The discovery of potent and efficacious intraocular pressure (IOP) lowering PGF2α analogs 1–5 (Figure 1) has revolutionized the treatment of ocular hypertension, a major risk factor for progression of the irreversible blinding disease glaucoma [1,2]. Until recently, free isopropyl ester prodrugs, i.e., unoprostone (1, Rescula), latanoprost (2, Xalatan) and travoprost (3, Travatan), and one amide prodrug bimatoprost (4, Lumigan) of ω-chain modified PGF2α analogs have been used as first line therapy for the treatment of open angle glaucoma and ocular hypertension due to their potent IOP-lowering efficacy, low likelihood of systemic adverse effects, once-daily dosing and good patient adherence [3,4,5,6]. However, all the available commercial preparations of 1–4 contain preservatives to maintain the sterility of the solutions, which may impair topical tolerance of these IOP-lowering agents during long-term use . New preservative-free agents with greater IOP-lowering efficacy and milder local side effects are therefore still needed.
Tafluprost (5) is the newest 15-deoxy-15,15-difluorinated PGF2α receptor agonist possessing effective IOP-reducing effects, and also the first hypotensive drug released in a preservative-free formulation under trade names Taflotan or Saflutan in Europe and Zioptan in USA [8,9,10,11,12]. A preservative-free tafluprost formulation is as potent as a preserved one (BAK-Tapros, Osaka, Japan), but it has fewer and milder toxic effects on the eye. Therefore, it may be particularly beneficial for patients who are sensitive to preservatives, such as patients with dry eyes, and patients who discontinue medication early because of adverse events. Compared with other PGF2α analogs 2–4, the major modification of tafluprost is the substitution of the C-15 hydrogen and hydroxyl group with two fluorine atoms. Tafluprost is a prodrug containing an isopropyl ester moiety that is rapidly hydrolyzed by corneal esterases to the pharmacologically active metabolite tafluprost free acid (AFP-172, 6) . Tafluprost acid has 12 times the affinity for the prostaglandin F receptor as does latanoprost acid, while having almost no affinity for related receptors, such as prostaglandin D or E .
PGF2α analogs 1–5 are mostly synthesized by one of the known variants of the Corey method, in which lower and upper side chains are sequentially attached in a specific order to a derivative of the commercially available (−)-Corey aldehyde/lactone [9,15,16,17,18,19,20,21,22,23,24,25,26]. In 2004, Matsumura et al.  published the first synthesis of tafluprost (5) from Corey aldehyde derivative 7 comprising a sequence of the following reactions: (a) the installation of the lower side chain (ω-chain) via a Horner–Wadsworth–Emmons condensation of the Corey aldehyde 7 with dimethyl (2-oxo-3-phenoxypropyl)phosphonate; (b) difluorination of the ω-chain 15-one 8 with morpholinosulfur trifluoride to a 15-deoxy-15,15-difluoro derivative 9; (c) removal of benzoyl group with potassium carbonate; (d) reduction of the lactone moiety to a lactol 11 with diisobutylaluminum hydride; (e) addition of the α-chain by Wittig olefination of 11 with an ylide prepared from 4-carboxybutyltriphenylphpsphonium bromide, leading to a cis-5,6-alkene; and (f) esterification of the acid 6 with isopropyl iodide in the presence of 1,8-diazabicyclo[5.4.0]undec-7-ene (Scheme 1). Thus far, however, there was no convergent methodology that would allow effective and highly diastereoselective preparation of a whole series of antiglaucoma PGF2α analogs 2–5 from a common and structurally advanced prostaglandin intermediate.
In 2007, Martynow et al.  reported preparation of the structurally advanced prostaglandin phenylsulfone (5Z)-16 (Scheme 2) with the attached α-chain possessing a carboxyl group protected by 4-methyl-2,6,7-trioxabicyclo[2.2.2]octane (methyl-OBO). The choice of the methyl-OBO carboxyl-masking group was predicted by its high stability under basic reaction conditions commonly applied during prostaglandin synthesis. On the other side, the compounds of the 4-methyl-2,6,7-trioxabicyclo[2.2.2]octane structure are easily hydrolyzed under acidic conditions to corresponding 2,2-bis(hydroxymethyl)-1-propyl esters, which can be in turn converted into other alkyl esters, like for example alkyl esters, salts or carboxylic acids. Thus, the ω-chain elongation of the phenylsulfone 16 by SN2 alkylation with enantiomerically pure alkyl halides allowed the synthesis of 13,14-dihydro-15-ol PGF2α analogs, such as latanoprost (2) . The Julia–Lythgoe olefination of the sulfone 16 with enantiomerically pure aldehyde ω-chain synthons allowed the synthesis of trans-13,14-en-15-ol PGF2α analogs, such as travoprost (3) and bimatoprost (4) [28,29]. One of the main reasons for choosing this method of phenylsulfone 16 ω-chain elongation was formation of trans-13,14-ene as the only product of olefination. Compared to triphenylphosphonium or phosphonate derivatives, used in competitive Wittig or Horner-Wadsworth-Emmons olefinations, phenylsulfones are mostly stable and crystal compounds that could be easily purified to pharmaceutical grade at early stage of prostaglandin synthesis. Recently, we envisaged that the prostaglandin key intermediate 16 could also find the application in the convergent synthesis of tafluprost (5), depending on the structure of the aldehyde ω-chain synthon and the sequence of chemical reactions needed for the ω-chain elongation and functionalization (Scheme 2). The use of the same structurally advanced prostaglandin phenylsulfone 16, as a starting material in parallel syntheses of all commercially available antiglaucoma PGF2α analogs 2–5, would significantly reduce manufacturing costs resulting from its synthesis on an industrial scale and development of technological documentation. Since the orthoester 16 is stable at about 0 °C in the presence of a trace of pyridine, large amounts of 16 can be manufactured and stored in a pure form for a long time allowing the synthesis of the desired PGF2α analog 2–5 at any time.
Glaucoma is one of the most common causes of irreversible visual impairment and blindness. It is estimated that over 60 million individuals were afflicted with open-angle and angle-closure glaucoma as of 2010, which will increase to almost 80 million by 2020 . The disease affects all ethnicities, and 6.7 million people are bilaterally blind consequently. As a result of a growing demand for prostaglandin antiglaucoma drugs in the pharmaceutical market, we have developed a novel synthesis of tafluprost (5) from the advanced prostaglandin phenylsulfone 16 and α-hydroxy protected aldehyde 17 (Scheme 2) based on the reversed order of side chain attachment to the Corey lactone via Wittig and Julia–Lythgoe olefinations. Since the fourth commercially available PGF2α drug tafluprost (5) contains the trans-13,14-en-15-deoxy-15,15-difluoro moiety in the ω-chain, when compared to latanoprost (2) travoprost (3) and bimatoprost (4), some modifications in an earlier developed approach for the highly diastereoselective construction of the ω-chain allylic moiety to incorporate the desired C-15 difluorinated center needed to be developed.
Most common PGF2α analogs, latanoprost (2), travoprost (3), bimatoprost (4) and tafluprost (5), are licensed for the reduction of elevated intraocular pressure in patients with open angle glaucoma and ocular hypertension, but their non approved use as eyelash enhancers is becoming popular, especially in patients with eyelashes hypotrichosis . There are other prostaglandin derivatives, such as tafluprost methyl ester (32) [32,33] and tafluprost ethyl amide (33) , that could be used in antiglaucoma ophthalmic compositions or cosmetics because they appear capable of IOP reduction and influencing eyelash growth; however, their efficacy and safety has not been fully studied and evaluated yet. Therefore, there is also a strong demand for a convenient process that is suitable for preparation of high purity tafluprost methyl ester (32) and tafluprost ethyl amide (33) on a commercial scale.
2. Results and Discussion
Based on our previous work on ω-chain elongation of the prostaglandin phenylsulfone 16 with the aldehyde ω-chain synthons of travoprost and bimatoprost [28,29], we envisaged that the Julia–Lythgoe olefination [35,36,37] of the sulfone 16 with the α-hydroxy protected aldehyde 17 should give a mixture of β-hydroxysulfones 18 (Scheme 2), which via reductive dehydroxy-desulfonylation and deprotection/protection of methyl-OBO and hydroxyl groups followed by oxidation of the 15-ol and deoxydifluorination of 13,14-en-15-one could enable construction of the ω-chain trans-13,14-en-15-deoxy-15,15-difluoro moiety. Irrespective of stereochemical configuration of the C-2 carbon atom in the aldehyde 17, corresponding to the C-15 carbon atom in tafluprost (5), the deprotection of tert-butyldimethylsilyl protected hydroxyl group and its oxidation to a 15-one followed by its deoxydifluorination should lead to the non-chiral C-15 gem-difluorinated carbon atom in the final prostaglandin analog 5. Therefore, for economic reasons, the 2-(tert-butyl-dimethylsilyloxy)-3-phenoxypropanal (17) used for the construction of the ω-chain of tafluprost (5) should be prepared as a racemate.
2.1. Synthesis of the Aldehyde ω-Chain Synthon 17
To the best of our knowledge, the aldehyde ω-chain synthon 17 has never been obtained in a racemic form. In 2010, Lin et al. reported the preparation of optically active (R)-17 based on epoxide ring-opening of (R)-benzyl glicydyl ether with phenol in the presence of organic base . The protection of the resulting secondary alcohol with tert-butyldimethylsilyl trifluoromethanesulfonate followed by removing of benzyl group and Dess-Martin oxidation of the (S)-25 gave the optically active aldehyde (R)-17. Although the route described above could be applied to supply of the racemic aldehyde 17, the cost of benzyl glicydyl ether and the use of heavy metal catalyzed hydrogenation process limit its large-scale application in the pharmaceutical industry. Therefore, a more industrially applicable procedure for synthesis of racemic 2-(tert-butyldimethylsilyloxy)-3-phenoxypropanal (17) (Scheme 3) was developed based on a synthetic route applied earlier for the preparation of the aldehyde ω-chain synthon of travoprost .
2.2. Synthesis of Tafluprost (5), Tafluprost Methyl Ester (32) and Tafluprost Ethyl Amide (33)
The efficient ω-chain elongation of phenylsulfone 16 with sterically hindered aldehyde 17 to the mixture of diastereoisomeric β-hydroxysulfones 18 was achieved by using freshly prepared LDA (Scheme 4). Since significant losses of 18 were observed during column chromatography due to limited stability of the methyl-OBO protecting group on silica gel, the crude mixture of β-hydroxysulfones 18 was subjected to reductive elimination with a fresh 20% sodium amalgam to afford the methyl-OBO protected prostaglandin precursor 26. It is well documented that sodium amalgam reduction of aliphatic α-hydroxysulfones furnishes olefins having the trans configuration of the ω-chain 13,14-double bond . The phosphate-buffered conditions of Trost  were used to ensure the optimal pH conditions and prevent deprotection of the cyclopentane hydroxyl groups. Similarly to the methyl-OBO protected β-hydroxysulfones 18, the methyl-OBO protected prostaglandin intermediate 26 was not purified by means of silica gel column chromatography. Significantly, there was no need for purification of the prostaglandin intermediates 18 and 26, which was a major advantage for any future scale-up. The triethylsilyl and methyl-OBO masking groups in prostaglandin precursor 26 were all removed under the action of pyridinium p-toluenesulfonate (PPTS) in a mixture of CH2Cl2-MeOH, immediately followed by esterification of free hydroxyl groups with acetic anhydride to the diastereoisomeric 2,2-bis(acetoxymethyl)propyl ester 28a/b. The compound 28a/b was purified by silica gel flash chromatography; however, only a sample of one of the diastereoisomers was isolated and identified by spectroscopic methods. The unique stereochemical features of the Julia–Lythgoe olefination allowed preserving the relative cis/trans stereochemistry of side chains in the isolated prostaglandin intermediate. Additionally, the appropriately functionalized aldehyde ω-chain synthon 17 with bulky tert-butyldimethylsilyl group at C-2, in agreement with literature data for olefination with aldehydes with steric encumbrance , afforded trans-13,14-alkenes 28a/b as the only products of Julia–Lythgoe olefination. The tert-butyldimethylsilyl protecting group was removed with camphor-10-sulfonic acid (CSA) to afford a mixture of diastereoisomeric 15-OH prostaglandin derivatives 29a/b. Similarly to the ester 28a/b, silica gel flash chromatography allowed isolation and identification by spectroscopic methods of only one C-15 epi-isomer from the mixture of diastereoisomeric alcohols 29a/b. Most importantly, since the chirality of C-15 carbon atom disappears during tafluprost (5) final synthesis, there was no need for purification and separation of the diastereoisomeric prostaglandin intermediates 27a/b, 28a/b and 29a/b, which additionally was a major advantage for any future scale-up. A mixture of diastereiosmeric alcohols 29a/b was then oxidized with Dess-Martin periodinane  to a 15-keto intermediate 30. The use of Dess-Martin periodinane avoided some difficulties encountered with other methods, e.g., oxidation of 13,14-en-15-ols with toxic and unstable DDQ , such as adverse products, long reaction times, difficult workup procedures or the need to apply a large excess of the oxidizing agent. Additionally, the product of the oxidation route could be easily purified with short column chromatography. The structure of 30 was confirmed by spectroscopic methods.
Although a gem-difluoride unit has been drawn more attention recently in medicinal chemistry, there is no general method to prepare allyl difluorides from the corresponding 13,14-en-15-ones efficiently [42,43]. Therefore, we studied the reaction using some commercially available fluorinating agents, such as (diethylamino)sulfur trifluoride (DAST), (diethylamino)-difluorosulfonium tetrafluoroborate (XtalFluor-E), difluoro(morpholino)sulfonium tetrafluoro-borate (XtalFluor-M), bis(2-methoxyethyl)aminosulfur trifluoride (Deoxo-Fluor) and 4-tert-butyl-2,6-dimethylphenylsulfur trifluoride (Fluolead). After exploring various possibilities, Deoxo-Fluor was found to be the best reagent of choice. Thus, the fluorination reaction of 13,14-en-15-one 30 with Deoxo-Fluor in boiling CH2Cl2 for 24 h afforded geminal difluoride 31 with 78% yield.
The deprotection of acetyl and 2,2-bis(acetoxymethyl)propyl groups in 31 by means of potassium carbonate in MeOH yielded tafluprost methyl ester (33) in 76% yield. The ester 31 was hydrolyzed with lithium hydroxide monohydrate and, after acidification with citric acid, tafluprost acid (6) was isolated in almost quantitative yield. In the final step of the synthesis, acid 6 was successfully esterified with i-Pr/DBU in acetone to afford tafluprost (5). The crude 5 was purified by silica gel flash chromatography and identified by spectroscopic methods. The α-chain allylic carbons C-4 and C-7 of tafluprost are seen in the 13C-NMR spectrum as two singlets located at 25.7 and 26.6 ppm, which is characteristic for 5,6-cis diastereoisomers of PGF2α analogs . The fluorination of C-15 carbonyl group in 30 with Deoxo-Fluor leads to several significant effects in NMR spectra, which are important from the point of view of structure conformation/determination. The biggest change related to the replacement of oxygen with two fluorine atoms leads to strong shielding increase of C-15 nucleus by ca. 80 ppm and appearance of this signal in 13C-NMR spectrum as a triplet with relative large 1J(C-F) coupling of ca. 240 Hz. Similar but rather smaller effects are observed for neighboring protons and carbons. In the case of 1H-NMR spectrum, transition from 30 to 5 leads to shielding increase of all protons in ω-chain by ca. 0.3–0.5 ppm. The same is true for 13
|Preferred IUPAC name|
|Other names |
3D model (JSmol)
|Molar mass||102.13 g·mol−1|
|Melting point||−73 °C (−99 °F; 200 K)|
|Boiling point||89 °C (192 °F; 362 K)|
Solubility in water
|4.3 g/100 mL (27 °C)|
|Vapor pressure||42 mmHg (20 °C)|
Magnetic susceptibility (χ)
EU classification (DSD) (outdated)
|R-phrases(outdated)||R11, R36, R66, R67|
|S-phrases(outdated)||(S2), S16, S26, S29, S33|
|Flash point||2 °C (36 °F; 275 K)|
|460 °C (860 °F; 733 K)|
|Lethal dose or concentration (LD, LC):|
LC50 (median concentration)
|11,918 ppm (rat, 8 hr)|
|US health exposure limits (NIOSH):|
|TWA 250 ppm (950 mg/m3)|
IDLH (Immediate danger)
Except where otherwise noted, data are given for materials in their standard state (at 25 °C [77 °F], 100 kPa).
|N verify (what is YN ?)|
Isopropyl acetate is an ester, an organic compound which is the product of esterification of acetic acid and isopropanol. It is a clear, colorless liquid with a characteristic fruity odor.
Isopropyl acetate is a solvent with a wide variety of manufacturing uses that is miscible with most other organic solvents, and moderately soluble in water. It is used as a solvent for cellulose, plastics, oil and fats. It is a component of some printing inks and perfumes.
Isopropyl acetate decomposes slowly on contact with steel in the presence of air, producing acetic acid and isopropanol. It reacts violently with oxidizing materials and it attacks many plastics.
Isopropyl acetate is quite flammable in both its liquid and vapor forms, and it may be harmful if swallowed or inhaled.
The Occupational Safety and Health Administration has set a permissible exposure limit (PEL) of 250 ppm (950 mg/m3) over an eight-hour time-weighted average for workers handling isopropyl acetate.