Nisha
,
Kewal Kumar
and
Vipan Kumar
*
Department of Chemistry, Guru Nanak Dev University, Amritsar-143005, Punjab, India. E-mail: vipan_org@yahoo.com; Fax: +91-183-2258819-20; Tel: +91-183-2258802 ext. 3320
First published on 6th January 2015
Despite recent developments in combinatorial chemistry and related techniques for facilitating drug discovery and development, natural products continue to play a prominent and evolving role for the development of new therapeutic agents. Pyrrole containing natural products constitute an integral part of this strategy. The structure and reactivity of pyrrole along with its propensity to polymerize renders it a relative speciality and certainly not something for the faint of heart. Besides, the well known tetrapyrrolic “Pigments of life,” other fascinating natural products incorporating multiple copies of the pyrrole ring system are attracting the attention of organic medicinal chemists and must be acknowledged.
Past decade has witnessed the emergence of a number of reports on the impressive biological properties such as immunosuppressive,4 antimalarial,5,6 antimicrobial,7 antitumor,7a,b,8 anticancer,9 cell pH regulation by H+/Cl− symport/antiport activity,10 phosphatase inhibition11 and DNA-interchelation activities12 exhibited by natural and synthetic prodigiosins. Most interesting are their immunosuppressive activities at doses that are not cytotoxic. In vivo studies further suggested that the prodigiosins act synergistically with cyclosporine A or FK506,4c,13,14 which are presently the dominant drugs in clinical immunosuppressive regimens. Its derivative Obatoclax (GX15-070), commercially developed by the pharmaceutical company Gemin X Pharmaceuticals (recently acquired by Cephalon), is also involved in phase I/II clinical trials on leukemia, lymphoma, and solid tumor malignancies with promising anticancer potential.
Given the immense medicinal potential of prodigiosins and the complexities of their structural assignments, various synthetic chemists have developed novel synthetic approaches for their total synthesis in order to unambiguously confirm their assigned structures. The present review article is an attempt to focus on the developments, the past decade has witnessed, in the total synthesis of prodigiosin alkaloids along with their medicinal potential. The aim will be to provide an inspiration to the marvels and pit falls of constructing the polypyrrole heterocycles with in the complex systems.
Patil et al.16 evaluated the larvicidal potential of microbial pigment prodigiosin produced by Serratia marcescens NMCC46 against Aedes aegypti and Anopheles stephensi. These results confirmed that these species would be more useful against vectors responsible for diseases of public health importance. Thompson and co-workers17 synthesized and evaluated prodigiosin complexes with tin, cobalt, boron and zinc (11–14) (Fig. 2). The antimalarial activities of these prodigiosins were evaluated in vitro against the 3D7 strain of P. falciparum. The presence of a nitrogen atom in the A-ring is needed for antimalarial activity while the presence of an alkyl group at the β-position of the C-ring seems detrimental. Dibutyl tin complexes exhibited IC50 values in the nanomolar range with equal or improved activity compared to the free-base prodigiosin ligand, despite the fact that the general toxicity of tin complex is lower than that of the free bases.
Mahajan et al.18 synthesized 53 prodigiosins and assayed their in vitro anti-malarial activity against P. falciparum pansensitive D6, with chloroquine (CQ) as a reference drug. These synthetic prodigiosins having various substituents like –F, –Cl, alkyl, –NH2, etc. at different positions were also explored in the CoMSIA (Comparative Molecular Similarity Indices Analysis) model in order to explore the role of structural features on the antimalarial activity. The analyses revealed that the lipophilicity, hydrogen donor/acceptor and steric factors of the synthesized prodiginines play crucial role in the design of new analogues. The most active compound of the series 15 displayed an IC50 value of 0.9 nM against D6 strain of P. falciparum (Fig. 3).
Sainis and co-workers20 in a recent communication investigated the mechanism of cell death induced by the N-alkylated prodigiosin analogue viz. 2,2′-[3-methoxy-1′-amyl-5′-methyl-4-(1′′-pyrryl)]dipyrryl-methene (MAMPDM) 20 (Fig. 5) in S-180 and EL-4 tumour cell lines. Investigations into the mechanism of cell death by MAMPDM in S-180 cells showed the absence of characteristics of apoptotic cell death such as activation of caspase 3, DNA fragmentation and presence of cells with sub-diploid DNA content. However, rapid loss of membrane integrity was observed as assessed by the uptake of propidium iodide, which is a characteristic of necrosis. In contrast to the induction of necrosis in S-180 cells, MAMPDM has also shown to induce apoptotic cell death in EL-4 cells as evident by activation of caspase 3, fragmentation of DNA and sub-diploid DNA containing cells.
Fig. 5 N-Alkylated prodigiosin analogue, 2,2′-[3-methoxy-1′-amyl-5′-methyl-4-(1′′-pyrryl)]dipyrryl-methene (MAMPDM). |
Antiproliferative activities of prodigiosins derived from Serratia marcescens against HT-29 and T47D cancer cell lines was reported by Samadi et al.21 using MTT assay. The evaluation studies clearly elucidated the potential of PG as potent apoptotic agents in human colon adenocarcinoma exhibiting an IC50 value of 400 nM; better than doxirubicin.
Thomson22 and co-workers synthesized prodigiosin analogues 21 bearing an additional methyl and a carbonyl group at the C-ring. In vitro anticancer activity (NCI) and the study of modes of action (copper-mediated cleavage of double-stranded DNA and transmembrane transport of chloride anions) showed that the presence of the methyl group is not detrimental to their activity (Fig. 6). Although the presence of an ester conjugated to the prodigiosin C-ring have shown to decrease both pKa and chloride transport efficacy compared to the natural product, these analogues still exhibited a high rate of chloride ion transport. All synthesized analogues exhibited good in vitro anticancer activity and reduced toxicity as compared to the natural product with an acute systemic toxicity of 100 mg kg−1 in mice vs. 4 mg kg−1 for prodigiosin suggesting a larger therapeutic window of synthetic analogues than for the natural product.
Thompson et al.9 further synthesized a novel series of prodigiosin analogues 22 incorporating pendent functional esters and β-carbonyl substituents on the C-ring and evaluated for their anticancer activities (Fig. 7). The synthesized prodigiosin analogues retained the activity of prodigiosin in 60 human cancer cell lines with no reduction in efficacy being observed by the introduction of conjugated β-carbonyl group or the pendent ester.
Fig. 7 Anticancer Prodigiosin analogues with pendent functional esters and β-carbonyl substituents on the C-ring. |
Sainis26 and co-workers have also reported immunosuppressive activity of N-alkylated prodigiosin analogue, 2,2′-[3-methoxy-1′-amyl-5′-methyl-4-(1′′-pyrryl)]dipyrryl-methene (MAMPDM) 20 in mitogen stimulated spleen cells (Fig. 5). An increase in the accumulation of interlukin-2 (IL-2) and induction of apoptotic cell death was observed in these studies. Since IL-2 regulates both cellular proliferation and activation induced cell death (AICD), the effect of MAMPDM on the expression of IL-2 regulated genes involved in these two opposite processes was further investigated. The mitogen stimulated mouse spleen cells did not undergo a single division in presence of proliferation inhibitory concentrations of MAMPDM. An increase in the percentage of apoptotic cells was observed in the undivided cell population. The cells were arrested in G1 phase independent of the p53 expression. Expression of IL-2 regulated genes such as CD71, proliferating cell nuclear antigen (PCNA) and cyclin D was suppressed while the expression of Fas remain unchanged. MAMPDM therefore selectively inhibited the pro-mitogenic signaling without affecting proapoptotic signaling by IL-2. The induction of apoptosis in presence of MAMPDM was the effect rather than cause for the antiproliferative activity.
Kim et al.27 compared the inhibitory potency and mode of action of Prodigiosin with cyclosporine A (CsA) in a mouse model. PG efficiently inhibited T cell proliferation with an IC50 of 3.37 ng mL−1, while CsA exhibited an IC50 of 2.71 ng mL−1. PG has shown to inhibit only IL-2Ra expression and not IL-2 expression, whereas CsA inhibited both. Exogenously added IL-2 reversed the suppressive activity of CsA, but not that of PG. Further although both PG and CsA markedly reduced mortality rates in lethal acute graft-versus-host disease (GVHD), the combined treatment was shown to be more effective than either drug alone. These results clearly demonstrated that PG and CsA have similar inhibitory potencies, but different modes of action suggesting the potential use of PG as a supplementary immunosuppressant in combination with CsA for the treatment of GVHD.
Fig. 8 High-field shift of an aliphatic proton as a characteristic signature of the 1H NMR spectrum of compound with a rigid meta-pyrrolophane structure. |
Retrosynthetically, the assembly of the pyrrolopyrromethane 2 was envisaged from the corresponding building blocks 24–26 by successive condensation and cross-coupling reactions as shown in Fig. 9.33,34 Because of ease of availability of 25 and 26, the success of this synthesis entirely hinged upon the synthesis of aldehyde 24.
For the preparation of aldehyde 24, cyclononadienylacetone 27 was considered as the optimal substrate because of (1) the presence of two double bonds increasing the number of reactive encounters during Narasaka–Heck cyclization; (2) the remaining double bond that should provide a handle for the introduction of the butyl side chain; and (3) the symmetrical structure, facilitating its large scale preparation.
The total synthesis of desired butylcycloheptylprodigiosin 2 started with an initial reduction of (Z,Z)-cyclononadienone5a,28 27 using diisobutyl-aluminium hydride (DIBAL-H) with subsequent acetylation to result in the corresponding acetate 28 in quantitative yields (Scheme 2). The treatment of 28 with methyl acetoacetate in the presence of NaH and catalytic amount of [Pd(PPh3)4] interestingly led to the isolation of Z-configured 29, advocating the notion that the two allylic sites in 28 were uncoupled during Tsuji–Trost reaction.35 Krapcho decarboxylation36 of 29 yielded the cyclononadienylacetone 30 which was converted to pentafluorobenzoyl oxime ester 31 under standard conditions. The synthesized oxime ester 31 underwent Narasaka–Heck cyclization37 in the presence of a catalyst prepared in situ from Pd(OAc)2 and [P(o-tolyl)3] in DMF at 110 °C, to yield the unsaturated bicyclic imine 32 in good yields. In order to aromatize the product 32, potassium 3-aminopropylamide38 (KAPA) in 1,3-diaminopropane was used. This, through a series of thermodynamic deprotonation/reprotonation resulted in the selective shift of the 9,10 double bond to deliver the highly sensitive pyrrole 33, which was immediately N-Boc protected to yield 34.39
The functionalization of alkene in 34 by using BH3·THF followed by stepwise oxidation with H2O2 and Dess–Martin periodinane yielded ketone 35 as anti-Markovnikov isomer (Scheme 3).40,41 Wittig olefination of 35 delivered the product 36 which was subsequently selectively reduced with [Ir(pyridine)(cod)(PCy3)]PF6 (ref. 42) to yield 37. Subsequent oxidation of methyl substituent in 37 was optimized by the use of cerium ammonium nitrate (CAN) in CHCl3/H2O mixture using a small amount of 1,2-dimethoxy ethane (DME) as a phase transfer catalyst to yield the product 38.
Base promoted condensation of aldehyde 38 with commercially available lactam 25 with subsequent cleavage of the N-Boc protecting group led to the formation of 39. The reaction of 39 with Tf2O (Tf = trifluoromethanesulfonyl) induced a re-organization of the π-system yielding the corresponding triflate 40 as the substrate for Suzuki coupling. The treatment of boronic acid 26 with 40 in the presence catalytic amounts of [Pd(PPh3)4] and LiCl afforded butycycloheptylprodigiosin 2 in 70% yield as a deeply red–pink colored solid.
The synthetic methodology involved an initial ring expansion of cyclooctanone 44 using ethyl diazoacetate in the presence of Meerwein salt (Et3O+BF4−)48 resulting in the isolation of ketoester 45 (Scheme 4). Krapcho decarboxylation36 of ketoester 45 ensured the synthesis of cyclononanone 46. Cyclononanone 46 was converted to acetal 47 which when reacted with Br2 led to the isolation of corresponding dibromide 48 as the major product. Potassium tert-butoxide promoted dehydrobromination of 48 transformed it into (Z,Z)-configured di-unsaturated ketal 51 along with the isomeric diene 50 as a minor product. trans-Deacetalization of this reaction mixture with acetone in the presence of catalytic amount of pyridinium p-toluene sulfonate afforded the desired (Z,Z)-cyclononadiene 27.
Treatment of 27 with diisobutylaluminium hydride (Dibal-H) afforded the corresponding doubly allylic alcohol 52 which was subsequently acetylated to yield 28 (Scheme 5). Reaction of 28 with methylacetoacetate in the presence of NaH and catalytic amount of [Pd(PPh3)4] afforded 29 as the major isomer along with small amount of conjugated diene isomer 53. Krapcho decarboxylation36 of 29 with subsequent conversion of the resulting cyclononadienylacetone 30 into the pentafluorobenzoyl oxime ester 31 via the corresponding oxime 54 set the stage for Narasaka–Heck cyclization (Scheme 5).37 Pd(OAc)2 and P(o-tolyl)3 promoted transformation of 31 delivered the unsaturated bicyclic imine 32. Reluctancy of 32 to undergo spontaneous aromatization promoted the workers to deprotonate the compound at the bridge head position α to nitrogen which will form a stable aza-pentadienylanion. Thus, the treatment of 32 with KAPA (potassium 3-aminopropylamide)38 led to the formation of labile 33 via a series of thermodynamic deprotonation/re-protonation events. The highly labile 33 was immediately N-Boc protected to yield the compound 34.
Hydrogenation of 34 with BH3·THF followed by stepwise oxidation with H2O2 and subsequent treatment with Dess–Martin periodinane41 furnished ketone 35 and the un-conjugated regio-isomer 55 (Scheme 6). Wittig olefination of 35 in boiling toluene delivered the corresponding olefin 36 as a mixture of both stereoisomers. The synthesized tri-substituted alkene was hydrogenated using [(cod)(pyridine)Ir(Pcy3)]PF6 (ref. 42) to give the desired ortho-pyrrolophane 37 without reducing the pyrrole ring.
The oxidation of 37 with CAN49 in the presence of dimethoxyethane (DME) afforded the desired aldehyde 38a in good yields along with minor quantities of alcohol 38b. Base promoted condensation of 38a with commercially available lactam 25 with subsequent deprotection resulted in the synthesis of 39. Treatment of 39 with Tf2O induced re-organization of the π-system resulted in the corresponding triflate 40 as the substrate for the final Suzuki coupling.50 Treatment of 40 with boronic acid 26 in the presence of catalytic amounts of [Pd(PPh3)4] and LiCl afforded prodigiosin 2 in 61% yield. A comparison of the spectrum of the synthesized prodigiosin 2 and authentic sample showed an excellent match confirming the fact that butylcyclohepytylprodigiosin 2 is a natural product distinct from streptorubin B 4.
Total synthesis of 2 was initiated by the oxidation of commercially available cyclononanone with IBX (o-iodoxybenzoate) to yield cyclononenone 58 as reported by Nicolaou and co-workers.51b CuI-catalyzed addition of n-BuMgCl to 58 proceeded efficiently in THF at −40 °C (Scheme 7).52 The resulting enolate was trapped with 59, obtained by partial reduction of commercially available ethyl-4-oxazole carboxylate, to give 57 as a single diastereomer. The treatment of 57 with MsCl/Et3N in THF resulted in the synthesis of desired formyl pyrrole 24 in good yields. The elaboration of 24 into 2 was done in three steps viz. the condensation of 24 with commercially available pyrrolinone 25 to yield 56, the treatment of 56 with Tf2O to yield the corresponding triflate 40, Suzuki cross coupling of 40 with boronic acid 26 with subsequent hydrolysis to yield (±)butylcycloheptyl prodigiosin 2 in good yields (5 steps, 23% overall yield).31a,32,34
Following these synthetic studies, Challis and co-workers53 isolated a cyclic prodigiosin from S. coelicolor M511 and assigned it as streptorubin B (4). The findings by Challis and co-workers did not match with the report by Floss and the structural confirmation by both Furstner and Reeves, thus creating a doubt whether BCHP (2) is a natural product or not. Thomson and co-workers,54 in a recent communication reported the detailed studies regarding the electron impact (EI) mass spectra of synthetic BCHP (2) and streptorubin B (4). These studies motivated by a proposed evolutionary hypothesis have concluded that BCHP (2) was not the compound isolated from S. coelicolor A3 by Floss and infact was streptorubin B, as indicated by identical EI-MS fragmentation patterns with report of Challis and co-workers. The combination of mass spectral comparisons with genetic and biochemical data provided evidence that BCHP (2) is not a natural product produced by S. coelicolor.
Thus the treatment of enantiopure 69 with allyl magnesium bromide,58 under refluxing in the THF59 led to the synthesis of 70 which was subsequently N-tosylated to yield 71. The lactam 71 was converted to vinyl triflate 72 which was subjected to Pd-catalyzed coupling with α-ethoxydienyl boronate 67 to yield ethoxytriene 73 in good yields (Scheme 8).60 The hydrolysis of 73 to furnish desired 63 did not proceed under mild acidic conditions while the harsh conditions (80 °C with 20% H2SO4) were not compatible with the delicate moieties involved, thus abandoning the developed approach.61
Another approach developed by the same workers included a linear sequence starting from enantiopure (R)-pyrrolidine 74 which after O-TBDMS-protection was transformed into the corresponding vinyl triflate and carbonylated (10% Pd(OAc)2, Ph3P, Et3N, CO)62 in the presence of methanol to yield 76 and 77 (Scheme 9). The synthesized esters viz. 76 and 77 were transformed into corresponding Horner–Emmons–Wadsworth reagents 79 and 80 as per the protocol developed by Chiu for related carbacyclic systems olefination63 yielding the corresponding dienone 82 and 83 which were directly used for electrocyclization in cold TFA. The reaction on completion furnished cyclopentafused 84 and 85 albeit in 40% and 27% yields respectively with subsequent oxidation with DDQ64 to form 63. The key intermediate 63 in the synthesis of roseophilin was obtained in eight steps from compound 69 in 3% overall yield.
The methodology initiated with the synthesis of requisite bicycle via Buono's enamine bis-allylation protocol.66 Addition of LDA resulted in an efficient conversion of 91 to 92. Rubottom oxidation67 of 92 provided an easy access to 93 which was subjected to subsequent epoxidation using m-CPBA to yield 94 in good yields.
The oxidative cleavage of 94 using lead tetraacetate in methanol afforded the corresponding ketoester 95 which served as an aldehyde equivalent for pyrrole condensation (Scheme 10). Treatment of 95 with ammonium acetate under Paal–Knorr conditions with subsequent saponification of the methyl ester afforded the desired model system 98 via an intramolecular Friedel–Crafts acylation reaction.
The synthesis involved an initial desymmetrization of cyclohexene 102 with subsequent Jones oxidation of the intermediate aldehyde to provide carboxylic acid 103 (Scheme 11).71 The refluxing of 103 with trifluoroacetic anhydride and N-tosylpyrrole resulted in the selective acylation to form ketopyrrole.72 Reductive deoxygenation of ketopyrrole using zinc iodide and sodium cyanoborohydride provided the corresponding ester 104.73 The treatment of 104 with DIBAL-H with subsequent Swern oxidation provided the corresponding aldehyde 105 which was converted to α,β-unsaturated ester 106 via reaction with methyldiethyl phosphonoacetate using Horner–Wadsworth–Emmons74 conditions.
The synthesized 106 upon Vilsmeier–Haack formylation75 led to the formation of pyrrolyl carboxaldehyde 107 which upon Corey–Fuchs transformation76 afforded gem-dibromoalkene 108 (Scheme 12).
Reduction of 108 with DIBAL-H and subsequent silylation of alcohol 109 yielded gem-dibromoalkene 110 which was converted to corresponding alkyne 111. Corey–Fuchs sequence selective deprotonation of 111 with lithium hexamethyldisilazide and subsequent addition of methylchloroformate yielded alkynyl ester 112. [1,3]-Dipolar cycloaddition reaction of 112 with nitrone 113 provided the corresponding isoxazole 114.
The synthesis of Nazarov cyclization precursor was affected by treatment of 114 with a slight excess of m-chloroperbenzoic acid (m-CPBA) at 0 °C affording the β-ketoester 115 (Scheme 13). The silyl protecting group in the Nazarov substrate 115 was exchanged with an acetyl protecting group to yield 116 which was subsequently heated with the catalytic amount of scandium(III)triflate and 1 eq. of perchlorate providing the Nazarov product 117.77
The addition of sodium enolate of 117 to a refluxing solution of tetrakis(triphenylphosphine)palladium provided a 4:1 mixture of macrocycle 118 to a product resulting from β-hydride elimination A as shown in Scheme 14.
Recrystallization led to the isolation of 118 which upon hydrogenation and subsequent deprotection of the pyrrole nitrogen furnished the macrocyclic β-ketoester 119 (Scheme 15). Krapcho dealkoxy carbonylation of 119 in the final step delivered 60 in good yields.
The methodology involved a regioselective acylation of N-tosylpyrrole 126 with 6-heptenoic acid79 in the presence of TFAA to yield acylpyrrole 127.72 Reduction of carbonyl in 127 using borane-tert-butylamine complex in the presence of aluminium trichloride80 led to the synthesis of 2-(6′-hetenyl)-pyrrole 123.
Another precursor 2-methoxy carbonyl-4-methyl pentenoic acid 122 was obtained via Knoevenagel condensation81 between tert-butyl methyl malonate and isobutyraldehyde with subsequent removal of protecting group with TFA (Scheme 16). Tandem pyrrole acylation–Nazarov cyclization between 122 and 123 using TFAA resulted in the formation of variety of products. A variety of Lewis acids were employed to improve the yield of desired cyclopenta[b]pyrrole derivative 121. FeCl3 proved to be the most useful in the formation of 121 in 75% isolated yield. Cross olefin metathesis reaction of 121 with allylacetate gave 120 whose palladium catalyzed intramolecular Tsuji–Trost reaction56e,h resulted in 128 in moderate yields.
The synthetic approach initiated with lithiation of methoxyfuran 133 with its subsequent ZnBr2–Pd catalyzed carboxylation to yield the corresponding carboxylic acid 134 (Scheme 17).83 Condensation of 134 with 1-(methanesulfonyl)-1H-benzotriazole afforded the corresponding amide which was acylated with 2-(8-nonenyl)pyrroleby using TiCl4 (ref. 84) to yield the corresponding bis-heteroaryl ketone 135 in high yields. Treatment of 135 with KH and diethylchlorophosphite gave the N-phosphinyl derivative which was oxidized to corresponding phosphoramide 136.85 Metathesis of 136 with isopropyl ketone 137 (ref. 86) with subsequent in situ Pd-catalyzed hydrosilylation87 gave the ketone 138 as an amber oil. The treatment of 138 with crown ether/KHMDS combination at 55 °C resulted in the gradual formation of pyrrolophane 141, probably via a kinetic enolate 139 in equilibrium with hindered aldol salt 140. The elimination of potassium diethyl phosphate from 141 afforded 142. Hydrogenation of 142 in the presence of catalyst generated from Rh(cod)2OTf and a Josiphos ligand led to the isolation of cis-β-pyrrolyl ketone 144 with high diastereoselectivity (>25:1). Cyclo-dehydration of 144 using [ReBr-(CO)3(thf)]2 smoothly afforded the unstable 2-azafulvene 145. The unstable 2-azafulvene 145 was not isolated and treated in situ with dry HCl and substiochiometric amounts of t-BuOH to yield roseophilin hydrochloride 3 in 32% over all yields.
Synthetic protocol initiated with the synthesis of prolinol 152 from trans-4-hydroxyproline 151 via a sequence of four step reaction including esterification, tosylation, silylation and reduction. Prolinol 152, thus obtained was transformed into α,β-unsaturated ethyl ester 153 via Swern oxidation and subsequent Wittig olefination (Scheme 18). Hydrogenation of 153 using hydrogen and catalytic amount of 10% palladium on carbon followed by reduction using lithium aluminium hydride resulted in the isolation of alcohol 154. Pyridinium chlorochromate promoted oxidation of 154 with subsequent olefination with methyltriphenylphosphonium iodide gave the corresponding olefin 155. The olefin 155 was desilylated with tetra-n-butylammonium fluoride and oxidized with pyridinium chlorochromate to result in the synthesis of ketone 150. Grignard addition of 1-nonenyl-5-magnesium bromide 156 to the ketone 150 with subsequent ring closing metathesis using second generation Grubbs catalyst resulted in 157 in 58% yield. The product 157 was then transformed to known Furstner's intermediate 148 via hydrogenation with 10% Pd on carbon with subsequent dehydration by using boron triflouride etherate.
To solve this problem, Thomson and co-workers90 has recently described the enatioselective total synthesis of streptorubin B involving a one pot enatioselective aldol cyclization/Wittig reaction and an anionic oxy-cope rearrangement as the key steps. The retrosynthesis devised for the preparation of streptorubin involved an initial disconnection of the bis-pyrrole side arm to generate the pyrrolophane core 158 (Fig. 19). Paal–Knorr simplification of 158 with subsequent functional group interconversions led to the cyclodecanone 159, containing the full retron for the anionic oxy-cope rearrangement.91 The functionalized cyclohexanol precursor 160 could be assessed via a proline-catalysed enantioselective desymmetrizing intramolecular aldol reaction of dialdehyde 162 (ref. 92) with subsequent in situ Wittig reaction to form 161.
Thus the treatment of 164, obtained in a single step93 from commercially available cycloheptene 163, with 10 mol% (s)-proline with subsequent addition of ylide 166 resulted in the isolation of homoallylic alcohol 161 as a major diastereomer (98:2 mixture of enantiomers). 161 upon oxidation followed by addition of the vinyl anion 167 (ref. 94) gave the precursor 160 required for anionics oxy-Cope rearrangement with an 97:3 ee. Treatment of alcohol 160 with KHDMS and 18-crown-6 yielded the desired 10-membered ring 159 with an enantiopurity of 97:3. The alkene reduction in 159 with concomitant benzyl ether cleavage, oxidation of the liberated alcohol to the aldehyde and the Paal–Knorr pyrrole synthesis afforded the pyrrole core 158. Acid promoted condensation between pyrrole 158 and aldehyde 169,95 with subsequent removal of the Boc group via methanolysis yielded 4 in an enantioselective manner (Scheme 19). The streptorubin B 4 was prepared in nine steps from 163 in 20% overall yield. The comparison between CD spectra of synthesized and natural sample of streptorubin B coupled with X-ray crystallography confirmed the absolute stereochemistry of this prodigiosin.
The synthesis of metacycloprodigiosin was initiated with the treatment of ethyl magnesium bromide to enone 170 using 6 mol% (R,S)-Josi Phos leading to an intermediate which was trapped with chlorosilane 171 yielding silyl bis-enol ether 172 (Scheme 20). 172 was directly subjected to the oxidative bond formation using ceric ammonium nitrate and di-tert-bu-pyridine affording dione 173 as a mixture of diastereomers. The treatment of dione 173 with 10 mol% of Grubbs second generation catalyst98 led to the synthesis of 12-membred ring 174. Subsequent hydrogenation with H2/Pd(OH)2 gave the fully saturated system which was converted to pyrrole 175 upon treatment with ammonium acetate. Trimethyl silyltriflate-mediated aldol coupling of 176 with 25 (ref. 99) gave ether 177 which upon treatment with HCl in THF afforded the requisite lactam 178. Triflation of 178 with subsequent Suzuki cross-coupling with pyrrole 26 afforded metacycloprodigiosin 5 in 76% yields.
The synthetic methodology involved an initial Vilsmeier–Haack haloformylation of 4-methoxy-3-pyrrolin-2-one 25 to yield the bromoenamine 183. Suzuki coupling of 183 with Boc-1H-pyrrol-2yl-boronic acid 26 afforded the Boc-protected analogue 169 in 48% yields (Scheme 21).
Another intermediate 179, was prepared by a sequence of synthetic steps as shown in Scheme 22 involving the addition of Grignard 185 to pyrrole–aldehyde 184 to yield the corresponding secondary alcohol 186. Ley oxidation of 186 yielded the ketone 187 which upon Muchowski's one-pot cascade synthesis led to the isolation of 188.102 Cross-metathesis of 188 with 189 in the presence of Grubbs II catalyst103 resulted in the isolation of desired 179 along with the conjugate addition products 190 and 191. Interestingly, increasing catalyst loading and lowering the temperature from 0.5 to 30 mol% improved the yield of cross-metathesis product 179 (40% yield).
Acid promoted condensation of biosynthetic fragments 179 and 169 delivered the C1–C25 acyclic precursor 180, required for the proposed inverse-electron-demand hetero-Diels–Alder reaction. However, the use of varied reaction conditions (heat, microwave, photochemical, Lewis acid catalysis, mineral acid, solvent and additives) to carry out the inverse-electron-demand hetero-Diels–Alder reaction failed to deliver 181 from 180, which was further supported by modelling studies (Scheme 23).
A Paal–Knorr pyrrole synthesis with subsequent ring closing metathesis (RCM) would facilitate the formation of macrocycle 194 from 1,4-diketone 195. 195 in turn, would be obtained via key setter reaction from 196 which is a critical intermediate derived from Evan's auxillary phosphonate 197, vinyl magnesium bromide and (s)-propylene oxide as depicted in Scheme 24. The synthetic protocol involved a copper catalyzed Grignard addition to (s)-propylene oxide 198 with subsequent in situ silylation of the resulting alcohol to yield olefin 200. Ozonolysis of 200 resulted in the corresponding aldehyde 201 which upon Horner–Wadsworth–Emmons olefination with Evan's auxillary phosphonate, prepared in two steps from (R)-oxazolidinone 204, yielded acyloxazolidinone 205. Cu-promoted conjugate addition of allyl magnesium bromide to 205 delivered 206 with >20:1 dr.
Another intermediate 207 was prepared via an initial mono-PMB protection of cis-butene-1,4-diol to yield alcohol 208 with subsequent oxidation using MnO2 to yield 209. TiCl4-mediated aldol reaction of 206 with 209 under Crimmin's conditions delivered Evan's syn product 210 with 10:1 dr as shown in Scheme 25.105,106 Hydrolysis of the auxillary with LiBH4 generated corresponding alcohol which was immediately protected as TIPS silyl ether 211 (Scheme 25). VO(acac)2-promoted epoxidation of 211 yielded the oxirane 212 as a single stereoisomer.107 The secondary alcohol functionality in 212 was protected as the benzyl ether while subsequent removal of PMB group using DDQ led to the formation of primary alcohol 213. The ring opening of epoxide using Red-Al yielded 1,3-diol 214 with >20:1 ratio over the 1,2-diol congener.108 The primary hydroxyl group in 214 was subsequently protected as a pivalate while secondary alcohol was converted to methyl ester affording 197 as a key intermediate.
Deprotection of TIPS in 197 using BF3·OEt2 resulted in the formation of primary alcohol 215 which upon oxidation by using Parikh–Doering condition led to the aldehyde 216 (Scheme 26).109,110 A two step sequence viz. addition of vinyl Grignard reagent with subsequent Dess–Martin periodinane oxidation resulted in the corresponding α,β-unsaturated ketone 217. Reaction of 217 with 6-heptenal under Stetter conditions yielded 218 which upon ring closing metathesis using Grubbs I catalyst (30%) afforded the desired RCM product 219. Microwave promoted reaction of 219 using ammonium acetate in methanol delivered the desired macrocyclic pyrrole moiety 194 of marineosin 8 in 5.1% overall yield.
The synthetic protocol was initiated by the addition of vinyl magnesium bromide to the readily available lactone 223 to yield the hydroxyketone 224.112 The hydroxyketone 224 was subsequently protected as its triethyl silyl ether to afford the corresponding enone 225 as shown in Scheme 27. Reaction of 225 with benzaldehyde oxide, N-chlorosuccinimide and triethylamine provided the corresponding isoxazoline 226a which upon hydrogenolysis over RANEY® nickel led to the formation of hemi-iminal 227a as a mixture of isomers. Sodium hydride promoted methylation of 227a gave the methyl ether iminal 228a with subsequent hydrolysis of triethyl silyl ether to result in the desired spiroiminals 230a, 231a and 233a. The major isomer 230a showed equilibration in 2 weeks to give 19:1 mixture of 230a and 232a, establishing the identical stereochemistry at C-4 and C-7 (Scheme 28). The methodology developed for the synthesis of phenyl-substituted spiroiminals was further extended towards the synthesis of spiroaminals with a pyrrole substituent. Thus, the treatment of 225 with the oxime of N-SEM-pyrrole-2-carboxaldehyde113 NCS and Et3N at −78 °C in THF afforded isoxazoline 226c in <30% yield. However, the reaction of N-SEM-pyrrole-2-carboxaldehyde oxime with 5% aqueous NaOCl114 generated the nitrile N-oxide which gave 226c in 73% yield. Hydrolysis of triethyl silyl ether functionality in 228c with 2 M aqueous hydrochloric acid gave the protected spiroaminals 230c in an inseparable equilibrium mixture with 231c and 233c. Deprotection of 230c with TBAF and molecular sieves in THF at 60 °C afforded spiroaminal 230b in 54% yields.
The synthetically strenuous prodigiosins with enthralling biological potential will always be an attraction for synthetic organic chemists. The examples cited in the review, summarizes both the achievements and contribution of organic synthesis in total synthesis of bacterial prodigiosins. Note-worthy are the explicit assignment of structures to prodigiosins and the remarkable control of stereoselectivity demonstrated in some synthesis.
One of the crucial factors impeding the clinical development of prodigiosins is their high synthetic cost and therefore the development of simple and concise routes for the enantioselective synthesis of prodigiosins and their analogues with biological relevance is indeed desirable.
This journal is © The Royal Society of Chemistry 2015 |