 8 (7.98 g, 13.9 mmol), stirred under
Ar at 0 °C in anhydrous THF (25 mL), tBuOK (3.85 g, 34 mmol,
Fluka Ͼ 97%) was added in a few portions. After 5 min, the cool-
ing bath was removed, and anhydrous THF (10 mL) was added.
The mixture was vigorously stirred and warmed to 20 °C over
20 min. The reaction mixture was cooled to 0 °C, and a solution
of hemiacetal 5 (mixture of epimers, 2.70 g, 6.54 mmol) in anhy-
drous THF (10 mL) was added. After 15 min, the cooling bath was
removed, and the mixture warmed to 20 °C. After stirring for
80 min, Al(tBuO)3 (420 mg, 1.7 mmol) was added, and stirring at
20 °C was continued for another 5 h. The reaction mixture was
cooled to 0 °C, and 3% aqueous pyridine (10 mL) and saturated
aqueous NaHCO3 (70 mL) were slowly added. The mixture was
extracted with EtOAc (3ϫ50 mL). The organic layers were com-
(S)-1,2-Epoxy-4-phenylbutane (17). Chiral-Pool Synthesis: Crystal-
line compound 16 (1.61 g, 5.84 mmol, 99.85% ee) was dissolved in
THF (12 mL). H2O (1.5 mL) and solid KOH (1.95 g) were added.
The mixture was stirred at 20 °C under Ar for 7 h, and then trans-
ferred to a separatory funnel containing brine (50 mL), H2O
(100 mL), and Et2O (120 mL). After extraction, the phases were
separated, and the organic phase was extracted with H2O (150 mL)
and then dried with Na2SO4 (20 g). The drying agent was filtered
and washed with Et2O (20 mL). The filtrates were combined and
concentrated in vacuo at 0 °C and then dried (10 °C, 5 Torr,
15 min), affording epoxide 17 (0.70 g, 81% yield, 99.85% ee). [α]2D0
= –22.5 (c = 1.00, CHCl3); 1H NMR (200 MHz, CDCl3): δ = 1.79–
1.92 (m, 2 H), 2.47 (dd, J = 4.9, 2.6 Hz, 1 H), 2.66–2.87 (m, 3 H),
2.91–3.00 (m, 1 H), 7.15–7.34 (m, 5 H) ppm; 13C NMR (50 MHz,
CDCl3): δ = 32.25, 34.29, 47.25, 51.79, 126.02, 128.38, 128.45,
141.26 ppm; C10H12O (148.2) calcd. C 81.04, H 8.16; found C
81.25, H 8.20.
(2R/S,4S)-4-(2-Phenylethyl)-1,3,2-dioxathiolane 2-Oxide (18). Chi-
ral-Pool Synthesis: Diol 12 (13.4 g, 80.6 mmol, 99.92% ee) was dis-
solved in CH2Cl2 (600 mL). Diisopropylethylamine (28.2 mL, bined and washed with brine (2ϫ50 mL). Pyridine (3 drops) was
20.92 g, 161.9 mmol) was added, and the mixture was cooled to
added, and the solution was dried with Na2SO4 (25 g) at 4 °C and
5 °C. Subsequently, thionyl chloride (6.20 mL, 10.11 g, 85 mmol)
then concentrated in vacuo to give an oil (8.2 g). This sample was
Eur. J. Org. Chem. 2007, 689–703
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
699

J. G. Martynow, O. Achmatowicz et al.
FULL PAPER
purified by flash chromatography on silica gel (250 g). Elution with
EtOAc/hexanes (70:30) + 0.15% pyridine gave a mixture (35:65) of
20b and 20c (584 mg, 15.4% yield) as a colorless glass; H NMR
it was warmed to room temperature. After 5 h of stirring, the mix-
ture was cooled to –78 °C, and 1 % pyridine in brine/saturated
aqueous NaHCO3 (1:1, 4 mL) was added dropwise, followed by
EtOAc/CH2Cl2 (6:1, 50 mL). The cooling bath was changed to a
H2O bath at 10 °C. 1 % Pyridine in brine/saturated aqueous
NaHCO3 (1:1, 50 mL) was added, and the mixture was extracted
twice with EtOAc/CH2Cl2 (1:1, 60 mL and 30 mL). The organic
phases were combined, washed with 1% pyridine in brine/saturated
aqueous NaHCO3 (1:1, 50 mL), dried with anhydrous Na2SO4
(15 g), filtered, concentrated, and dried under vacuum to give a
yellow oil (6.92 g), which was purified by flash chromatography on
silica gel (250 g, 0.1% pyridine in 5–99% EtOAc/hexanes). This
procedure gave iodide 15 [3.76 g, 99.2% ee, 1H NMR (200 MHz,
CDCl3) was identical with that of an original sample], sulfone 22
[1.456 g, 46.2% yield, a 5:2 mixture of isomers as a colorless glassy
solid. The composition of this mixture was established based on the
integration in the 1H NMR (500 MHz, CDCl3) spectrum; NMR
spectroscopic data for the individual components (13R)-22 and
(13S)-22 are cited below. C52H88O8SSi3 (957.6): calcd. C 65.22, H
9.26, S 3.35; found C 65.58, H 9.37, S 3.55], recovered substrate 21
[165 mg, 7.2 % yield as a colorless glass; 1H NMR (200 MHz,
CDCl3) was identical with that of an original sample], and ester 23
[453 mg, 13.5% yield, a 5:2 mixture of isomers as a yellowish, thick
oil; 1H NMR (200 MHz, CDCl3): δ = 0.57 (m, 18 H), 0.83 (s, 3 H),
0.94 (m, 27 H), 1.40–1.82 (m, 5 H), 1.90 (m, 2 H), 1.98–2.45 (m,
10 H), 3.10 (br. s, 1 H), 3.22 (m, 0.29 H), 3.47 (m, 0.71 H), 3.55
(br. s, 4 H), 3.87 (m, 1 H), 3.89 (br. s, 1 H), 4.18 (m, 4 H), 4.41 (m,
0.29 H), 4.53 (m, 0.71 H), 5.26–5.53 (m, 2 H), 7.08–7.33 (m, 5 H),
7.46–7.72 (m, 3 H), 7.84–7.94 (m, 2 H) ppm]. A sample of 22
(90 mg, 5:2 mixture of epimers) was separated by repeated
chromatography on silica gel LiChroprep (25–40 µm, 7 g). Elution
with 10% EtOAc/hexanes with 0.1% pyridine gave the more mobile
major isomer, 1-{(Z)-6-[(1R,2R,3R,5S)-2-[(1R,3S)-3-(triethylsilyl-
oxy)-5-phenyl-1-(phenylsulfonyl)pentyl]-3,5-bis(triethylsilyloxy)-
cyclopentyl]hex-4-enyl}-4-methyl-2,6,7-trioxabicyclo[2.2.2]octane
{[(13R)-22, 26.6 mg) as a colorless viscous oil; 1H NMR (500 MHz,
CDCl3 + 1% C5D5N): δ = 0.572 (m, 18 H), 0.798 (s, 3 H), 0.940
(m, 27 H), 1.560 (m, 4 H), 1.654 (m, 3 H), 1.730 (m, 2 H), 1.992
(m, 1 H), 2.074–2.238 (m, 5 H), 2.362 (ddd, J = 10.1, 8.2, 1.9 Hz,
1 H), 2.494 (ddd, J = 13.6, 11.0, 5.2 Hz, 1 H), 2.584 (ddd, J = 13.6,
1
(200 MHz, CDCl3): δ = 0.57 (q, J = 8.0 Hz, 6 H), 0.79 (s, 1.95 H),
0.80 (s, 1.05 H), 0.93 (t, J = 8.0 Hz, 9 H), 1.34–1.54 (m, 3.5 H),
1.56–1.78 (m, 3 H), 1.82–2.20 (m, 6.6 H), 3.01 (dd, J = 14.4,
11.2 Hz, 0.65 H), 3.32 (dd, J = 14.4, 2.6 Hz, 0.65 H), 3.56 (m, 0.65
H), 3.88 (s, 3.9 H), 3.89 (s, 2.1 H), 3.90 (m, 0.35 H), 4.13 (m, 1.30
H), 5.15–5.38 (m, 2 H), 7.54–7.74 (m, 3 H), 7.93–7.99 (m, 2 H)
ppm. Further elution with EtOAc/hexanes (84:16) + 0.15% pyri-
dine gave 3.597 g of a glassy solid, slightly contaminated with
Ph3PO. Trituration of this solid with Et2O twice (8 mL and 4 mL)
1
afforded compound 20a (2.0 g, 65.5% yield) as a glassy solid. H
NMR (200 MHz, CDCl3): δ = 0.80 (s, 3 H), 1.36–1.52 (m, 3 H),
1.61 (m, 2 H), 1.8–2.29 (m, 8 H), 3.05 (dd, J = 14.3, 11.0 Hz, 1 H),
3.33 (dd, J = 14.3, 2.6 Hz, 1 H), 3.53 (br. s, 1 H), 3.88 (s, 6 H),
4.09 (m, 1 H), 4.31 (m, 1 H), 5.17–5.39 (m, 2 H), 7.47–7.74 (m, 3
H), 7.93–7.99 (m, 2 H) ppm.
1-<(Z)-6-{(1R,2R,3R,5S)-2-[(Phenylsulfonyl)methyl]-3,5-bis(tri-
ethylsilyloxy)cyclopentyl}hex-4-enylϾ-4-methyl-2,6,7-trioxabicyclo-
[2.2.2]octane (21): (a) A solution of cyclopentanediol 20a (279 mg,
0.58 mmol) in anhydrous DMF (10 mL) was stirred under Ar at
0 °C. Imidazole (160 mg, 2.34 mmol) and Et3 N (300 µL,
2.15 mmol) were added, followed by chlorotriethylsilane (420 µL,
2.5 mmol). After 2.5 h, pyridine (0.5 mL) was added, followed by
saturated aqueous NaHCO3 (60 mL). The mixture was extracted
with EtOAc/hexanes (1:1, 3ϫ20 mL). The solutions were com-
bined, washed with saturated aqueous NaHCO3 (60 mL), dried
with anhydrous Na2SO4 (12 g), and concentrated in vacuo. The res-
idue was purified by flash chromatography on silica gel (60 g). Elu-
tion with 18% EtOAc/hexanes +0.007% pyridine, followed by vac-
uum drying (1 Torr, 25 °C, 3 h) gave compound 21 (361 mg, 87%
yield) as a glassy solid. [α]2D2 = +9.3 (c = 1.00, CHCl3); H NMR
1
(200 MHz, CDCl3): δ = 0.55 (q, J = 8.0 Hz, 12 H), 0.80 (s, 3 H),
0.92 (t, J = 8.0 Hz, 18 H), 1.39–1.72 (m, 6 H), 1.84 (m, 1 H), 1.98
(m, 1 H), 2.16 (m, 4 H), 3.19 (br. d, J = 5.3 Hz, CH2SO2Ph, 2 H),
3.89 (s, 6 H), 4.14 (m, 2 H), 5.33 (m, 2 H), 7.50–7.68 (m, 3 H),
7.89–7.95 (m, 2 H) ppm; 13C NMR (50 MHz, CDCl3): δ = 4.69 (3
C), 4.82 (3 C), 6.77 (3 C), 6.82 (3 C), 14.49, 23.15, 25.18, 26.94,
30.12, 36.16, 43.65, 46.73, 47.15, 57.40, 71.40, 72.47 (3 C), 74.74,
108.95, 127.95 (2 C), 128.54, 129.08 (2 C), 130.28, 133.38, 140.27 11.0, 6.1 Hz, 1 H), 3.471 (ddd, J = 7.2, 5.2, 2.0 Hz, 1 H), 3.898 (s,
ppm; C36H62O7SSi2 (695.1): calcd. C 62.20, H 8.99; found C 62.04,
H 8.68. (b) The mixture of 20b and 20c described above (558 mg,
0.961 mmol) was dissolved in DMF and treated with imidazole,
Et3N, and chlorotriethylsilane as described in section (a) above. Af-
ter workup, the crude product was stored at 4 °C overnight, after
which time it was subjected to chromatography, as described above,
giving compound 21 (67% yield); 1H NMR (200 MHz, CDCl3)
data was identical with the data cited above for 21.
6 H), 3.92 (m, 1 H), 4.168 (m, 1 H), 4.582 (m, 1 H), 5.402 (m, 2
H), 7.132 (br. d, J = 7.1 Hz, 2 H), 7.181 (br. t, J = 7.3 Hz, 1 H),
7.276 (br. t, J = 7.6 Hz, 2 H), 7.532 (br. t, J = 7.8 Hz, 2 H), 7.594
(ddd, J = 7.4, 2.0, 1.6 Hz, 1 H), 7.886 (bdd, J = 7.2, 1.5 Hz, 2 H)
ppm; 13C NMR (125 MHz, CDCl3 + 1% C5D5N): δ = 4.96 (3 C),
5.04 (3 C), 5.22 (3 C), 6.87 (6 C), 6.93 (3 C), 14.51, 23.21, 25.69,
27.24, 30.16, 31.11, 34.35, 36.30, 39.07, 44.14, 45.78, 50.44, 60.74,
69.04, 71.09, 72.52 (3 C), 73.03, 109.02, 125.67, 128.22 (2 C), 128.28
(2 C), 128.37 (2 C), 128.54, 128.91 (2 C), 130.20, 133.10, 140.13,
142.17 ppm] and the minor isomer, 1-{(Z)-6-[(1R,2R,3R,5S)-2-
[(1S,3S)-3-(triethylsilyloxy)-5-phenyl-1-(phenylsulfonyl)pentyl]-3,5-
bis(triethylsilyloxy)cyclopentyl]hex-4-enyl}-4-methyl-2,6,7-trioxabi-
cyclo[2.2.2]octane {[(13S)-22, 17.9 mg] as a colorless viscous oil;
1H NMR (500 MHz, CDCl3 + 1% C5D5N): δ = 0.580 (m, 18 H),
0.783 (s, 3 H), 0.950 (m, 17 H), 1.420–1.609 (m, 6 H), 1.677 (m, 3
H), 1.805 (quintet, J = 7.1 Hz, 1 H), 1.924–2.374 (m, 7 H), 2.475
(ddd, J = 13.8, 11.4, 4.8 Hz, 1 H), 3.240 (ddd, J = 7.0, 4.8, 1.9 Hz,
1 H), 3.882 (s, 6 H), 3.896 (m, 1 H), 4.183 (dd, J = 9.9, 5.1 Hz, 1
H), 4.484 (ddd, J = 12.6, 7.8, 5.9 Hz, 1 H), 5.265 (bdd, J = 5.0,
4.4 Hz, 2 H), 7.027 (dddd, J = 7.0, 2.2, 1.8, 1.4 Hz, 2 H), 7.161
(dd, J = 7.4, 1.4 Hz, 1 H), 7.239 (dddd, J = 7.4, 7.0, 2.2, 1.6 Hz, 2
H), 7.494 (dddd, J = 7.6, 7.1, 1.6, 1.2 Hz, 2 H), 7.550 (dd, J = 7.5,
1-{(Z)-6-[(1R,2R,3R,5S)-2-[(1R/1S,3S)-3-Triethylsilyloxy-5-phenyl-
1-(phenylsulfonyl)pentyl]-3,5-bis(triethylsilyloxy)cyclopentyl]hex-4-
enyl}-4-methyl-2,6,7-trioxabicyclo[2.2.2]octane (22) and 2,2-Bis-
(hydroxymethyl)propyl-(Z)-7-[(1R,2R,3R,5S)-2-[(1R/1S,3S)-3-tri-
ethylsilyloxy-5-phenyl-1-(phenylsulfonyl)pentyl]-3,5-bis(triethylsilyl-
oxy)cyclopentyl]hept-5-enoate (23): To a solution of sulfone 21
(2.29, 3.29 mmol) in anhydrous THF (15 mL), vigorously stirred
under Ar at –78 °C, lithium bis(trimethylsilyl)amide (LHMDS,
1.0  solution in THF, 14 mL, 14 mmol) was added over 3 min.
After 20 min, the mixture was warmed to 0 °C and stirred for
10 min. The mixture was then cooled to –78 °C, and a solution of
iodide 15 (4.96 g, 12.7 mmol, 99.2% ee) in anhydrous THF (4 mL)
was added dropwise. After 10 min, the temperature was raised to
0 °C, and the mixture was stirred at 0 °C for 80 min, at which time
700
www.eurjoc.org
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2007, 689–703

A New Synthetic Approach to High-Purity (15R)-Latanoprost
FULL PAPER
1.3 Hz, 1 H), 7.912 (dddd, J = 7.1, 2.0, 1.5, 1.2 Hz, 2 H) ppm; 13C
NMR (125 MHz, CDCl3 + 1% C5D5N): δ = 4.97 (3 C), 4.99 (3 C),
5.05 (3 C), 6.86 (3 C), 6.92 (3 C), 6.94 (3 C), 14.53, 23.17, 25.70,
26.18, 27.07, 29.65, 30.16, 30.37, 34.93, 36.29, 37.74, 43.57, 46.32,
51.51, 61.23, 69.49, 71.86, 72.53, 73.64, 109.05, 125.61, 128.13,
128.16 (2 C), 128.19 (2 C), 128.90 (2 C), 128.92 (2 C), 130.48,
133.31, 139.44, 142.10 ppm}.
ter 25 (120 mg, 0.243 mmol) in MeOH (5 mL), stirred under Ar,
H2O (0.5 mL) was added, followed by LiOH·H2O (240 mg,
5.72 mmol). After 8 h, the mixture was transferred to a separatory
funnel containing saturated aqueous NH4Cl (50 mL), aqueous
NaHSO4 (2 , 30 mL), and EtOAc (50 mL). The phases were sepa-
rated, aqueous NaHSO4 (2 , 20 mL) was added to the aqueous
phase, and the mixture was extracted with EtOAc (30 mL). The
EtOAc phases were combined, washed with a mixture of saturated
aqueous NH4Cl (20 mL) and aqueous NaHSO4 (2 , 10 mL) and
dried with anhydrous Na2SO4 (10 g). The drying agent was filtered
and washed with EtOAc (10 mL). The filtrates were combined, con-
centrated, and dried in vacuo (1 Torr, 25 °C, 3 h), affording acid 26
(94.1 mg, 99% yield) as a thick, pale yellow oil. [α]2D0 = +29.7 (c =
1.00, MeOH); 1H NMR (200 MHz, CDCl3): δ = 1.35 (m, 2 H),
1.45–1.86 (m, 10 H), 2.07–2.37 (m, 7 H), 2.71 (m, 2 H), 3.66 (m, 1
H), 3.94 (m, 1 H), 4.14 (m, 1 H), 5.42 (m, 2 H), 5.70 (br. s, 3 H),
7.21 (m, 5 H) ppm; 13C NMR (50 MHz, CDCl3): δ = 24.82, 26.50,
26.68, 29.23, 32.10, 33.49, 35.38, 38.86, 42.44, 51.62, 52.32, 71.33,
74.28, 78.42, 125.74, 128.35 (2 C), 128.41 (2 C), 129.44, 129.50,
142.21, 177.10 ppm; HRMS (ESI): calcd. for C23H34O5Na [M +
Na]+ 413.2298; found 413.2279.
1-{(Z)-6-[(1R,2R,3R,5S)-2-[(R)-3-(Triethylsilyloxy)-5-phenylpentyl]-
3,5-bis(triethylsilyloxy)cyclopentyl]hex-4-enyl}-4-methyl-2,6,7-tri-
oxabicyclo[2.2.2]octane (24): Sulfone 22 (116 mg, 0.121 mmol, 5:2
mixture of epimers) was dissolved in anhydrous MeOH (5 mL), and
NaHPO4 (150 mg, 1.06 mmol) was added. The mixture was stirred
under Ar at room temperature for 20 min, cooled to 0 °C, and 10 %
Na/Hg amalgam (360 mg, 1.5 mmol Na) was added. Stirring at
0 °C was continued for 1 h. The cooling bath was removed, and the
reaction mixture was warmed to 20 °C with continued stirring. Af-
ter another 40 min, the mixture was cooled to 0 °C and, with vigor-
ous stirring, saturated aqueous solution of NH4Cl (3.0 mL) was
added dropwise. The mixture was stirred for 10 min, H2O (3 mL)
was added, and stirring was continued for another 15 min. The
mixture was transferred to a separatory funnel containing saturated
aqueous NH4Cl (40 mL) and 1 % pyridine/EtOAc (40 mL). The
phases were separated (mercury was carefully collected). The aque-
ous phase was extracted with EtOAc (30 mL). The combined or-
ganic phases were dried with Na2SO4 (10 g), filtered, concentrated,
and dried in vacuo (1 Torr, 40 min, 25 °C), affording compound 24
(106 mg, 99% yield) as a glassy solid. [α]2D0 = +10 (c = 1.00, CHCl3
Isopropyl (Z)-7-{(1R,2R,3R,5S)-3,5-Dihydroxy-2-[(R)-3-hydroxy-5-
phenylpentyl]cyclopentyl}hept-5-enoate (Latanoprost, 1): Acid 26
(20.0 mg, 0.051 mmol) was dissolved in acetone (2.0 mL). To this
solution, stirred under Ar, DBU (100 µL, 102 mg, 0.67 mmol) was
added, followed after 3 min by 2-iodopropane (100 µL, 170 mg,
1.0 mmol). Stirring was continued at 20 °C for 14 h. Subsequently,
the mixture was concentrated to about 0.5 mL, 4% aqueous citric
acid (4 mL) and EtOAc (10 mL) were added, and the mixture was
transferred to a separatory funnel containing EtOAc (20 mL), brine
(30 mL), and 4% aqueous citric acid (10 mL). The phases were
separated, and the aqueous phase was re-extracted with EtOAc
(10 mL). The organic phases were combined and extracted with a
mixture of brine (20 mL) and saturated aqueous NaHCO3 (20 mL).
The organic phase thus obtained was dried with anhydrous Na2SO4
(7 g). The drying agent was filtered and washed with EtOAc
(5 mL). The organic phases were combined, concentrated in vacuo,
and dried in vacuo (1 Torr, 20 °C, 1 h), giving a colorless, thick oil,
which was chromatographed on LiChroprep silica gel (25–40 µm,
4.0 g). Elution with 20% hexanes/EtOAc gave 1 as an oil, showing
91% HPLC purity [ca. 9% (5E)-1]. This sample was further puri-
fied by preparative HPLC on silica gel to give latanoprost (1,
14.0 mg, 63.2 % yield) as a thick, colorless oil. HPLC purity
99.83%; [α]2D0 = +32.0 (c = 1.00, CHCl3); 1H NMR (500 MHz,
CDCl3) and 13C NMR (125 MHz, CDCl3) fully assigned spectra
of compound 1 are given in Table S1 in the Supporting Infor-
mation. HRMS (ESI): calcd. for C26H40O5Na [M + Na]+ 455.2768;
found 455.2756; C26H40O5 (432.6): calcd. C 72.19, H 9.32; found
C 71.96, H 9.27.
1
+ 0.1% Et3N); H NMR (200 MHz, CDCl3 + 0.1% C5H5N): δ =
0.60 (m, 18 H), 0.79 (s, 3 H), 0.96 (m, 27 H), 1.30–1.61 (m, 8 H),
1.65–1.82 (m, 5 H), 2.01–2.32 (m, 5 H), 2.68 (m, 2 H), 3.73 (m, 2
H), 3.88 (s, 6 H), 4.09 (bdd, J = 11.5, 5.8 Hz, 1 H), 5.39 (m, 2 H),
7.20 (m, 3 H), 7.28 (m, 2 H) ppm; 13C NMR (50 MHz, CDCl3 +
0.1% C5H5N): δ = 4.87 (3 C), 4.90 (3 C), 5.09 (3 C), 6.85 (3 C),
6.86 (3 C), 6.97 (3 C), 14.49, 23.22, 25.74, 27.00, 27.93, 30.11, 31.66,
34.35, 36.21, 39.08, 44.19, 48.15, 50.10, 71.72, 72.34, 72.47 (3 C),
76.23, 108.96, 125.51, 128.23 (2 C), 128.26 (2 C), 129.32, 129.71,
142.73 ppm; C46H84O6Si3 (817.4): calcd. C 67.59, H 10.36; found
C 67.62, H 10.40.
2,2-Bis(hydroxymethyl)propyl (Z)-7-{(1R,2R,3R,5S)-3,5-Dihydroxy-
2-[(R)-3-hydroxy-5-phenylpentyl]cyclopentyl}hept-5-enoate (25): To
a solution of orthoester 24 (378 mg, 0.462 mmol) in acetone
(11 mL), stirred under Ar at 18 °C, H2O (1.5 mL) was added, fol-
lowed by PPTS (210 mg, 0.836 mmol). Stirring was continued for
5 h, and then the mixture was concentrated to 1.5 mL, and EtOAc
(40 mL), brine (30 mL), and saturated aqueous NaHCO3 (20 mL)
were added. After extraction, the aqueous layer was washed with
EtOAc (2ϫ20 mL). The combined organic layers were washed with
brine (50 mL), dried with anhydrous Na2SO4 (7 g), filtered, and
concentrated in vacuo. The residue was subjected to flash
chromatography on silica gel (14 g). Elution with 5 % MeOH/
EtOAc gave ester 25 (217 mg, 95% yield) as a colorless glass. [α]2D0
Supporting Information (see also the footnote on the first page of
this article): Experimental details and characterization data for all
compounds reported in this manuscript not included in the Experi-
mental Section, experimental details for larger scale experiments,
1
= +29.7 (c = 1.8, CHCl3); H NMR (200 MHz, CDCl3): δ = 0.85
(s, 3 H), 1.36 (m, 2 H), 1.48–1.88 (m, 11 H), 2.02–2.39 (m, 6 H),
2.71 (m, 2 H), 3.42 (br. s, 4 H), 3.55 (s, 4 H), 3.63 (m, 1 H), 3.95
(m, 1 H), 4.12 (br. s, 2 H), 5.41 (m, 2 H), 7.18 (m, 3 H), 7.28 (m,
2 H) ppm; 13C NMR (50 MHz, CDCl3 + 1% C5D5N): δ = 16.88,
24.80, 26.52, 27.02, 29.61, 32.15, 33.54, 35.74, 39.05, 40.51, 42.51,
51.70, 52.63, 61.41, 66.55, 67.00, 71.15, 74.53, 78.53, 125.78, 128.39
(2 C), 128.42 (2 C), 129.21, 129.77, 142.23, 174.56 ppm; HRMS
(ESI): calcd. for C28H44O7Na [M + Na]+ 515.2979; found 515.2962.
1
and tables with H and 13C NMR assignments for latanoprost (1),
(15S)-1, and (5E)-1.
Acknowledgments
The authors wish to express their sincere gratitude to Professor
Martin J. O’Donnell and Professor Robert E. Minto, both of the
Department of Chemistry and Chemical Biology, IUPUI, for the
very helpful discussions and suggestions regarding this manuscript.
(Z)-7-{(1R,2R,3R,5S)-3,5-Dihydroxy-2-[(R)-3-hydroxy-5-
phenylpentyl]cyclopentyl}hept-5-enoic Acid (26): To a solution of es-
Eur. J. Org. Chem. 2007, 689–703
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
701

J. G. Martynow, O. Achmatowicz et al.
FULL PAPER
19900331, Priority: PL 1984-259545 19841207; Chem. Abstr.
1991, 115, 28974.
a) T. K. Schaaf, E. J. Corey, J. Org. Chem. 1972, 37, 2921; b)
J. S. Amburgey, J. A. Wos, M. A. Delong, B. De, G. Haiyan,
D. L. Soper, WO 9912899 A1 19990318, Priority: US 97-58253
19970909; Chem. Abstr. 1999, 130, 252191.
a) P. J. Kocienski, A. Bell, P. R. Blakemore, Synlett 2000, 365,
and references cited therein; b) A. Kutner, J. Martynow, M.
Chodynski, W. Szelejewski, H. Fitak, M. Krupa, WO
2003087048 A2 20031023, Priority: PL 2002-353328 20020411,
PL 2002-353832 20020512; Chem. Abstr. 2003, 139, 323699.
a) E. J. Corey, N. Raju, Tetrahedron Lett. 1983, 24, 5571; b) P. J.
Kocienski, Protecting Groups, Georg Thieme Verlag, Stuttgart,
1994, chapter IV; c) T. W. Green, P. G. Wuts, Protective Groups
in Organic Synthesis, John Wiley & Sons, Inc., New York, NY,
1999, chapter V; d) U. Pindur, J. Mueller, C. Flo, H. Witzell,
Chem. Soc. Rev. 1987, 16, 75.
a) O. Mitsunobu, Synthesis 1981, 1; b) D. J. Cundy, P. A. Gurr,
Org. Prep. Proced. Int. 2000, 32, 461; c) For the alternative
PhSSPh/Bu3P/C5H5N conditions see: H. Miyaoka, T. Baba, H.
Mitome, Y. Yamada, Tetrahedron Lett. 2001, 42, 9233.
a) K. R. Guertin, A. S. Kende, Tetrahedron Lett. 1993, 34,
5369; MMPP hexahydrate: b) M.-Y. Chen, L. N. Patkar, C.-C.
Lin, J. Org. Chem. 2004, 69, 2884 and references cited therein;
c) M. Therien, J. Y. Gauthier, Y. Leblanc, S. Leger, H. Perrier,
P. Prasit, Z. Wang, Synthesis 2001, 1778.
[1]
a) T. Nogrady, Medicinal Chemistry. A Biochemical Approach,
Oxford University Press, Inc., New York, NY, 1985, chapter 5;
b) Advances in Prostaglandins, Thromboxane, and Leukotriene
Research (Eds.: B. Samuelsson, R. Paoletti), Raven: New York,
1983–1989, vol. 11–19; c) P. W. Collins, S. W. Djuric, Chem.
Rev. 1993, 93, 1533 and references cited therein.
a) E. J.Corey, X.-M. Cheng, The Logic of Chemical Synthesis,
John Wiley & Sons, Inc. New York, NY, 1989, chapter XI; b)
E. J. Corey, Angew. Chem. Int. Ed. Engl. 1991, 30, 455; c) J.-
H.Fuhrhop, G. Li, Organic Synthesis – Concepts and Methods,
Wiley-VCH, Weinheim, 2003, chapter II; d) R. Noyori, Asym-
metric Catalysis in Organic Chemistry, John Wiley & Sons, Inc.,
New York, NY, 1994, chapter VI; e) S. Okamoto, Y. Kobaya-
shi, H. Kato, K. Hori, T. Takahashi, J. Tsuji, F. Sato, J. Org.
Chem. 1988, 53, 5590; f) C. J. Sih, R. G. Salomon, P. Price, R.
Sood, G. Peruzzotti, J. Am. Chem. Soc. 1975, 97, 857; g) U. F.
Axen, US 3901923 19750826 Chem. Abstr. 1975, 83, 192668.
[9]
[10]
[2]
[11]
[12]
[13]
[3]
[4]
D. G. Hall, S. Manku, F. Wang, J. Comb. Chem. 2001, 3, 125.
a) B. Resul, J. Stjernschantz, K. No, C. Liljebris, G. Selen, M.
Astin, M. Karlsson, L. Z. Bito, J. Med. Chem. 1993, 36, 243;
b) B. Resul, J. Stjernschantz, K. No, C. Liljebris, G. Selen, M.
Astin, M. Karlsson, L. Z. Bito, J. Med. Chem. 1993, 36, 2242;
c) J. W. Stjernschantz, B. Resul, WO 9002553 A1 19900322,
Priority: SE 88-3110 19880906, SE 88-3855 19881028. Chem.
Abstr. 1990, 113, 205515.
a) C. B. Toris, C. B. Camras, M. E. Yablonski, Ophthalmology
1993, 100, 1297; b) L. D. Goldberg, J. Walt, PharmacoEconom-
ics 2006, 24, 251; c) S. Holmstrom, P. Buchholz, J. Walt, J.
Wicxtrom, M. Aagren, Curr. Med. Res. Op. 2005, 21, 1875; d)
M. F. Sugrue, J. Med. Chem. 1997, 40, 2793.
a) A. K. Greenwood, D. McHattie, D. G. Thompson, D. W.
Clissold, WO 2002096898 A2 20021205, Priority: GB 2001-
12699 20010524; Chem. Abstr. 2002, 138, 24587; b) A. Gutman,
G. Nisnevich, M. Etinger, I. Zaltzman, L. Yudovich, B. Pertsi-
kov, B. Tishin, US 2005209337 A1 20050922; Chem. Abstr.
2005, 143, 326127; c) A. L. Gutman, G. Nisnevich, M. Etinger,
I. Zaltzman, L. Yudovitch, B. Pertsikov, WO 2002096868 A2
20021205, Priority: IL 2001-143477 20010531; Chem. Abstr.
2002, 138, 13979; d) A. Gutman, G. Nisnevich, M. Etinger,
I. Zaltzman, L. Yudovich, B. Pertsikov, WO 2001055101 A2
20010802, Priority: IL 2000-134241 20000127; Chem. Abstr.
2001, 135, 152659; e) K. E. Henegar, WO 2001087816 A1
20011122, Priority: US 2000–204242 20000515; Chem. Abstr.
2001, 136, 5851; f) K. E. Henegar, WO 2003008368 A1
20030130, Priority: US 2001-306026 20010717; Chem. Abstr.
2003, 138, 137085; g) K. E. Henegar, US 2003187071 A1
20031002, Priority: US 2001–306026 20010717; Chem. Abstr.
2003, 139, 292090; h) J. Ivanics, T. Szabo, I. Hermecz, G. Dal-
madi, J. Ivanics, G. Kovacs, B. Resul, WO 9300329 A1
19930107, Priority: HU 91–2092 19910624; Chem. Abstr. 1993,
118, 254622; i) B. Resul, WO 9202496 A1 19920220, Priority:
SE 90-2596 19900808; Chem. Abstr. 1992, 116, 235335; j) D. W.
Clissold, S. W. Craig, R. R. Gadikota, M. He, J. F. Jurayj, S.
Kazerani, E. Rannala, P. K. Sharma, WO 2005058812 A2
20050630, Priority: GB 2003–29379 20031219; Chem. Abstr.
2005, 143, 97203; k) I. Obadalova, T. Pilarcik, M. Slavikova,
J. Hajicek, Chirality 2005, 17, S109–S113 and literature cited
therein.
[14]
[15]
Y. Matsumura, T. Nakano, M. Makino, Y. Morizawa, EP
669329 A1 19950830; Chem. Abstr. 1996, 124, 8498.
[5]
[6]
a) A. Maeycker, Organic Reactions, John Wiley & Sons, New
York, NY, 1965, vol. 14, p. 270; b) J. A. Secrist III, S. R. Wu,
J. Org. Chem. 1979, 44, 1434.
G. H. Verdoorn, C. W. Holzapfel, J. M. S. Koekemoer, Afr. J.
Chem. 1987, 40, 134
[16]
[17]
a) B. B. Lohray, Synthesis 1992, 1035; b) H.-S. Byun, L. He,
R. Bittman, Tetrahedron 2000, 56, 7051; c) Sulfite 17: H. Yosh-
ida, A. Takada, O. Mitsunobu, Tetrahedron Lett. 1998, 39,
3007; d) Sulfate 18: H. Zhao, Y. L. Wu, Chin. Chem. Lett. 1994,
5, 367; e) epi-16: C. P. Baird, P. C. Taylor, J. Chem. Soc., Chem.
Commun. 1995, 893; f) E. J. Corey, C. J. Helal, Tetrahedron
Lett. 1993, 34, 5227; g) Epoxides in alkylation of sulfones: B.
Achmatowicz, S. Marczak, J. Wicha, J. Chem. Soc., Chem.
Commun. 1987, 1226.
a) H. C. Kolb, M. S. VanNieuwenhze, K. B. Sharpless, Chem.
Rev. 1994, 94, 2583, and references cited therein; b) H. Becker,
K. B. Sharpless, Angew. Chem. Int. Ed. Engl. 1996, 35, 448; c)
T. J. Hodgkinson, M. Shipman, Synthesis 1998, 1141; d) Z.-M.
Wang, X.-L. Zhang, K. B. Sharpless, Tetrahedron Lett. 1993,
34, 2267; e) J. Hasegawa, M. Ogura, S. Tsuda, S. Meamoto, H.
Kutsuki, T. Ohashi, Agric. Biol. Chem. 1990, 54, 1819; f) T.
Tsujigami, T. Sugai, H. Ohta, Tetrahedron: Asymmetry 2001,
12, 2543; g) B. P. Branchaud, H. S. Blanchette, Tetrahedron
Lett. 2002, 43, 351; h) T. Ishida, R. Akiyama, S. Kobayashi,
Adv. Synth. Catal. 2003, 345, 576; i) M. Rezaei, T. M. Harris,
C. J. Rizzo, Tetrahedron Lett. 2003, 44, 7513.
(R)-2,3-O-Isopropylideneglyceraldehyde: a) C. R. Schmid, J. D.
Bryant, M. J. McKennon, A. I. Meyers, Org. Synth., Coll. Vol.
1998, 9, 450, and references cited therein; b) B. C. Saha, US
2003215931 A1 20031120, Priority: US 2002–146616 A
20020515; Chem. Abstr. 2003, 139, 380110; c) H. Azuma, R.
Takao, H. Niiro, K. Shikata, S. Tamagaki, T. Tachibana, K.
Ogino, J. Org. Chem. 2003, 68, 2790; d) A recent application
of (S)-2,3-O-isopropylideneglyceraldehyde for chain elong-
ation: P. J. L. M. Quaedflieg, B. R. R. Kesteleyn, P. B. T. P.
Wigerinck, N. M. F. Goyvaerts, R. J. Vijn, C. S. M. Liebregts,
J. H. M. H. Kooistra, S. Cusan, Org. Lett. 2005, 7, 5917.
[18]
[19]
[7]
[8]
a) J. Hutton, M. Senior, N. C. A. Wright, Synth. Commun.
1979, 9, 799; b) J. Hutton, Synth. Commun. 1979, 9, 483; c) M.
Node, K. Nishide, Y. Shigeta, H. Shiraki, K. Obata, J. Am.
Chem. Soc. 2000, 122, 1927; d) R. Noyori, I. Tomino, M. Yam-
ada, M. Nishizawa, J. Am. Chem. Soc. 1984, 106, 6717.
a) E. J. Corey, H. Shirahama, H. Yamamoto, S. Terashima, A.
Venkateswarlu, T. K. Schaaf, J. Am. Chem. Soc. 1971, 93, 1490;
b) B. Achmatowicz, E. Baranowska, A. R. Daniewski, J. Pan-
kowski, J. Wicha, Tetrahedron 1988, 44, 4989; c) B. Achmatow-
icz, A. R. Daniewski, J. Pankowski, J. Wicha, PL 149389 B1
[20]
a) G.-J. Boons, G. H. Castle, A. Clase, P. Grice, S. V. Ley, C.
Pinel, Synlett 1993, 913; b) M. J. Martinelli, E. D. Moher, WO
9809942 A1 19980312, Priority: US 96-25523 19960906. Chem.
Abstr. 1998, 128, 230129.
702
www.eurjoc.org
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2007, 689–703

A New Synthetic Approach to High-Purity (15R)-Latanoprost
FULL PAPER
[21] (3S)-Hydroxytosylate: a) J. M. Klunder, T. Onami, K. B.
Sharpless, J. Org. Chem. 1989, 54, 1295.
Calculations were run on a Silicon Graphics Octane work sta-
tion.
[22] H. Yamamoto, in: Organometallics in Synthesis, John Wiley & [26] a) B. M. Trost, H. C. Arndt, P. E. Strege, T. R. Verhoeven, Tet-
Sons, Inc., New York, NY, 1994, chapter VII.
rahedron Lett. 1976, 17, 3477; b) Y. Liu, Y. Zhang, Org. Prep.
Proced. Int. 2001, 33, 372, and literature cited therein.
[27] Portions of this work have been disclosed as Polish patent ap-
plication P 374461 (2005.04.18).
[23] a) B. M. Trost, Bull. Chem. Soc. Jpn. 1988, 61, 107; b) N. S.
Simpkins, Tetrahedron 1990, 46, 6951; c) I. R. Baldwin, R. J.
Whitby, Chem. Commun. 2003, 2786.
[24] J. Mulzer, U. Kaselow, K.-D. Graske, H. Kühne, A. Sieg, H. J.
Martin, Tetrahedron 2004, 60, 9599.
[25] MacroModel® Interactive Molecular Modeling System. Ver-
sion 7.0. MacroModel User Manual, 1999 Schrödinger, Inc.
[28] Pharma Tech International, Inc., 21 Just Road, Fairfield, NJ
07004, USA.
Received: August 24, 2006
Published Online: November 23, 2006
Eur. J. Org. Chem. 2007, 689–703
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
703