Stereoselective synthesis of MaR2n-3 DPA.

The first total synthesis of the n-3 docosapentaenoic derived oxygenated product MaR2n-3 DPA has been achieved. The 13R and 14S stereogenic centers were introduced using 2-deoxy-d-ribose in a chiral pool strategy. The geometry of the Z,E,E-triene moiety was prepared using highly E-selective Wittig- and Takai-olefination reactions as well as the Z-stereoselective Lindlar reduction. LC/MS-MS data of synthetic MaR2n-3 DPA matched data for the biosynthetic formed product that enabled the configurational assignment of this oxygenated natural product to be (7Z,9E,11E,13R,14S,16Z,19Z)-13,14-dihydroxydocosa-7,9,11,16,19-pentaenoic acid.

Recent studies have demonstrated that polyunsaturated fatty acids (PUFAs) derived specialized pro-resolving mediators (SPMs) actively govern and promote the resolution of inflammation [1]. PUFAs are enzymatically converted into different families of SPMs, e.g. the lipoxins, resolvins, protectins and maresins [2]. Maresin 1 (MaR1) is biosynthesized [3] from docosahexaenoic acid (DHA) in the presence of 12-lipoxygenase and was the first member of the maresin family of SPMs to be reported [4] and prepared by total synthesis [5].

In 2013 Dalli and co-workers reported several new SPMs biosynthesized from n-3 docosapentaenoic acid (n-3 DPA) [6]. n-3 DPA, consisting of 22 carbons and five all-Z double bonds, is an elongated product of eicosapentaenoic acid and an intermediate in the biosynthesis of DHA [7]. Using a self-limited model of inflammation and targeted metabololipidomics during the onset and resolution of acute inflammation, Dalli and co-workers [6] uncovered several novel n-3 DPA SPMs that are potent bioactive molecules. The structures of MaR1n-3 DPA (1), MaR2n-3 DPA (2) and MaR3n-3 DPA (3) are shown in Fig. 1.

Fig. 1.

Structures of MaR1n-3 DPA (1), MaR2n-3 DPA (2) and MaR3n-3 DPA (3). The absolute configuration is presented where established.

Based on their novel pro-resolving and anti-inflammatory bioactions, SPMs have attracted significant interest from the biomedical, pharmacological and synthetic organic communities [8]. SPMs act as agonists on individual GPCRs [9] exhibiting nanomolar pro-resolution and anti-inflammatory bioactions [10]. Some SPMs have entered initial clinical trial development programs [11]. These endogenously formed products are available in minute amounts from their natural sources and contain several stereogenic centers and conjugated E- and Z-double bonds. Hence, stereoselective synthesis for configurational assignment and extensive biological testing becomes necessary.

A few of the n-3 DPA-derived SPMs have recently been prepared [12] and subjected to biological evaluations [13], but MaR2n-3 DPA (2) has not been synthesized to date and its absolute configuration at C-13 remained to be determined. These facts, as well as the high demand for sufficient material for biological and pharmacological testing, inspired us to report the first total synthesis of MaR2n-3 DPA (2).

The three key intermediates 4, 5 and 6 in our retrosynthetic analysis are depicted in Scheme 1. The stereogenic centers at C13 and C14 were assumed to be R and S, respectively, based on biosynthetic considerations [6]. Hence, 2-deoxy-d-ribose (7) was deemed a suitable commercially available starting material for preparing MaR2n-3 DPA (2). This carbohydrate has been used in the stereoselective total synthesis of other SPMs [14].

Scheme 1.

Retrosynthetic analysis of MaR2n-3 DPA (2).

The phosphonium salt 8 was synthesized from Z-hex-3-en-1-ol (9) as previously described [12c]. Intermediate 11 was obtained from known TBS-protected aldehyde 10 [12d] using a highly Z-selective Wittig reaction with the in situ generated ylide of 8 (Scheme 2). This produced 11 as one diastereomer in 84% yield (ESI). Next, selective deprotection of the primary TBS-group in 11 was achieved with para-toluene sulfonic acid (PTSA) in MeOH at −20 °C giving alcohol 12 that was oxidized (Dess-Martin periodinane (DMP), NaHCO3, CH2Cl2) to its aldehyde 13 in 40% yield over the two steps. Aldehyde 13 was dissolved in toluene and 1.3 equiv. of the stabilized ylide (triphenyl-phosphoranylidene)acetaldehyde was added. The reaction mixture was heated at reflux for 19 h to afford the E-configured α,β-unsaturated aldehyde 14 in 60% yield after purification by column chromatography (ESI). To complete the formation of fragment 4, a Takai reaction was performed on aldehyde 14. After acidic work-up and purification by column chromatography the sensitive E,E-vinylic iodide 4 was isolated in 73% yield (Scheme 2).

Scheme 2.

Synthesis of vinylic iodide 4. Reagents and conditions: i) NaHMDS, CH2Cl2, − 78 °C; ii) para-toluene sulfonic acid (PTSA), MeOH, −20 °C; iii) DMP, NaHCO3, CH2Cl2; iv) toluene, (triphenyl-phosphoranylidene)acetaldehyde, Δ; v) CrCl2, dioxane, THF, CHI3, 0 °C.

Terminal alkyne 5 was conveniently prepared in a four-step sequence, starting from cycloheptanone (15), see Scheme 3. Bayer-Villiger oxidation on 15 followed by Fischer-esterification gave hydroxyl-ester 16 that was oxidized to aldehyde 17 and reacted in the Ohira-Bestmann reaction affording alkyne 5 in 12% yield from 15.

Scheme 3.

Synthesis of alkyne 5. Reagents and conditions: i) a) m-CPBA, CH2Cl2; b) MeOH, H2SO4; ii) DMP, NaHCO3, CH2Cl2; iii) dimethyl(1-diazo-2-oxopropyl) phosphonate; K2CO3, MeOH.

The Sonogashira coupling reaction with key fragments 4 and 5 produced alkyne 18 in 50% isolated yield after careful chromatographic purification (Scheme 4). Next, removal of the two-TBS groups in 18 with excess TBAF in THF produced diol 19. Reduction of the internal alkyne in 19 using the Lindlar-reduction (Pd-CaCO3, EtOAc/pyridine/1-octene, H2 1 atm) gave the methyl ester of MaR2n-3 DPA (20) in 55% isolated yield over the two steps and with >95% chemical purity (HPLC, ESI). Finally, careful saponification (LiOH, H2O, MeOH, 0 °C) of 20 gave MaR2n-3 DPA (2) in 97% yield (Scheme 4). Data from NMR, LC/MS-MS and UV experiments (ESI) confirmed the structure of 2.

Scheme 4.

Total synthesis of MaR2n-3 DPA (2). Reagents and conditions: i) CuI, Et2NH, Pd(PPh3)4 (5%); ii) TBAF, THF; iii) Pd/CaCO3, EtOAc/pyridine/1-octene, H2; iv) LiOH, H2O, MeOH, 0 °C.

We next tested whether synthetic 2 matched the endogenous MaR2n-3 DPA (2) prepared from human samples. We first isolated material from human serum and the retention time of the endogenous mediator using RP-HPLC-MS-MS lipid mediator profiling experiments [15]. Using multiple reaction monitoring (MRM) of the parent ion with m/z 361 and the daughter ions m/z 223 or m/z 193, we obtained a sharp peak with retention time (RT) of 14.4 min (Fig. 2A). Of note, a similar retention time of 14.4 min was obtained with synthetic 2 (see Fig. 2A). Moreover, co-injection (2 μL) of a homogenous sample of biological MaR2n-3 DPA (2) with synthetic 2 in a 1:10 M ratio, respectively, gave a single sharp peak in MRM experiments, with RT 14.4 min (Fig. 2A). Similar findings were made with platelet rich plasma, where endogenous MaR2n-3 DPA (5) gave a RT of 14.4 min that co-eluted with synthetic 2 (Fig. 2B). To obtain further evidence that the chemical structure for synthetic 2 matches that of endogenous MaR2n-3 DPA we next assessed the MS/MS fragmentation spectra. Here we found that, in accordance with published findings [6], MaR2n-3 DPA from both human serum and platelet rich plasma gave the following ions m/z 361 = M−H, m/z 344 = M−H−H2O, m/z 325 = M−H−2H2O, m/z 317 = M−H−Co2, m/z 299 = M−H−H2O−CO2, m/z 281 = M−H−2H2O−CO2, m/z 179 = 223-CO2, m/z 161 = 223-H2-O−CO2, m/z 149 = 193-CO2, ions that were also found in the MS/MS spectrum of synthetic 2 (Fig. 3).

Fig. 2.

Synthetic 2 matches endogenous MaR2n-3 DPA in human serum and cells. (A) human serum (B) platelet rich plasma were collected, placed in ice-cold methanol, lipid mediators were extracted and MaR2n-3 DPA was identified using lipid mediator profiling. Panels depict representative MRM chromatograms for m/z 361 > 223 (human serum) or m/z 361 > 193 (platelet rich plasma). Top panels depict the chromatograms obtained with biological material, center panels depict chromatograms obtained with synthetic 2 and bottom panels depict chromatograms obtained with the biological material co-injected with synthetic 2. Results are representative of three determinations for A and n = 3 distinct human donors for B.

Fig. 3.

MS/MS fragmentation spectra for synthetic 2 and MaR2n-3 DPA from human serum and platelet rich plasma. Lipid mediators were extracted from (A) human serum and (B) platelet rich plasma and MS/MS spectra for endogenous MaR2n-3 DPA, together with those of (C) synthetic 2, were obtained using lipid mediator profiling. Results are representative of n = 3 determination for A and C and 3 volunteers for B.

The SPMs are among the most exciting small and naturally occurring molecules currently undergoing investigations towards drug development of new anti-inflammatory drugs [1,16]. The stereoselective synthesis of 2 using the Lindlar reaction, the Sonogashira coupling reaction and the Takai olefination produced multi milligram quantities of 2 that is now available for further biological and pharmacological evaluations to be conducted.