Biology and Total Synthesis of n-3 Docosapentaenoic Acid-Derived Specialized Pro-Resolving Mediators

Abstract

Research over the last 25 years related to structural elucidations and biological investigations of the specialized pro-resolving mediators has spurred great interest in targeting these endogenous products in total synthesis. These lipid mediators govern the resolution of inflammation as potent and stereoselective agonists toward individual G-protein-coupled receptors, resulting in potent anti-inflammatory activities demonstrated in many human disease models. Specialized pro-resolving mediators are oxygenated polyunsaturated products formed in stereoselective and distinct biosynthetic pathways initiated by various lipoxygenase and cyclooxygenase enzymes. In this review, the reported stereoselective total synthesis and biological activities of the specialized pro-resolving mediators biosynthesized from the polyunsaturated fatty acid n-3 docosapentaenoic acid are presented.

1. Introduction

1.1. Polyunsaturated Fatty Acids and Health Effects

A high dietary intake of ω-3 long-chain polyunsaturated fatty acids (PUFAs) is associated with many beneficial health effects [1]. These PUFAs include eicosapentaenoic acid (EPA, 1), docosahexaenoic acid (DHA, 2), and n-3 docosapentaenoic acid (n-3 DPA, 3) (Figure 1). The ω-3 PUFAs are essential nutrients that cannot be biosynthesized by the human body in sufficient amounts and must therefore be obtained from the diet.

Moreover, the dietary ω-3 PUFAs have been shown to be associated with preventing a variety of inflammatory disorders [2], including cardiovascular diseases [3], rheumatoid arthritis [4], Alzheimer’s disease [5], asthma [6], and type 2 diabetes [7]. Until recently, no molecular basis or cellular mechanisms have been established for the health effects accounted for the ω-3 PUFAs. However, the diligent and continuous efforts led by Professor Charles N. Serhan and collaborators over the last 25 years have demonstrated that the three ω-3 PUFAs 1–3, but also the ω-6 PUFA arachidonic acid (AA, 4, Figure 1), are precursors for the enzymatically formed oxygenated products named specialized pro-resolving mediators (SPMs). SPMs potently down-regulate the inflammatory process and possess nanomolar pro-resolving bioactions [8,9,10]. Since uncontrolled inflammation is a common theme for the human diseases listed above, the pro-resolving and anti-inflammatory bioactions reported for the SPMs have attracted great interest in biomedical research [11]. SPMs are also highly interesting targets for stereoselective total synthesis, enabling drug discovery projects [12].

1.2. Inflammation, Resolution of Inflammation, and Lipid Mediators in Inflammation

The inflammatory process is an essential part of the protective response to tissue injury and infection by invading microbial pathogens [13]. The inflammatory response may be divided into acute and chronic inflammation, which are defined according to the nature of the inflammatory cells appearing in tissue [13]. Acute inflammation is further divided into the initiation phase and the resolution phase of inflammation. The former has the classic cardinal signs such as rubor (redness), calor (heat), tumor (swelling), and dolor (pain), described by Celsus in the 1st century [9], in addition to the loss of function, which was added by Rudolf Virchow in the 19th century [14]. Although the primary goal of the inflammation phase is to regain homeostasis [15], if kept uncontrolled, it may result in the development of a chronic state of inflammation.

The course from initiation to the resolution of acute inflammation is illustrated in Figure 2 and shows the most central cell types in the different stages of inflammation. In the early stages of the inflammatory process, activated endothelial cells start to produce pro-inflammatory mediators, such as cytokines and chemokines, as well as chemoattractants like histamine and bradykinin. Additionally, pro-inflammatory lipid mediators, such as the leukotrienes (LTs) and prostaglandins (PGs), are biosynthesized from AA (4) after its release from the phospholipid membrane. These mediators increase vessel permeability, vasodilation, and the recruitment of leukocytes to the site of injury, leading to the classic cardinal signs of inflammation [9,13].

Basophils, eosinophils, neutrophils, monocytes, and lymphocytes are examples of leukocytes. The neutrophils, also known as the polymorphonuclear neutrophils (PMNs), are the most predominant of these cell types in the early stages of inflammation [16]. The PMNs are among the first cells to appear at the site of injury or infection and act as the host’s first line of defense, as these cells swarm to the inflamed tissue. The PMNs are essential for further acceleration of the inflammatory process by the production of inflammatory mediators, including cytokines, chemokines, lipid mediators, and growth factors [13]. The inflammatory response needs to be terminated once the incoming stimulus, caused by tissue injury or invasion of pathogens, has been defeated. During the termination of ongoing inflammation, also referred to as the resolution phase, PMNs are gradually being replaced by mononuclear cells, mainly monocytes (Figure 2). These cells differentiate into macrophages in the tissue and are often referred to as phagocytic cells, meaning they ingest foreign material and cell debris in a process named phagocytosis. Macrophages also clear apoptotic cells and cellular debris in a process named efferocytosis. Phagocytosis and efferocytosis are typical processes associated with SPMs during the resolution of inflammation. Another important event in the resolution phase of inflammation is the timely lipid mediator class switch, illustrated in Figure 3. Herein, the biosynthesis of the pro-inflammatory LTs and PGs, such as leukotriene B₄ (LTB₄, 5) and prostaglandin E₂ (PGE₂, 6), are diminished, and replaced by an enhanced biosynthesis of SPMs, such as lipoxin A₄ (LXA₄, 7), protectin D1 (PD1, 8), resolvin D1 (RvD1, 9) and maresin 1 (MaR1, 10), thus initiating resolution of inflammation [17]. The release of AA (4) from the cell membrane by cytosolic phospholipase A₂ (cPLA₂) is the first and overall rate-determining step in the biosynthesis of eicosanoids by effector and immune cells, as illustrated in Figure 3. This unbound form of intracellular AA (4) is rapidly converted in a cell type-specific manner by the cyclooxygenase (COX) or lipoxygenase (LOX) enzymes to generate lipid mediators, such as LTB₄ (5) and PGE₂ (6) [13], with pro-inflammatory properties.

Failure to reduce further neutrophil recruitment and clearance of apoptotic cells by macrophages may result in the development of a chronic state of inflammation [9,18]. Chronic inflammation is defined according to the accumulation of lymphocytes, macrophages, and plasma cells in the tissue and not by the duration of the inflammatory process [19]. When present in extravascular sites, these cells may lead to the secretion of a variety of factors, one example being tumor necrosis factor-β (TNF-β). Such factors activate fibroblasts and result in the production of cross-linked collagen, which may, in turn, lead to extensive collagenous scars [13]. This highly undesirable outcome of the acute inflammatory response has proved to be a part of the pathogenesis of various disorders [2,3,4,5,6,7,20]. As stated above, the ideal outcome of an inflammation is resolution, a process that is governed by SPMs. Thus, the active process of resolution of inflammation by SPMs biosynthesized from the ω-3 PUFAs EPA (1), DHA (2), and n-3 docosapentaenoic acid (n-3 DPA, 3) is a dynamic and detailed programmed response, and not just a means of passive dilution of chemoattractants, as previously thought [13]. These active processes leading to the resolution of inflammation are considered a biomedical paradigm shift [21,22]. Resolving inflammatory exudate converts the ω-3 PUFAs 1–3 to families of structurally distinct signaling molecules, named resolvins, protectins, and maresins, while the ω-6 PUFA AA (4) forms lipoxins [8]. SPMs are agonists in the resolution phase of inflammation and exert their bioactions by stereoselective interaction with G-protein-coupled receptors (GPCRs), hence limiting the infiltration of PMNs and enhancing the clearance of apoptotic cells by phagocytosis [18]. The importance of SPMs in the resolution phase of inflammation may lead to the development of small organic molecular drugs that are not immunosuppressive [18,21] or constitute a new way to treat inflammation in the future based on the principles of resolution pharmacology [22,23,24].

1.3. Overview of Specialized Pro-Resolving Mediators

Research over the last three decades has led to the structural elucidations and biological investigations of the SPM families outlined in Figure 4. SPMs have shown potent and interesting biological effects in many human disease models and cell systems [25,26,27,28].

The resolvins biosynthesized from DHA (2) are termed D-series resolvins (RvDs), while EPA (1) gives rise to the E-series resolvins (RvEs). DHA (2) also gives rise to the protectins and maresins, including the sulfido-conjugated compounds called protectin conjugates in tissue regeneration (PCTRs), resolvin conjugates in tissue regeneration (RCTRs), and maresin conjugates in tissue regeneration (MCTRs) (Figure 4). The term “resolvins” reflects their resolving features during inflammation [29], and the “protectins” got their name due to their potent protective activity in inflammatory and neuronal systems [30]. The “maresins” were assigned this name due to their role as macrophage mediators that resolve inflammation [31]. Hamberg, Serhan, and Samuelsson reported in 1984 that AA (4) was biosynthesized into lipoxins A₄ and B₄ [32,33]. The n-3 DPA (3)-derived SPMs include the n-3 DPA resolvins, n-3 DPA maresins, n-3 DPA protectins [34], and the most recently identified 13-series resolvins (RvTs) [35].

1.4. Specialized Pro-Resolving Mediators Derived from n-3 DPA

In 2013, Dalli, Collas, and Serhan reported that n-3 DPA (3), like EPA (1) and DHA (2), is a substrate for 5-, 12-, and 15-LOX enzymes in mice and human leukocytes, leading to the discovery of the novel n-3 DPA-derived SPMs [34,35]. These SPMs share structural resemblance with the DHA-derived protectins, maresins, and resolvins, except for the absence of the C₄–C₅ Z-olefin in the n-3 DPA-derived respective protectins, maresins, and resolvins. Hence, these 12 SPMs are congeners of the DHA-derived SPMs. In addition, Dalli, Chiang, and Serhan reported investigations of the transcellularly biosynthesis using n-3 DPA (3) and COX-2 during neutrophil–endothelial cell interactions. These studies resulted in the identification of the four-membered 13-resolvins (RvTs) [35]. The naming originates from the first and common oxygenation step occurring at carbon thirteen [35]. The hitherto reported stereoselective total syntheses of the n-3 DPA SPMs are discussed below, but first, the individual families of the n-3 DPA-derived SPMs are presented.

1.4.1. Biosynthesis of n-3 DPA Protectins

As their names suggest, PD1_{n-3 DPA} (11) and PD2_{n-3 DPA} (12) are both biosynthesized from n-3 DPA (3) [34,36,37]. 15-LOX activity on n-3 DPA (3) yields the known 17(S)-HpDPA (13) intermediate that is further subjected to 15-LOX and 15-LOX-2 to yield the epoxide intermediate 14 [36,37]. Different epoxide hydrolase activities lead to the formation of either PD1_{n-3 DPA} (11) or PD2_{n-3 DPA} (12) [37]. The biosynthetic pathways of n-3 DPA protectins, in contrast to the other n-3 DPA families of SPMs, have been studied in detail [36,37], as illustrated in Scheme 1.

1.4.2. Biosynthesis of n-3 DPA Maresins

The maresins derived from n-3 DPA (3) comprise MaR1_{n-3 DPA} (15), MaR2_{n-3 DPA} (16), and MaR3_{n-3 DPA} (17). n-3 DPA (3) is a substrate for 12-LOX, and this oxygenase forms the common and known intermediate 14(S)-HpDPA (18), as shown in Scheme 2. Enzymatic epoxidation of 18 by 12-LOX most likely forms an epoxide that undergoes ring opening by different epoxide hydrolases, producing either MaR1_{n-3 DPA} (15) or MaR2_{n-3 DPA} (16). Alternatively, the insertion of molecular oxygen at C₂₁ in 18 and additional peroxidase activity forms MaR3_{n-3 DPA} (17). Of note, these biosynthetic pathways are presented based on prior knowledge [8,34] and have not been established by experiments.

1.4.3. Biosynthesis of n-3 DPA Resolvins

The n-3 DPA-derived resolvins consist of RvD1_{n-3 DPA} (19), RvD2_{n-3 DPA} (20), and RvD5_{n-3 DPA} (21). The proposed biosynthesis suggests the insertion of molecular oxygen to n-3 DPA (3) by 15-LOX to yield the common and known intermediate 17(S)-HpDPA (13), as shown in Scheme 3. Then, 5-LOX catalyzes the formation of another hydroperoxide in the C₇ position to yield the intermediate 7(S),17(S)-dihydroperoxy-RvD_{n-3 DPA} (22). An enzymatic conversion yields an epoxy intermediate that is further hydrolyzed by different epoxide hydrolase enzymes to afford either RvD1_{n-3 DPA} (19) or RvD2_{n-3 DPA} (20). Alternatively, direct peroxidase activity on 22 gives rise to RvD5_{n-3 DPA} (21) [34]. As of today, no direct evidence is available for the biosynthesis presented in Scheme 3, but it should share great similarities to the biosynthesis of the DHA-congener resolvin D1 (9) [29,38] (see Figure 3).

A self-limited model of inflammation was applied to investigate tissue levels of n-3 DPA products during onset and resolution of inflammation [34]. In these studies, the concentration of RvD1_{n-3 DPA} (19) showed a bi-phasic profile by reaching a maximum during peak neutrophil infiltration and late into resolution. The peak level of RvD2_{n-3 DPA} (20) accorded with the onset of resolution, i.e., the point where PMN levels reach ~50% of transport maximum (T_max). The level of RvD5_{n-3 DPA} (21) gradually increased over the course of inflammation resolution, with a maximum in the late stages of the resolution phase.

1.4.4. Biosynthesis of 13-Series Resolvins

In 2015, a new series of n-3 DPA-derived resolvins was reported and termed 13-series resolvins [35]. Their names are supported by the common 13(R)-alcohol moiety in all four structures [39]. The biosynthesis of the RvTs commences with the insertion of molecular oxygen at C₁₃ by COX-2 enzymes to yield 13(R)-HpDPA (24), as shown in Scheme 4. The peroxide intermediate 24 is then subjected to peroxidase enzymes to yield the 13(R)-alcohol 25 [40]. Different enzymatic activities on 25 form the four RvTs 26–29. Interestingly, the biosynthetic formation of these SPMs was increased by atorvastatin via S-nitrosylation of the COX-2 enzymes and reduced by COX-2 inhibitors [35].

n-3 DPA (3) is an intermediate in the biosynthesis of DHA (2) from EPA (1) [41,42,43], as shown in Scheme 5. EPA (1) is converted to n-3 DPA (3) by elongase enzymes. Further elongase and Δ⁶-desaturase activities on 3 yield the 24:6 n-3 fatty acid (30) via 31, which is finally converted to DHA (2) by β-oxidation. The absence of the C4-C5 Z-configured double bond in n-3 DPA makes this PUFA incapable of biosynthesizing the RvD3, RvD4, and RvD6 n-3 DPA congeners since this double bond is involved in the biosynthesis of RvD3, RvD4, and RvD6. The naming of the n-3 DPA maresins, protectins, and resolvins is based on the chemically similar structures as the DHA-derived SPMs [34].

2. Stereoselective Syntheses of n-3 DPA-Derived SPMs

Since SPMs are biosynthesized only on the nano- to picogram scale, direct NMR analyses for their individual structural elucidations are not possible. Hence, mass spectrometry-based identification using multiple reacting monitoring (MRM) is therefore used to establish the structures from biological sources [8]. The inconvenience of this method is that it can only provide the basic structures without stereochemistry. Hence, matching the biogenic product with the product obtained by stereoselective total synthesis with defined stereochemistry is necessary to establish the complete stereochemical assignment of the SPMs. Specialized pro-resolving mediators are interesting biotemplates in drug development efforts with the aim to provide new anti-inflammatory agents without immunosuppressive effects [21,23]. The n-3 DPA (3)-derived SPMs are no exception [44].

The following sections provide an overview of the hitherto published syntheses of the various SPMs derived from n-3 DPA (3). In addition, an overview of the biological studies reported is also presented.

2.1. Synthesis and Biological Studies of PD1_{n-3 DPA} (11)

The first total synthesis of PD1_{n-3 DPA} (11) was reported in 2014 [45]. This was a convergent synthesis with Wittig salt 32, aldehyde 43, and alkyne 45 as key fragments (Scheme 6). The Wittig salt 32 was prepared in four steps from cycloheptanone (33). A Baeyer–Villiger oxidation of 33 yielded the lactone 34 in 93% yield, which was treated with catalytic amounts of H₂SO₄ in MeOH to yield the ring-opened hydroxy methyl ester 35 in 87% yield. An Appel reaction using Ph₃P, I₂, and imidazole in CH₂Cl₂ was next applied to convert the alcohol moiety in 35 to the corresponding iodide 36. Finally, refluxing 36 with Ph₃P in MeCN yielded the desired Wittig salt 32 in 77% yield over the two steps. For the synthesis of the key fragment 43, the known aldehyde 37 was prepared as previously reported from commercially available pyridinium-1-sulphonate (38) using a two-step protocol [46,47]. The first treatment of salt 38 with aqueous potassium hydroxide at −20 °C yielded the glutaconaldehyde potassium salt 39, which was transformed further with the Br₂/PPh₃ complex to aldehyde 37 in 41% yield over the two steps. Then, an Evans–Nagao acetate aldol with chiral auxiliary 40 and aldehyde 37 was executed, which yielded the aldol product 41 in 15.3:1 dr, according to the procedure of Olivio and coworkers [48]. Protection as TBS-ether 42 and reductive removal of the auxiliary yielded aldehyde 43, as earlier reported in the literature [49,50]. Aldehyde 43 was then reacted in a highly Z-selective Wittig reaction with the ylide of 32 to furnish the desired Z-alkene 44 in 54% yield. The yield in this reaction was hampered by the elimination of the TBS-protected alcohol in 44 to yield an all-conjugated system. Alkyne 45 was prepared from commercially available 1-butyne and THP-protected (S)-glycidol, as reported earlier [51,52]. A Sonogashira cross-coupling reaction between the vinylic bromide in 44 and alkyne 45 was achieved to yield 46 an excellent 92% yield using catalytic amounts of Pd(PPh₃)₄ and CuI in Et₂NH. Deprotection of the two silyl ethers in 46 was achieved using TBAF in THF at 0 °C, which afforded 47. A Z-selective reduction of the internal alkyne in 47 was then carried out using Lindlar’s catalyst in a solvent system containing EtOAc/pyridine/1-octene (10:1:1) to obtain PD1_{n-3 DPA} methyl ester (48) in 50% yield. Finally, saponification of the methyl ester 48 using LiOH in H₂O/MeOH (1:1) at 0 °C afforded the natural product 11 in 71% yield and chemical purity >98% based on HPLC chromatography. The longest linear sequence of this synthesis was 10 steps, with an overall yield of 9%. Matching experiments between synthetic and endogenous PD1_{n-3 DPA} (11) provided evidence for the absolute configuration to be (7Z,10R,11E,13E,15Z,17S,19Z)-10,17-dihydroxydocosa-7,11,13,15,19-pentaenoic acid (11).

PD1_{n-3 DPA} (11) is the n-3 DPA SPM member that has been the subject of most biological studies. This SPM displays potent anti-inflammatory and pro-resolving bioactivities comparable to those of PD1 (8) in that it decreases neutrophil recruitment during peritonitis and increases macrophage phagocytosis of both zymosan A and apoptotic neutrophils [45]. A recent study revealed that PD1_{n-3 DPA} (11) regulated neuroinflammation and reduced weight loss and cognitive deficit during epileptogenesis, in addition to halting the ensuing epileptic seizures [53]. Another interesting study uncovered the protective effects of 11 against colitis and intestinal inflammation in mice [54].

2.2. Synthesis of PD2_{n-3 DPA} (12)

The hitherto only reported total synthesis of PD2_{n-3 DPA} (12) was published in 2020 by Primdahl, Tungen, and Hansen [55]. This convergent synthesis relied on the two key fragments, aldehyde 56 and Wittig salt 57 (Scheme 7). For the synthesis of key aldehyde 56, compound 49 [56] was prepared from 2-deoxy-D-ribose (50). Conversion of 50 into its thioacetal 51 was followed by global protection into 52 and cleavage of the thioacetal to afford aldehyde 49. A Z-selective Wittig reaction of aldehyde 49 with the ylide of commercially available propyltriphenylphosphonium bromide, the latter obtained in situ after treatment with NaHMDS, afforded 53 (Scheme 7). Selective deprotection by PTSA in EtOH/MeOH at −20 °C revealed the primary alcohol 54, which was further partially oxidized to 55 using Dess–Martin periodinane. Aldehyde 55 was then reacted in a double E-selective Wittig reaction with (triphenylphosphoranylidene)acetaldehyde in toluene at elevated temperature to yield the key fragment 56. The other key fragment, Wittig salt 57, was obtained by following a four-step sequence. Firstly, a Z-selective Wittig reaction between aldehyde 59 and the ylide of 32, the latter obtained after reaction with NaHMDS, yielded Z-alkene 60. Then, deprotection yielded alcohol 61, and an Appel reaction and quaternization with triphenylphosphine afforded 57 (Scheme 7). The two key fragments, aldehyde 56 and Wittig salt 57, were joined by a Z-selective Wittig reaction to afford compound 58. The final steps included deprotection of the TBS-ethers in compound 58 with TBAF to yield the diol 62, which was hydrolyzed to PD2_{n-3 DPA} (12). This convergent synthesis provided the target molecule PD2_{n-3 DPA} (12) in 10 steps (longest linear sequence) and 5% overall yield. As of today, no biological studies with 12 have been reported.

2.3. Synthesis and Biological Evaluations of MaR1_{n-3 DPA} (15)

The only synthesis of MaR1_{n-3 DPA} (15) was reported in 2014 [57]. This convergent synthesis relied on a Sonogashira reaction between the two key fragments alkyne 69 (Scheme 8) and vinyl bromide 70, and a sp³–sp³ Negishi cross-coupling reaction between bromide 72 and 4-ethoxy-4-oxobutylzinc bromide (Scheme 9). The synthesis of alkyne 69 commenced with the protection of the alcohol moiety in commercially available (S)-(−)-α-hydroxy-γ-butyrolactone (63) using TBSOTf and 2,6-lutidine in CH₂Cl₂ to yield compound 64 in near quantitative yield, as shown in Scheme 8. Lactone 64 was then reduced to the corresponding lactol using DIBAL-H in CH₂Cl₂ at −78 °C, followed by a solvent switch to THF and reacted in a Colvin homologation reaction using LDA and trimethylsilyldiazomethane (TMSCHN₂) to afford alcohol 65 in 57% isolated yield. Swern oxidation yielded aldehyde 66, which was subjected to a Z-selective Wittig reaction with the ylide of 67. This yielded the desired alkyne 69 in 83% yield. The Wittig salt 67 was prepared in 90% yield from commercially available (Z)-3-hexen-1-ol (68) over two steps using a literature procedure [58].

The vinyl bromide 70 was prepared by reduction of known 42 (see Scheme 6 for the synthesis of 42) with LiBH₄ in a solvent system containing Et₂O and MeOH. A Sonogashira cross-coupling reaction of alkyne 69 and the vinylic bromide in 70 yielded the product 71 in an acceptable 68% yield. The primary alcohol in 71 was transformed to the corresponding bromide 72 to be further reacted in a sp³–sp³ Negishi cross-coupling reaction using the commercial palladium-based PEPPSI™-IPr catalyst, which yielded the alkyne 73. This is an example of an early application of the sp³–sp³ Negishi cross-coupling reaction in the total synthesis of an advanced natural product. Removal of the two TBS-ethers in 73 using fluoride anions yielded the diol 74 in near quantitative yield. The internal alkyne in 74 was reduced in a highly Z-selective fashion by employing Lindlar’s catalyst in a mixed solvent system containing EtOAc, pyridine, and 1-octene under a hydrogen atmosphere. Finally, saponification of the ethyl ester yielded the desired MaR1_{n-3 DPA} (15) in 86% yield and chemical purity >98% based on HPLC analyses. This convergent synthesis yielded MaR1_{n-3 DPA} (15) over 11 steps (longest linear sequence) and 12% overall yield.

Matching experiments between synthetic and human macrophage MaR1_{n-3 DPA} (15) demonstrated that the synthetic material co-elutes with the naturally occurring 15, establishing the absolute configuration to be (7S,8E,10E,12Z,14S,16Z,19Z)-7,14-dihydroxydocosa-8,10,12,16,19-pentaenoic acid. MaR1_{n-3 DPA} (15) was next assayed for its potential to enhance human macrophage efferocytosis, a highly important pro-resolution hallmark. The results provided clear evidence for the potent bioactions of 15 and its corresponding ethyl ester in stimulating efferocytosis of apoptotic human neutrophils by macrophages [57].

2.4. Synthesis of MaR2_{n-3 DPA} (16)

A stereoselective total synthesis of MaR2_{n-3 DPA} (16) was reported in 2020 [59]. The key fragments in this synthesis were alkyne 76 (Scheme 10), aldehyde 49, and Wittig salt 67 (Scheme 11). Alkyne 76 was synthesized in a four-step sequence, starting with a Bayer–Villiger oxidation of commercially available cycloheptanone (33), which was followed by Fischer esterification of the corresponding lactone of 33, as shown in Scheme 10, to yield compound 35 in 31% yield. Next, partial oxidation of the primary alcohol moiety in 35 using the Dess–Martin periodinane reagent afforded aldehyde 77 in 92% yield. A Seyferth–Gilbert homologation reaction, using the Ohira–Bestmann reagent, with aldehyde 77 produced the terminal alkyne 76 in 41% yield.

Aldehyde 49 was prepared as previously reported [56] (see Scheme 7). A Z-selective Wittig reaction with the ylide of Wittig salt 67, the latter obtained in situ after reacting 67 with NaHMDS at −78 °C and aldehyde 49, yielded Z-alkene 78. Selective deprotection of the primary TBS-ether in 78 using para-toluene sulfonic acid (PTSA) in MeOH at −20 °C revealed the primary alcohol 79, which was further oxidized to the corresponding aldehyde 80 using the Dess–Martin periodinane (DMP) reagent. Next, an E-selective Wittig between aldehyde 80 and (triphenyl-phosphoranylidene)acetaldehyde in toluene at elevated temperature afforded the α,β-unsaturated aldehyde 81 in 60% yield. A Takai olefination reaction was then performed on aldehyde 81 to yield the E,E-vinylic iodide 82 in 73% isolated yield after chromatographic purifications. A palladium-mediated Sonogashira cross-coupling reaction between vinylic iodide 82 and alkyne 76 yielded the key intermediate 83 in 50% yield. Diol 84 was prepared by deprotection of the two TBS-ethers in 83 using TBAF in THF. A Lindlar hydrogenation protocol afforded the desired methyl ester 85 in 72% yield and chemical purity >95% based on HPLC analysis. Finally, a mild hydrolysis of the methyl ester 85 yielded the desired natural product MaR2_{n-3 DPA} (16) in 97% isolated yield. Matching experiments of synthetic and endogenous 16 revealed that the right stereoisomer was prepared, thus establishing the absolute configuration of MaR2_{n-3 DPA} (16) to be (7Z,9E,11E,13R,14S,16Z,19Z)-13,14-dihydroxydocosa-7,9,11,16,19-pentaenoic acid (16). MaR2_{n-3 DPA} (16) has not been utilized in any biological studies.

2.5. Synthesis and Biological Evaluations of RvD1_{n-3 DPA} (19)

In 2019, a total synthesis of RvD1_{n-3 DPA} (19) was reported [56]. This convergent synthesis relied on a Sonogashira cross-coupling reaction between the two key fragments, vinyl iodide 90 and alkyne 91 (Scheme 12). Vinyl iodide 90 was prepared in six steps from the known aldehyde 49 (Scheme 7). First, aldehyde 49 was reduced to the corresponding alcohol with NaBH₄ in MeOH, followed by an Appel halogenation to yield the bromide 86. This bromide was subjected to a sp³-sp³ Negishi cross-coupling reaction with 4-ethoxy-4-oxobutylzinc bromide using the palladium-based PEPPSI^TM-IPr catalyst to furnish compound 87 in 54% isolated yield. Selective removal of the primary TBS-ether using PTSA revealed the primary alcohol 88. Partial oxidation of 88 using Dess–Martin periodinane, followed by an E-selective Wittig reaction, yielded the α,β-unsaturated aldehyde 89 in 58% yield over the two steps. A Takai olefination reaction converted aldehyde 89 to the vinyl iodide 90. The other key fragment, alkyne 91, was prepared from the previously prepared compound 45 [51,52] in a three-step sequence, including a zirconation/iodination reaction, a Sonogashira coupling, and a deprotection reaction. A Sonogashira cross-coupling was then performed to unite 90 and 91. The product herein, 92, was next assumed to be reduced in a Z-selective manner; however, both the trusted Boland and Lindlar reductions on 92 and its triol failed to yield the desired ethyl ester 93. The Karstedt alkyne hydrosilylation/protodesilylation protocol [60] was then utilized to convert internal alkyne 92 to the corresponding Z-alkene, which gratefully yielded, via 94a and 94b, the desired RvD1_{n-3 DPA} ethyl ester (93) in 78% yield over the two steps and with chemical purity > 97% based on HPLC analysis. A mild saponification of the ethyl ester yielded RvD1_{n-3 DPA} (19). Metabololipidomics LC-MS/MS experiments were used to determine if the synthetic and authentic material of RvD1_{n-3 DPA} (19) matched. Identical retention time and co-elution of synthetic and authentic 19 provided evidence for the right stereoisomer to be synthesized, thus establishing the absolute configuration of 19 to be (7S,8R,9E,11E,13Z,15E,17S,19Z)-7,8,17-trihydroxydocosa-9,11,13,15,19-pentaenoic acid.

The synthetic material of RvD1_{n-3 DPA} (19) showed potent agonism toward the human receptor GPR32 in an impedance assay, and this SPM displayed nanomolar anti-inflammatory, pro-resolution, and anti-bacterial effects [56].

2.6. Synthesis and Biological Evaluations of RvD2_{n-3 DPA} (20)

The stereoselective total synthesis of RvD2_{n-3 DPA} (20) was reported as a convergent synthesis, relying on a Sonogashira cross-coupling reaction between the two key fragments vinyl iodide 95 and alkyne 97 (Scheme 13 and Scheme 14) [61]. For the synthesis of vinyl iodide 95, aldehyde 55 was first prepared from commercially available and cheap 2-deoxy-D-ribose (50) as previously reported in the literature [55] (Scheme 7), and then 55 was reacted in an E-selective Wittig reaction with commercially available (triphenylphosphoranylidene)acetaldehyde to yield α,β-unsaturated aldehyde 96 in 71% yield. A Takai olefination reaction on 96 furnished the desired vinyl iodide 95 in an excellent 82% yield.

For the preparation of the terminal alkyne 97 needed for the subsequent Sonogashira cross-coupling reaction with 95, commercially available and cheap diester 98 was first selectively hydrolyzed using aqueous NaOH in THF at a lowered temperature. The carboxylic acid 99 was then converted to its acid chloride in situ, followed by a Friedel–Crafts acylation with bis(trimethylsilyl)acetylene (BTMSA) to yield 100 in acceptable 59% yield. The Midland (S)-Alpine-borane reagent was then applied to synthesize (S)-alcohol 101 in 93% enantiomeric excess and 95% yield. The alcohol in 101 was protected using TBSCl and imidazole in CH₂Cl₂ to yield 102, followed by removal of the TMS group using K₂CO₃ in MeOH to obtain terminal alkyne 103. Hydrostannylation of 103 yielded 104, which was subjected to an in situ iodination protocol to prepare vinyl iodide 105. Vinyl iodide 105 was in fact first reported in 2020 by Rodriguez and Spur for the total synthesis of RvT1 (26) and RvT4 (29) [62]. The greatest difference between the two syntheses of 105 was the use of the Midland Alpine-borane reagent for the asymmetric reduction of the acetylenic ketone herein rather than the ruthenium-catalyzed asymmetric reduction. Also, the availability of reagents was crucial for choosing different reaction conditions for the synthesis of vinyl iodide 105 herein. Next, a Sonogashira cross-coupling between 105 and trimethylsilylacetylene yielded crude 106, which was directly subjected to a TMS-deprotection step to obtain terminal alkyne 97 in 81% yield. Vinyl iodide 95 and alkyne 97 were reacted in a Sonogashira cross-coupling reaction with Pd(PPh₃)₄ (3 mol%) and CuI (9 mol%) as the catalysts of choice to afford the internal alkyne 107. The protection groups were then removed with TBAF in THF to yield triol 108 in an excellent 93% yield.

For the Z-selective reduction of the internal alkyne in 108, several different strategies were attempted [63,64] that proved problematic. No product formation was observed using the Lindlar catalyst. The Karstedt platinum-catalyzed alkyne hydrosilylation/protodesilylation protocol [60] was successful for the synthesis of structurally similar RvD1_{n-3 DPA} (19) [56]; hence, this reaction was attempted next for the Z-selective reduction of 107. Using this procedure, the protodesilylation step could also remove the silyl ethers in one pot. Unfortunately, the two-step reaction afforded a mixture of the desired product 109 and inseparable by-products. HPLC analysis revealed a chemical purity of disappointingly 81% after extensive purification by flash chromatography utilizing different combinations of eluent systems. Finally, a Z-selective hydrogenation protocol using potassium cyanide and zinc in 1-propanol/H₂O [65] yielded the natural product 20 (Scheme 15). However, due to issues in the purification step, re-esterification with TMS-diazomethane in toluene/MeOH was needed, which afforded the RvD2_{n-3 DPA} methyl ester (109) in 59% yield over the two steps and in >96% chemical purity based on HPLC analysis. Hydrolysis of 109 to RvD2_{n-3 DPA} (20) was performed just prior to matching experiments and biological testing due to the inherent chemical-sensitive nature of this SPM. MRM LC-MS/MS matching experiments were conducted that revealed that the synthetically produced material 20 was indeed identical to that of biologically produced RvD2_{n-3 DPA} (20). Also, these studies confirmed the structure of 20 to be (7S,8E,10Z,12E,14E,16R,17S,19Z)-7,16,17-trihydroxydocosa-8,10,12,14,19-pentaenoic acid (20).

RvD2_{n-3 DPA} (20) potently increased the uptake of the Gram-positive S. aureus bacteria in macrophages [61]. The effects were dose-dependent, between 0.01 and 10 nM. Moreover, using the same doses, macrophages pretreated with synthetic 20 also showed a statistically significant increase in the digestion of zymosan A bioparticles, a type of macromolecules derived from the yeast wall of Saccharomyces cerevisiae, thus providing evidence for its anti-fungal activity as well [61]. The clearance of such pathogens and inflammatory molecules by macrophages is a key step in the resolution phase of inflammation [66].

2.7. Synthesis and Biological Evaluations of RvD5_{n-3 DPA} (21)

The stereoselective synthesis of RvD5_{n-3 DPA} (21) was reported very recently [67]. This convergent synthesis relied on the two key fragments, vinyl iodide 110 (Scheme 16) and alkyne 115 (Scheme 17), which were combined in a Sonogashira cross-coupling reaction. The synthesis of vinyl iodide 110 was achieved in seven steps starting from commercially available and affordable (Z)-4-heptenal (111), as previously reported [68]. The most prominent step herein was the Macmillan organocatalytic α-oxyamination reaction, which afforded the diol 112 in 82% yield and 98% ee. Diastereomerically pure 110 was obtained from diol 112 after an E-selective Takai olefination reaction as the most pivotal step.

For the synthesis of the other fragment, alkyne 115, the known vinyl iodide 105 [61] was reacted in a Sonogashira cross-coupling reaction with commercially available 113 to afford the coupled product 114 in quantitative yields (Scheme 17). Next, compound 114 was reacted with AgNO₃ and KCN to gently remove the TMS-protection group and reveal the terminal alkyne 115. Alkyne 115 was further reacted in another palladium-mediated Sonogashira coupling with vinyl iodide 110 to yield the carbon backbone 116 in an acceptable 62% yield. The alkyne 115 is most likely highly unstable and prone to rapid decomposition, hence the slightly modest yield in this step, although using 2.5 equivalents of 115. For the Z-selective reduction of the two internal alkynes in 116, the Lindlar hydrogenation protocol was first attempted, but no reduction of the triple bonds was observed. A Z-selective hydrogenation protocol utilizing Pd/BaSO₄/quinoline was then applied, which showed rapid and selective conversion of the two internal alkynes to the respective Z-alkenes and furnished compound 117 in 71% yield. Finally, a mild removal of the TBS-ethers using catalytic amounts of acetyl chloride in dry MeOH afforded the RvD5_{n-3 DPA} methyl ester (118) in 68% yield and chemical purity of 97% based on HPLC analysis. This convergent synthesis achieved the methyl ester of RvD5_{n-3 DPA} in 8% overall yield over 12 steps (longest linear sequence).

Hydrolysis of the methyl ester 118 to RvD5_{n-3 DPA} (21) was conducted just prior to analyses due to the inherent chemically sensitive nature of this SPM. Firstly, matching experiments with synthetic and biogenic 21 provided evidence for the right stereoisomer to be synthesized. Next, agonism studies of RvD5_{n-3 DPA} (21) with the human receptor GPR101 were measured. These results confirmed the nanomolar agonism of RvD5_{n-3 DPA} (21) toward this GPCR. Additionally, the anti-inflammatory, pro-resolution, and anti-bacterial effects of 21 [67] were evaluated and confirmed [69].

2.8. Synthesis and Biological Actions of RvT1 (26) and RvT4 (29)

Rodriguez and Spur reported the only total synthesis of RvT1 (26) and RvT4 (29) so far in 2020 [62]. These two SPMs are structurally quite similar (see Scheme 4), and the authors came up with a clever solution where RvT1 (26) could be produced from synthetic RvT4 (29) by a lipoxygenation of the latter. One of the two key fragments in this synthesis, vinylic iodide 105, was prepared in a nine-step sequence (Scheme 18), starting from dicarboxylic acid 119. The diacid 119 was first esterified to yield the diester 98, followed by selective hydrolysis of one of the methyl esters using porcine pancreatic lipase. The resulting product 99 was next transformed into the corresponding acid chloride 120, which was reacted in a Friedel–Crafts acylation with bis(trimethylsilyl)acetylene (BTMSA) in the presence of AlCl₃ to yield the propargyl ketone 100 in 65% yield over the two steps. A Noyori asymmetric hydrogenation protocol was applied to introduce the (S)-alcohol at C₇ in an excellent 97% yield and >94% ee. TBS-protection of the alcohol, followed by TMS-deprotection of the terminal acetylene, produced compound 103 in 91% yield over the two steps. Finally, a standard hydrostannylation/iodination protocol afforded the vinyl iodide 105 in 83% yield.

The other key fragment in this synthesis, terminal alkyne 121, was prepared starting from commercially available and optically pure glycidol derivative 122, which was reacted with trimethylsilylacetylene, n-BuLi, and BF₃·Et₂O to yield the secondary alcohol 123, as shown in Scheme 19. The secondary alcohol in 123 was then protected using TBSCl, imidazole, and 4-DMAP, followed by selective deprotection of the primary TBS-ether using CSA in CH₂Cl₂/MeOH. Treatment of the resulting alcohol 125 with Dess–Martin periodinane yielded the corresponding aldehyde 126, which was reacted in an E-selective Wittig reaction with (triphenylphosphoranylidene)acetaldehyde to obtain the α,β-unsaturated aldehyde 127. A Z-selective Wittig reaction between 127 and known phosphorane 67 provided compound 128 in 82% yield. Removal of the TMS-protection group revealed the terminal acetylene 121, which was next reacted with vinyl iodide 105 in a Sonogashira cross-coupling reaction, providing the whole carbon skeleton of the target compounds. The two TBS-protection groups in 129 were then removed using acetyl chloride in MeOH to yield diol 130. A Boland reduction protocol was applied to selectively reduce the internal triple bond in 130 to predominantly yield the Z-olefin 131. Hydrolysis of the methyl ester 131 afforded the desired natural product RvT4 (29). To obtain RvT1 (26), RvT4 (29) was subjected to a lipoxygenation using lipoxidase type I-B from soybean. After the reduction of the resulting hydroperoxide with tris(2-carboxyethyl)phosphine hydrochloride (TCEP-HCl), pure RvT1 (26) was obtained after HPLC purification and desalting.

Patients with rheumatoid arthritis have an increased risk of developing cardiovascular diseases, including atherosclerosis. To this end, RvT4 (29) recently proved to enhance macrophage cholesterol efflux in arthritic mice to reduce vascular diseases and thus limit morbidity in inflammatory arthritis [70].

2.9. Synthesis of RvT2 (27) and Biological Actions of the RvTs

Rodriguez and Spur have disclosed the hitherto only synthesis of RvT2 (27) [71]. The synthesis commenced with the preparation of aldehyde 132 (Scheme 20). Firstly, commercially available S-(−)-1,2,4-butanetriol was transformed to the corresponding crystalline phosphonium iodide 133 and thus reacted with aldehyde 134 in a Wittig reaction to afford olefin 135. The double bond was next reduced using platinum on carbon under a hydrogen atmosphere to yield the saturated compound 136 in quantitative yield. Removal of the acetonide-protection group using diluted HCl, followed by TBS-protection, yielded the desired silylated compound 138. Selective removal of the TBS-ether of the primary alcohol to yield 139 was achieved in an acceptable 49% yield. Finally, oxidation of the primary alcohol 139 using Dess–Martin periodinane afforded compound 132.

The synthesis of the other key fragment in this synthesis, the Wittig salt 140, began with the reaction of L-(+)-ribose (141) with methyl (triphenylphosphoranylidene)acetate, as shown in Scheme 21. This E-selective Wittig reaction yielded 142, which was used directly in the next step to yield the diacetonide-protected compound 143. Selective removal of the terminal acetonide using amberlyst^® 15 yielded compound 144. Cleavage of the diol using NaIO₄ afforded aldehyde 145, which was reacted in an E-selective Wittig reaction with (triphenylphosphoranylidene)acetaldehyde to yield 146 and then further subjected to a Z-selective Wittig reaction with the ylide prepared from phosphonium iodide 67. The product herein, 147, was reduced with DIBAL-H at low temperature to convert the ester moiety to the corresponding alcohol. The resulting alcohol 148 was then transformed to the Wittig salt 140 and reacted with aldehyde 132 in a new Z-selective Wittig reaction to provide 149 in 29%, containing the carbon backbone of the target compound. Finally, acidic removal of the acetonide-protection group, followed by basic hydrolysis of the methyl ester, provided the natural product RvT2 (27). Also, the synthesis of the 13(R)-epimer of RvT2 (27) was reported by Rodriguez and Spur by following the same synthetic route as for 27, starting from d-(−)-arabinose instead of l-(+)-ribose (141) [71].

Recently, the SPMs 26–29 were reported to reduce neutrophil extracellular traps (NETs) in human blood [72]. The formation of NETs is primarily through a cell death process called NETosis [73] and is a way for neutrophils to protect the host against invading pathogens. NETs can trap microbes [74]; however, excessive formation is known to be a source of collateral tissue damage in the pathology of an array of diseases [75,76,77,78,79,80]. This phenomenon is especially known in SARS-CoV-2 infections [81] and acute respiratory distress syndrome (ARDS) [82]. Hence, the role of the RvTs in decreasing NETosis could be utilized to find a new approach for treating such infections in the future. As of today, no details on the total synthesis of RvT3 (28) have been reported.

2.10. Synthesis and Biological Actions of the ω-22 Monohydroxylated Metabolite 22-OH-PD1 (151)

At the current time, few reports exist on the further metabolism of n-3 DPA-derived SPMs. However, the further metabolism of 11 has been studied, which showed that the monohydroxy metabolite named 22-OH-PD1_{n-3 DPA} (151) was formed in human serum and neutrophils (Scheme 22). The Hansen group exploited the similarity with PD1_{n-3 DPA} (11) to synthesize its ω-oxidation further metabolite 151 [83]. Known aldehyde 66 [57] was reacted in a Z-selective Wittig reaction with the ylide of commercially available 152, the latter obtained after reaction with NaHMDS, to obtain Z-olefin 153 in 74% yield. Compound 153 was next reacted in a palladium-mediated Sonogashira cross-coupling reaction with compound 44 to nicely yield the coupled product 154. The three TBS-ethers in 154 were then removed using excess TBAF to afford triol 155, which was further subjected to a Boland reduction protocol to yield 22-OH-PD1_{n-3 DPA} methyl ester (156) in an acceptable 46% yield. Finally, a saponification of the methyl ester in 156 yielded the metabolite 151 in 90% yield and 94% chemical purity (based on HPLC).

Since SPMs are formed in nano- to picogram amounts at the site of injury, SPM metabolites are even more challenging to isolate and characterize. Hence, LC/MS-MS data were attained that showed that the biosynthetic and synthetic materials of 151 matched data from MRM experiments [83]. Biosynthetic studies with human neutrophils and human monocytes revealed in both experiments the direct formation of 22-OH-PD1_{n-3 DPA} (151) from PD1_{n-3 DPA} (11) [83]. Moreover, studies adding n-3 DPA to human neutrophils also allowed the detection of the metabolite 151 and its precursor PD1_{n-3 DPA} (11) [83].

3. Conclusions

n-3 DPA is a PUFA that has gained an increased interest in biomedical and life science research [84]. Biosynthetic studies in the presence of this PUFA and several oxygenase enzymes enabled the discovery of seven resolvins, three maresins, and two protectins [85], whereas ten of these have been prepared and confirmed by stereoselective total synthesis [45,55,56,57,59,61,62,67,71]. However, based on the chemical structure of n-3 DPA and the catalytic mechanisms of LOXs, several additional new DPA-derived SPMs are envisioned. Among those, sulfido-conjugated n-3 DPA-derived SPMs, similar to their original congeners, should be formed since epoxides are intermediates [36,37]. Moreover, the n-3 DPA-derived families of SPMs should also be attractive as substrates for receptor and metabolism studies, where less knowledge is available at this point. As of today, only the metabolite of PD1_{n-3 DPA} (11), named 22-OH-PD1_{n-3 DPA} (151), has been reported and studied [83]. The biological evaluations of simpler chemical synthetic analogs of the 12 n-3 DPA-derived SPMs known will also be of future interest. Some studies have emerged recently [86,87]. However, success in such endeavors is dependent on stereoselective synthesis of the native SPMs highlighted herein, but also the synthesis of isomers [88] and analogs [86], in particular since SPMs are produced in nanogram amounts in living systems, making NMR studies for their exact structural elucidation impossible to perform [89].