Skip to content
Publicly Available Published by De Gruyter February 12, 2019

Phosphonate reagents and building blocks in the synthesis of bioactive compounds, natural products and medicines

  • Marian Mikołajczyk EMAIL logo

Abstract

This account outlines the results obtained in the author’s laboratory on the application of phosphonates in the synthesis of various classes of biologically active cyclopentenones and cyclopentanones. In the first place two general methods for the synthesis of mono-, 1,2- and 1,4-dicarbonyl compounds are presented. The first is based on the use of α-phosphoryl sulfides in conjunction with the Horner reaction while in the second method the oxygenation reaction of α-phosphonate carbanion is a key step. The utility of these two approaches to 1,4-diketones as precursors of cyclopentenones was exemplified by the synthesis of dihydrojasmone and (Z)-jasmone. The use of simple phosphonates, α-phosphoryl sulfides and β- and γ-ketophosphonates as starting reagents in the synthesis of cyclopentanoid antibiotics (methylenomycin B, racemic desepoxy-4,5-didehydromethylenomycin, enantiomeric sarkomycins) is presented. The synthesis and reactivity of achiral 3-(phosphorylmethyl)cyclopent-2-enone and chiral diastereoisomeric camphor protected 3-(phosphorylmethyl)-4,5-dihydroxycyclopent-2-enones as building blocks is discussed as a platform for developing a new access to a variety of bioactive cyclopentenones. The utility and value of achiral phosphonate building block is demonstrated by the synthesis of racemic and enantiopure prostaglandin B1 methyl esters and enantiomeric phytoprostanes B1 type I and II. The range of biologically active compounds prepared from chiral diastereoisomeric cyclopentenone phosphonates is wider. Herein the total syntheses of the following target compounds are presented: enantiomeric isoterreins, natural (−)-neplanocin A and its unnatural (+)-enantiomer, anticancer prostaglandin analogues (enantiomers of TEI-9826, NEPP-11, iso-NEPP-11). The design and synthesis of racemic and four enantiopure stereoisomers of an antiulcer drug rosaprostol is also described.

Introduction

Organic compounds containing heteroatoms like phosphorus, sulfur or silicon occupy a unique role in the arsenal of reagents widely applied in organic synthesis [1], [2], [3]. There is no doubt that the Wittig reaction of phosphorus ylides with carbonyl compounds as well as the Horner P(O)-olefination reaction using α-phosphonate carbanions and the Peterson reaction of α-silyl carbanions play the most important role in the synthesis of organic compounds containing carbon–carbon double bond. Our interest in organophosphorus and organosulfur chemistry led us to a broad study on the synthesis, properties and reactions of organic compounds containing both phosphorus and sulfur in one molecule [4]. Typical examples of such mixed P- and S-compounds obtained in our laboratory are α-phosphoryl sulfides and their oxidized analogues, α-trimethylsilyl substituted α-phosphoryl sulfides, α-(methylsulfenyl)phosphonoacetic acid and α-(methylsulfenyl)-δ-oxoalkanephosphonates. Their general structures are shown below (Scheme 1). In all these structures the α-hydrogen atom is relatively acidic and the corresponding α-phosphonate carbanions may be easily generated under basic conditions and reacted with carbonyl compounds and with a variety of other electrophiles [4].

Scheme 1: General structures of selected sulfenylated phosphonates.
Scheme 1:

General structures of selected sulfenylated phosphonates.

α-Methylsulfenyl substituted phosphonates are particularly useful reagents in reversal (Umpolung) of the normal reactivity of nucleophilic and electrophilic centres. For example, α-phosphoryl sulfides are starting materials in the two-step synthesis of aliphatic [5] and aromatic ketones [6] shown in Scheme 2.

Scheme 2: Synthesis of aliphatic and aromatic ketones.
Scheme 2:

Synthesis of aliphatic and aromatic ketones.

In terms of Umpolung concept [7] the synthesis of ketones shown above may be described as a nucleophilic acylation with elaboration of the electrophilic centre and the α-phosphoryl sulfides moiety may be regarded as a synthetic equivalent of an acyl anion. A similar reaction sequence with α-(methylsulfenyl)phosphonoacetic acid as a starting reagent afforded phenylpyruvic acid demonstrating thus a new approach to the synthesis of 1,2-dicarbonyl compounds (Scheme 3) [8]. Interestingly, the same bifunctional reagent was also converted in two simple steps (esterification, intramolecular Horner reaction) into unsaturated five- and six-membered cyclic esters (Scheme 3).

Scheme 3: Conversion of α-(methylsulfenyl)phosphonoacetic acid into pyruvic acid and cyclic unsaturated lactones.
Scheme 3:

Conversion of α-(methylsulfenyl)phosphonoacetic acid into pyruvic acid and cyclic unsaturated lactones.

As the α-phosphoryl sulfide moiety is a synthetic equivalent of a carbonyl group, the easily accessible diethyl 1-(methylsulfenyl)-4-oxopentanephosphonate (1) may be regarded as a masked form of the 1,4-dicarbonyl compounds which, in turn, may undergo intramolecular base-catalyzed cyclization to cyclopent-2-enones. In this context, it is important to note that the cyclopent-2-enone and cyclopentanone units are contained in many natural products such as jasmonoids, cyclopentanoid antibiotics, terreins, prostaglandins and their analogues, and some medicines. In spite of the fact that the structures of these compounds are deceptively simple, their synthesis is not trivial because of the low chemical stability and in many cases of specific functionalization of the five-membered ring. Therefore, based on our early basic studies on the synthesis and chemistry of phosphorus compounds containing sulfur, silicon and selenium we initiated a broad long-term project on the application of phosphorus and sulfur compounds in organic synthesis with a special emphasis on the invention and development of general methods for the synthesis of functionalized cyclopentenones and cyclopentanones.

The aim of the present account is to summarize the results obtained in the author’s laboratory in the frame of the program mentioned above and to show briefly new directions of its future development. For the sake of brevity of this short review, the discussion on contributions of other research groups to the synthesis of our target compounds will be limited to a minimum since it may be found together with appropriate references in our original papers.

α-Phosphoryl sulfides in the synthesis of 1,4-diketones and cyclopent-2-enones: synthesis of dihydrojasmone

A simple retrosynthetic analysis (Scheme 4) revealed two possible ways for the construction of a 1,4-dicarbonyl system from α-phosphoryl sulfides. In the first route A diethyl 1-(methylsulfenyl)-4-oxopentanephosphonate 1 is used as a starting reagent. Its Horner reaction with mono-carbonyl component should give the corresponding vinyl sulfide which after hydrolysis may be converted into the desired 1,4-diketone. The second, more interesting, route B starts with the Horner reaction of a simple α-phosphoryl sulfide and 1,3-dicarbonyl compound. The isomeric vinyl sulfide formed in this reaction should give the same 1,4-diketone. This route shows also a new strategy for the conversion of 1,3-dicarbonyl compounds into 1,4-dicarbonyl homologues [9].

Scheme 4: Synthetic strategy for the preparation of 1,4-dicarbonyl compounds from α-phosphoryl sulfides.
Scheme 4:

Synthetic strategy for the preparation of 1,4-dicarbonyl compounds from α-phosphoryl sulfides.

The value and utility of this novel strategy for the construction of 1,4-dicarbonyl compounds was exemplified by the total synthesis of dihydrojasmone 2 which is one of the components of the essential oil of Jasminum grandiflorum and is responsible for the characteristic jasmine fragrance [10]. The synthesis of 2 is shown in Scheme 5.

Scheme 5: Synthesis of dihydrojasmone 2.
Scheme 5:

Synthesis of dihydrojasmone 2.

The reaction of the lithium salt of the keto-protected 1 with n-hexanal gave unexpectedly very stable addition product which after isolation and treatment with sodium hydride in the presence of small amounts of crown ether, 18-c-6, gave the corresponding vinyl sulfide in high yield. Its acidic hydrolysis afforded undecane-2,5-dione. In an alternative synthesis of 2, the Horner reaction of α-phosphoryl sulfide 3 with protected 3-oxobutanal was also carried out stepwise and isomeric vinyl sulfide formed was converted into the same diketone. As expected, the unsymmetrical diketone prepared here underwent cyclization to dihydrojasmone 2 in a regioselective way i.e. giving the product derived from the attack of the more substituted α-carbonyl carbanion on a second carbonyl group. In summary, starting from 1 dihydrojasmone 2 was obtained in 66% overall yield. The synthesis of 2 from 3 was less efficient and afforded the desired product in 47% overall yield.

Oxygenation of α-phosphonate carbanions directed towards synthesis of 1,4-diketones and cyclopent-2-enones: synthesis of cis-jasmone

The oxidation of phosphorus ylides and phosphinoxide carbanions investigated by Bestmann and Horner and their coworkers [11], [12] was found to occur with oxidative cleavage of the phosphorus–carbon bond giving, however, instead of the expected carbonyl compounds symmetrical olefins. The latter were formed in a subsequent reaction (Wittig or Horner) of the primary produced carbonyl compounds with the starting phosphorus reagents.

The reaction of α-phosphonate carbanions with oxygen was shown by us [13] to be bidirectional in course and afforded the corresponding α-hydroxyphosphonates (direction a) and/or carbonyl compounds resulting from the oxidative P–C bond cleavage (direction b). The reaction course can be controlled by the reaction conditions and the formation of carbonyl compounds is preferred by the use of lithium as a counterion (see Scheme 6).

Scheme 6: Oxidation of α-phosphonate carbanions.
Scheme 6:

Oxidation of α-phosphonate carbanions.

Our results on the oxygenation of α-phosphonate carbanions [14] together with those reported at the same time by Zimmer et al. [15] and Vedejs et al. [16] confirmed its general character and synthetic utility. Interestingly, as in the case of α-phosphoryl sulfides, the α-phosphonate carbon atom may be treated as a synthetic equivalent of a carbonyl anion and used for synthesis of other carbonyl compounds (Scheme 7).

Scheme 7: Synthesis of ketones by oxidative cleavage of the phosphorus-carbon bond.
Scheme 7:

Synthesis of ketones by oxidative cleavage of the phosphorus-carbon bond.

Based on this property of the phosphonate moiety a new synthetic approach to 1,4-diketones and cyclopent-2-enones was developed. The four-step synthesis of (Z)-jasmone 4 shown below (Scheme 8) illustrates the usefulness of this method [17]. It started with the alkylation of the readily accessible ethylene ketal of diethyl 4-oxopentanephosphonate 5 with (Z)-hex-3-enyl iodide. The product of alkylation was oxygenated to give the corresponding half-protected 1,4-diketone which after deprotection and subsequent base-catalyzed cyclization afforded (Z)-jasmone 4 in an overall yield of 34%.

Scheme 8: Synthesis of (Z)-jasmone 4.
Scheme 8:

Synthesis of (Z)-jasmone 4.

According to the same reaction sequence a series of 2,3-disubstituted cyclopentenones have been obtained from 5, among them also dihydrojasmone 2 [14].

Scheme 9: Mono- and disubstituted cyclopentenones prepared from phosphonates 5 and 7.
Scheme 9:

Mono- and disubstituted cyclopentenones prepared from phosphonates 5 and 7.

The preparation of 2-(6-carboxyhexyl)cyclopent-2-enone 6 from the phosphonate 7 exemplifies the synthesis of monosubstituted cyclopentenones by this strategy (Scheme 9).

In summary, it should be emphasized that both approaches to 1,4-diketones and cyclopent-2-enones involving as a key step the Horner reaction of α-phosphoryl sulfides and oxygenation of α-phosphonate carbanions are complementary and compare favorably in terms of the brevity, use of simple reagents and efficiency with the methods described in the literature.

Synthesis of cyclopentanoid antibiotics using phosphonate reagents

As mentioned above, the cyclopent-2-enone and cyclopentanone units are important components of structures of a wide range of natural products exhibiting diverse biological activity. In the early stage of realization of our project a considerable attention of synthetic chemists was paid to cyclopentanoid antibiotics [18]. Among them a family of methylenomycins is representative of this class of compounds. All its members (methylenomycin A 8, desepoxy-4,5-didehydromethylenomycin A 9 and methylenomycin B 10) have been isolated from the culture broth of Streptomyces species and show antibacterial activity. A simplified structure-activity relationship of methylenomycins is depicted in Scheme 10.

Scheme 10: Biological activity of methylenomycin A and derivatives.
Scheme 10:

Biological activity of methylenomycin A and derivatives.

The most interesting observation from this bioactivity investigation was that the presence of the exocyclic methylene moiety is mainly responsible for antibacterial activity of methylenomycins. On the other hand, due to a high reactivity of this conjugated carbon–carbon double bond methylenomycins were found to be chemically unstable. Therefore, their synthesis was challenging not only for our group. Having in our hands two elaborated and tested strategies for the synthesis of cyclopentenones, we decided to synthesize at first methylenomycin B 10 – the simplest structure exhibiting antibacterial activity [10]. Scheme 11 shows the synthesis of 10 starting from α-phosphoryl sulfide 1-p and alternatively from α-(methylsulfenyl)propanephosphonate 11. The Horner olefination reaction of 1-p with acetaldehyde carried out stepwise, as in the case of dihydrojasmone 2, afforded n-heptane-2,5-dione. The same diketone was prepared from α-phosphoryl sulfide 11 and 3,3-ethylenedioxabutanal. The base-catalyzed cyclization of the latter gave 2,3-dimethylcyclopent-2-enone. In a final step, the exocyclic methylene group was introduced according to the procedure described earlier by Jernow et al. [19] and methylenomycin B 10 was obtained in 22% overall yield from 1-p and 16% overall yield from 11 (Scheme 12).

Scheme 11: Synthesis of methylenomycin B 10 from α-phosphoryl sulfides 1 and 11.
Scheme 11:

Synthesis of methylenomycin B 10 from α-phosphoryl sulfides 1 and 11.

In an attempt to improve efficiency of the synthesis of methylenomycin B 10, the phosphonate 5 was converted in four steps into 2,3-dimethylcyclopent-2-enone using oxygenation as a key reaction [14]. Then, the latter upon treatment with formaldehyde and subsequent water elimination gave 10 in an overall yield of 14%.

Scheme 12: Synthesis of methylenomycin B 10 from phosphonate 5.
Scheme 12:

Synthesis of methylenomycin B 10 from phosphonate 5.

Being disappointed with a low efficiency of the syntheses of 10 presented above our attention was directed to β-ketophosphonates as possible starting reagents in the preparation of methylenomycins. In this case, the corresponding α-phosphonate carbanions are very easily generated from β-ketophosphonates upon treatment even with very week bases, as for example potassium carbonate. Moreover, these carbanions are stabilized not only by the phosphoryl group but also by a β-carbonyl moiety.

A simple retrosynthetic analysis indicated that methylenomycin B 10 can be obtained from β-ketophosphonates 12 and 13. In the devised synthesis of 10 the construction of the 1,4-diketone skeleton (a precursor of the cyclopentenone ring) may be accomplished according to Umpolung concept through 1,4-addition of an acyl anion equivalent to vinyl ketone formed as intermediate at the early stage of the synthesis. Another characteristic feature of the planned synthesis is that the methylsulfenylmethylene substituent is introduced at the beginning of the synthesis and transformed into exo-methylene function at its last step. Scheme 13 shows the synthesis of methylenomycin B 10 carried out according to this strategy [20].

Scheme 13: Synthesis of methylenomycin B 10 from β-ketophosphonate 12.
Scheme 13:

Synthesis of methylenomycin B 10 from β-ketophosphonate 12.

Thus, the Horner reaction of 12 with (methylsulfenyl)acetaldehyde in the presence of potassium carbonate as a base gave the corresponding vinyl ketone to which the lithium salt of acetaldehyde S,S-diethyl dithioacetal S-monooxide as a synthetic equivalent of acetyl anion was added. The addition product formed was converted into the corresponding 1,4-diketone and then cyclized to 2,3-dimethyl-5-(methylsulfenylmethyl)cyclopent-2-enone. Its final conversion to methylenomycin B 10 was achieved stepwise but in a one-pot procedure by oxidation of the sulfide moiety to the sulfonyl group followed by elimination of methanesulfinate in the presence of sodium hydrogen carbonate. This four steps synthesis afforded methylenomycin B 10 in 26% overall yield.

With the β-ketophosphonate 13 as a starting reagent the same strategy required six steps and allowed to obtain methylenomycin B 10 in 13% overall yield (see Scheme 14) [21].

Scheme 14: Synthesis of methylenomycin B from β-ketophosphonate 13.
Scheme 14:

Synthesis of methylenomycin B from β-ketophosphonate 13.

However, using the same β-ketophosphonate 13 it was possible to elaborate much shorter (four steps) and, even at present, most effective (39% overall yield) synthesis of 10. It is shown in Scheme 15. Interestingly, in this approach to 10 the exo-methylene function was formed via the Horner olefination reaction under very mild conditions [22].

Scheme 15: Synthesis of methylenomycin B 10 from β-ketophosphonate 13.
Scheme 15:

Synthesis of methylenomycin B 10 from β-ketophosphonate 13.

Later on, Bałczewski [23] in his independent studies reported a free-radical version of the synthesis of 4-phosphorylheptan-2,5-dione – the precursor of 10 (see Scheme 16). The essence of this approach was the use of α-phosphonate carbon radical generated from 13 for the addition to isopropenyl acetate. Then, the addition product was oxidized to the desired diketone, however, in a low yield (15%).

Scheme 16: Synthesis of phosphorylated 1,4-diketone from α-substituted β-ketophosphonate 13.
Scheme 16:

Synthesis of phosphorylated 1,4-diketone from α-substituted β-ketophosphonate 13.

The synthesis of 10 was also completed with a comparable efficiency with the best one (Scheme 15) when γ-ketophosphonate 14 was used as a substrate [22]. The key step in this synthesis (Scheme 17) comprised the addition of n-propanal to the α-phosphonate carbanion derived from the protected 14. After oxidation of the hydroxy adduct the corresponding phosphorylated 1,4-diketone formed was converted in four already known steps into methylenomycin B 10 in 34% overall yield.

Scheme 17: Synthesis of methylenomycin B 10 from γ-ketophosphonate 14.
Scheme 17:

Synthesis of methylenomycin B 10 from γ-ketophosphonate 14.

In all our syntheses of functionalized and bioactive cyclopentenones discussed so far the cyclopentenone ring was constructed according to a general methodology elaborated in this laboratory including in the first place the preparation of suitably substituted 1,4-dicarbonyl compounds using structurally diverse phosphonate reagents followed by a ring closure. A completely different strategy was devised for the synthesis of other cyclopentanoid antibiotics that is desepoxy-4,5-didehydromethylenomycin A methyl ester 9 and sarkomycin. However, this conceptually new approach to cyclopentanoids was first examined using methylenomycin B 10 as a test compound. In this case the formation of 2,3-dimethyl-5-phosphorylcyclopent-2-enone was accomplished by intramolecular carbenoid cyclization of α-diazo-β-ketophosphonate [24]. The four-step synthesis of 10 from diethyl methanephosphonate 15 is shown in Scheme 18.

Similarly, the synthesis of methylenomycin methyl ester 9 starts from diethyl methanephosphonate 15 and involves in a crucial step the Nazarov cyclization of the properly substituted dienone [25]. It was obtained from the lithium-copper salt of the starting phosphonate 15 in three consecutive reactions: acylation with tiglic acid chloride, alkylation with methyl bromoacetate and bromination-dehydrobromination reaction. The Nazarov reaction was carried out in the presence of iron(III) chloride at low temperature (−30°C) and gave the corresponding cyclopentenone as the only reaction product which was converted into the desired target 9via the Horner reaction with gaseous formaldehyde. In this way racemic methylenomycin methyl ester 9 was obtained in 31% overall yield (Scheme 19).

Scheme 18: Synthesis of methylenomycin B 10via an intamolecular carbenoid cyclization.
Scheme 18:

Synthesis of methylenomycin B 10via an intamolecular carbenoid cyclization.

In continuation of our work on the synthesis of cyclopentanoid antibiotics, sarkomycin 16, belonging to this class of compounds, was selected as a next target. Sarkomycin 16 was first isolated by the group of Umezawa from the soil microorganisms which produce the levorotatory enantiomer having the absolute configuration R. It exhibits not only antibacterial and antiphage properties but also shows antitumor activity. In spite of chemical instability due to the presence of the exocyclic methylene group in the structure of 16 and some problems in its storage pharmacological studies led to marketing in Japan and USA of a preparation containing 16 as an antitumor drug. Based on the experience in the synthesis of methylenomycin B 10via an intramolecular carbenoid cyclization, a similar strategy was adapted for the synthesis of racemic sarkomycin 16 from diethyl 2-oxopropanephosphonate 12 as a starting reagent [26]. Scheme 20 shows how this strategy was put into practice.

Scheme 19: Synthesis of racemic methyl ester 9via the Nazarov reaction.
Scheme 19:

Synthesis of racemic methyl ester 9via the Nazarov reaction.

Thus, our synthesis started with dianion generated from 12 and its alkylation with homoallyl bromide. The corresponding β-ketophosphonate formed was transformed into α-diazophosphonate under typical diazo-transfer reaction conditions. Then, it was decomposed in the presence of rhodium(II) acetate to give the desired 2-phosphoryl-3-vinylcyclopentanone. The conversion of the vinyl moiety into carboxylic group was effected by ozonolysis and subsequent oxidation of the aldehyde formed with Jones reagent. In the last step, the Horner reaction of 2-phosphoryl-3-carboxycyclopentanone with gaseous formaldehyde afforded racemic sarkomycin 16 in 9% overall yield.

At the same time the second, shorter synthesis of (±)-16 was worked out according to a different strategy [27]. It started from the easily available 2-(phenylsulfenylmethyl)cyclopent-2-enone 18. Conjugated addition of nitromethane to 18 gave the corresponding addition product which was converted into cyclopentanone aldehyde under the Nef reaction conditions. Then, the Jones oxidation resulted in the formation of the transient cyclopentanone containing carboxylic acid and sulfone groups. The latter upon basic conditions eliminated benzenesulfinic acid and afforded racemic sarkomycin 16 in 13% overall yield (Scheme 21).

Scheme 20: Synthesis of racemic sarkomycin 16via intramolecular carbenoid cyclization.
Scheme 20:

Synthesis of racemic sarkomycin 16via intramolecular carbenoid cyclization.

Scheme 21: Synthesis of racemic sarkomycin 16 from cyclopentenone 18.
Scheme 21:

Synthesis of racemic sarkomycin 16 from cyclopentenone 18.

Having in our hands racemic sarkomycin 16 and being stimulated by the report of Helmchen and coworkers [28] on the synthesis of optically active sarkomycin methyl ester, the next our task was the synthesis of both enantiomers of sarkomycin [26]. As the cyclopentenone 17 (see Scheme 20) is a quite stable compound and contains carboxylic and carbonyl groups as possible resolving handles, its resolution was carried out in a classical way via separation of the diastereoisomeric products formed with (−)-(S)-1-(naphthyl)ethylamine. However, in contrast to our expectations, instead of diastereoisomeric salts two diastereoisomeric enamines were formed. They were easily separated by column chromatography and converted in a simple way into the enantiomerically pure (−)-(R)- and (+)-(S)-sarkomycin as shown in Scheme 22.

Scheme 22: Synthesis of enantiomeric sarkomycins 16.
Scheme 22:

Synthesis of enantiomeric sarkomycins 16.

In regard to determination of enantiomeric purity of the cyclopentanone 17, it is interesting to point out that the partially resolved samples of 17 show enantiomeric non-equivalence in 31P NMR spectra measured in benzene. This new example of chiral self-discrimination of enantiomers was rationalized by formation of short-lived diastereoisomeric homo- and heterodimers (Scheme 23) exhibiting different chemical shift and intensity.

Scheme 23: Hydrogen bonded dimeric structures of (+)- and (−)-17.
Scheme 23:

Hydrogen bonded dimeric structures of (+)- and (−)-17.

Another simple approach to optically active cyclopentanone carboxylic acids 17 was based on enzymatic hydrolysis of racemic methyl ester 19 performed to ca. 50% conversion (Scheme 24). The most efficient kinetic resolution was observed with α-chymotripsin. In this case, the enzymatic hydrolysis gave the acid (+)-17 with 77 ee.

Scheme 24: Kinetic resolution of racemic cyclopentanone 19 in hydrolysis catalyzed by α-chymotripsin.
Scheme 24:

Kinetic resolution of racemic cyclopentanone 19 in hydrolysis catalyzed by α-chymotripsin.

Terreins belong also to the family of cyclopentanoids. However, their synthesis will be presented in the next part of this account because a new phosphonate-based strategy was devised for their preparation.

Achiral and chiral cyclopentanone phosphonates as new building blocks: synthesis and reactivity of 3-(phosphorylmethyl)cyclopent-2-enones

The syntheses of jasmonoids and cyclopentanoid antibiotics described above were completed in a quite satisfactory way starting from simple phosphonate reagents such as α-unsubstituted phosphonates, α-phosphoryl sulfides, β- and γ-ketophosphonates. However, depending on the structure of a starting phosphonate and of a desired target compound it was necessary to devise in each case a suitable synthetic strategy. Various approaches to methylenomycin B discussed herein best illustrate our attempts to prepare it in a good yield and in a shortest way. To overcome this shortcoming our attention was turned out on the building block strategy for the synthesis of biologically active compounds and natural products. In the field of prostaglandins [29], for example, the use of the chiral optically active Corey’s lactone or 4,5-dihydroxycyclopent-2-enone acetonide as building blocks for their synthesis was very successful. Moreover, the one-pot, three component strategy of the prostaglandin synthesis from cyclopentenones developed by Noyori opened a new efficient way also for the synthesis of other classes of bioactive compounds.

Having this in mind, the reaction of a-phosphonate carbanions with dicarboxylic acid diesters was investigated. As expected, it afforded the corresponding bis-β-ketophosphonates which were found to undergo under basic conditions intramolecular cyclization to 3-(phosphorylmethyl)cycloalk-2-enones (Scheme 25) [30].

Scheme 25: General synthesis of 3-(phosphorylmethyl)cycloalk-2-enones.
Scheme 25:

General synthesis of 3-(phosphorylmethyl)cycloalk-2-enones.

The structure of these cyclic products indicates that they are formed by the intramolecular Horner reaction. However, our detailed investigations of the reaction course revealed that the kinetically controlled cyclization product first formed results from the reversible Knovenagel reaction. Under basic reaction conditions it undergoes irreversible conversion to the thermodynamically controlled Horner olefination product as depicted in Scheme 26.

Scheme 26: The cyclization course of bis-β-ketophosphonate.
Scheme 26:

The cyclization course of bis-β-ketophosphonate.

The most interesting from the view-point of realization of our synthetic program was 3-(phosphorylmethyl)cyclopent-2-enone 20 as potential building block. The structure of 20 offers many possibilities for further elaboration. In the first instance, the reactivity of the anion derived from 20 was investigated. In was found that, it exhibits typical ambivalent reactivity as a consequence of the negative charge distribution between three atoms i.e. α-phosphonate carbon atom, α-carbonyl carbon atom and carbonyl oxygen. It was demonstrated that alkylation occurs mainly at the α-carbonyl carbon atom, acylation, silylation, and phosphorylation take place at the carbonyl oxygen while the reaction with carbonyl compounds gives the Horner olefination product (Scheme 27).

Scheme 27: Reactivity of anion derived from 20.
Scheme 27:

Reactivity of anion derived from 20.

It should be added that the exchange of hydrogen for deuterium was also observed at the methylene carbon atoms 4 and 5 when the exchange was carried out at thermodynamically controlled conditions. The unique reactivity of cyclopentenone phosphonate 20 shown above indicated that it can easily be functionalized, especially at the positions 2 and 3 of the cyclopentenone ring, and can serve as a building block in the synthesis of diverse target compounds. The another advantage of 20 is its easy availability. In addition to the first synthesis of 20 from cyclopentenone described by Oehler and Zbiral [31] (method a, Scheme 28) and our synthesis based on intramolecular cyclization of bis-β-ketophosphonates, Modro synthesized 20 from 3-(methoxy)cyclopent-2-enone [32] (method b, Scheme 28). Later on, two other approaches to 20 were elaborated in our group utilizing the heteroatom containing phosphonates as reagents (method c and d, Scheme 28) [33].

Scheme 28: Synthetic approaches to cyclopentenone phosphonate 20.
Scheme 28:

Synthetic approaches to cyclopentenone phosphonate 20.

A similar strategy was applied in the synthesis of optically active 3-(phosphorylmethyl)-4,5-dihydroxycyclopentenones 21a and 21b [34]. Here, in the first important step a complete desymmetrization of optically inactive meso-tartaric acid was successfully performed when it was reacted with (+)-camphor and methyl ortoformate under acidic conditions. The single diastereoisomer of the diester formed in this reaction had the R and S absolute configuration at the diol carbon atoms as determined by X-ray analysis. Then, this key intermediate was treated with dimethyl lithiomethanephosphonate yielding, however, a separable mixture of the diastereoisomeric cyclopentenone phosphonates 21a and 21b (Scheme 29).

Scheme 29: Desymmetrization of meso-tartaric acid and synthesis of camphor protected 3-(phosphorylmethyl)-4,5-dihydroxycyclopent-2-enones 21a and 21b.
Scheme 29:

Desymmetrization of meso-tartaric acid and synthesis of camphor protected 3-(phosphorylmethyl)-4,5-dihydroxycyclopent-2-enones 21a and 21b.

The synthesis of achiral cyclopentanone phosphonate 20 and its chiral congeners 21a and 21b extended our synthetic possibilities and allowed to undertake the synthesis of a wide range of biologically active compounds and natural products.

Use of achiral 3-(phosphorylmethyl)cyclopent-2-enone in the synthesis of biologically active cyclopentenones and cyclopentanones

The title building block 20 may be easily functionalized by alkylation of the anion derived from 20 and Horner olefination reaction taking place at the position 2 and 3 of the cyclopentanone ring, respectively. Therefore, combination of these two reactions paved the way for the synthesis of 2,3-disubstituted cyclopentenones. This simple methodology was exploited in the synthesis of many interesting target compounds. A few examples presented below illustrate the usefulness of our approach to this class of cyclopentenones.

Synthesis of racemic and optically active prostaglandin B1 methyl ester

The discovery of prostaglandins (PGs), which are the most attractive class of bioactive compounds, stimulated the development of synthetic methods for their construction Therefore, it was also interesting to apply the phosphonate strategy for the synthesis of prostaglandin B1 (PGB1) methyl ester since its structure represents a disubstituted cyclopentanone [35]. Thus, starting cyclopentanone phosphonate 20 was converted into the desired PGB1 methyl ester 22 in two steps involving regioselective alkylation at C(2) with methyl 7-iodoheptanoate and the subsequent Horner olefination reaction with the dimer of 2-hydroxyheptanal. This two-step synthesis produced racemic PGB1 methyl ester 22 in 42% overall yield (Scheme 30).

Scheme 30: Preparation of racemic and enantiomeric PGB1 methyl esters.
Scheme 30:

Preparation of racemic and enantiomeric PGB1 methyl esters.

To obtain both enantiomers of PGB1 methyl ester 22 the 2-alkylated cyclopentanone 20 was subjected to the Horner olefination reaction with the enantiopure (+)-(R)- and (−)-(S)-(tert-butyldimethylsilyloxy)heptanals. Desilylation of the olefination products with fluoride anion afforded the enantiomeric (−)-(S)- and (+)-(R)-methy ester 22, respectively, in 28% overall yield. Their full enantiomeric purity was confirmed by 1H NMR spectra measured in the presence of (+)-(R)-t-butylphenylthiophosphinic acid (t-BuPhP(S)OH), as a chiral solvating agent. In this context, it is interesting to point out that the use of DBU as a base in this Horner olefination reaction is crucial because the same reaction with the use of sodium hydride for the a-phosphonate carbanion generation was accompanied by partial racemization of enantiopure heptanals and gave enantiomerically enriched prostaglandins 22.

Synthesis of enantiomeric phytoprostanes B1 type I and II

Phytoprostanes are botanical analogues of prostaglandins and are formed in plants from a-linolenic acid via a free-radical-catalyzed nonenzymatic mechanism. Both title phytoprostanes (PPB1) exhibit potent biological activity including activation of mitogen-activated protein kinase (MAPK) and the induction of glutathione-S-transferase (GTS), defence genes and phytoalexins. They are regarded to be potentially useful as markers of oxidative degradation of plant-derived food.

As phytoprostanes B1 structurally belong to 2,3-disubstituted cyclopentenones (see Scheme 31), they were ideal target molecules which could be synthesized by our strategy starting from cyclopentenone phosphonate 20 as a building block. Some details of the synthesis of enantiomeric forms of 23 are outlined in Scheme 32 and briefly discussed below [36].

Scheme 31: Structures of phytoprostanes B1 type I and II.
Scheme 31:

Structures of phytoprostanes B1 type I and II.

Scheme 32: Synthesis of enantiomeric phytoprostanes B1 type I methyl esters.
Scheme 32:

Synthesis of enantiomeric phytoprostanes B1 type I methyl esters.

As in the case of the synthesis of enantiomeric prostaglandin B1 methyl esters 22, the first step involved alkylation of the anion generated from cyclopentenone phosphonate 20 with methyl 8-bromooctanoate. The product of the C(2)-alkylation was obtained in a moderate yield (45%) due to competitive alkylation at the carbonyl oxygen. Fortunately, the Horner reaction with the enantiomeric benzoyl protected α-hydroxy butanals, (+)-(R) and (−)-(S), occurred in an excellent yield giving the olefination products in a straightforward way. After deprotection of the hydroxyl groups the enantiopure phytoprostanes (−)-(R)- and (+)-(S)-23 were obtained in ca. 25% overall yield.

The synthesis of enantiomeric phytoprostane B1 type II methyl ester 24 was completed in ca. 30% overall yield from 20via the two reaction sequence including ethylation of the ring carbon C(2) and introduction of a proper substituent at the ring carbon C(3) in the Horner olefination reaction of the C(2)-alkylated cyclopentenone with enantiomeric methyl 9-(tert-butyldimethylsilyloxy)-9-formylnonanoates 25. Since the latter optically active aldehydes have not been described in the literature, it was necessary to prepare them before the synthesis of enantiomeric esters 24. Both enantiomers of the required aldehydes 25 were easily prepared from racemic methyl 9-hydroxy-10-undecenoate via the asymmetric Sharpless epoxidation performed under kinetic resolution conditions followed by ozonolysis of the double bond. The synthesis of aldehydes 25 is shown in Scheme 33.

Scheme 33: Synthesis of silylated aldehydes 25.
Scheme 33:

Synthesis of silylated aldehydes 25.

With both enantiomerically pure aldehydes 25 in hand, the synthesis of enantiomeric phytoprostanes B1 methyl esters 24 was accomplished according to our general strategy for the synthesis of 2,3-disubstituted cyclopentenones [37]. Thus, in the first step, the anion generated from 20 was treated with ethyl iodide and the corresponding C(2)-ethylated cyclopentenones were reacted with aldehydes (−)-(S)-25 and (+)-(R)-25 using DBU-LiClO4 as a mild basic system to avoid racemization of aldehydes 25. Final desilylation of the Horner reaction product gave the desired target compounds (+)-(S)-24 and (−)-(R)-24 in ca. 30% overall yield (Scheme 34).

Scheme 34: Synthesis of enantiomeric phytoprostanes B1 type II methyl esters 24.
Scheme 34:

Synthesis of enantiomeric phytoprostanes B1 type II methyl esters 24.

Use of diastereoisomeric (+)-camphor protected 3-[(dimethoxyphosphoryl)methyl-4,5-dihydroxycyclopent-2-enones in the synthesis of biologically active cyclopentenones and cyclopentanones

In contrast to achiral cyclopentenone phosphonate building block 20, the title diastereoisomerically pure cyclopentenone phosphonates 21a and 21b contain, in addition to (+)-camphor, the enantiomerically pure diol moiety in the five membered ring at the carbon atoms 4 and 5. Therefore, it was possible to use them for the synthesis of functionalized cyclopentenones containing not only enantiomerically pure exocyclic substituents but also the cyclopentanone ring carbon atoms. Moreover, it was expected that further transformations of the ring and generation of other stereogenic centres may occur in an asymmetric manner under steric control of the camphor protected diol moiety. Below is a concise description of bioactive compounds and natural products that have been synthesized by the use of chiral building blocks 21a and 21b.

Synthesis of terreins

(+)-Terrein 26, a metabolite of Aspergillus terreus, has been isolated in 1935 by Raistrict and Smith and belongs to the family of cyclopentanoids. The structure of (+)-26, determined independently by Grove and Barton and Miller, has the trans-configuration of the diol moiety while the racemic isoterrein 27 obtained latter by Barton and Hulshof has cis-orientation of the hydroxyl groups. The structures of the enantiomeric forms of terrein and isoterrein are depicted in Scheme 35.

Scheme 35: Enantiomeric terreins and isoterreins.
Scheme 35:

Enantiomeric terreins and isoterreins.

Although the synthesis of terreins attracted a considerable attention, at the beginning of our studies only two papers appeared on the synthesis of (+)- and (−)-terrein 26. In connection with our independent studies on bis-β-ketophosphonates the most interesting synthesis of (+)-26 was reported by Altenbach and Holzapfel [38]. They found that the reaction of the protected diethyl (+)-tartrate with dimethyl lithiomethanephosphonate gave a mixture of two cyclopentenone phosphonates. The first of them, readily separated chromatographically, was the intramolecular Horner reaction product of the transiently formed bis-β-ketophosphonate. Its reaction with acetaldehyde under basic conditions followed by deprotection of the chiral diol moiety resulted in the formation of (+)-terrein 26 in 26% overall yield (Scheme 36).

Scheme 36: Synthesis of (+)-terrein 26.
Scheme 36:

Synthesis of (+)-terrein 26.

In our independent attempt to synthesize (+)-terrein 26 the benzylidene protected dimethyl (+)-tartrate 28 was used as a starting material. Upon treatment with dimethyl lithiomethanephosphonate it was converted efficiently (95% yield) into the corresponding bis-β-ketophosphonate 29 and then cyclized to 3-(phosphorylmethyl)cyclopentenone 30. The Horner reaction of the latter intermediate with acetaldehyde and deprotection of the diol moiety resulted in the unexpected formation of (−)-isoterrein 27 with a rather low optical rotation, [α]D=−14.4, and the ee value ca. 12% as determined by NMR (Scheme 37). The overall yield of this four-step synthesis of (−)-27 was 32% [39].

Scheme 37: Unexpected synthesis of (−)-isoterrein 27.
Scheme 37:

Unexpected synthesis of (−)-isoterrein 27.

The synthesis of (−)-27 shown above requires two comments with regard to the observed trans- to cis-isomerization of the chiral diol unit and formation of the highly racemized (−)-isoterrein 27. Most probably, these two undesirable processes accompany the conversion of bis-β-ketophosphonate 29 to bicyclic 3-(phosphorylmethyl)cyclopentenone 30, and may be rationalized in terms of a reversibility of the discrete cyclization steps with the exception of the last one and of a fast proton transfer between the bis-β-ketophosphonate anions transiently formed. The proposed course of the cyclization of 29–30 is shown in Scheme 38.

Scheme 38: Possible course of the 29–>30 conversion involving trans- to cis–isomerization of the diol unit and racemization.
Scheme 38:

Possible course of the 29–>30 conversion involving trans- to cis–isomerization of the diol unit and racemization.

As the first synthesis of (−)-isoterrein 27 was unsatisfactory for the reasons discussed above, our attention was paid to the diastereoisomeric building blocks 21a and 21b as ideal substrates since their structures contain the chiral cis-diol unit in the cyclopentenone ring. In fact, starting from 21a and 21b the corresponding enantiomerically pure (−)- and (+)-isoterreins 27 were obtained in two steps and in ca. 50% overall yield (Scheme 39) [34].

Scheme 39: Synthesis of enantiomeric isoterreins 27.
Scheme 39:

Synthesis of enantiomeric isoterreins 27.

Synthesis of natural (−)-neplanocin A and its unnatural (+)-enantiomer

Carbocyclic nucleosides form an important group of structural analogues of nucleosides arising from replacement of the sugar oxygen by the methylene carbon [40]. They exhibit potent antiviral activity as well as anticancer properties. Some carbocyclic nucleosides like AZT or abacavir are well known drugs against HIV (Scheme 40).

Scheme 40: Structures of selected carbocyclic nucleosides.
Scheme 40:

Structures of selected carbocyclic nucleosides.

Neplanocin A 31 is a typical member of the family of carbocyclic nucleosides. It is naturally occurring carba-nucleoside produced by certain prokaryotic organism. The isolation and X-ray structure of natural (−)-neplanocin A 31 have been reported in 1981. Due to its potential use as a therapeutic agent, efficient synthetic approaches to neplanocin A have been the subject over the last decades. Neplanocin A attracted also our attention as a new target compound, which could be synthesized using chiral diastereoisomeric building blocks 21a and 21b. This possibility was suggested by comparison of the structure of neplanocin A with that of the already obtained by us isoterrein 27. Both structures contain two the same structural units: the cyclopentene ring and chiral cis-diol moiety. Furthermore, it was expected that the exocyclic propene substituent could eventually be elaborated into the hydroxymethylene group and the carbonyl oxygen in isoterrein may be replaced by nucleobase. Although realization of this preliminary plan was in practice slightly modified, the synthesis of both enantiopure (−)-neplanocin A 31 and its unnatural (+)-enantiomer from (+)-21b and (−)-21a, respectively, have been successfully accomplished and patented [41]. The seven step synthesis of (−)-neplanocin A 31 from (+)-21b is shown in Scheme 41. In the first step, the Horner reaction of (+)-21b with n-pentanal was carried out using DBU and LiClO4 as a basic system. The dienone obtained was subjected to ozonolysis to give the corresponding aldehyde which in a crucial step was selectively reduced to the corresponding 3-hydroxymethyl cyclopentenone by triacetoxysodium borohydride. Then, protection of the hydroxy group by silylation and reduction of the carbonyl group by sodium borohydride in the presence of cerium(III) chloride afforded the desired alcohol in which the three oxygen atoms in the cyclopentene ring were in cis relation. In a final step, the hydroxyl group was replaced by adenine under the Mitsunobu reaction conditions and after deprotection of the triol (−)-neplanocin A 31 was obtained in 22% overall yield. It is necessary to point out that all the synthetic steps with the exception of the Mitsunobu reaction were occurring efficiently giving the intermediate products in yields over 85%. The optical rotation of (−)-31 so obtained, [α]D=−150.0 was a little lower than that reported by the Japanese group due to a very small contamination with triphenylphosphine oxide difficult to remove completely.

Scheme 41: Synthesis of natural (−)-neplanocin A 31 and unnatural (+)-neplanocin A
Scheme 41:

Synthesis of natural (−)-neplanocin A 31 and unnatural (+)-neplanocin A

Starting from the chiral diastereoisomeric building block (−)-21a (+)-neplanocin A 31 was synthesized according to the same reaction sequence in 23% overall yield. It is worth of note that it was the second synthesis of enantiomerically pure unnatural (+)-neplanocin A 31 since in the light of our results the synthesis of (+)-31 reported by Hegedus and Geisler [42] in 2000 gave highly racemized (+)-31 as indicated by its optical rotation [α]D=+35. Both enantiomerically pure enantiomers of neplanocin A 31 were reported in 2001 by Chu and collaborators [43]. Recently the synthesis of the enantiopure (+)-neplanocin A was also published by Jung and coworkers [44].

Synthesis of anticancer cyclopentenone prostaglandin analogues

Continuing investigation on the synthesis of biologically active cyclopentenones and cyclopentanones we recently became interested in the structural analogues of prostaglandins of the A and J types containing a cross-conjugated dienone unit. Some representatives of this class of compounds are shown in Scheme 42.

Scheme 42: Structures of cross-conjugated cyclopentadienones.
Scheme 42:

Structures of cross-conjugated cyclopentadienones.

As other members of this class of cyclopentadienones, TEI-9826 shows potent anticancer activity tested on a number of cancer lines. It is also active against cis-platin resistant tumors. In view of a possible use of TEI-9826 as a therapeutic agent and the reported syntheses of racemic and enantiomeric forms of 32, its synthesis from the diastereoisomeric building blocks 21 was challenging. A simple retrosynthetic analysis for the synthesis of 32 indicated that the introduction of a proper substituent (ω chain) at C(12) (prostaglandin numbering) could be achieved via the Horner reaction followed by hydrogenation of the exocyclic double bond formed. In this step, the endocyclic cyclopentenone double bond should also be reduced in an asymmetric way under the control of the chiral diol moiety generating the desired stereocenter at C(12). The next steps of the synthesis of 32 would be the conversion of the protected diol moiety into the double bond at C(10) and installation of the α-chain by aldol condensation/elimination reaction. The practical execution of the above synthetic plan is outlined in Scheme 43 [45].

Scheme 43: Synthesis of (−)-TEI-9826.
Scheme 43:

Synthesis of (−)-TEI-9826.

The synthesis of 32 commenced with the Horner reaction of (−)-21a with heptanal using DBU/ LiClO4 as a mild basic system. The dienone formed in this reaction was hydrogenated in the presence of Pd/C to give the corresponding n-octyl substituted cyclopentanone. It was gratifying to find that a simultaneous hydrogenation of the endocyclic double bond occurred with complete diastereoselectivity leading to a single diastereoisomer with an (S)-configuration of a newly formed stereogenic center at C(4). The conversion of the camphor protected cis-diol groups into a double bond was carried out stepwise. At first, the Johnson selective deoxygenation using aluminum amalgam was performed and then the remaining 3-hydroxy group was eliminated under acidic conditions to give 4-octylcyclopent-2-enone as a key intermediate product. Its transformation into the final product (−)-32 was achieved in three consecutive steps i.e. aldol reaction, mesylation and elimination. The overall yield of this seven step synthesis of (−)-TEI-9826 32 was 42%.

The (+)-enantiomer of TEI-9826 was obtained according to the same strategy, however, essentially in four steps since some synthetic steps were combined and performed as a one-pot reaction. Thus, the two-step conversion of the protected diol moiety into the endocyclic olefinic bond as well as the three-step transformation of 4-octylcyclopent-2-enone into (+)-TEI-9826 32 were carried out as one-pot reactions (see Scheme 44). This shorter procedure afforded the dextrorotatory enantiomer of TEI-9826 in 44% overall yield.

Scheme 44: Synthesis of (+)-TEI-9826.
Scheme 44:

Synthesis of (+)-TEI-9826.

In an extension of our work on cross-conjugated cyclopentadienones and their bioactivity, the synthesis of two other analogues of PGA and PGJ 33 and 34, designated as NEPP-11 and iso-NEP11 (see Scheme 45), was undertaken. In spite of a great concentration on neuroprotective and neuroregenerative properties of both compounds, no experimental details on their synthesis and information on the expected anticancer activities have been reported.

Scheme 45: Structures of NEPP-11 and iso-NEPP-11.
Scheme 45:

Structures of NEPP-11 and iso-NEPP-11.

The enantiomers of NEPP-11 33 and iso-NEPP-11 34 were obtained according to the strategy devised for the synthesis of enantiomeric TEI-9826. Generally, it involved four basic operation i.e. installation of a suitable ω-chain, asymmetric hydrogenation of the endocyclic carbon-carbon double bond, conversion of the chiral diol function into the olefinic bond and introduction of a proper α-chain. An efficient synthesis of (−)-NEPP-11 33 from (−)-21a is shown in Scheme 46. In spite of seven steps (−)-33 was obtained in ca. 50% overall yield [46], [47].

Scheme 46: Seven-step synthesis of (−)-NEPP-11 33.
Scheme 46:

Seven-step synthesis of (−)-NEPP-11 33.

The synthesis of (+)-iso-NEPP-11 34 carried out in a similar way is depicted in Scheme 47. In this case (+)-iso-NEPP-11 was obtained from (+)-21b in 22% overall yield.

Scheme 47: Seven-step synthesis of (+)-iso-NEPP-11 34.
Scheme 47:

Seven-step synthesis of (+)-iso-NEPP-11 34.

The cytotoxicity of (−)-NEPP-11 and (+)-iso-NEPP-11 as well as their precursor i.e. anti- and syn-aldols was investigated against cancer lines (HeLa, K562, HL-60) and for comparison against of the normal human cells (HUVEC). It was found that (−)-NEPP-11 and its anti-aldol exhibit a high anticancer activity against HeLa and HL-60 human cancer lines whereas (+)-iso-NEPP-11 and its anti- and syn-aldols were much less toxic or nontoxic against the cancer and normal cells.

Rosaprostol – an antiulcer drug: synthesis of racemic and enantiomerically pure stereoisomeric forms

Rosaprostol 35 is a trade name for 7-(2-hexyl-5-hydroxycyclopentyl)heptanoic acid which belongs to a series of 19,20-dinorprostanoic acid derivatives. The sodium salt of racemic 35 as a mixture of racemic trans,trans and trans,cis isomers has been launched in Italy as Rosal® for the treatment of gastric and duodenal ulcers. Rosaprostol exhibits gastric antisecretory activity and cytoprotective action without the undesired side effects common to other prostanoids.

The structure of rosaprostol 35 is interesting from the stereochemical view-point. Although in the structure of 35 are three stereogenic centers [at C(1), C(2) and C(5)] it can exist in the form of only four stereoisomers (Scheme 48), beacause of the trans-arrangement of substituents at C(1) and C(5). For the same reason, racemic rosaprostol is a mixture of two racemic forms having different configuration of the hydroxyl group at C(5) as shown in Scheme 48.

Scheme 48: Structure, numbering system and four enantiomeric stereoisomers of rosaprostol.
Scheme 48:

Structure, numbering system and four enantiomeric stereoisomers of rosaprostol.

At the end of the last century, racemic rosaprostol became an interesting target of biological and synthetic studies. Its synthesis was also included into our synthetic program with the aim to further demonstrate the great utility of the phosphonate chemistry and efficiency of strategies devised for the synthesis of prostaglandin analogues.

In the first our synthesis of racemic rosaprostol 35 [48], dimethyl methanephosphonate 36 was used as a substrate. It was converted via α-phosphonate carbanion into the corresponding β-ketophosphonate by treatment with methyl decanoate. The intramolecular cyclization α-diazo-β-ketophosphonate afforded trans-2-dimethoxyphosphoryl-3-hexylcyclopentanone as a key intermediate. Its conversion into rosaprostol 35 was carried out in four simple steps including the Horner reaction, reduction of the α-ethylenic moiety, hydrolysis of the methyl ester and final reduction of the carbonyl group. This seven-step synthesis afforded rosaprostol 35 as a 1:1 mixture of racemic 1,2-trans-1,5-cis and 1,2-trans-1,5-trans diastereoisomers in 42% overall yield (Scheme 49).

Scheme 49: Synthesis of racemic rosaprostol 35.
Scheme 49:

Synthesis of racemic rosaprostol 35.

In the second approach to (±)-35 achiral building block 20 was used as a starting material [49]. Introduction of α- and β-substituents to the cyclopentenone ring was done according to a general strategy i.e. α-alkylation of the mesomeric anion derived from 20 and the subsequent Horner reaction with a carbonyl component. In Scheme 50 the five-step synthesis of (±)-35 from 20 is depicted. In spite of the fact that it is shorter than the first one, the overall yield of the 20 ->35 conversion was lower (36%). This is due to a rather low yield of the first α-alkylation step of the mesomeric anion occurring in 47% yield. However, it is necessary to point out that at the turn of 20th century the two syntheses of racemic rosaprostol developed in our laboratory were the most efficient among those published at that time.

Scheme 50: Conversion of achiral building block 20 into racemic rosaprostol 35.
Scheme 50:

Conversion of achiral building block 20 into racemic rosaprostol 35.

Recently, inspired by the trend in pharmaceutical sciences and industry to replace racemic by enantiomeric drugs (so-called “chiral shift”) we came back to rosaprostol with the aim to study the stereostructure-bioactivity relationship in this drug. Therefore, our first task was the synthesis of all four enantiomerically pure stereoisomers of rosaprostol 35a–d which were unknown and not characterized. Herein the two different approaches to 35a–d are presented.

In the first approach their synthesis was accomplished according to the elaborated strategy for the synthesis of cross-conjugated cyclopentadienones where chiral building blocks 21a and 21b were used as starting reagents [50]. The reaction sequence leading to stereoisomeric rosaprostols 35a and 35b and some experimental details are shown in Scheme 51. The conversion of (−)-21a into (−)-35a was accomplished in nine steps in 18% overall yield and does not require special comments. The second stereoisomer 35b was obtained in two steps from 35a involving esterification of (−)-35a and one-pot Mitsunobu esterification-hydrolysis [51]. The latter step was accompanied by inversion of configuration at C(5). It was obvious that the two remaining stereoisomers of rosaprostol i.e. 35c and 35d could be prepared from the chiral building block (+)-21b according to the same procedure.

Scheme 51: Synthesis of enantiomerically pure stereoisomers 35a and 35b.
Scheme 51:

Synthesis of enantiomerically pure stereoisomers 35a and 35b.

However, in our opinion the first synthesis was not fully satisfactory from the view point of efficiency and atom and step economy. Therefore, we sought a shorter and more efficient approach to our target stereoisomers. In devising a new approach to stereoisomeric rosaprostols we took advantage of our experience in the synthesis of racemic rosaprostol from dimethyl methanephosphonate via 2-dimethoxyphosphoryl-3-hexylcyclopentanone 37 as a key intermediate (see Scheme 49). The latter was converted into racemic rosaprostol in four steps and in 58% yield. Therefore, it was expecting that starting from the enantiomers of 37 the four stereoisomers 35a–d should easily be prepared. Thus, in the first step cyclopentanone phosphonate (±)-37 was resolved into enantiomers via the diastereoisomeric enamines formed with (+)-(R)-1-(1-naphthyl)ethylamine. The resolution of (±)-37 and some experimental details are depicted in Scheme 52.

Scheme 52: Resolution of racemic 2-(dimethoxyphosphoryl)-3-hexylcyclopentanone 37.
Scheme 52:

Resolution of racemic 2-(dimethoxyphosphoryl)-3-hexylcyclopentanone 37.

In the second step, the conversion of (+)-37 into the rosaprostol stereoisomer (−)-35a was accomplished in four already known steps (see Scheme 49) in 56% overall yield. In the same way the second stereoisomer (+)-35c was prepared from (−)-37 in 54% yield. Finally, both stereoisomers (−)-35a and (+)-35c were converted in ca. 70% yield into the corresponding stereoisomers (−)-35b and (+)-35d with inverted configuration at C(5) by a two-reaction sequence involving esterification and one-pot Mitsunobu esterification/hydrolysis. The synthetic pathway to all four stereoisomers of rosaprostol 35 is outlined in Scheme 53 [52].

Scheme 53: Synthesis of four enantiomerically pure stereoisomers of rosaprostol 35a–d.
Scheme 53:

Synthesis of four enantiomerically pure stereoisomers of rosaprostol 35a–d.

This more efficient approach to the stereoisomeric rosaprostols 35a–d allows their preparation in gram quantities for a detailed evaluation of the biological activity and preclinical studies.

Concluding remarks

The aim of this article was to demonstrate a great utility of simple phosphonate reagents and cyclopentenone phosphonate building blocks for effecting a variety of synthetic transformations allowing the synthesis of several naturally occurring and biologically active compounds. The phosphonate-based methods compare favourably in terms of efficiency and the use of simple reagents and operations with the majority of the previously reported syntheses. Another advantage of the use of phosphonate reagents as substrates and the presence of phosphorus atom in some reaction intermediates is that the course of each synthetic step can easily be followed by mean of 31P NMR spectroscopy.

As in some cases the performed total syntheses required many steps, the present investigations are directed towards the synthesis of the so-called second generation phosphonate building blocks (see for example [53]) which should allow to complete the planned syntheses in a shorter and more efficient way. Moreover, the future work should be more concentrated on the synthesis and anticancer activity of new prostaglandin analogues and carried out in a closer collaboration with pharmaceutical departments and companies.


Article note

A collection of invited papers based on presentations at the 22nd International Conference on Phosphorous Chemistry (ICPC-22) held in Budapest, Hungary, 8–13 July 2018.


Acknowledgement

This research program was financially supported by the Polish Academy of Sciences and Ministry of Science and Higher Education in Poland. The author is indebted to the group of talented coworkers and colleagues, whose names appear in the list of references, for their invaluable contribution to the realization and development of the reviewed investigations.

References

[1] J. I. G. Cadogan (Ed.). Organophosphorus Reagents in Organic Synthesis, Academic Press, London (1979).Search in Google Scholar

[2] E. Block. Reaction of Organosulfur Compounds, Academic Press, New York (1978).Search in Google Scholar

[3] E. W. Colvin. Silicon in Organic Synthesis, Butterworths, London (1980).Search in Google Scholar

[4] M. Mikołajczyk. Rev. Heteroatom Chem.2, 19 (1989).Search in Google Scholar

[5] E. J. Corey, J. I. Shulman. J. Org. Chem. 35, 777 (1970).10.1021/jo00828a052Search in Google Scholar

[6] M. Mikołajczyk, S. Grzejszczak, A. Chefczyńska, A. Zatorski. J. Org. Chem. 44, 2967 (1979).10.1021/jo01331a003Search in Google Scholar

[7] B.-T. Groebel, D. Seebach. Synthesis 357 (1977).10.1055/s-1977-24412Search in Google Scholar

[8] M. Mikołajczyk, S. Grzejszczak, A. Zatorski. J. Org. Chem. 40, 1979 (1975).10.1021/jo00901a023Search in Google Scholar

[9] M. Mikołajczyk. In Current Trends in Organic Synthesis, H. Nozaki (Ed.), pp. 347–356, Pergamon Press, Oxford and New York (1983).10.1016/B978-0-08-029217-5.50034-6Search in Google Scholar

[10] M. Mikołajczyk, S. Grzejszczak, P. Łyżwa. Tetrahedron Lett. 23, 2237 (1982).10.1016/S0040-4039(00)87310-4Search in Google Scholar

[11] H. J. Bestman, H. Haberlein, O. Kratzer. Angew. Chem., Int. Ed. Engl. 4, 830 (1965).10.1002/anie.196508301Search in Google Scholar

[12] L. Horner, H. Hoffmann, G. Klare, V. G. Toscano, H. Ertel. Chem. Ber. 94, 1987 (1961).10.1002/cber.19610940809Search in Google Scholar

[13] M. Mikołajczyk, S. Grzejszczak, W. Midura, M. Popielarzyk, J. Omelańczuk. In Phosphorus Chemistry, L. D. Quin (Ed.), ACS Symposium Series, USA, 171, 55 (1981).10.1021/bk-1981-0171.ch011Search in Google Scholar

[14] M. Mikołajczyk, W. Midura, S. Grzejszczak. Tetrahedron Lett. 25, 2489 (1984).10.1016/S0040-4039(01)81212-0Search in Google Scholar

[15] H. Zimmer, R. E. Koenikramer, R. L. Cepulis, D. M. Nene. J. Org. Chem. 45, 2018 (1980).10.1021/jo01298a060Search in Google Scholar

[16] E. Vedejs, H. Mastalerz, G. P. Meier, D. W. Powell. J. Org. Chem. 46, 5233 (1981).10.1021/jo00338a052Search in Google Scholar

[17] M. Mikołajczyk, W. Midura. Zeit. Naturfsch. 41b, 263 (1986).10.1515/znb-1986-0220Search in Google Scholar

[18] M. Mikołajczyk. In Antibiotics and Antiviral Compounds. Chemical Synthesis and Modifications, K. Krohn, H. A. Kirst and H. Maag (Eds.), pp. 205–213, VCH Verlaggeselschaft, Weinheim, FRG (1993).Search in Google Scholar

[19] J. Jernow, W. Tautz, P. Rosen, T. H. Wiliams. J. Org. Chem. 44, 4212 (1979).10.1021/jo01337a049Search in Google Scholar

[20] M. Mikołajczyk, P. Bałczewski. Synthesis 659 (1987).10.1055/s-1987-28041Search in Google Scholar

[21] M. Mikołajczyk, P. Bałczewski. Synthesis 691 (1984).10.1055/s-1984-30937Search in Google Scholar

[22] M. Mikołajczyk, A. Zatorski. J. Org. Chem. 56, 1217 (1991).10.1021/jo00003a051Search in Google Scholar

[23] P. Bałczewski. Heteroat. Chem. 8, 67 (1997).10.1002/(SICI)1098-1071(1997)8:1<67::AID-HC11>3.0.CO;2-2Search in Google Scholar

[24] M. Mikołajczyk, R. Zurawinski. Synlett2, 575 (1991).10.1055/s-1991-20803Search in Google Scholar

[25] P. Bałczewski, M. Mikołajczyk. Org. Lett. 2, 1153 (2000).10.1021/ol005742bSearch in Google Scholar

[26] M. Mikołajczyk, R. Zurawinski, P. Kiełbasinski, M. W. Wieczorek, J. Błaszczyk. Synthesis 356 (1997).10.1055/s-1997-1193Search in Google Scholar

[27] M. Mikołajczyk, P. Bałczewski. Tetrahedron45, 7023 (1989).10.1016/S0040-4020(01)89170-5Search in Google Scholar

[28] G. Helmchen, K. Ihrig, H. Schindler. Tetrahedron Lett. 28, 187 (1987).10.1016/S0040-4039(00)95682-XSearch in Google Scholar

[29] S. Das. S. Chandrasekhar, J. S. Yadav, R. Gre. Chem. Rev. 107, 3286 (2007).10.1021/cr068365aSearch in Google Scholar PubMed

[30] M. Mikołajczyk, M. Mikina. J. Org. Chem. 59, 6760 (1994).10.1021/jo00101a041Search in Google Scholar

[31] E. Oehler, E. Zbiral. Synthesis 357 (1991).10.1055/s-1991-26465Search in Google Scholar

[32] M. J. Mphahlele, T. A. Modro. J. Org. Chem. 60, 8236 (1995).10.1021/jo00130a024Search in Google Scholar

[33] M. Mikołajczyk, W. Perlikowska. Synthesis 1225 (2007).10.1055/s-2007-965982Search in Google Scholar

[34] M. Mikołajczyk, M. Mikina, M. W. Wieczorek, J. Błaszczyk. Angew. Chem., Int. Ed. Engl. 35, 1560 (1996).10.1002/anie.199615601Search in Google Scholar

[35] M. Mikołajczyk, M. Mikina, A. Jankowiak. J. Org. Chem. 65, 5127 (2000).10.1021/jo0001129Search in Google Scholar PubMed

[36] W. Perlikowska, M. Mikołajczyk. Synthesis 2715 (2009).10.1055/s-0029-1216883Search in Google Scholar

[37] W. Perlikowska, M. Mikołajczyk. Tetrahedron: Asym22, 1767 (2011).10.1016/j.tetasy.2011.10.005Search in Google Scholar

[38] H.-J. Altenbach, W. Holzapfel. Angew. Chem., Int. Ed. Engl. 29, 67 (1990).10.1002/anie.199000671Search in Google Scholar

[39] M. Mikina, M. Mikołajczyk. Phosphorus Sulfur Silicon75, 39 (1993).10.1080/10426509308037359Search in Google Scholar

[40] X.-F. Zhu. Nucleosides Nucleotides Nucleic Acids19, 651 (2000).10.1080/15257770008035015Search in Google Scholar PubMed

[41] M. Mikołajczyk, M. Mikina, A. Jankowiak. P Appl., P-366531 (23.03.2004); PL 202753 B1 (31.07.2009).Search in Google Scholar

[42] L. S. Hegedus, I. Geisler. J. Org. Chem. 65, 4200 (2000).10.1021/jo000154xSearch in Google Scholar PubMed

[43] G. Y. Song, V. Paul, H. Choo, J. Morrey, R. W. Sidwell, R. F. Schinazi, C. K. Chu. J. Med. Chem.44, 3985 (2001).10.1021/jm010256vSearch in Google Scholar PubMed

[44] Y. H. Jung, S. I. Kim, Y. J. Hong, S. J. Park, K. T. Kang, S. Y. Kim, I. S. Kim. Tetrahedron71, 1068 (2015).10.1016/j.tet.2014.12.093Search in Google Scholar

[45] R. Zurawinski, M. Mikina, M. Mikołajczyk. Tetrahedron: Asym21, 2794 (2010).10.1016/j.tetasy.2010.11.007Search in Google Scholar

[46] R. Zurawinski, M. Mikołajczyk, M. Cieslak, K. Królewska, J. Kaźmierczak-Barańska. Patent PL 218907 (2015).Search in Google Scholar

[47] R. Zurawinski, M. Mikołajczyk, M. Cieslak, K. Królewska, J. Kaźmierczak-Barańska. Org. Biomol. Chem.13, 7000 (2015).10.1039/C5OB00550GSearch in Google Scholar

[48] M. Mikołajczyk, R. Zurawinski. J. Org. Chem. 63, 8894 (1998).10.1021/jo981120gSearch in Google Scholar

[49] M. Mikołajczyk, M. Mikina, A. Jankowiak, M. Mphahlele. Synthesis 1075 (2000).10.1055/s-2000-6316Search in Google Scholar

[50] B. Łukasisk, W. Perlikowska, R. Zurawinski, M. Mikołajczyk. J. Org. Chem. 80, 9798 (2015).10.1021/acs.joc.5b01749Search in Google Scholar PubMed

[51] J. Skarżewski, Ł. Sidorowicz. Synthesis5, 708 (2011).10.1055/s-0030-1259483Search in Google Scholar

[52] W. Perlikowska, R. Zurawinski, M. Mikolajczyk. Beilstein J. Org. Chem.12, 2234 (2016).10.3762/bjoc.12.215Search in Google Scholar PubMed PubMed Central

[53] B. Łukasik, M. Mikołajczyk, G. Bujacz, R. Zurawinski. Org. Biomol. Chem. 13, 807 (2015).10.1039/C4OB01535ESearch in Google Scholar

Published Online: 2019-02-12
Published in Print: 2019-05-27

©2019 IUPAC & De Gruyter. This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License. For more information, please visit: http://creativecommons.org/licenses/by-nc-nd/4.0/

Downloaded on 4.3.2024 from https://www.degruyter.com/document/doi/10.1515/pac-2018-1117/html
Scroll to top button