Menaquinones, Lipoquinones and Other Vitamin K Derivatives

molecules

Review Synthesis of Naphthoquinone Derivatives: Menaquinones, Lipoquinones and Other Vitamin K Derivatives

Margaret Braasch-Turi 1 and Debbie C. Crans 1,2,*

1 Chemistry Department, Colorado State University, Ft. Collins, CO 80525, USA; [email protected] 2 Cell & Molecular Biology Program, Colorado State University, Ft. Collins, CO 80525, USA * Correspondence: [email protected]

Academic Editors: Roman Dembinski and Igor Alabugin Received: 13 September 2020; Accepted: 27 September 2020; Published: 29 September 2020

Abstract: Menaquinones are a class of isoprenoid molecules that have important roles in human biology and bacterial electron transport, and multiple methods have been developed for their synthesis. These compounds consist of a methylnaphthoquinone (MK) unit and an isoprene side chain, such as found in vitamin K1 (phylloquinone), K2, and other lipoquinones. The most common naturally occurring menaquinones contain multiple isoprene units and are very hydrophobic, rendering it difficult to evaluate the biological activity of these compounds in aqueous assays. One way to overcome this challenge has been the application of truncated MK-derivatives for their moderate solubility in water. The synthesis of such derivatives has been dominated by Friedel-Crafts alkylation with BF OEt . This attractive method occurs over two steps from commercially available starting 3· 2 materials, but it generally produces low yields and a mixture of isomers. In this review, we summarize reported syntheses of both truncated and naturally occurring MK-derivatives that encompass five different synthetic strategies: Nucleophilic ring methods, metal-mediated reactions, electrophilic ring methods, pericyclic reactions, and homologation and side chain extensions. The advantages and disadvantages of each method are discussed, identifying methods with a focus on high yields, regioselectivity, and stereochemistry leading to a detailed overview of the reported chemistry available for preparation of these compounds.

Keywords: menaquinone; lipoquinone; synthesis; Friedel-Crafts alkylation; nucleophilic substitution; metal-mediated; electrophilic; pericyclic; and homologation

1. Introduction Menaquinones are hydrophobic, isoprenoid molecules containing a methylated naphthoquinone unit and an isoprene side chain which constitutes a subgroup of lipoquinones [1–7]. There are two major structural subgroups of lipoquinones. Ubiquinones (or benzoquinones, U) are generally found in eukaryotes and Gram-negative prokaryotes, and menaquinones (or naphthoquinones, MK) generally found in Gram-positive prokaryotes. In humans, menaquinones have several biological properties, including facilitating blood coagulation. In bacteria, menaquinones are essential molecules that shuttle electrons between the membrane-bound protein complexes acting as electron acceptors and donors in the respiratory electron transport system and consequently resulting in ATP synthesis. Menaquinones are also referred to as vitamin K2, a subgroup of the class of compounds categorized as vitamin K. The preparation and studies of some of these compounds have been extensive and include the preparation both by fermentation and synthesis [8–10]. The major structural variation in menaquinones involves the isoprene side chain; that is the number of isoprene units and saturation in the side chain [11]. Although these minor structural changes are deceptively simple, the speciﬁc

Molecules 2020, 25, 4477; doi:10.3390/molecules25194477 www.mdpi.com/journal/molecules Molecules 2020, 25, 4477 2 of 37 stereochemistry of each is required to maintain their biological action. Hence, the synthesis of these compounds requires attention to detail and often meticulous purification, which consequently leads to a decrease in yields of the desired compounds. In the following review, we will summarize reported methods and discuss the advantages and disadvantages of each, as well as identify the most suitable methods within each synthetic strategy. Vitamin K2 and vitamin K1 (also referred to as phylloquinone) are natural vitamins and together make up the family of compounds known as vitamins K. The variety of the structures of vitamin K2 depend on the number of isoprene units in the side chain. In Figure1, we show the structures of menadione (vitamin K3) 1, menadiol 2, vitamin K1, and menaquinone (MK) derivatives with 1 to 9 isoprene groups in the side chain. The abbreviation used for the menaquinones in this manuscript will indicate the number of isoprene units in the side chain. For example, MK-1 describes a menaquinone with one isoprene unit, and MK-9, the major MK-derivative found in the Mycobacteria, contains nine isoprene units [12]. If the MK-derivative contains isoprene units that are saturated, it will be indicated by the addition of a Roman numeral to specify the location of the isoprene unit numbered from the naphthoquinone. For example, MK-9(II-H2) is the major MK-derivative active as the electron transport agent in the Mycobacterium tuberculosis. It is a MK-9 derivative with the second isoprene group saturated [6]. When more than one isoprene units are saturated, such as the three units in vitamin K1, the nomenclature will identify the location of the saturations using Roman numerals and the number of H-atoms added, such as MK-4(II,III,IV-H6) indicating the first isoprene unit is still unsaturated [11]. The biological activities of these compounds have been studied in-depth, and most reviews that been reported have a biological-biomedical focus. In general, terpenes and isoprenoid compounds have been widely reviewed for many uses, including pharmaceutical, flavor fragrance, and possible applications in biofuel industries [13]. Naphthoquinones have been assessed for their biological activity against cancer [14], cardiovascular disease [15], tuberculosis [16,17], diabetes [18], kidney function [19], and age-related diseases [20]. The most well-known members of the naphthoquinone family are the compounds known as vitamin K. Much work has been done in this area comparing the health effects of synthetic vitamin K analogs to naturally-derived analogs [21]. The biosynthesis of vitamin K analogs have been mapped out within bacteria [22], particularly with respect to intestinal bacteria in relation to coagulation homeostasis [23], and discovery of possible drug targets to inhibit electron transport systems [24]. Another important effect caused by vitamin K1 is the regulation of calcium uptake, particularly in bone of humans and other mammals [25]. In addition, menaquinones are hydrophobic molecules that shuttle electrons between the membrane-bound protein complexes acting as electron acceptors and donors in the respiratory electron transport system facilitating oxidative phosphorylation in bacteria [16,22]. In Mycobacteria, a menaquinone headgroup is required, whereas in a bacterium such as E. Coli, both menaquinones and ubiquinones (U-9, Figure1) are able to serve this function [ 6]. In the case of Mycobacterium smegmatis, MK-9(II-H2) was proposed as a contextual virulent factor [6], but this suggestion was later rejected when the presence of inactive MenJ protein was shown to prevent infection by the bacterium. Thus, the protein rather than the substrate was found to be the contextual virulent factor [26]. Development of inhibitors for the biosynthesis of menaquinones or other isoprene-derivatives has been explored as potential treatments [27,28].

1.1. Properties and Biological Function of Menaquinones The structural and redox properties of menaquinones are key for their biological function. The quinone group of the menaquinones is reduced through a two-electron transport system to form a radical anion after the addition of the ﬁrst electron [29–31]. After addition of the second electron, a dianion catecholate is formed, which then forms a catechol in the presence of a protic solvent (Figure2). The two intermediates are su ﬃciently long-lived to be observed in aprotic solvents because the proton transfer step is slow. However, in a protic solvent the proton transfer is faster, and protonation of the intermediate radical ion occurs quickly, and the second intermediate is not Molecules 2020, 25, 4477 3 of 37

Molecules 2020, 25, x FOR PEER REVIEW 3 of 39 Molecules 2020, 25, x FOR PEER REVIEW 3 of 39 long-lived enough to be observed using electrochemistry. For MK-derivatives in cellular systems, to be observed using electrochemistry. For MK-derivatives in cellular systems, the presence of H2O the presenceto be observed of H2 Ousing renders electrochemistry. the redox chemistryFor MK-derivatives a one step in cellular process. systems, Even the though presence menaquinones of H2O renders the redox chemistry a one step process. Even though menaquinones are located in the are locatedrendersin the the redox membrane, chemistry H 2 aO one is an step accessible process. Even proton though source. menaquinones It is highly are likely located that in the the redox membrane, H2O is an accessible proton source. It is highly likely that the redox reaction involving reactionmembrane, involving H2O fewer is an intermediatesaccessible proton is importantsource. It is for highly the biologicallikely that propertiesthe redox reaction of these involving compounds. fewer intermediates is important for the biological properties of these compounds. Our group has Our groupfewer intermediates has been investigating is important thefor the redox biological properties properties of some of these menaquinone compounds. systems Our group and has found been investigating the redox properties of some menaquinone systems and found distinctly different distinctlybeen diinvestigatingﬀerent properties the redox in properties organic solvents of some wheremenaquinone the MK-derivatives systems and found are solubledistinctly [12 different], compared properties in organic solvents where the MK-derivatives are soluble [12], compared to aqueous to aqueous systems where the menaquinone will be interacting with a membrane or a model membrane systems where the menaquinone will be interacting with a membrane or a model membrane system, system, such as a liposome [32]. such as a liposome [32].

FigureFigure 1. 1.Common Common menaquinonesmenaquinones and and ubiquinones. ubiquinones. Figure 1. Common menaquinones and ubiquinones.

Figure 2. The redox reaction of menaquinones showing the two steps observed in aprotic organic

solvent; in the presence of a protic solvent it occurs in only one step, but the overall reaction is conserved. Molecules 2020, 25, 4477 4 of 37

The location of menaquinones in biological systems is associated with the membrane due to their hydrophobicity. These compounds have been presumed to reside inside the membrane; however, few studies have been done to investigate the specific interaction. Some computational studies have examined the association with the membrane because it is critical to understand the interaction for the electron transport action of menaquinones in bacterial systems [33]; however, less work has been done experimentally. A range of studies have been performed with ubiquinone derivatives, and evidence for the membrane association has been reported both in computational and experimental systems [34–36]. On the other hand, significantly less work has been done with the menaquinone system, but so far, the reports support the presumptions often that the menaquinone is associated with the membrane. Despite the polarity of the quinone group, these molecules are very hydrophobic, even though this property does vary with the length and the nature of the isoprene side group. We have recently found by synthesis of truncated MK-derivatives that only MK-1, MK-2, and MK-3, including derivatives with fully or partially saturated counterparts, are soluble in aqueous solution [12,37,38]. This means that assays with MK-4, even though it is known as an enzyme substrate in vivo, may not demonstrate enzyme activity even if the aqueous assay includes surfactants [39]. This experiment was attempted for MenJ, the membrane bound enzyme reported to stereospecifically hydrogenate the second isoprene unit of MK-9 in Mycobacterium tuberculosis and Mycobacterium smegmatis. The inability of the isolated enzyme to saturate MK-4 was attributed to the poor solubility of MK-4 in vitro since it was following the study that demonstrated in vivo activity. These experiments document the need for use of the truncated MK-derivatives and caution the use of computational methods for evaluation of these systems. For example, the anticancer properties of MK-1 through MK-7 were investigated with a series of five cancer cell lines [40]. MK-1, MK-2, and MK-3 were shown to be cytotoxic, however, MK-4, MK-5, MK-6, and MK-7 did not exhibit any cytotoxic activity against any of the cell five cell lines. Based on the studies with MenJ, it seems likely that the anticancer effects of longer MK-derivatives are a consequence of the physical properties of these derivatives, and any correlations and conclusions regarding the measured activities are not based on the true cytotoxicity of these compounds, but a reflection of their hydrophobicity. Due to the distinct hydrophobicity of menaquinones, assaying these substrates in vitro is often critical and or convenient in biological studies. Since assays are often done in aqueous solution, this can cause problems if the substrates are menaquinones with longer isoprene side chains, because the longer naturally occurring MK-derivatives are often insoluble in aqueous solution. Contrary to common belief, the addition of surfactants in such assays is often insufficient to solubilize these compounds. The approach we have used to overcome this challenge is to synthesize water soluble, truncated MK-substrates [41]. To this end, the small, truncated MK-derivatives, MK-1 through MK-3, are somewhat soluble in aqueous assays with added surfactants, which makes these MK-derivatives excellent substrates for biological studies. We have recently used this approach in studies of MenJ documenting the effectiveness of this approach [26].

1.2. Synthetic Strategy for the Preparation of Menaquinones Contemplating the many synthetic strategies available, consideration of the structure and the application of the MK-derivatives may be important when choosing the appropriate method for synthesis. So far, our group has focused on the preparation of truncated MK-derivatives, where the syntheses are relatively short and direct [12,37,38]. Therein, we used Friedel-Crafts alkylation using BF OEt as a Lewis acid catalyst. This use of electrophilic aromatic substitution is one of the most 3· 2 common across the literature for the synthesis of this family of compounds. This is most likely because it occurs over two steps from commercially available starting materials. For example, for our synthesis of MK-2, we begin by reducing menadione 1 with 10% Na2S2O4 at room temperature for 30 min to produce menadiol 2 in ~84% yield (Scheme1)[ 12]. Then menadiol 2 underwent alkylation with commercially available geraniol in the presence of BF OEt , producing MK-2 in 20.1% yield. 3· 2 Molecules 2020, 25, x FOR PEER REVIEW 5 of 39 Molecules 2020, 25, 4477 5 of 37 min to produce menadiol 2 in ~84% yield (Scheme 1) [12]. Then menadiol 2 underwent alkylation with commercially available geraniol in the presence of BF3∙OEt2, producing MK-2 in 20.1% yield. This approachThis approach may may be direct,be direct, but but it generallyit generally results results in low yields yields (< ( <30%)30%) and and a mixture a mixture of isomers. of isomers. AlthoughAlthough biological biological studies studies do do not not require require very very muchmuch sample, consistently consistently low low yielding yielding reactions reactions are are not suitablenot suitable for multi-step for multi-step syntheses syntheses at largeat large scales, scales such, such as as longer longer MK-derivatives MK-derivatives wherewhere thethe key step is a Friedel-Craftsstep is a Friedel alkylation.-Crafts alkylation.

Scheme 1. Friedel-Crafts alkylation using BF3 OEt2 and commercially available materials [12]. Scheme 1. Friedel-Crafts alkylation using BF3·∙OEt2 and commercially available materials [12].

Two reviewsTwo reviews [42, 43[42,43] have] have been been published published that that focusfocus on on the the synthesis synthesis of ofthe the biologically biologically relevant relevant vitaminvitamin K1. InK1 this. In this review, review, we we summarize summarize the the synthesessyntheses of of MK MK-derivatives-derivatives reported reported in recent in recent history. history. This includesThis includes menaquinones menaquinones of of varying varying side chain chain lengths, lengths, which which include include many many types typesof vitamin of vitamin K2 K2 derivatives,derivatives, as as well well as as vitamin vitamin K1.. T Thishis review review will will highlight highlight the the five ﬁve most most common common synthetic synthetic strategiesstrategies used forused preparation for preparation of truncated of truncated and full-length and full-length MK-derivatives: MK-derivatives: nucleophilic nucleophilic ring methods, ring metal-mediatedmethods, metal and-mediated radical reactions, and radical electrophilic reactions, electrophilic ring methods, ring methods, pericyclic pericyclic reactions, reactions, and side and chain extensionsside chain (Figure extensions3). Each (Figure class of3) reactions. Each class will of be reactions summarized, will be identifying summarized, the identifying advantages the and advantages and disadvantages of each. This will allow an evaluation of each strategy based on the disadvantages of each. This will allow an evaluation of each strategy based on the overall yields overall yields of the synthesis, regioselectivity, and the stereoretention of the first isoprene (α) unit of the synthesis, regioselectivity, and the stereoretention of the ﬁrst isoprene (α) unit from the ring. from the ring. Other comparisons concerning the number of steps, competing side reactions, and Othersafety comparisons will also be concerning presented, thewhich number will result of steps, in a summary competing identifying side reactions, the most andattractive safety synthetic will also be presented,strategies which from will each result category ina for summary preparation identifying of this class the of mostcompounds. attractive synthetic strategies from eachMolecules category 2020 for, 25 preparation, x FOR PEER REVIEW of this class of compounds. 6 of 39

FigureFigure 3. 3. 55 Main Main synthetic synthetic strategies strategies from across from the across literature the for literature the synthesis for of the menaquinone synthesis of menaquinonederivatives. derivatives.

2. Nucleophilic Ring Methods

2.1. Enolate Alkylation In 1974, Snyder and Rapoport published a comprehensive report detailing their attempts to synthesize menaquinones [44]. The main goal of their syntheses was to retain the stereochemistry of α-isoprene double bond. The first approach used enolate chemistry to alkylate the C3 position of menadiol 2. Beginning with menadiol 3a, the potassium salt 4a was formed using potassium hydride or potassium methoxide (Scheme 2). It was suspected that Claisen alkylation had occurred, but upon further analysis it was concluded that the reaction proceeded via enolate alkylation to form 5a with 97:3 E/Z. Oxidation of the ring in the presence of Ag2O formed MK-2 in an average of 20% yield over three steps. The authors postulated the lower yield was due to competing Friedel-Crafts alkylation occurring on the C2 position. To prevent competition, the authors redeveloped the route using C1 methyl ether protected menadiol 3b. Conversion of 3b to MK-2 yielded 45% with 97% E alkene (Scheme 2). Despite the improved yield, the synthesis of 3b was more complicated than the authors originally thought. Therefore, this route was abandoned for one with more accessible starting materials.

Molecules 2020, 25, 4477 6 of 37

2. Nucleophilic Ring Methods

2.1. Enolate Alkylation In 1974, Snyder and Rapoport published a comprehensive report detailing their attempts to synthesize menaquinones [44]. The main goal of their syntheses was to retain the stereochemistry of α-isoprene double bond. The ﬁrst approach used enolate chemistry to alkylate the C3 position of menadiol 2. Beginning with menadiol 3a, the potassium salt 4a was formed using potassium hydride or potassium methoxide (Scheme2). It was suspected that Claisen alkylation had occurred, but upon further analysis it was concluded that the reaction proceeded via enolate alkylation to form 5a with 97:3 E/Z. Oxidation of the ring in the presence of Ag2O formed MK-2 in an average of 20% yield over three steps. The authors postulated the lower yield was due to competing Friedel-Crafts alkylation occurring on the C2 position. To prevent competition, the authors redeveloped the route using C1 methyl ether protected menadiol 3b. Conversion of 3b to MK-2 yielded 45% with 97% E alkene (Scheme2). Despite the improved yield, the synthesis of 3b was more complicated than the authors originally thought. Therefore,Molecules this2020, 2 route5, x FOR PEER was REVIEW abandoned for one with more accessible starting7 of materials.39

SchemeScheme 2. Synthesis 2. Synthesis of of MK-2 MK-2 via via enolate enolate alkylation alkylation [44]. [44].

A few yearsA later, few years Tabushi later, Tabushi et al. et al. published published on theon allylation the allylation of naphthoquinones of naphthoquinones using β- using cyclodextrin as an inclusion catalyst, especially for synthesis of vitamin K1 [45,46]. Menadiol 2 was β-cyclodextrinalkylated as an inclusionat the C3 position catalyst, in mild especially basic conditions, for borate synthesis buffer (pH of 9), vitamin with and without K1 [45 β,-46]. Menadiol 2 was alkylatedcyclodextrin at the C3 (Scheme position 3A). The in yields mild of MK basic-1 were conditions, found to be 40% borate and 15%, bu respectively,ﬀer (pH showing 9), with and without β a significant decrease in yield without β-cyclodextrin. The only byproduct observed was menadione -cyclodextrin (Scheme1 in 49% and3 A).28% yield, The respectively. yields of MK-1 were found to be 40% and 15%, respectively, showing a signiﬁcant decreaseThe results in yield strongly without indicateβ that-cyclodextrin. β-cyclodextrin played The significant only byproduct role catalyzing observed this reaction. was menadione 1 in 49% and 28%The yield, authors respectively. noted the selectivity of the alkylation behaved as if it were a ligase and or oxidase. Due Moleculesto the semi 2020-, conical25, x FOR structure PEER REVIEW of β -cyclodextrin, menadiol is surrounded by the hydrophobic cavity.8 of 39 The C1 and C4 hydroxy groups interact with the hydrophilic exterior (Scheme 3B). The C1 hydroxy group hypothesized to be hydrogen bonding on the narrower end of the scaffold. The hydrogen bonding interactions were found to lower the pKa of the opposite hydroxy group at the C4 position. Unbound menadiol has a pKa of 9.45, and bound menadiol was calculated to have a pKa of ~8.90 [46]. The decrease in pKa was said to enhance the nucleophilicity of the partially charged carbanion on the C3 position. Deprotonation of the C4 hydroxy group in pH 9 medium created a more nucleophilic enolate, alkylating at the C3 position. The authors did not observe any C2-alkylated product in their trials. They postulated the sterically hindered, semi-conical shape of the β- cyclodextrin scaffold prevented C2 alkylation (Scheme 3B). Menadione 1 was the only other product formed. In both trials, 1 was produced in higher yields than MK-1. In protic solvents, menadiol 2 will spontaneously oxidize to form menadione 1. In the presence of an enzyme-like cavity of β- cyclodextrin, the transformation was most likely accelerated in the buffer solution. Although menadione 1 was formed in higher yields, it can be recovered and recycled for subsequent use.

Scheme 3. (A)Scheme Synthesis 3. (A) Synthesis of MK-1 of MKusing-1 usingβ-cyclodextrin β-cyclodextrin (β-CD) (β-CD) as an inclusionas an inclusion catalyst. (B) catalyst.Steric (B) Steric hinderance (shownhinderance in bold) (shown of in thebold)β of-cyclodextrin the β-cyclodextrin scascaffoldﬀold preventing preventing C2 alkylation C2 alkylation [45,46]. [45,46]. 2.2. Transmetalation After marginal success with enolate alkylations, Snyder and Rapoport shifted their focus towards more direct nucleophilic methods [44]. Bismethyl ether 2-bromomenadiol 6 was transformed into 2-metallo derivatives using lithium, magnesium, and copper to react with a variety of electrophiles (Scheme 4). The authors were originally interested in using aldehydes as electrophiles; however, their attempts were unsuccessful. Removal of the resultant benzyl alcohol led to the formation of a vinyl alkene or isomerization of the α-isoprene double bond. To avoid this, prenyl halide substrates were used instead. Preliminary reactions were performed to assess the stereoretention of the α-isoprene double bond for each 2-metallo derivative. All three 2-metallo derivatives left the α-isoprene double bond virtually unaffected. Organolithium yields were 10% and 65% for geranyl chloride and bromide, respectively, leading to the exclusive use of prenyl bromide electrophiles. Lithium organocuprate and organocuprate yields were 74% and 82%, respectively. The Grignard reagent was formed in 92% yield. Snyder and Rapoport continued their studies using the Grignard reagent. To examine the utility of this method, the authors synthesized MK-2 and MK-9 (Scheme 4). The Grignard reagent was produced by stirring 6 with Mg turnings in dry THF. Geranyl bromide and solanesyl bromide were added to the solution, resulting in the respective alkylated products, 7a and 7b. Removal of the bismethyl ethers and oxidation of the ring were achieved with AgO in acidic conditions yielding 80% and 73% for MK-2 and MK-9, respectively. Both products retained the stereochemistry of the α-isoprene double bond, each with ≥ 98% E alkene.

Scheme 4. Synthesis of MK-2 and MK-9 using Grignard reagents [44].

Unlike Snyder and Rapoport, Saá and coworkers were successful using aldehydes as electrophiles in the early 1990s. This challenge was overcome with the use of stereocontrolled Birch hydrogenolysis conditions (BIHY) to selectively remove silyl ether protected (E)- or (Z)-α-

Molecules 2020, 25, x FOR PEER REVIEW 8 of 39

Molecules 2020, 25, 4477 7 of 37

The results strongly indicate that β-cyclodextrin played significant role catalyzing this reaction. The authors noted the selectivity of the alkylation behaved as if it were a ligase and or oxidase. Due to the semi-conical structure of β-cyclodextrin, menadiol is surrounded by the hydrophobic cavity. The C1 and C4 hydroxy groups interact with the hydrophilic exterior (Scheme3B). The C1 hydroxy group hypothesized to be hydrogen bonding on the narrower end of the scaffold. The hydrogen bonding interactions were found to lower the pKa of the opposite hydroxy group at the C4 position. Unbound menadiol has a pKa of 9.45, and bound menadiol was calculated to have a pKa of ~8.90 [46]. The decreaseScheme 3 in. ( pKaA) Synthesis was said of to MK enhance-1 using the β- nucleophilicitycyclodextrin (β-CD) of the as partially an inclusion charged catalyst. carbanion (B) Steric on the C3 position.hinderance Deprotonation (shown in bold) of of the the C4β-cyclodextrin hydroxy group scaffold in preventing pH 9 medium C2 alkylation created [45,46 a more]. nucleophilic enolate, alkylating at the C3 position. The authors did not observe any C2-alkylated product in 2.2.their Transmetalation trials. They postulated the sterically hindered, semi-conical shape of the β-cyclodextrin scaffold preventedAfter C2 marginal alkylation success (Scheme with3B). enolate Menadione alkylations,1 was the Snyder only other and Rapoportproduct formed. shifted In their both focustrials, 1towardswas produced more direct in higher nucleophilic yields than methods MK-1. [44] In.protic Bismethyl solvents, ether menadiol 2-bromomenadiol2 will spontaneously 6 was transformed oxidize intoto form 2-metallo menadione derivatives1. In the using presence lithium, of an enzyme-like magnesium, cavity and of copperβ-cyclodextrin, to react withthe transformation a variety of waselectrophiles most likely (Scheme accelerated 4). The in authors the buffer were solution. originally Although interested menadione in using1 was aldehydes formed as in electrophiles; higher yields, however,it can be recovered their attempts and recycled were unsuccessful. for subsequent Removal use. of the resultant benzyl alcohol led to the formation of a vinyl alkene or isomerization of the α-isoprene double bond. To avoid this, prenyl 2.2. Transmetalation halide substrates were used instead. Preliminary reactions were performed to assess the stereoretentionAfter marginal of the success α-isoprene with enolate double alkylations, bond for Snyder each 2- andmetallo Rapoport derivative. shifted All their three focus 2- towardsmetallo derivamore directtives left nucleophilic the α-isoprene methods double [ 44bond]. virtually Bismethyl unaffected. ether 2-bromomenadiol Organolithium yields6 was were transformed 10% and 65%into for 2-metallo geranyl derivativeschloride and using bromide, lithium, respectively, magnesium, leading and to the copper exclusive to react use withof prenyl a variety bromide of electrophiles.electrophiles (SchemeLithium 4organocuprate). The authors and were organocuprate originally interested yields w inere using 74% aldehydesand 82%, respectively. as electrophiles; The Grignardhowever, theirreagent attempts was formedwere unsuccessful. in 92% yield. Removal Snyder ofand the Rapoport resultant continued benzyl alcohol their led studies to the formationusing the Grignardof a vinyl alkenereagent. or isomerization of the α-isoprene double bond. To avoid this, prenyl halide substrates wereTo used examine instead. the Preliminary utility of this reactions method, were the performed authors synthesized to assess the MK stereoretention-2 and MK-9 of(Scheme the α-isoprene 4). The Grignarddouble bond reage fornt each was 2-metallo produced derivative. by stirring All 6 with three Mg 2-metallo turnings derivatives in dry THF. left theGeranylα-isoprene bromide double and solanesylbond virtually bromide una ffwereected. added Organolithium to the solution, yields resulting were 10% in andthe 65%respective for geranyl alkylated chloride products, and bromide, 7a and 7brespectively,. Removal leadingof the bismethyl to the exclusive ethers use and of oxidation prenyl bromide of the electrophiles.ring were achieved Lithium with organocuprate AgO in acidic and conditionsorganocuprate yielding yields 80% were and 74% 73% and 82%,for MK respectively.-2 and MK The-9, Grignardrespectively reagent. Both was products formed retainedin 92% yield. the stereochemistrySnyder and Rapoport of the continuedα-isoprene their double studies bond using, each the with Grignard ≥ 98% E reagent. alkene.

Scheme 4 4.. SynthesisSynthesis of of MK MK-2-2 and and MK MK-9-9 using Grignard reagents [44] [44]..

UnlikTo examinee Snyder the utilityand Rapoport, of this method, Saá and the coworkers authors synthesized were successful MK-2 and using MK-9 aldehydes (Scheme 4 as). electrophilesThe Grignard in reagent the early was 1990s. produced This bychallenge stirring was6 with overcome Mg turnings with inthe dry use THF. of stereocontrolled Geranyl bromide Birch and hydrogenolysolanesyl bromidesis conditions were added (BIHY) to the solution, to selectively resulting remove in the respective silyl ether alkylated protected products, (E)- 7aor and(Z)-7bα-. Removal of the bismethyl ethers and oxidation of the ring were achieved with AgO in acidic conditions yielding 80% and 73% for MK-2 and MK-9, respectively. Both products retained the stereochemistry of the α-isoprene double bond, each with 98% E alkene. ≥ Unlike Snyder and Rapoport, Saá and coworkers were successful using aldehydes as electrophiles in the early 1990s. This challenge was overcome with the use of stereocontrolled Birch hydrogenolysis conditions (BIHY) to selectively remove silyl ether protected (E)- or (Z)-α-alkenylbenzyl alcohols [47–49]. The authors applied this method towards the synthesis of MK-2 and MK-4 (Scheme5). Using the Molecules 2020, 25, 4477 8 of 37 Molecules 2020, 25, x FOR PEER REVIEW 9 of 39

alkenylbenzyl alcohols [47–49]. The authors applied this method towards the synthesis of MK-2 and same starting material as Snyder and Rapoport, bismethyl ether 2-bromomenadiol 6 underwent MK-4 (Scheme 5). Using the same starting material as Snyder and Rapoport, bismethyl ether 2- lithium-bromide exchange to form the organolithium reagent. Commercially available aldehydes, citral bromomenadiol 6 underwent lithium-bromide exchange to form the organolithium reagent. 8a (68:32 E/Z) and geranylgeranial 8b ( 95% E), were used without further puriﬁcation to assess the Commercially available aldehydes, citral≥ 8a (68:32 E/Z) and geranylgeranial 8b (≥ 95% E), were used stereoretentivewithout further abilities purification of BIHY to conditions. assess the Upon stereoretentive nucleophilic abilities attack, of the BIHY resultant conditions benzylic. Upon alcohols 9anucleophilicand 9b were attack, formed the in resultant ~65% and benzylic 58% yield, alcohols respectively. 9a and 9b The were alcohols formed were in ~65% protected and 58% with yield, TMSCl inrespectively. the presence The of alcohols HMDS, were as used protected in previous with TMSC reportsl in [ 48the,49 presence]. In the of literature, HMDS, as the used authors in previous used the abbreviationreports [48,49 “HMDSZ”]. In the literature, instead the of“HMDS”, authors used which the isabbreviation now more “HMDSZ” the commonly instead used of abbreviation.“HMDS”, Thewhich protected is now alcohol more the was commonly then reduced usedin abbreviation the presence. The of lithiumprotected metal alcohol and liquidwas then ammonia reduced to in produce the thepresence free methylene of lithium10 metal. With and this liquid method, ammonia the to resulting produce stereochemistrythe free methylene of 1010a. Withand this10b method,α-isoprene doublethe resulting bonds stereochemistry were determined of to10 bea and 2:1 10bE/Z αand-isoprene95% Edoublealkene, bonds respectively. were determined These results to be 2:1 reﬂect E/Z the ≥ stereochemistryand ≥95% E alkene, of the respectively. aldehyde precursorsThese results and reflect demonstrate the stereochemistry the remarkable of the stereoretentionaldehyde precursors of BIHY. Theand methyl demonstrate ether protectingthe remarkable groups stereoretention were removed of usingBIHY. CANThe methyl to produce ether MK-2protecting and groups MK-4. Onlywere the reportedremoved yield using was CAN for to MK-4 produce in 86%. MK-2 and MK-4. Only the reported yield was for MK-4 in 86%.

Scheme 5. Synthesis of MK-2 and MK-4 using BIHY conditions [47–49].

Further improvementScheme 5.upon Synthesis the of strategies MK-2 and established MK-4 using BIHY by Snyder conditions and [47 Rapoport–49]. was reported by Swenton and coworkers in 1977 [50]. Therein, preliminary results were published using electrolysis to Further improvement upon the strategies established by Snyder and Rapoport was reported by protect bismethyl 2-bromomenadiol 6 as bismethyl ketals. The authors used lithium organocuprate Swenton and coworkers in 1977 [50]. Therein, preliminary results were published using electrolysis nucleophiles instead of Grignard reagents. In 1980, the authors published a complete study towards to protect bismethyl 2-bromomenadiol 6 as bismethyl ketals. The authors used lithium organocuprate thenucleophiles synthesis ofinstead menaquinones of Grignard [51 reagents]. Bismethyl. In 1980, ether the 2-bromomenadiolauthors published a6 completeunderwent study electrolysis towards in athe divided synthesis cell of with menaquinones 2% KOH and [51]. methanol, Bismethyl ether producing 2-bromomenadiol bisketal 11 6(Scheme underwent6). Lithium-bromideelectrolysis in a exchangedivided incell THF with and 2% subsequent KOH and transmetalation methanol, producing with cuprous bisketal iodide11 (Scheme produced 6). Lithium the desired-bromide lithium organocuprateexchange in THF dimer and 12subsequent. The corresponding transmetalation electrophiles, with cuprous prenyl iodide bromide produced13a theand desired geranyl lithium bromide 13borganocuprat, were addede dimer to the 12 solution. The corresponding and immediately electrophiles, carried prenyl forward bromide to hydrolysis 13a and without geranyl puriﬁcation.bromide MK-113b, were and MK-2added wereto the formed solution in and 92% immediately and 96% yield, carried respectively. forward to hydrolysis The authors without did not purification. observe any evidenceMK-1 and of MK the-Z2 isomerwere formed in NMR; in 92% however, and 96% its yield, absence respectively. could not The be concluded. authors did not observe any evidence of the Z isomer in NMR; however, its absence could not be concluded.

Molecules 2020, 25, 4477 9 of 37 Molecules 2020, 25, x FOR PEER REVIEW 10 of 39

Scheme 6. Synthesis of MK-1 and MK-2 featuring electrolysis as a protection method [50,51]. Scheme 6. Synthesis of MK-1 and MK-2 featuring electrolysis as a protection method [50,51].

2.3. Friedel-Crafts2.3. Friedel-Crafts Alkylation Alkylation Throughout all the strategies described in this report, the Friedel-Crafts alkylation is by far the Throughout all the strategies described in this report, the Friedel-Crafts alkylation is by far the most most popular across the literature. The most common Lewis acid catalyst used for this specific popular across the literature. The most common Lewis acid catalyst used for this specific transformation transformation is BF3∙OEt2, which became popular after Lindlar’s 1953 patent detailing its usage[52]. is BF OEt , which became popular after Lindlar’s 1953 patent detailing its usage [52]. Two more Two3· more2 reports of its use specifically towards the synthesis of vitamin K1 were published in the reportsfollowing of its use year. specifically First, Isler towardsand Doebel the detailed synthesis the ofsynthesis vitamin of K vitamin1 were K published1 and various in theother following racemic year. First,derivatives Isler and Doebel[53]. Second, detailed Hirschmann the synthesis et al. reported of vitamin the comparison K1 and various of different other racemicLewis and derivatives Brønsted- [53]. Second,Lowry Hirschmann acid catalysts et towards al. reported the synthesis the comparison of vitamin K of1 [54] diff erentin an effort Lewis to andimprove Brønsted-Lowry upon existing acid catalystsreported towards methods the atsynthesis the timeof [55 vitamin–61]. Most K1 notably[54] in an among effort those to improve methods, upon Fieser existing reported reported the methodscondensation at the time of menadiol [55–61]. and Most phytol notably in the among presence those of oxalic methods, acid with Fieser overall reported yields the of 25 condensation–30% [55]. of menadiolMuch and of the phytol yield in was the lostpresence to undesired of oxalic side acid products, with overall such yields as phytadiene of 25–30% and [55 the]. Much C2-alkylated of the yield product [62]. Many of the early reports do not use protecting groups or other functional handles to was lost to undesired side products, such as phytadiene and the C2-alkylated product [62]. Many of induce regioselectivity. the early reportsTo address do not the use regioselectivity protecting groups challenge, or other Hirschmann functional et al. handles designed to induce monoacetate regioselectivity. 14 to Toinfluence address C3 the alkylation regioselectivity (Scheme challenge, 7). Monoacetate Hirschmann 14 underwent et al. designed condensation monoacetate with phytol14 to in influence the C3 alkylationpresence of (Scheme an acid 7catalyst). Monoacetate to form alkylated14 underwent product condensation15. Potassium acid with sulfate, phytol oxalic in the acid, presence Duolite of an acid catalystC-60 cation to exchange form alkylated resin, and product BF3∙OEt152 produced. Potassium varied acid results sulfate,, as outlined oxalic acid,in Table Duolite 1. Removal C-60 of cation exchangethe acetate resin, protecting and BF OEtgroupproduced was achieved varied using results, Claisen as alkali outlined conditions in Table (dilute1. Removal KOH in methanol), of the acetate 3· 2 protectingand oxidation group was with achieved Ag2O wo usingrk up Claisenformed vitamin alkali conditions K1. (dilute KOH in methanol), and oxidation withMolecules Ag2 O2020 work, 25, x upFOR formed PEER REVIEW vitamin K1. 11 of 39 Table 1. Yields for different acid catalysts used in the synthesis of vitamin K1 [54].

Acid Catalyst % Yield 1 KHSO4 55% Oxalic Acid N/A 2 Duolite C-60 8% BF3∙OEt2 66.5% 1 Over two steps. 2 No yield reported.

SchemeScheme 7. 7.Synthesis Synthesis ofof vitaminvitamin KK11 using Friedel Friedel-Crafts-Crafts alkylation alkylation with with different diﬀerent acid acid catalysts catalysts [54] [54. ].

In 1990, Schmid et al. published a comprehensive report on the synthesis and analysis of all four stereoisomers of (E)-vitamin K1 [63]. The synthesis of the naturally occurring stereoisomer showcased a unique transformation using BF3∙OEt2 as the catalyst (Scheme 8). Starting with monoacetate 14, the free phenol was alkylated with phytyl chloride 16 and potassium carbonate in 79–87% yield with 99.5% E alkene. In the presence of BF3∙OEt2, the O-alkylated product 15 was thought to have undergone a Claisen rearrangement to form C-alkylated product 17, but upon further analysis, the reaction was determined to proceed via and intramolecular Friedel-Crafts mechanism. The C- alkylated product 17 was formed in 76–80% yield and the E/Z ratio was found to be 97:3. This kind of intramolecular transformation was first reported by Yoshizawa et al. in 1982 on ubiquinone derivatives [64]. Removal of the acetyl group in basic conditions produced vitamin K1 in 96.5% yield based on HPLC. The E/Z ratio of the α-isoprene double bond was determined to be 97:3, unchanged from the previous step. In 2003, the authors of a vitamin K1 syntheses review [42] commented that this route has received little attention.

Scheme 8. Synthesis of vitamin K1 featuring an intramolecular Friedel-Crafts alkylation [63].

Allyl alcohols have a reputation for being intrinsically unstable towards alcohol rearrangement. To circumvent this issue, Min et al. used prenyl chlorides instead [65]. Menadiol dimethyl ether 18 was reacted with prenyl chloride 19 in presence of BF3∙OEt2 to form sulfonyl intermediate 20 (Scheme 9). Truncated prenyl chain 19 allowed for further functionalization of the side chain to make more diverse analogs, which will be discussed in detail later in Section 6.2. The authors assessed the efficacy of a large scope of Lewis acids, which are summarized in Table 2. Based on these results, AlCl3 performed the best with 72% yield and all E configuration of the α-isoprene double bonds.

Molecules 2020, 25, x FOR PEER REVIEW 11 of 39 Molecules 2020, 25, 4477 10 of 37

Table 1. Yields for diﬀerent acid catalysts used in the synthesis of vitamin K1 [54].

Acid Catalyst % Yield 1

KHSO4 55% Oxalic Acid N/A 2 Duolite C-60 8% BF OEt 66.5% 3· 2 1 Over two steps. 2 No yield reported.

In 1990, Schmid et al. published a comprehensive report on the synthesis and analysis of all four stereoisomers ofScheme (E)-vitamin 7. Synthesis K of1 vitamin[63]. The K1 using synthesis Friedel-Crafts of the alkylation naturally with different occurring acid catalysts stereoisomer [54]. showcased a unique transformation using BF OEt as the catalyst (Scheme8). Starting with monoacetate 14, 3· 2 the free phenolIn was1990, alkylatedSchmid et al. with published phytyl a compr chlorideehensive16 reportand potassiumon the synthesis carbonate and analysis in 79–87%of all four yield with stereoisomers of (E)-vitamin K1 [63]. The synthesis of the naturally occurring stereoisomer showcased 99.5% E alkene. In the presence of BF3 OEt2, the O-alkylated product 15 was thought to have undergone a unique transformation using BF·3∙OEt2 as the catalyst (Scheme 8). Starting with monoacetate 14, the a Claisen rearrangementfree phenol was alkylated to form with C-alkylated phytyl chloride product 16 and17 potassium, but upon carbonate further in analysis,79–87% yield the with reaction was determined99.5% to proceedE alkene. via In the and presence intramolecular of BF3∙OEt2,Friedel-Crafts the O-alkylated productmechanism. 15 was The thought C-alkylated to have product 17 was formedundergone in 76–80% a Claisen yieldrearrangement and the toE form/Z ratio C-alkylated was found product to 17 be, but 97:3. upon This further kind analysis, of intramolecular the reaction was determined to proceed via and intramolecular Friedel-Crafts mechanism. The C- transformationalkylated was product ﬁrst reported17 was formed by Yoshizawa in 76–80% yield et al. and in the 1982 E/Z onratio ubiquinone was found to derivatives be 97:3. This kind [64]. Removal of the acetylof intramolecular group in basic transformation conditions was produced first reported vitamin by Yoshizawa K1 in 96.5% et al. yield in 1982 based on ubiquinone on HPLC. The E/Z ratio of thederivativesα-isoprene [64]. Removal double of bond the acetyl was group determined in basic conditions to be 97:3, produced unchanged vitamin K from1 in 96.5% the yield previous step. based on HPLC. The E/Z ratio of the α-isoprene double bond was determined to be 97:3, unchanged In 2003, the authors of a vitamin K1 syntheses review [42] commented that this route has received from the previous step. In 2003, the authors of a vitamin K1 syntheses review [42] commented that little attention.this route has received little attention.

Molecules 2020, 25, x FOR PEER REVIEW 12 of 39

Table 2. Friedel-Crafts alkylation of 18 with sulfonyl 19 to form 20 [65]. Scheme 8.SchemeSynthesis 8. Synthesis of vitamin of vitamin K1 Kfeaturing1 featuring an an intramolecular intramolecular Friedel Friedel-Crafts-Crafts alkylation alkylation [63]. [63]. Lewis Acid 1 % of 20 (E/Z) Allyl alcohols have a reputation for being intrinsically unstable towards alcohol rearrangement. Allyl alcohols have a reputationBF3∙OEt for2 being intrinsically0 (-) unstable towards alcohol rearrangement. To circumvent this issue, Min et al. used prenyl chlorides instead [65]. Menadiol dimethyl ether 18 To circumvent this issue, Min et al. usedMgBr2prenyl chlorides0 (- instead) [65]. Menadiol dimethyl ether 18 was was reacted with prenyl chlorideTiCl 19 in4 presence of BF3∙OEt- 2(-2) to form sulfonyl intermediate 20 (Scheme reacted with9). Truncated prenyl chloride prenyl cha19in in19 allowed presenceFeCl3 for offurther BF functionalizationOEt55 (4:1)to form sulfonylof the side intermediate chain to make more20 (Scheme 9). 3· 2 Truncateddiverse prenyl analogs, chain which19 allowed will be discussed forEt2AlCl further in detail functionalization later in56 S(7:1)ection 6.2. of The the authors side chain assessed to the make efficacy more diverse SnCl4 56 (E) analogs, whichof a large will scope be discussed of Lewis acids, in detail which arelater summarized in Section in 6.2 Table. The 2. Based authors on these assessed results, the AlCl eﬃ3 cacy of a performed the best with 72% yieldZnBr and2 all E configuration60 (7:1) of the α-isoprene double bonds. large scope of Lewis acids, which areZnCl summarized2 in Table67 (7:1)2 . Based on these results, AlCl 3 performed the best with 72% yield and all E conﬁgurationAlCl3 of the α72-isoprene (E) double bonds. 1.1.2 equiv of Lewis acid was used. 2.Decomposition of the starting materials was observed.

SchemeScheme 9. Synthesis 9. Synthesis of truncated of truncated MK-derivatives MK-derivatives using using Friedel Friedel-Crafts-Crafts alkylation alkylation [65]. [65].

Despite the historical popularity of BF3∙OEt2, the competition between C2/C3 alkylation continued to be a persistent challenge. Various types of protecting groups have been used to prevent competition with varying success. Due to competing side reactions, its use in industrial applications has been accused of being wasteful, inspiring researchers to identify new, more sustainable Lewis acids to minimize byproducts and maximize yield. Coman et al. developed a new class of heterogenous, partly hydroxylated magnesium and aluminum fluorides to address such concerns [66]. The authors predict that this class of catalysts will replace homogenous BF3∙OEt2 in industrial applications. Vitamin K1 was synthesized to illustrate this concept starting with menadiol 2 and isophytol 21 (Scheme 10). The reactions resulted in 100% conversion of the starting material with each catalyst. The yields were low and showed considerable formation of byproducts, mainly chromanol and C2-alkylated product, as shown in Table 3. The catalyst, MgF2-57 provided the best results with respect to vitamin K1, yielding 26.5%, which is comparable to BF3∙OEt2.

Table 3. The catalytic results in the synthesis of vitamin K1, K1-chromanol, and C2-alkylated product from menadiol [66].

Catalyst % of Vitamin K1 % of K1 Chromanol % of C2 Product MgF2-40 21.2 5.9 58.8 MgF2-57 26.5 21.0 43.7 MgF2-71 15.6 20.8 52.9 MgF2-87 1 0 0 0 AlF2-50 7.6 42 41.2 1 0% conversion.

Scheme 10. Synthesis of vitamin K1 using partly hydroxylated magnesium and aluminum fluorides [66].

Molecules 2020, 25, x FOR PEER REVIEW 12 of 39

Table 2. Friedel-Crafts alkylation of 18 with sulfonyl 19 to form 20 [65].

Lewis Acid 1 % of 20 (E/Z) BF3∙OEt2 0 (-) MgBr2 0 (-) TiCl4 - 2(-) FeCl3 55 (4:1) Et2AlCl 56 (7:1) SnCl4 56 (E) ZnBr2 60 (7:1) ZnCl2 67 (7:1) AlCl3 72 (E) 1.1.2 equiv of Lewis acid was used. 2.Decomposition of the starting materials was observed.

Molecules 2020, 25, 4477 11 of 37

Table 2. Friedel-Crafts alkylation of 18 with sulfonyl 19 to form 20 [65].

Scheme 9. Synthesis of truncatedLewis Acid MK-1derivatives% of using 20 (E /FriedelZ) -Crafts alkylation [65].

BF3 OEt2 0 (-) Despite the historical popularity · of BF3∙OEt2, the competition between C2/C3 alkylation MgBr2 0 (-) continued to be a persistent challenge. Various types of protecting2 groups have been used to prevent TiCl4 - (-) competition with varying success. DueFeCl to competing3 side55 (4:1)reactions, its use in industrial applications has been accused of being wasteful, inspiringEt2AlCl researchers 56 (7:1) to identify new, more sustainable Lewis acids to minimize byproducts and SnCl maximize4 yield. 56 Coman (E) et al. developed a new class of heterogenous, partly hydroxylated magnesiumZnBr2 and aluminum60 (7:1) fluorides to address such concerns ZnCl2 67 (7:1) [66]. The authors predict that this class of catalysts will replace homogenous BF3∙OEt2 in industrial AlCl3 72 (E) applications. Vitamin K1 was synthesized to illustrate this concept starting with menadiol 2 and 1. 1.2 equiv of Lewis acid was used. 2. Decomposition of the starting materials was observed. isophytol 21 (Scheme 10). The reactions resulted in 100% conversion of the starting material with each catalyst.Despite The theyields historical were low popularity and showed of BF3 considerableOEt2, the competition formation between of byproducts, C2/C3 alkylation mainly chromanol continued · toand be C2 a persistent-alkylated challenge. product, as Various shown types in Table of protecting 3. The catalyst, groups MgF have2- been57 provided used to preventthe best competitionresults with withrespect varying to vitamin success. K1, yielding Due to competing26.5%, which side is reactions,comparable its to use BF3 in∙OEt industrial2. applications has been accused of being wasteful, inspiring researchers to identify new, more sustainable Lewis acids to minimizeTable byproducts3. The catalytic and results maximize in the synthesis yield. Coman of vitamin et al.K1, K developed1-chromanol a, and new C2 class-alkylated of heterogenous, product partlyfrom hydroxylated menadiol [66] magnesium. and aluminum ﬂuorides to address such concerns [66]. The authors predict that this class of catalysts will replace homogenous BF3 OEt2 in industrial applications. Vitamin Catalyst % of Vitamin K1 % of K1 Chromanol· % of C2 Product K1 was synthesizedMgF to illustrate2-40 this21.2 concept starting with5.9 menadiol 2 and58.8 isophytol 21 (Scheme 10). The reactions resultedMgF in2-57 100% conversion26.5 of the starting21.0 material with each43.7 catalyst. The yields were low and showed considerableMgF2-71 formation15.6 of byproducts,20.8 mainly chromanol52.9 and C2-alkylated product, 1 as shown in TableMgF3. The2-87 catalyst, MgF0 2-57 provided0 the best results with0 respect to vitamin K1, AlF2-50 7.6 42 41.2 yielding 26.5%, which is comparable to BF3 OEt2. ·1 0% conversion.

Scheme 10. Synthesis of vitamin K using partly hydroxylated magnesium and aluminum ﬂuorides [66]. Scheme 10. Synthesis of vitamin K1 using partly hydroxylated magnesium and aluminum fluorides [66].

Table 3. The catalytic results in the synthesis of vitamin K1,K1-chromanol, and C2-alkylated product from menadiol [66].

Catalyst % of Vitamin K1 % of K1 Chromanol % of C2 Product

MgF2-40 21.2 5.9 58.8 MgF2-57 26.5 21.0 43.7 MgF2-71 15.6 20.8 52.9 1 MgF2-87 0 0 0 AlF2-50 7.6 42 41.2 1 0% conversion.

Recently, we synthesized fully and partially saturated MK-derivatives in an effort to understand their structural and electrochemical properties in model membranes [12,37,38]. For the synthesis of MK-2(II-H2), the condensation of isophytol 22 and menadiol 2 was accomplished in 11% yield using MgF2-48, a Coman et al. inspired catalyst (Scheme 11). Before purification, the crude yield was determined to be 60%, consisting of regio- and stereoisomers. The regioisomers separated readily via column chromatography; however, the E/Z isomers of the C3-product required thorough purification using preparative TLC. Therefore, these stereoisomers are purified on demand for the biological studies. Molecules 2020, 25, x FOR PEER REVIEW 13 of 39

Recently, we synthesized fully and partially saturated MK-derivatives in an effort to understand their structural and electrochemical properties in model membranes [12,37,38]. For the synthesis of MK-2(II-H2), the condensation of isophytol 22 and menadiol 2 was accomplished in 11% yield using MgF2-48, a Coman et al. inspired catalyst (Scheme 11). Before purification, the crude yield was determined to be 60%, consisting of regio- and stereoisomers. The regioisomers separated readily via column chromatography; however, the E/Z isomers of the C3-product required thorough purification

usingMolecules preparative2020, 25, 4477 TLC. Therefore, these stereoisomers are purified on demand for the biological12 of 37 studies.

Scheme 11. Synthesis of MK-2(II-HMK-2(II-H2)) using using using MgF2--4848 as as a a catalyst catalyst [38] [38].. 2.4. Summary 2.4. Summary The three main nucleophilic ring methods throughout the literature include enolate alkylation, The three main nucleophilic ring methods throughout the literature include enolate alkylation, transmetalation of bromonaphthoquinone derivatives, and Friedel-Crafts alkylation. The advantages transmetalation of bromonaphthoquinone derivatives, and Friedel-Crafts alkylation. The advantages and disadvantages of each have been outline in Table4. Out of all nine methods presented, only three and disadvantages of each have been outline in Table 4. Out of all nine methods presented, only three of them reported overall yields greater than 80%. These methods were Snyder and Rapoport’s of them reported overall yields greater than 80%. These methods were Snyder and Rapoport’s aryl- aryl-Grignard reaction with prenyl bromides, Swenton and coworker’s use of electrolytically protected Grignard reaction with prenyl bromides, Swenton and coworker’s use of electrolytically protected lithium organocuprate, and Schmid et al.’s unique intramolecular Friedel-Crafts alkylation. The ﬁrst lithium organocuprate, and Schmid et al.’s unique intramolecular Friedel-Crafts alkylation. The first two methods also demonstrated exemplary regiocontrol due to lithium-bromide exchange with of two methods also demonstrated exemplary regiocontrol due to lithium-bromide exchange with of bismethyl ether 2-bromomenadiol 6. The other methods produced regioisomers owing to prominent bismethyl ether 2-bromomenadiol 6. The other methods produced regioisomers owing to prominent competition between C2 and C3 alkylation. Many methods did not utilize protecting groups or directing competition between C2 and C3 alkylation. Many methods did not utilize protecting groups or group manipulations to inﬂuence regiocontrol. For most of these methods, the chosen electrophiles directing group manipulations to influence regiocontrol. For most of these methods, the chosen were prenyl halides, which left the α-isoprene double bond virtually undisturbed. However, in the electrophiles were prenyl halides, which left the α-isoprene double bond virtually undisturbed. case of Tabushi et al., the authors only synthesized MK-1; therefore, the stereochemical implications of However, in the case of Tabushi et al., the authors only synthesized MK-1; therefore, the the method were not addressed. stereochemical implications of the method were not addressed. Table 4. Summary of nucleophilic ring methods. Table 4. Summary of nucleophilic ring methods. Methods Advantages Disadvantages Methods Advantages Disadvantages Section 2.1. Enolate Alkylations Section 2.1. Enolate Alkylations -Low yields (20–45%) Snyder and -Stereoretention of α-isoprene double -Low yields (20–45%) -Stereoretention of α-isoprene double bond (97% E- -C2 alkylation competition via Snyder and Rapoport bond (97% E-alkene) Rapoport -C2Friedel-Crafts alkylation competition alkylation via Friedel- Enolate Alkylation [44] -3 step synthesisalkene) (not including starting Enolate -Unviable synthesisCrafts alkylation of starting -3 step synthesis (not includingmaterial) starting material) Alkylation [44] -Unviable synthesismaterial of starting material -Low-Low yields yields (40% (40% with inclusionwith inclusion Tabushi et al. -Regiocontrol via sterically hindered Tabushi et al. -Regiocontrol via sterically hindered nature of β- catalyst) β-cyclodextrin nature of β-cyclodextrin catalyst) β-cyclodextrin cyclodextrin -Competition between C3 alkylation inclusion catalyst -Menadione is the only byproduct -Competition between C3 alkylation inclusion catalyst -Menadione is the only byproduct and C3 protonation [45,46] -1 step synthesis and C3 protonation [45,46] -1 step synthesis -Only synthesized MK-1 -Only synthesized MK-1 Section 2.2. Transmetalations Section 2.2. Transmetalations -Regiocontrol through lithium-bromide -Regiocontrol through lithium-bromide exchange Snyder and exchange -Stereoretention of-Stereoretention the α-isoprene of thedoubleα-isoprene bond (≥ RapoportSnyder and Rapoport 95%)double bond ( 95%) -Need-Need to prepare to prepare starting starting material material 6 6 Grignard reactionGrignard reaction [44] ≥ -Alkylation-Alkylation step is high step yielding is high yielding (> 95%) (> 95%) [44] -3 step synthesis-3 step(not synthesis including (not starting including material) starting material) Saá and -Stereoretention of the α-isoprene double bond -Moderate yields for nucleophilic -Moderate yields for nucleophilic coworkers during BIHY reduction addition (58–65%) and BIHY Saá and coworkers addition (58–65%) and BIHY -Stereoretention of the α-isoprene BIHY Reduction reduction (53–70%) double bond during BIHY reduction [47–49] -5 step synthesis (not including starting material) Molecules 2020, 25, 4477 13 of 37

Table 4. Cont.

Methods Advantages Disadvantages -Unique use of electrolysis as a protection method -High yields for all reported steps -Lithium organocuprate nucleophile ( 85%) only used one of two bisketal Swenton and coworkers ≥ -Regiocontrol through lithium bromide rings—poor atom economy Electrolysis & Lithium exchange -Difficult purification because of Organocuprate -Stereoretention of α-isoprene double unreacted starting materials [50,51] bond (<5% Z-alkene estimated) -5 step synthesis (not including -Deprotection of bisketals to starting material) menaquinone ring structure via hydrolysis, no oxidation required Section 2.3. Friedel-Crafts Alkylation -Favors C3 alkylation over C2 due to monoacetate 14 -Low to moderate yields (8–66.5%) -Avoided formation of undesired Hirschmann et al. depending on acid catalyst used byproducts (phytadiene and chromanol) Friedel-Crafts Alkylation (Table1) -Monoacetate 14 was the only Lewis Acid Analysis [54] -Stereoretention of the α-isoprene recoverable byproduct double bond was not discussed -2 step synthesis (not including starting material) -Features unique intramolecular Friedel-Crafts alkylation at C3 position Schmid et al. -High yields (76–96.5%) throughout all Intramolecular steps -Need to prepare starting material 22 Friedel-Crafts -Stereoretention of α-isoprene double [63] bond -3 step synthesis (not including starting material) -Stereoretention of α-isoprene double Min et al. bond with AlCl3 Friedel-Crafts Alkylation -Produced a functional handle for chain -Low to moderate yields (0–72%) Lewis Acid Analysis extension methods depending on Lewis acid used [65] -1 step synthesis (not including starting material) -Universally low yields (0–26.5%) -Predicted industrial benefit to replace -Poor regiocontrol to prevent C2 Coman et al. [66] and BF OEt 3· 2 alkylation Koehn et al. [38] -Performed without protecting groups, -Difficult purification Heterogenous but could benefit from them -Synthesis of partly hydroxylated Lewis Acid Catalysts -1 step synthesis (not including starting metal fluorides requires the use of material) dangerous aqueous HF [67,68]

3. Metal-Mediated and Radical Reactions

3.1. Cross-Coupling After the incorporation of transmetalation, synthetic eﬀorts were shifted to investigate cross-coupling reactions. In 1973, Sato et al. used π-allylnickel chemistry to synthesize vitamin K1 and MK-9 after initial success with MK-1 [69]. For this method, π-allylnickel complexes 24 and 28 were formed in situ using allyl bromides 23 and 27 with Ni(CO)4 (Schemes 12A and 13A). Vitamin K1 was synthesized from diacetate 25 and π-allyl complex 24 in MPD at 50 ◦C for 5 h to form the expected alkylated product 26 in 75% yield (Scheme 12B). The acetate protecting groups were removed in mild basic conditions, and the oxidation was achieved using FeCl3, producing vitamin K1 in 93% over two steps. The observed E/Z ratio was 8:2 with respect to the α-isoprene double bond. For MK-9, the same starting material diacetate 25 and π-allyl complex 28 were heated to 50 ◦C for 16 h to form alkylated product 29 in 52% yield (Scheme 13A). MK-9 was produced in 85% yield and 7:3 E/Z using the same methods of deprotection and oxidation (Scheme 13B). Molecules 2020, 25, x FOR PEER REVIEW 15 of 39 Molecules 2020, 25, 4477 14 of 37 Molecules 2020, 25, x FOR PEER REVIEW 15 of 39

Scheme 12. (A) Formation of π-allylnickel complex. (B) Synthesis of vitamin K1 using π-allylnickel Scheme 12. (A) Formation of π-allylnickel complex. (B) Synthesis of vitamin K1 using π-allylnickel crossScheme-coupling 12. (A )[69] Formation. of π-allylnickel complex. (B) Synthesis of vitamin K1 using π-allylnickel cross-coupling [69]. cross-coupling [69].

Scheme 13. A π B π Scheme 13. (A() Formation) Formation of ofπ-allylnickel-allylnickel complex. complex. (B) Synthesis ( ) Synthesis of MK of-9 MK-9using usingπ-allylnickel-allylnickel cross- cross-coupling [69]. couplingScheme 13.[69] (A. ) Formation of π-allylnickel complex. (B) Synthesis of MK-9 using π-allylnickel cross- coupling [69]. Stille et al. published the synthesis of vitamin K1 using trimethylstannane derivatives to cross Stille et al. published the synthesis of vitamin K1 using trimethylstannane derivatives to cross couple phytyl bromide in 1983 [70]. 2-Bromomenadiol 30 was protected with TBSCl in 95% yield coupleStille phytyl et al. bromide published in 1983the synthesis [70]. 2-Bromomenadiol of vitamin K1 using 30 was trimethylstannane protected with TBSClderivatives in 95% to yieldcross (Scheme 14). Lithiation with t-BuLi and transmetalation with trimethyltin chloride formed 31 in 77% (Schemecouple phytyl 14). Lithiation bromide with in 1983 t-BuLi [70] and. 2- Bromomenadioltransmetalation with30 was trimethyltin protected chloride with TBSCl formed in 95%31 in yield 77% yield over two steps. Phytyl bromide was then coupled with 31 in the presence of ZnCl2 forming 32. yield(Scheme over 14). two Lithiation steps. Phy withtyl bromidet-BuLi and was transmetalation then coupled withwith 31trimethyltin in the presence chloride of ZnClformed2 forming 31 in 77% 32. PCC was used to deprotect and oxidize the ring to form vitamin K1 in 40% yield over two steps. PCCyield was over used two tosteps. deprotect Phytyl and bromide oxidize was the then ring coupled to form withvitamin 31 inK1 the in 40%presence yield ofover ZnCl two2 forming steps. 32. PCC was used to deprotect and oxidize the ring to form vitamin K1 in 40% yield over two steps.

Molecules 2020, 25, x FOR PEER REVIEW 16 of 39 Molecules 2020, 25, 4477 15 of 37 Molecules 2020, 25, x FOR PEER REVIEW 16 of 39

Scheme 14. Synthesis of vitamin K1 using Stille organostannane chemistry [70].

Scheme 14. Synthesis of vitamin K1 using Stille organostannane chemistry [70]. In 1990, LiebeskindScheme 14. andSynthesis Foster of discoveredvitamin K1 using an unexpected Stille organostannane transformation chemistry that [70] appeared. useful towardsIn 1990, the Liebeskindsynthesis of and MK Foster-derivative discovereds [71], an which unexpected was previously transformation mentioned that appeared in a review useful of In 1990, Liebeskind and Foster discovered an unexpected transformation that appeared useful synthesestowards the of synthesis vitamin ofK MK-derivativesand analogs [42] [71. MK], which-1 was was synthesized previously mentionedusing ring- instrained a review dione of syntheses 33 and towards the synthesis of MK-derivatives [71], which was previously mentioned in a review of propyne,of vitamin forming K and analogs alcohol [ 4234]. (Scheme MK-1 was 15). synthesized Then alcohol using 34 ring-strained underwent dione a Liebeskind33 and propyne,-Moore syntheses of vitamin K and analogs [42]. MK-1 was synthesized using ring-strained dione 33 and rearrangementforming alcohol in34 the(Scheme presence 15). of Then Bu3 alcoholSnOMe34 tounderwent form the stannylated a Liebeskind-Moore product rearrangement35 in 49% over in two the propyne, forming alcohol 34 (Scheme 15). Then alcohol 34 underwent a Liebeskind-Moore steps.presence From of Buthere,3SnOMe MK-1 to was form synthesized the stannylated in 90% product yield 35usingin 49% Stille over cross two-coupling steps. From condition there,s MK-1with rearrangement in the presence of Bu3SnOMe to form the stannylated product 35 in 49% over two Pd(0)was synthesized and prenyl inbromide. 90% yield using Stille cross-coupling conditions with Pd(0) and prenyl bromide. steps. From there, MK-1 was synthesized in 90% yield using Stille cross-coupling conditions with Pd(0) and prenyl bromide.

Scheme 15. Synthesis of MK-1 featuring Liebeskind-Moore rearrangement to ring expansion [71]. Scheme 15. Synthesis of MK-1 featuring Liebeskind-Moore rearrangement to ring expansion [71] . 3.2. Coordination Complex 3.2. CoordinationScheme 15. Complex Synthesis of MK-1 featuring Liebeskind-Moore rearrangement to ring expansion [71]. An intriguing approach to the synthesis of MK-derivatives was developed by Dötz et al. in 1986. Using3.2. CoordinationAn pentacarbonyl(methoxyphenylcarbene)chromium(0) intriguing Complex approach to the synthesis of MK-derivatives complex was develope36 andd an by alkyne Dötz et (Scheme al. in 1986. 16), Using pentacarbonyl(methoxyphenylcarbene)chromium(0) complex 36 and an alkyne (Scheme 16), the quinoneAn intriguing ring was approach formed to via the carbonylation synthesis of MK [43,-72derivatives,73]. This was approach develope providedd by Dötz access et al. to in several 1986. the quinone ring was formed via carbonylation [43,72,73]. This approach provided access to several diUsingfferent pentacarbonyl(methoxyphenylcarbene)chromium(0) naphthoquinone derivatives using functionalized alkynes. complex The 36 authorsand an alkyne synthesized (Scheme vitamin 16), different naphthoquinone derivatives using functionalized alkynes. The authors synthesized vitamin Kthe1 andquinone MK-1 ring through was formed MK-3 to via illustrate carbonylation this transformation. [43,72,73]. This Phenyllithium approach provided reacted access with oneto several of the K1 and MK-1 through MK-3 to illustrate this transformation. Phenyllithium reacted with one of the carbonyldifferent naphthoquinone ligands (CO) on Cr(CO)derivatives6, and using then functionalized was methylated alkynes. by trimethyloxonium The authors synthesized tetrafluoroborate vitamin carbonyl ligands (CO) on Cr(CO)6, and then was methylated by trimethyloxonium tetrafluoroborate toK1 formand MK complex-1 through36. Vitamin MK-3 to K 1illustrateand MK-1 this through transformation. MK-3 were Phenyllithium formed by using reacted alkynes with 37aoneor of37b the, to form complex 36. Vitamin K1 and MK-1 through MK-3 were formed by using alkynes 37a or 37b, respectively.carbonyl ligands Upon (CO) addition, on Cr(CO) the6,alkyne and then displaced was meth anotherylated by molecule trimethyloxonium of CO, to form tetrafluoroborate intermediate respectively. Upon addition, the alkyne displaced another molecule of CO, to form intermediate complexto form complex38. The resulting36. Vitamin menadiol K1 and ringMK-39a1 through/b was coordinatedMK-3 were formed to Cr(CO) by3 .using To make alkynes this route37a or more 37b, complex 38. The resulting menadiol ring 39a/b was coordinated to Cr(CO)3. To make this route more sustainable,respectively. the Upon authors addition, determined the alkyne Cr(CO) displaced6 could beanother regenerated molecule by pressurizingof CO, to form the intermediate system with sustainable, the authors determined Cr(CO)6 could be regenerated by pressurizing the system with COcomplex to displace 38. The Cr(CO) resulting3 and menadiol liberate ring the alkylated39a/b was product coordinated40a/b to. Oxidation Cr(CO)3. To with make standard this route oxidizing more CO to displace Cr(CO)3 and liberate the alkylated product 40a/b. Oxidation with standard oxidizing agentssustainable, afforded the theauthors menaquinone determined products Cr(CO) accordingly.6 could be regenerated The yields of by each pressurizing reaction were the notsystem reported with agents afforded the menaquinone products accordingly. The yields of each reaction were not inCO Rüttimann’s to displace Cr(CO) 1986 review3 and [liberate43]. the alkylated product 40a/b. Oxidation with standard oxidizing reported in Rüttimann’s 1986 review [43]. agents afforded the menaquinone products accordingly. The yields of each reaction were not reported in Rüttimann’s 1986 review [43].

Molecules 2020, 25, 4477 16 of 37 Molecules 2020, 25, x FOR PEER REVIEW 17 of 39 Molecules 2020, 25, x FOR PEER REVIEW 17 of 39

Scheme 16. SynthesisSynthesis of of MK MK-derivatives-derivatives using using unique Cr(CO) 6 mediatedmediated ring ring formation formation [43,72,73 [43,72,73]].. Scheme 16. Synthesis of MK-derivatives using unique Cr(CO)6 6mediated ring formation [43,72,73]. In 1980, Liebeskind et al. designed a synthesis of naphthoquinones using InIn 1980, 1980, Liebeskind Liebeskind et et al. al. designed designed a a synthesis synthesis of of naphthoquinones naphthoquinones using using bis(triphenylphosphine)phthaloylcobalt complex 41 (Scheme 17A). Ring formation was achieved bis(triphenylphosphine)phthaloylcobaltbis(triphenylphosphine)phthaloylcobalt complex complex 4141 (Scheme(Scheme 1717AA).). Ring Ring formation formation was was achieved achieved uponupon additio additionn of an alkyne in the presence of AgBF 4 [74][74].. In In 1986, 1986, an an update update was was published published wherein wherein upon addition of an alkyne in the presence of AgBF4 [74]. In 1986, an update was published wherein the authors describe an improved cobalt complex to increase the yield of the reaction and minimize the thethe authorsauthors describedescribe anan improvedimproved cobaltcobalt complexcomplex toto increaseincrease thethe yieldyield of of thethe reactionreaction andand minimizeminimize theamount amount of AgBFof AgBF4 required4 required [75 [75]]. The. The updated updated complex complex42 42replaced replaced thethe triphenylphosphinetriphenylphosphine ligands the amount of AgBF4 required [75]. The updated complex 42 replaced the triphenylphosphine ligands with pyridine and dimethylglyoxime (Scheme 17B). withwith pyridine pyridine and and dimethylglyoxime dimethylglyoxime (Scheme (Scheme 1717BB).).

Scheme 17. (A) Original cobalt complex synthesized by Liebeskind et al. for this transformation. SchemeScheme 17. 17. ( (AA)) Original Original cobalt cobalt complex complex synthesized synthesized by by Liebeskind Liebeskind et et al. al. for for this this transformation. transformation. ( (BB)) (B) Synthesis of MK-1 and MK-2 using the updated catalyst [74,75]. SynthesisSynthesis of of MK MK--11 and and M MKK--22 using using the the updated updated catalyst catalyst [74,75 [74,75]]. . This new complex was more tolerant towards diﬀerent Lewis acids as well as hydrated salts, ThisThis newnew complexcomplex waswas moremore toleranttolerant towardstowards differentdifferent LewisLewis acidsacids asas wellwell asas hydratedhydrated saltssalts,, which was demonstrated by the formation of 44 using complex 42, as shown in Table5. MK-1 and whichwhich waswas demonstrateddemonstrated byby thethe formationformation ofof 4444 usingusing complexcomplex 4242, , asas shownshown inin TableTable 5.5. MKMK--11 andand MKMK--22 were were synthesized synthesized using using alkynesalkynes 4343aa andand 43b43b, ,cobalt cobalt complex complex 4242 andand Lewis Lewis acidacid in in the the formform of of CoCl2∙6H2O at 80 °C producing MK-1 and MK-2 in 94% and 86% yield, respectively. CoCl2∙6H2O at 80 °C producing MK-1 and MK-2 in 94% and 86% yield, respectively.

Molecules 2020, 25, 4477 17 of 37

MK-2 were synthesized using alkynes 43a and 43b, cobalt complex 42 and Lewis acid in the form of CoClMolecules6H O 20 at20, 8025, x CFOR producing PEER REVIEW MK-1 and MK-2 in 94% and 86% yield, respectively. 18 of 39 Molecules2· 20202 , 25, x FOR◦ PEER REVIEW 18 of 39 Table 5. Effects of Additives on 2,3-diethyl-1,4-naphthoquinone 44 formation at 80 °C [75]. TableTable 5. EffectsEﬀects of Additives on 2,3 2,3-diethyl-1,4-naphthoquinone-diethyl-1,4-naphthoquinone 44 formation at 80 ◦°CC[ [7755]]..

GC Yield % of 44 AdditiveAdditive GC Yield % of 44 (1 Equiv(1) Equiv)2 hr 2 h5 hr 5 h 1818 h hr Additive GC Yield % of 44 None None25 2552 52 77 77 (1 EquivAgBF)4 2 80hr 825 hr - 18 hr AgBF4 80 82 - NoneBF3∙OEt 2 2574 7952 82 77 BF3 OEt2 74 79 82 AgBFSnCl4 2 · 8041 3982 70 - SnCl2 41 39 70 CoCl2∙6H2O 59 83 91 BF3∙OEtCoCl2 2 6H274O 5979 83 91 82 CoCl2(anhyd) · 61 86 86 SnCl2CoCl 2(anhyd)41 6139 86 86 70 p-CH3PhSO3H 14 31 76 CoCl2∙6Hp-CH2O 3PhSO593H 1483 31 76 91 CH3CO2H 23 47 74 CoCl2(anhyd)CH 3CO2H61 2386 47 74 86 3.3. Radical Reactions p-CH3PhSO3H 14 31 76 CH3CO2H 23 47 74 3.3. Radical Reactions 3.3.1. Metal-Mediated Radical Reactions 3.3. Radical Reactions 3.3.1. Metal-MediatedIn 1972, Jacobsen Radical and Torssell Reactions reported the use of allyl radicals produced via decarboxylation of carboxylic acids to alkylate quinones. Upon mixing silver nitrate and ammonium peroxodisulfite, 3.3.1.In Metal 1972,-Mediated Jacobsen Radical and Torssell Reactions reported the use of allyl radicals produced via decarboxylation Ag+ and S2O82 produce the radical, Ag2+ as shown in Equation (1). Then Ag2+ abstracts an electron of carboxylic acids to alkylate quinones. Upon mixing silver nitrate and ammonium peroxodisulﬁte, fromIn 1972, the carboxylicJacobsen andacid Torssell to produce reported CO2 and the a radicaluse of allylspecies radicals R as shown produced in Equation via decarboxylation (2) [76]. The of Ag+ and S O 2 produce the radical, Ag2+ as shown in Equation (1). Then Ag2+ abstracts an electron carboxylicauthors 2acids were8 toconcerned alkylate about quinones. possible Upon rearrangement mixing silver of the nitrateposition and of isoprene ammonium double peroxodisulfite, bond [77]. from4- theMethyl carboxylic-3-pentenoic acid acid to 45 produce formed COthe 23,3and-dimethylallyl a radical radical species 46R, whichas shown resonate ins Equationbetween α,α (2) [76]. Ag+ and S2O82 produce the radical, Ag2+ as shown in Equation (1). Then Ag2+ abstracts an electron The authorsand γ,γ positions were concerned (Scheme about18A). possible rearrangement of the position of isoprene double bond [77]. from the carboxylic acid to produce CO2 and a radical species R as shown in Equation (2) [76]. The 4-Methyl-3-pentenoic acid 45 formed+ the2 3,3-dimethylallyl- 2+ ∙- radical2- 46, which resonates between α,α authors were concerned about possibleAg + S2 Orearrangement8  Ag + SO4 of + SOthe4 position of isoprene double bond(1) [77]. γ γ 4and-Methyl, positions-3-pentenoic (Scheme acid 4518 A).formed the 3,3-dimethylallyl radical 46, which resonates between α,α Ag2+ + RCOOH  R∙ + CO2 + H+ + Ag+ (2) and γ,γ positions (Scheme 18A). + 2 2+ 2 Ag + S O − Ag + SO ·− + SO − (1) 2 8 → 4 4 Ag+ + S2O82-  Ag2+ + SO4∙- + SO42- (1) 2+ + + Ag + RCOOH R· + CO + H + Ag (2) → 2 Ag2+ + RCOOH  R∙ + CO2 + H+ + Ag+ (2)

Scheme 18. (A) Radical formation via decarboxylation. (B) Synthesis of MK-1 using this method [76,77].

The more stable tertiary radical, α,α-dimethylallylquinone, was expected; however, only γ,γ- dimethylallylquinone was observed, favoring the more stable alkene. MK-1 was produced in 70% yield using 45 (Scheme 18B), leaving the question of E/Z alkene isomerization unanswered: S Schemecheme 18. ((A)) Radical Radical formation formation via via decarboxylation. decarboxylation (B.) Synthesis(B) Synthesis of MK-1 of MK using-1 this using method this method [76,77]. [76,77].

Molecules 2020, 25, 4477 18 of 37

The more stable tertiary radical, α,α-dimethylallylquinone, was expected; however, only γ,γ-dimethylallylquinone was observed, favoring the more stable alkene.

MK-1Molecules was 20 produced20, 25, x FOR inPEER 70% REVIEW yield using 45 (Scheme 18B), leaving the question of E19/Z of 39alkene isomerizationMolecules 2020 unanswered:, 25, x FOR PEER REVIEW 19 of 39 TheThe question question of αof-isoprene α-isoprene double double bond bond stereoretentionstereoretention was was later later answered answered by Yama by Yamagogo et al. etin al. in 20002000 with The withpreliminary question preliminary of results α- isoprene results of the of double radical the radical bond coupling stereoretention coupling of quinones of quinones was with later with organotelluriumanswered organotellurium by Yama reagentsgo reagents et al. in [78 ,79]. [78,792000 with]. Geranyl preliminary bromide results (73:23 ofE/Z the) 47 radical was converted coupling to of the quinones corresponding with organotellurium tolyltelluride 48 reagents in 41% Geranyl bromide (73:23 E/Z) 47 was converted to the corresponding tolyltelluride 48 in 41% yield with yield[78,79 ] with. Geranyl complete bromide retention (73:23 ofE/Z stereochemistry) 47 was converted (Scheme to the corresponding19A). Then the tolyltelluride tolyltelluride 48 48 in was41% complete retention of stereochemistry (Scheme 19A). Then the tolyltelluride 48 was photochemically photochemicallyyield with complete coupled retention to menadione of stereochemistry 1 to produce (Scheme MK-2 in 19 44%A). yield Then with the complete tolyltelluride retention 48 was of coupledstereochemistryphotochemically to menadione (Scheme coupled1 to produce19toB menadione). This MK-2reaction 1 to in wasproduce 44% previously yield MK with-2 in mentioned 44% complete yield in with retention a review complete by of Daines stereochemistryretention et al.of (Schemeinstereochemistry 2003 19B). [42] This. reaction(Scheme was19B).previously This reaction mentioned was previously in areview mentioned by Dainesin a review et al. by in Daines 2003 [et42 al.]. in 2003 [42].

SchemeScheme 19. (19.A) ( SynthesisA) Synthesis of of geranyltelluride geranyltelluride reagents.reagents. (B (B) )Synthesis Synthesis of ofMK MK-2-2 with with organotelluride organotelluride radicalradicalScheme alkylation alkylation 19. (A [)78 Synthesis ,[78,7979]. ]. of geranyltelluride reagents. (B) Synthesis of MK-2 with organotelluride radical alkylation [78,79]. 3.3.2.3.3.2. Non-Metal-Mediated Non-Metal-Mediated Radical Radical Reactions Reactions 3.3.2. Non-Metal-Mediated Radical Reactions In 2019,In 2019, we we continued continued our our pursuit pursuit toto synthesizesynthesize menaquinone menaquinone analogs analogs with with various various levels levels of of saturationsaturationIn within 2019, within we the continuedthe side side chain chain our [38 [38]pursuit].. For For analogs atonalogs synthesize with with the menaquinone the first ﬁrst isoprene isoprene analogs unit unit saturated, with saturated, various we employed welevels employed of chemistrychemistrysaturation developed developedwithin the by sideby Coppa Coppa chain et [38]et al. al.. inFor in 1991. 1991. analogs Therein, with the different di firstﬀerent isoprene methods methods unit for saturated, forhomolytic homolytic we methylation employed methylation of quinonesofchemistry quinones with developed with alkyl alkyl iodides by iodides Coppa were were et al. discusseddiscussed in 1991. Therein,[80] [80.]. In In one different one such such method, methods method, menadione for menadionehomolytic 1 reacted methylation1 reacted with a with saturatedof quinones alkyl with iodide alkyl iodides in the presencewere discussed of benzoyl [80]. In peroxide one such to method, form the menadione C3-alkylated 1 reacted product with 49 a a saturated alkyl iodide in the presence of benzoyl peroxide to form the C3-alkylated product 49 (Schemesaturated 20 alkylA). iodide in the presence of benzoyl peroxide to form the C3-alkylated product 49 (Scheme 20A). (Scheme 20A).

Scheme 20. (A) General synthesis of menaquinones with alkyl iodides [80]. (B) Synthesis of MK-2 Scheme(I,IIScheme- 20.H4) [38] (20.A) . ( GeneralA) General synthesis synthesis of of menaquinones menaquinones with with alkyl alkyl io iodidesdides [80] [80. (].B) (SynthesisB) Synthesis of MK of-2 MK-2 (I,II-H(I,II4)[-H384)]. [38].

Molecules 2020, 25, 4477 19 of 37

The results varied, with yields ranging from 68 to 93%. Prominent competition between the C3-alkylated product and the C3-self-coupling aryl product was observed, showing 32–94% the C3 aryl product in some trials. We saw the parallels between the substrates used by Coppa et al. and the saturated prenyl side chains required for our studies. Menadione 1 was coupled with alkyl iodide 50 to synthesize MK-2(I,II-H4) in 17% yield (Scheme 20B).

3.4. Summary The three main metal-mediated and radical methods reported in the literature included organometallic cross-coupling, the use of coordination complexes, and both metal and non-metal mediated radical reactions. The advantages and disadvantages of each have been outlined in Table6. Only four methods reported moderate to high yields (65–100%) across the synthesis. Liebeskind and Foster used Stille coupling for the key alkylation step (90%). Liebeskind et al., again, reported high yields (77%) using the updated cobalt complex to aid the cycloaddition of alkynes to the coordinated phthaloyl group. Jacobsen and Torssell achieved moderate yields (70%) in the alkylation of quinones with radicals generated by the decarboxylation of prenyl carboxylic acids. Lastly, Coppa et al. also used radical alkylation of quinones using short-chain alkyl iodides in the presence of benzoyl peroxide. Regiocontrol for these methods was achieved in three ways: (1) transmetalation of organolithium reagents to diﬀerent organometallates; (2) asymmetric alkynes coordinating to symmetric complexes; and 3) selective abstraction of aryl hydrogens adjacent to the carbonyl. Complete stereoretention was observed during Stille et al.’s cross-coupling of arylstannanes and phytyl bromide in the presence of ZnCl2, Dötz et al.’s cycloaddition using stereopure alkynes, and

Table 6. Summary of metal-mediated reactions.

Methods Advantages Disadvantages Section 3.1. Cross-Coupling -No coordination complex synthesis required-π-allyl complex is formed in situ -The yields drop at the cross-coupling, -Moderate to high yields (52–93%) across the especially for the much longer prenyl Sato et al. syntheses side chains, MK-9 (52%) π-Allylnickel -E/Z ratio of the α-isoprene double bond (7:3 E/Z -Authors note E/Z ratio is tunable Cross Coupling [69] for MK-9) depending on the solvent, but the -3 step synthesis (not including starting yields drop as a result material) -High yields (77%) for the formation of the -Low yield for cross-coupling (40% arylstannane Stille et al. over two steps) -The regiochemistry of the system is controlled Aryl Stannane -Requires the use of t-BuLi by transmetalation at C3 position Cross Coupling [70] -5 step synthesis (not including -Allylic transposition was not observed in starting material) analogous syntheses of myrcene [70] Section 3.2. Coordination Complex Liebeskind and Foster -Stille coupling achieved high yields (90%) Low yield for key Liebeskind-Moore Ring Expansion -3 step synthesis (not including starting rearrangement (49% over two steps) to Stille Coupling [71] material) -Only synthesized MK-1 -No coordination complex synthesis required -Known adverse health eﬀects related Dötz et al. -E/Z ratio of the α-isoprene double bonds was to hexavalent chromium Chromium Complex retained throughout the synthesis -No yields reported in Rüttimann’s Carbonylation -The regiochemistry of the system is controlled 1986 review [43] [72,73] by the alkynes 37a and 37b -5 step synthesis (not including -Cr(CO)6 is recyclable starting material) Molecules 2020, 25, 4477 20 of 37

Table 6. Cont.

Methods Advantages Disadvantages -High yields (>86%) -Simple coordination complex synthesis Liebeskind et al. required using commercially available -The authors did not address Cobalt Complex Cycloaddition materials α-isoprene double bonds [75] -The regiochemistry of the system is isomerization controlled by the alkynes 43a and 43b -1 step synthesis (not including catalyst) Section 3.3. Radical Reactions -Moderate yields (70%) -Regiocontrolled through aryl hydrogen Jacobsen & Torssell abstraction -Only synthesized MK-1 Radical Decarboxylation [77] -Selective for γ,γ-alkene product of MK-1 -1 step synthesis -Low yields for both formation of -Regiocontrolled through aryl hydrogen tolyltelluride and radical coupling Yamago et al. abstraction (~40%) Radical Organotelluride -Stereoretention of the α-isoprene -Known adverse health eﬀects related [78,79] double bond across all steps to working with tellurium and -2 step synthesis tellurium compounds -Koehn et al. reported very low yields (17%) for this transformation with a Coppa et al. [80] -Moderate to high yields of straight branched alkane & Koehn et al. chain alkyl iodides (68–93%) [80] -Substantial α-isoprene double bond [38] -1 step synthesis (not including starting isomerism Benzoyl Peroxide Initiated material) -Competing reactions interfere with Radical Alkylation C3-alkylated product (C3-C3, and C2 alkylation)

4. Electrophilic Ring Methods

4.1. 1,2-Addition versus 1,4-Addition In the late 1970s and early 1980s, Naruta published two reports independently and published one with Maruyuma, reporting the use of prenyl stannanes for the synthesis of biologically important quinones, focusing on vitamin K1 and MK-2 [81–83]. Using menadione 1, trans and cis isomers with respect to the α-isoprene double bond were synthesized using geranyl and neryl trimethylstannanes, 51a and 51b, respectively, in the presence of BF OEt (Scheme 21A). The yields were 30% and 41% for 3· 2 the respective trans and cis products with the conﬁguration of the double bond mostly maintained for each product. Trans-MK-2 was found to have 95% E alkene, and the cis isomer was found to have 76% Z alkene. Using the same conditions, vitamin K1 was synthesized in 48% using phytyl trimethylstannane 52 (> 95% E) (Scheme 21B). The product was found to have 96% E alkene conﬁguration, showing complete stereoretention.Molecules 2020, 25, x FOR PEER REVIEW 22 of 39

Scheme 21. The synthesis of (A) MK-2 and ( B) vitamin K1 using alkylstannanes [81–83].

Scheme 21. The synthesis of (A) MK-2 and (B) vitamin K1 using alkylstannanes [81–83].

Table 7. Summary of electrophilic ring methods.

Methods Advantages Disadvantages 4.1. 1,2-Addition vs. 1,4-Addition Naruta and -Stereoretention of the α-isoprene double Maruyuma. -Low yields for both formations (30–48%) bond Organostannane -Prominent competition between C2 and C3 -1 step synthesis (not including starting Michael Addition alkylation materials) [81–83]

5. Pericyclic Reactions

5.1. Diels-Alder Reactions In a review published by Rüttimann in 1986 [43], synthetic advancements of the preparation of vitamin K1 were presented, detailing methods from traditional substitutions, organometallic reactions, and pericyclic reactions. Therein, Rüttimann and coworkers explored the use of Diels- Alder reactions to form the naphthoquinone unit of vitamin K in a previously unpublished synthesis inspired by the work of Troll and Schmid. [86]. Dihydroisobenzofurane 53 was reacted with activated alkyne dienophile 54 (96:4 E/Z) at 80 °C overnight to form the Diels-Alder adduct 55 (Scheme 22). Deprotection of the silyl ethers was achieved using methanol, and reduction of C2 methyl ester to a methyl group with sodium bis(2-methoxyethoxy)aluminum hydride in toluene under reflux formed the substituted menadiol. Oxidation with air in slightly acidic conditions produced vitamin K1 in ~50% yield over four steps. The configuration of the isoprene double bond was not disturbed, resulting in 93–96% E alkene. Using insights gained from the previous route, Rüttimann continued to explore the use of Diels- Alder reactions to synthesize vitamin K1. He and Büchi designed an auxiliary-directed route using cyclopentadiene as the corresponding diene (Scheme 23) [87]. Endo-Diels-Alder adduct 56 was formed at room temperature using menadione 1 and cyclopentadiene in 93% yield. Formation of adduct 56 switched the C3 hybridization from sp2 to sp3, decreasing the pKa. Upon deprotonation by a strong base (e.g., potassium amide, sodium amide, or potassium t-butoxide (which is referred to as potassium t-butanolate in the source literature)) a stable carbanion is formed, allowing regioselective alkylation at the C3 position with a variety of electrophiles.

Molecules 2020, 25, 4477 21 of 37

Although the major product for all trials was the 1,4-addition product, 1,2-addition accounted for a large portion of the undesired byproducts. For cis-MK-2 and vitamin K1, the C2 isomer was isolated in 13% and 14% yield, respectively. The 1,2 addition is hypothesized to add the least hindered carbonyl carbon of menadione 1, which then undergoes a [3,3] sigmatropic rearrangement, similarly described by Araki et al. [84] and Evans and Hoﬀmann [85] in Section 5.3. The rearrangement places the prenyl chain on the C2 carbon with the preexisting methyl group. In addition to prominent mechanistic competition, the yields are low across all steps. This method does, however, feature complete retention of stereochemistry of the α-isoprene double bond, as detailed in Table7.

Table 7. Summary of electrophilic ring methods.

Methods Advantages Disadvantages Section 4.1. 1,2-Addition vs. 1,4-Addition Naruta and Maruyuma. -Low yields for both formations -Stereoretention of the α-isoprene double bond Organostannane (30–48%) -1 step synthesis (not including starting Michael Addition -Prominent competition between C2 and materials) [81–83] C3 alkylation

5. Pericyclic Reactions

Scheme 22. PreviouslyScheme 22. Previously unpublished unpublished synthesis synthesis of vitamin vitamin K1 inspired K1 inspired by Troll & by Schmid Troll [43] &. Schmid [43].

Using insightsAlkylat gainedion with from phytyl thebromide previous (≥ 98% E) gave route, the predicted Rüttimann product continued57. The authorsto found explore that the use of some O-alkylation product was formed in trace amounts, however it was cleaved with acidic aqueous Diels-Alder reactionswork up and to easily synthesize separated. vitamin Alkylated Kadduct1. He 57 andwas formed Büchi in designed approximately an 90% auxiliary-directed yield over route using cyclopentadienethree steps. Due as the to the corresponding intrinsic instability diene of adduct (Scheme 57, slow decomposition23)[87]. Endo was-Diels-Alder observed at room adduct 56 was temperature. Retro-Diels-Alder reaction was induced at high temperatures to remove the auxiliary formed at roomgroup temperature quickly, producing using vitamin menadione K1 in quantitative1 and yield cyclopentadiene [87]. in 93% yield. Formation of 2 3 adduct 56 switched the C3 hybridization from sp to sp , decreasing the pKa. Upon deprotonation by a strong base (e.g., potassium amide, sodium amide, or potassium t-butoxide (which is referred to as potassium t-butanolate in the source literature)) a stable carbanion is formed, allowing regioselective alkylation at the C3 position with a variety of electrophiles.

Scheme 23. Synthesis of vitamin K1 using Diels-Alder approach with cyclopentadiene auxiliary [43].

5.2. Anionic Diels-Alder Reactions In contrast to traditional Diels-Alder reactions where neutral species form adducts, anionic Diels-Alder reactions have been another useful method to form the naphthoquinone unit. In 1995, Tso and Chen. published a one-pot synthesis applicable towards the synthesis of vitamin K1, and MKs-1, 2, and 9 (Scheme 24) [88]. The dienophiles used were alkenyl sulfones 60a–d, which were readily prepared from the corresponding allyl phenylsulfone 58 and allyl bromides 59a–d for vitamin K1, MK-1, MK-2, and MK-9, respectively. Isobenzofuranone 61 was deprotonated with NaHMDS, then 60a-d were attacked forming the intermediate 62. Elimination of the benzenesulfonate produced the desired products vitamin K1, MK-1, MK-2, and MK-9 in moderate yields, 60–64%. The configuration of the α-isoprene unit was found to be > 98% E alkene for all substrates.

Molecules 2020, 25, x FOR PEER REVIEW 23 of 39

Scheme 22. Previously unpublished synthesis of vitamin K1 inspired by Troll & Schmid [43].

Alkylation with phytyl bromide (≥ 98% E) gave the predicted product 57. The authors found that some O-alkylation product was formed in trace amounts, however it was cleaved with acidic aqueous work up and easily separated. Alkylated adduct 57 was formed in approximately 90% yield over three steps. Due to the intrinsic instability of adduct 57, slow decomposition was observed at room Molecules 2020, 25, 4477 22 of 37 temperature. Retro-Diels-Alder reaction was induced at high temperatures to remove the auxiliary group quickly, producing vitamin K1 in quantitative yield [87].

Scheme 23. SynthesisScheme 23.of Synthesis vitamin of vitamin K1 using K1 using Diels-Alder Diels-Alder approach approach with cyclopentadiene with cyclopentadiene auxiliary [43]. auxiliary [43].

5.2. Anionic Diels-Alder Reactions Alkylation with phytyl bromide ( 98% E) gave the predicted product 57. The authors found that In contrast to traditional Diels≥ -Alder reactions where neutral species form adducts, anionic some O-alkylationDiels-Alder product reactions was have formed been another in trace useful amounts, method to howeverform the naphthoquinone it was cleaved unit. In with 1995, acidic aqueous work up andTso easily and Chen separated.. published Alkylated a one-pot synthesis adduct applicable57 was towards formed the synthesis in approximately of vitamin K1, and 90% yield over three steps. DueMKs to-1, the2, and intrinsic 9 (Scheme instability24) [88]. The dienophiles of adduct used57 ,were slow alkenyl decomposition sulfones 60a–d, which was observedwere at room readily prepared from the corresponding allyl phenylsulfone 58 and allyl bromides 59a–d for vitamin temperature. Retro-Diels-AlderK1, MK-1, MK-2, and MK reaction-9, respectively. was Isobenzofuranone induced at high 61 was temperatures deprotonated with to NaHMDS, remove the auxiliary group quickly,then producing 60a-d were attacked vitamin forming K1 in the quantitative intermediate 62. yieldElimination [87]. of the benzenesulfonate produced the desired products vitamin K1, MK-1, MK-2, and MK-9 in moderate yields, 60–64%. The 5.2. Anionic Diels-Alderconfiguration Reactions of the α-isoprene unit was found to be > 98% E alkene for all substrates.

In contrast to traditional Diels-Alder reactions where neutral species form adducts, anionic Diels-Alder reactions have been another useful method to form the naphthoquinone unit. In 1995, Tso and Chen. published a one-pot synthesis applicable towards the synthesis of vitamin K1, and MKs-1, 2, and 9 (Scheme 24)[88]. The dienophiles used were alkenyl sulfones 60a–d, which were readily prepared from the corresponding allyl phenylsulfone 58 and allyl bromides 59a–d for vitamin K1, MK-1, MK-2, and MK-9, respectively. Isobenzofuranone 61 was deprotonated with NaHMDS, then 60a-d were attacked forming the intermediate 62. Elimination of the benzenesulfonate produced the desired products vitamin K1, MK-1, MK-2, and MK-9 in moderate yields, 60–64%. The conﬁguration of the

α-isoprene unitMolecules was 2020 found, 25, x FOR to PEER be REVIEW>98% E alkene for all substrates. 24 of 39

Scheme 24.SchemeSynthesis 24. Synthesis of vitamin of vitamin K1, K MK-1,1, MK-1, 2,2, and and 9 using 9 using anionic anionic Diels-Alder Diels-Alder approach [88] approach. [88].

Nearly twentyNearly years twenty later, years later,Mal Mal et al. et al. suggested suggested be bettertter atom atom economy economy could be could obtainedbe in obtained in comparison to the work done by Tso and Chen. In an effort to improve the atom economy of anionic comparison toDiels the-Alder work reactions, done the by authors Tso and replaced Chen. the Inphenylsulfone an effort tomoiety improve with a nitrile the atom 63 (Scheme economy 25) of anionic Diels-Alder reactions,[89]. Using the this authors method, the replaced authors the synthesized phenylsulfone different C3 moiety-alkylated with MK a-derivatives nitrile 63 using(Scheme 25)[89]. Using this method,specifically the designed authors dienophiles. synthesized For the disynthesisfferent of C3-alkylatedMK-1 and MK-2, MK-derivativesmethyl acrylate derivatives using specifically 64a and 64b were used. It is important to note the synthesis of 64b produced a mixture of ~3:1 of E/Z isomers with respect to the methyl acrylate alkene, as determined by NMR. It is unclear to us whether that refers to stereochemical composition of the methyl acrylate alkene or the geranyl side chain of 64b. Isobenzofuranone 63 was deprotonated with LiOt-Bu at −78 °C in THF, and then formed the adduct 65a/b with dienophiles 64a/b in 73% and 65% yield, respectively. Demethylcarboxylation of adducts 65a/b was achieved using a second round of LiOt-Bu in THF under reflux to form MK-1 and MK-2 in 55% and 40% yield, respectively. The ratio of E/Z ratio of MK-2 was found to be 5:3.

Scheme 25. Synthesis of MK-1 and MK-2 using anionic Diels-Alder approach with improved atom economy [89].

Molecules 2020, 25, x FOR PEER REVIEW 24 of 39

Molecules 2020, 25, 4477Scheme 24. Synthesis of vitamin K1, MK-1, 2, and 9 using anionic Diels-Alder approach [88]. 23 of 37

Nearly twenty years later, Mal et al. suggested better atom economy could be obtained in comparison to the work done by Tso and Chen. In an effort to improve the atom economy of anionic designed dienophiles.Diels-Alder reactions, For the the synthesisauthors replaced of MK-1 the phenylsulfone and MK-2, moiety methyl with acrylatea nitrile 63 derivatives (Scheme 25) 64a and 64b were used.[89] It. is Using important this method, to note the authors the synthesis synthesized of different64b produced C3-alkylated a mixture MK-derivatives of ~3:1 using of E/Z isomers with respectspecifically to themethyl designed acrylatedienophiles. alkene, For the synthesis as determined of MK-1 and by MK NMR.-2, methyl It isacrylate unclear derivatives to us whether that 64a and 64b were used. It is important to note the synthesis of 64b produced a mixture of ~3:1 of E/Z refers to stereochemicalisomers with respect composition to the methyl acrylate of the alkene, methyl as determined acrylate by alkene NMR. It or is unclear the geranyl to us whether side chain of 64b. Isobenzofuranonethat refers63 to stereochemicalwas deprotonated composition with of the LiO methylt-Bu acrylate at 78 alkeneC in or THF, the geranyl and thenside chain formed of the adduct − ◦ 65a/b with dienophiles64b. Isobenzofuranone64a/b 63in was 73% deprotonated and 65% with yield, LiO respectively.t-Bu at −78 °C in DemethylcarboxylationTHF, and then formed the of adducts adduct 65a/b with dienophiles 64a/b in 73% and 65% yield, respectively. Demethylcarboxylation of 65a/b was achievedadducts 65a/b using was achieved a second using round a second of round LiOt -Buof LiO int-Bu THF in THF under under reﬂux reflux to to form form MK MK-1-1 and and MK-2 in 55% and 40%MK yield,-2 in 55% respectively. and 40% yield, respectively. The ratio ofTheE ratio/Z ratio of E/Z of ratio MK-2 of MK was-2 was found found to to be be 5:3. 5:3.

Scheme 25. Synthesis of MK-1 and MK-2 using anionic Diels-Alder approach with improved atom economy [89Scheme]. 25. Synthesis of MK-1 and MK-2 using anionic Diels-Alder approach with improved atom economy [89]. 5.3. [3,3] Sigmatropic Rearrangements-Cope

[3,3] Sigmatropic rearrangements were also used to synthesize menaquinones, as shown in Daines Molecules 2020, 25, x FOR PEER REVIEW 25 of 39 et al. 2003 review [42]. In 1976, Evans and Hoﬀmann took advantage of the Cope rearrangement to synthesize MK-15.3. (Scheme [3,3] Sigmatropic 26A) Rearrangements [85]. Using- Cope masked ketone 66, the unprotected carbonyl was attacked with prenylmagnesium[3,3] Sigmatropic bromide, rearrangements forming ketone were also67 usedin one to synthesize step. Interestingly, menaquinones, as the shown authors in discovered Daines et al. 2003 review [42]. In 1976, Evans and Hoffmann took advantage of the Cope that prenylmagnesiumrearrangement bromide to synthesize added MK to-1 the(Scheme carbonyl 26A) [85] in. Using a reverse masked prenylated ketone 66, the fashion unprotected68a (Scheme 26B). The C1-reversecarbonyl prenylated was attacked product with preny68almagnesiumset the bromide, transition forming state ketone68b 67 infor one astep. Cope Interestingly, rearrangement, 68b the authors discovered that prenylmagnesium bromide added to the carbonyl in a reverse prenylated to 69a, producingfashion C3-prenylated 68a (Scheme 26B). product The C1-reverse69a .prenylated Upon quenchingproduct 68a set with the transition saturated state 68b NH for4 aCl, as shown in intermediate 69bCope, the rearrangement C3 methoxy, 68b group to 69a, wasproducing removed C3-prenylated to reform product ketone 69a. Upon67 (Scheme quenching 26 withB). Deprotection saturated NH4Cl, as shown in intermediate 69b, the C3 methoxy group was removed to reform of the masked ketoneketone 67 was (Scheme achieved 26B). Deprotection using AgF of thein mild masked conditions ketone was achieved to form using MK-1 AgF in in 71% mild yield over two steps. Since MK-1conditions was to the form only MK-1 productin 71% yield synthesized, over two steps. Since there MK-E1 was/Z ratio the only of product the α synthesized,-isoprene double bond was not considered,there E/Z nor ratio were of the the α-isoprene eﬀects double of this bond synthesis was not considered, on longer nor prenyl were the chains. effects of this synthesis on longer prenyl chains.

Scheme 26. (A) Synthesis of MK-1 using prenylmagnesium bromide to Cope rearrangement. (B) Speciﬁc mechanism of theScheme Cope 26. rearrangement(A) Synthesis of MK- [185 using]. prenylmagnesium bromide to Cope rearrangement. (B) Specific mechanism of the Cope rearrangement [85].

In contrast to more conventional alkylating reagents, Araki et al. used organoindium reagents for the allylation of various quinones [84]. Throughout many trials with a wide scope of benzo- and naphthoquinone derivatives, organoindium reagents were found to selectively add to the least hindered carbonyl in 1,2-addition (Scheme 27A). Therefore, when choosing between the C1 and C4 carbonyl groups, 2-bromomenadione 70 was attacked at the less hindered C1 ketone with prenylindium 71 to form reverse prenylated intermediate 72a (Scheme 27B). Following addition, the reverse prenyl group underwent a Cope rearrangement, 72b to 73a, analogous to the system reported by Evans and Hoffmann. After oxidation with Ag2O, MK-1 was produced in 67% over two steps. This synthesis did not address the possibility of alkene isomerization, however, in other trials reported in the same paper [84], the authors reported geranylindium 74a and nerylindium 74b rearrangements on menadione 1, producing 75a and 75b in 85% and 95% yield, respectively (Scheme 27C). Each product showed a total loss of stereoretention, ~50/50 E/Z. The authors hypothesized the same trend would have been seen using 2-bromomenadione 70; therefore, stereocontrol of the α-isoprene double bond continues to be an important challenge with Cope rearrangements.

Molecules 2020, 25, 4477 24 of 37

In contrast to more conventional alkylating reagents, Araki et al. used organoindium reagents for the allylation of various quinones [84]. Throughout many trials with a wide scope of benzo- and naphthoquinone derivatives, organoindium reagents were found to selectively add to the least hindered carbonyl in 1,2-addition (Scheme 27A). Therefore, when choosing between the C1 and C4 carbonyl groups, 2-bromomenadione 70 was attacked at the less hindered C1 ketone with prenylindium 71 to form reverse prenylated intermediate 72a (Scheme 27B). Following addition, the reverse prenyl group underwent a Cope rearrangement, 72b to 73a, analogous to the system reported by Evans and Hoﬀmann. After oxidation with Ag2O, MK-1 was produced in 67% over two steps. This synthesis did not address the possibility of alkene isomerization, however, in other trials reported in the same paper [84], the authors reported geranylindium 74a and nerylindium 74b rearrangements on menadione 1, producing 75a and 75b in 85% and 95% yield, respectively (Scheme 27C). Each product showed a total loss of stereoretention, ~50/50 E/Z. The authors hypothesized the same trend would have been seen using 2-bromomenadione 70; therefore, stereocontrol of the α-isoprene double bond continuesMolecules to2020 be, 25 an, x FOR important PEER REVIEW challenge with Cope rearrangements. 26 of 39

Scheme 27. (A) Synthesis of MK-1 using organoindium reagents. (B) Closer look into the Cope rearrangement.Scheme 27. (A) ( C Synthesis) Attempted of MK synthesis-1 using organoindium of MK-2 using reagents. geranyl (B74a) Closerand neryl look into74b theorganoindium Cope rearrangement.rearrangement. [84 (C].) Attempted synthesis of MK-2 using geranyl 74a and neryl 74b organoindium rearrangement. [84]. 5.4. Summary 5.4. Summary Across the literature, the common types of pericyclic reactions are Diels-Alder cycloadditions and [3,3] sigmatropicAcross the rearrangements, literature, the common speciﬁcally types Copeof pericyclic rearrangements. reactions are The Diels advantages-Alder cycloadditions and disadvantages of eachand [3,3] method sigmatropic have been rearrangements, outlined in specifically Table8. Rüttimann Cope rearrangements. et al. achieved The advantages high yields and with a cyclopentadienedisadvantages of auxiliary-directed each method have Diels-Alderbeen outlined reaction.in Table 8. For Rüttimann all methods et al. achieved described high in yields this section, with a cyclopentadiene auxiliary-directed Diels-Alder reaction. For all methods described in this the asymmetry of certain reagents controlled the regioselectivity of the reaction. For example, section, the asymmetry of certain reagents controlled the regioselectivity of the reaction. For example, the asymmetry of the dienophiles involved in the Diels-Alder reactions controlled the regiochemistry the asymmetry of the dienophiles involved in the Diels-Alder reactions controlled the regiochemistry of theof the adduct. adduct. The The Cope Cope rearrangementrearrangement reactions reactions were were regiocontrolled regiocontrolled by using by using starting starting materials materials with leaving groups on the desired position, protecting groups, and by taking advantage of steric hinderance. For all Diels-Alder reactions described, except for the work done by Mal et al., complete retention of stereochemistry was observed for the α-isoprene double bond. In the case of Mal et al., the reported E/Z ratio for dienophile 64b was ~3:1 E/Z by NMR for the methyl acrylate dienophile. The E/Z ratio of the product MK-2 was found to be 5:3 E/Z. It is unclear to us whether this was a result of the Diels-Alder reaction or due to isomeric starting material. For all Cope rearrangements, it can be inferred that a complete loss of stereocontrol would be observed due to the reverse-prenyl addition to the carbonyl, as observed by Araki et al. using geranyl and nerylindium reagents with menadione 1.

Molecules 2020, 25, 4477 25 of 37 with leaving groups on the desired position, protecting groups, and by taking advantage of steric hinderance. For all Diels-Alder reactions described, except for the work done by Mal et al., complete retention of stereochemistry was observed for the α-isoprene double bond. In the case of Mal et al., the reported E/Z ratio for dienophile 64b was ~3:1 E/Z by NMR for the methyl acrylate dienophile. The E/Z ratio of the product MK-2 was found to be 5:3 E/Z. It is unclear to us whether this was a result of the Diels-Alder reaction or due to isomeric starting material. For all Cope rearrangements, it can be inferred that a complete loss of stereocontrol would be observed due to the reverse-prenyl addition to the carbonyl, as observed by Araki et al. using geranyl and nerylindium reagents with menadione 1.

Table 8. Summary of pericyclic reactions.

Methods Advantages Disadvantages Section 5.1. Diels-Alder -High regiocontrol through the -Overall low yields (~50% over four Rüttimann et al. symmetry of dihydroisobenzofurane steps) Diels-Alder Reaction inspired diene -Synthesis of starting materials by Troll & Schmid [43] -Stereoretention of α-isoprene double -4 step synthesis (not including bond ( 93%) starting material) ≥ -Uses commercially available starting materials (menadione and cyclopentadiene) Rüttimann et al. -High regiocontrol through adduct 66 -Slight competition between Auxiliary-Directed -Stereoretention of α-isoprene double C-alkylation and O-alkylation Diels-Alder [43] bond -5 step synthesis -Cyclopentadiene can be recycled -High yields throughout the synthesis ( 90%) ≥ Section 5.2. Anionic Diels-Alder -One-pot synthesis -High regiocontrol through asymmetry of dienophile -Moderate yields (60–64%) Tso and Chen -Stereoretention of α-isoprene double -Requires the synthesis of starting Anionic Diels-Alder [88] bond across all steps (>98%) materials -3 step synthesis (not including starting material) -Improved atom economy -Low to moderate yields (40–73%) Mal et al. -High regiocontrol through asymmetry -5:3 E/Z ratio Anionic Diels-Alder with of dienophile -Unclear if it is due to stereochemistry Improved Atom Economy [89] -2 step synthesis (not including starting of starting material or caused by the material) reaction Section 5.3. [3,3] Sigmatropic Rearrangements- Cope -Regiocontrol achieved through protected naphthoquinone Evans and Hoﬀmann -Cope rearrangement to achieve C3 -No consideration of the isomerization Grignard-Promoted alkylation of the isoprene double bond Cope Rearrangement [85] -Moderate yields (71% over two steps) -2 step synthesis (not including starting material) -Regiocontrol achieved through less hindered 1,2-addition of organoindium reagent Araki et al. -No protecting groups required No stereoretention observed in Cope Organoindium-Promoted -Cope rearrangement to achieve C3 rearrangement Cope Rearrangement [84] alkylation -Moderate yields (67% over two steps) -2 step synthesis (not including starting material) Molecules 2020, 25, 4477 26 of 37

6. Homologation & Side Chain Extension Methods

6.1. Homologation In 1998, Lipshutz et al. designed a route to install a one carbon functional handle at the C3 position to enable the synthesis of a wide variety of MK-derivatives [90]. Menadione 1 was reacted with formaldehyde and hydrogen chloride gas to form 3-chloromethylmenadione 76 in 87% yield (Scheme 28). The introduction of the chloromethyl group allows for reactions that were previously less approachable, MoleculesMolecules 20 2020, 2,5 2, 5x, FORx FOR PEER PEER REVIEW REVIEW 2828 of of39 39 like SN2 substitutions and organometallic cross-couplings. Using 3-chloromethylmenadione 76 as the starting material, the authors used Negishi cross-coupling conditions to take advantage of the chloromethylmenadionechloromethylmenadione 76 76 as as the the starting starting material, material, the the authors authors used used Negishi Negishi cross cross-coupling-coupling stereoselectiveconditionsconditions to to installation take take advantage advantage of alkenesof of the the stereoselective s basedtereoselective on the installation installation configuration of of alkenes alkenes of the based based organoalane on on the the configuration configuration species [91 –94]. For example,ofof the the organoalane phytyl organoalane alkyne species species79 underwent [91[91–94–94]. ] Negishi. For For example, example, carboalumination phytyl phytyl alkyne alkyne to form 7979 organoalane underwentunderwent 77 Negishi Negishi(Scheme 29). Vitamincarboaluminationcarboalumination K1 and MK-3 to to wereform form organoalane synthesized organoalane 77 using77 (Scheme (Scheme this 29). method29). Vitamin Vitamin (Scheme K 1K and1 and MK 28 MK).-3- Forw3 wereere the synthesized synthesized synthesis using ofusing vitamin this method (Scheme 28). For the synthesis of vitamin K1, 3-chloromethylmenadione 76 was coupled K1, 3-chloromethylmenadionethis method (Scheme 28). For 76thewas synthesis coupled of vitamin with phytylalane K1, 3-chloromethylmenadione77 in the presence 76 of was the nickelcoupled catalyst with phytylalane 77 in the presence of the nickel catalyst which is formed in situ using nickel (II) whichwith is phytylalane formed in situ77 in using the presence nickel (II)of the chloride, nickel catalyst triphenylphosphine, which is formed and in situn-BuLi. using Vitaminnickel (II) K 1 was chloride, triphenylphosphine, and n-BuLi. Vitamin K1 was formed in 88% with exclusively E formedchloride, in 88% triphenylphosphine, with exclusively andE configuration.n-BuLi. Vitamin MK-3 K1 was was formed prepared in 88% similarly, with exclusively but instead E with configuration. MK-3 was prepared similarly, but instead with farnesylalane 78, in 93% yield with E farnesylalaneconfiguration.78, MK in 93%-3 was yield prepared with Esimilarly,configuration but instead at the withα-isoprene farnesylalane double 78, in bond. 93% yield with E configurationconfiguration at at the the α -αisoprene-isoprene double double bond bond. .

SchemeSchemeScheme 28. 28. Synthesis28. Synthesis Synthesis of of ofMK-3MK MK-3- 3 and and vitamin vitamin K K1 K 1via1 via C3 C3 homologation homologation and and and Negishi Negishi Negishi cross cross- cross-couplingcoupling-coupling conditionsconditionsconditions [90 [90]]. [90]. .

Scheme 29. Negishi carboalumination of phytyl alkyne [91–94]. Scheme 29. Negishi carboalumination of phytyl alkyne [91–94]. In 2015, Mehta etScheme al. published 29. Negishi carboalumination a patent covering of phytyl the alkyne synthesis [91–94]. of stereospeciﬁc quinone derivativesInIn 2015, [95 2015,]. MehtaTherein, Mehta et et the al. al. authors published published described a a patent patent methods covering covering used the the synthesis for synthesis the synthesis of of stereospecific stereospecific of the various quinone quinone lengths of prenylderivativesderivatives side [95] chains [95]. Therein. Therein using, the, the a authors series authors ofdescribed described homologation methods methods used and used sidefor for the the chain synthesis synthesis extension of of the the various reactionsvarious lengths lengths featuring stereoselectiveofof prenyl prenyl side side alkene chains chains syntheses,using using a aseries series such of of homo as homo Wittig,logationlogation Horner-Wadsworth-Emmons, and and side side chain chain extension extension reactions reactions and featuring Still-Gennari.featuring Forstereoselective thestereoselective synthesis alkene of alkene MK-7, syntheses, syntheses, 1,4-dimethoxynaphthoquinol such such as as Wittig, Wittig, Horner Horner-Wadsworth-Wadsworth18 reacted-Emmons,-Emmons, with dichloro(methoxy)methane and and Still Still-Gennari.-Gennari. For For the synthesis of MK-7, 1,4-dimethoxynaphthoquinol 18 reacted with dichloro(methoxy)methane and andthe TiCl synthesis4 to form of MK the-7, C3 1,4 aldehyde-dimethoxynaphthoquinol80 in 95% yield 18 reacted (Scheme with 30 dichloro(methoxy)methane). Wittig homologation ofand80 was TiCl4 to form the C3 aldehyde 80 in 95% yield (Scheme 30). Wittig homologation of 80 was achieved achievedTiCl4 to using form ylidethe C381 aldehydefollowed 80 in by 95% mild yield acid (Scheme hydrolysis 30). Wittig to form homologation aldehyde 82of .80 Using was achieved phosphonate using ylide 81 followed by mild acid hydrolysis to form aldehyde 82. Using phosphonate ester 83, esterusing83, aldehydeylide 81 followed82 underwent by mild a acid Horner-Wadsworth-Emmons hydrolysis to form aldehyde reaction 82. Using to phosphonate form the desired ester 83E ,alkene aldehydealdehyde 82 82 underwent underwent a aHorner Horner-Wadsworth-Wadsworth-Emmons-Emmons reaction reaction to to form form the the d esireddesired E Ealkene alkene in in 85% 85% in 85% yield. The authors noted the use of Still-Gennari conditions for the synthesis of the Z alkene yield.yield. The The authors authors noted noted the the use use of of Still Still-Gennari-Gennari conditions conditions for for the the synthesis synthesis of of the the Z Zalkene alkene where where where appropriate. The ester was reduced to the alcohol using DIBAL-H in 91% yield, and then appropriate.appropriate. The The ester ester was was reduced reduced to to the the alcohol alcohol using using DIBAL DIBAL-H-H in in 91% 91% yield, yield, and and then then immediately immediately oxidizedoxidized to to aldehyde aldehyde 84 84 using using PCC PCC in in 88% 88% yield. yield. The The resulting resulting aldehyde aldehyde 84 84 was was protected protected as as dithiane dithiane 8585 in in 94% 94% yield. yield. Deprotonation Deprotonation of of the the methine methine hydrogen hydrogen of of 85 85 with with n -nBuLi-BuLi created created a astable stable anionic anionic nucleophile.nucleophile. Farnesylfarnesyl Farnesylfarnesyl bromide bromide 86 86 was was used used to to form form the the C3 C3-alkylated-alkylated product product in in 83%. 83%. The The dithianedithiane protected protected carbonyl carbonyl was was deprotected deprotected using using Dess Dess-Martin-Martin periodinane periodinane and and then then reduced reduced using using WolffWolff-Kishner-Kishner conditions, conditions, producing producing the the hydrocarbon hydrocarbon prenyl prenyl side side chain chain in in 82% 82% yield yield over over two two steps. steps.

Molecules 2020, 25, 4477 27 of 37 immediately oxidized to aldehyde 84 using PCC in 88% yield. The resulting aldehyde 84 was protected as dithiane 85 in 94% yield. Deprotonation of the methine hydrogen of 85 with n-BuLi created a stable anionic nucleophile. Farnesylfarnesyl bromide 86 was used to form the C3-alkylated product in 83%. The dithiane protected carbonyl was deprotected using Dess-Martin periodinane and then reduced Molecules 2020, 25, x FOR PEER REVIEW 29 of 39 using Wolﬀ-Kishner conditions, producing the hydrocarbon prenyl side chain in 82% yield over two steps.Oxidation Oxidation with CAN with formed CAN formed MK-7 MK-7in 80% in yield. 80% yield. The authors The authors achieved achieved the synthesis the synthesis of MK of- MK-77 in two in twodifferent diﬀerent iterations: iterations: (1) when (1) when the theentire entire chain chain was was added added in inone one step step (Scheme (Scheme 30), 30), and and ( (2)2) when the side chainchain was added on smaller segments using the same methodology,methodology, whichwhich is similarlysimilarly usedused in Masaki et al.’s work inin SectionSection 6.26.2..

Scheme 30. Synthesis of MK-7MK-7 using dithiane anion side chain extensions [[95]95].. 6.2. Side Chain Extensions 6.2. Side Chain Extensions Another popular method across the literature is the continued functionalization of truncated prenyl Another popular method across the literature is the continued functionalization of truncated chains that were installed using the previously described methods. In 1984, Masaki et al. developed prenyl chains that were installed using the previously described methods. In 1984, Masaki et al. a synthetic route to lengthen the prenyl chain starting with prenyl bromides (Scheme 31)[96,97]. developed a synthetic route to lengthen the prenyl chain starting with prenyl bromides (Scheme 31) This served as the starting material for the synthesis of vitamin K1 and MK-4. The common electrophiles [96,97]. This served as the starting material for the synthesis of vitamin K1 and MK-4. The common across all trials were prenyl bromides 92, synthesized in 70–90% yield from the corresponding prenyl electrophiles across all trials were prenyl bromides 92, synthesized in 70–90% yield from the alcohols 91 using PBr3 (Scheme 32). For the synthesis of vitamin K1 (Scheme 31), prenyl bromide 87 corresponding prenyl alcohols 91 using PBr3 (Scheme 32). For the synthesis of vitamin K1 (Scheme 31), was coupled to tosylate 88. Deprotonation of the tosylate methine hydrogen created a stable anion to prenyl bromide 87 was coupled to tosylate 88. Deprotonation of the tosylate methine hydrogen attack 87 to form the alkylated product 89 in 72% yield. Removal of the tosyl group was achieved using created a stable anion to attack 87 to form the alkylated product 89 in 72% yield. Removal of the tosyl modiﬁed Bauvault-Blanc desulfurization conditions in 70% yield. Deprotection of tosylate 89 obtained group was achieved using modified Bauvault-Blanc desulfurization conditions in 70% yield. 90 in 70% yield, and thus vitamin K1 was produced in 70% yield in the presence of CAN. For the Deprotection of tosylate 89 obtained 90 in 70% yield, and thus vitamin K1 was produced in 70% yield synthesis of MK-4, Masaki et al. approached the synthesis with two diﬀerent iterations (Scheme 33). in the presence of CAN. For the synthesis of MK-4, Masaki et al. approached the synthesis with two Starting with benzyl ether protected polyprenyl bromide 93a and 93b, the same alkylating conditions different iterations (Scheme 33). Starting with benzyl ether protected polyprenyl bromide 93a and were used to extend the chain using tosylates 94a and 94b to form products 95a and 95b in 86% and 93b, the same alkylating conditions were used to extend the chain using tosylates 94a and 94b to form 87% yield, respectively. Desulfurization of the prenyl chain produced the hydrocarbon side chain products 95a and 95b in 86% and 87% yield, respectively. Desulfurization of the prenyl chain of MK-4 in 68–73% yield. The authors noted HPLC analysis showed isomeric byproducts in 5–7%, produced the hydrocarbon side chain of MK-4 in 68–73% yield. The authors noted HPLC analysis likely formed during the desulfurization step. showed isomeric byproducts in 5–7%, likely formed during the desulfurization step.

Molecules 2020, 25, x FOR PEER REVIEW 30 of 39 Molecules 2020, 25, x FOR PEER REVIEW 30 of 39 Molecules 2020, 25, 4477 28 of 37 Molecules 2020, 25, x FOR PEER REVIEW 30 of 39

Scheme 31. Synthesis of vitamin K1 using tosylate extension methods [96,97]. Scheme 31. Synthesis of vitamin K 1 usingusing tosylate tosylate extension extension methods methods [96,97 [96,97]]..

Scheme 31. Synthesis of vitamin K1 using tosylate extension methods [96,97].

Scheme 32. Conversion of prenyl alcohols to respective prenyl bromides [96,97]. Scheme 32. Conversion of prenyl alcohols to respective prenyl bromides [96,97]. Scheme 32. Conversion of prenyl alcohols to respective prenyl bromides [96,97 ]. Scheme 32. Conversion of prenyl alcohols to respective prenyl bromides [96,97].

Scheme 33. Synthesis of MK-4 using tosylate extension methods [96,97].

Scheme 33. Synthesis of MK-4 using tosylate extension methods [96,97]. In contrast to Masaki et al.’s. approach with tosylates, Schmid et al. used organocuprate reagents as nucleophiles to install the remaining hydrocarbon chain for vitamin K1 [63]. Installation of the ﬁrst In contrast Schemeto Masa ki33. et Synthesis al.’s. approach of MK- with4 using tosylates, tosylate Schmidextension et methodsal. used organocuprate[96,97]. reagents prenyl group was achievedScheme 33. via Synthesis coupling of ofMK organocuprate-4 using tosylate reagent extension of methods bismethyl [96,97 ether]. 2-bromomenadiol as nucleophiles to install the remaining hydrocarbon chain for vitamin K1 [63]. Installation of the first 6 with isoprene oxide in a 1,4-addition (Scheme 34). The resulting alcohol was protected as ester 96. prenylIn Incontrast contrast group to to wasMasa Masa achievedkiki et et al.’s. al.’s. approachvia approach coupling withwith oftosylates, organocuprate Schmid Schmid et etreagent al. al. used used organocuprateof organocuprate bismethyl reagents ether reagents 2- This reaction yielded 90% over two steps with 93:7 E/Z conﬁguration. The rest of the phytyl chain was as asnucleophilesbromomenadiol nucleophiles to to install6 install with the the isoprene remaining remaining oxide hydrocarbonhydrocarbon in a 1,4-addition chain forfor(Scheme vitamin vitamin 34). K K1 [63]1 The [63]. Installation resulting. Installation alcohol of ofthe the first was first 97 prenylinstalledprenylprotected group using group as ester the was was organocuprate 96 achieved. achievedThis reaction viavia reagent yieldedcouplingcoupling formed 90% of over withorganocuprate two hexahydrofarnesyl steps with reagentreagent 93:7 E/Z bromideof ofconfiguration. bismethyl bismethyl, performing ether The ether rest 2 - 2 a- bromomenadiolSNbromomenadiol2of substitutionthe phytyl chain 6 with 6with with was the isoprene isopreneinstalled ester protecting oxide oxideusing inthein group, a aorganocuprate 1,4-addition forming (Scheme (Scheme thereagent extended formed 34). 34). The sideThe with resulting chainresulting hexahydrofarnesyl in alcohol 51–79% alcohol wasyield. was Oxidation of the ring was achieved with CAN with 77–86% yield to form vitamin K . protectedprotectedbromide as 97 as ester, performingester 96 96. This. This a reaction SreactionN2 substitution yielded yielded with 90%90% theover ester twotwo protecting stepssteps with with group, 93:7 93:7 E/Z E/Zforming configuration. configuration. the1 extended The The restside rest of ofthe the phytyl phytyl chain chain was was installed installed using using thethe organocuprate reagentreagent formed formed with with hexahydrofarnesyl hexahydrofarnesyl bromide 97, performing a SN2 substitution with the ester protecting group, forming the extended side bromide 97, performing a SN2 substitution with the ester protecting group, forming the extended side

Molecules 2020, 25, x FOR PEER REVIEW 31 of 39

chain in 51–79% yield. Oxidation of the ring was achieved with CAN with 77–86% yield to form Molecules 2020, 25, 4477 29 of 37 vitamin K1.

Scheme 34. Synthesis of vitamin K1 using organocuprate reagents to extend the length of the side Scheme 34. Synthesis of vitamin K1 using organocuprate reagents to extend the length of the side chain [63]. chain [63]. 6.3. Summary 6.3. Summary HomologationHomologation and and side side chain chain extension extension methodsmethods comprise comprise a aseparate separate category category from from the others the others despitedespite the usethe use of similarof similar techniques techniques because because theythey provide aa functional functional handle handle that that enables enables more more diversediverse reactions reactions that that were were previously previously less less accessible. accessible. Homologation Homologation describes describes the theaddition addition of one of one carbon-containingcarbon-containi groupng group to to the the C3 C3 position, position, whichwhich is then further further functionalized functionalized to synthesize to synthesize the the rest ofrest side of side chain. chain Side. Side chain chain extensionextension methods methods have have one oneor two or isoprene two isoprene units attached units attached to the ring to the ring whichwhich were were installed installed using using previously previously described described methods methods to then toadd then the addremainder the remainder of the chain of the chainusing using a a different diﬀerent method, method, allowing allowing for for continued continued iterative iterative additions. additions. The The advantages advantages and and disadvantagesdisadvantages of each of each method method have have been been outlined outlined inin Table 99.. Nearly Nearly all all described described methods methods produced produced moderate to high yields throughout all steps, except for the organocuprate substitution demonstrated moderate to high yields throughout all steps, except for the organocuprate substitution demonstrated by Schmid et al., which produced low yields compared to the rest of the synthesis. Regiocontrol for by Schmidthese methods et al., which was achieved produced by the low preinstalled yields compared carbons or to prenyl the rest chains of the connected synthesis. to the Regiocontrol ring which for thesemake methods the subsequent was achieved reactions by the chemoselective. preinstalled carbons Stereoretention or prenyl of chainsthe α-isoprene connected double to thebond ring was which makeachieved the subsequent in three ways:reactions (1) using chemoselective. stereospecific Stereoretention methodology, like of theNegishiα-isoprene carboalumination double bond and was achievedcross in-coupling; three ways: (2) stereoselective (1) using stereospeciﬁc alkene syntheses; methodology, and (3) takin likeg advantage Negishi of carboalumination the stereochemical and cross-coupling;outcome of S (2)N2 stereoselectivesubstitutions. alkene syntheses; and (3) taking advantage of the stereochemical outcome of SN2 substitutions.

Table 9. Summary of homologation and side chain extension methods.

Methods Advantages Disadvantages Section 6.1. Homologation -High yields throughout the synthesis (87–93%) -Method is applicable to a wide scope of benzo- and naphthoquinones -Stereochemistry of the α-isoprene double Lipshutz et al. bond is deﬁned by the conﬁguration of the -Requires the use of hydrogen Homologation to Negishi organoalane chloride gas Cross-Coupling [90] -Regiocontrolled by the installation of the chloromethyl group at the C3 position -No extraneous coordination complex synthesis required -3 step synthesis (including starting material) Molecules 2020, 25, 4477 30 of 37

Table 9. Cont.

Methods Advantages Disadvantages -High yields throughout the synthesis for all reported steps (80–95%) -Requires the use of protecting groups Mehta et al. -Strict use of stereoselective alkene and oxidation manipulations Stereoselective Alkene syntheses -11 step synthesis (not including Syntheses [95] -Methodology is applicable to full side starting material) chain extensions and smaller segments Section 6.2. Side Chain Extension -Moderate to high yields throughout the synthesis (68–90%) Masaki et al. Stereoretention of the α-isoprene double -4 step synthesis (not including Tosylate Substitution [96,97] bond with minor isomerization (5–7%) starting material) Methodology is applicable to full side chain extensions and smaller segments -Achieved regio- and stereocontrol -Low to moderate yields (51–79%) for Schmid et al. using isoprene oxide in a 1,4-addition alkylation step Organocuprate Substitution -Stereoretention of the α-isoprene -4 step synthesis (not including [63] double bond (97:3) starting material) -Iterative methodology

7. Conclusions Menaquinones are a biologically important class of isoprenoid compounds that comprise of a methylated naphthoquinone unit and an isoprene side chain of various lengths and levels of saturation. The intrinsic hydrophobicity of these compounds makes it exceedingly diﬃcult to analyze the activity in aqueous-based in vitro assays and growth studies. The moderately water soluble, truncated derivatives, MK-1 through MK-3, have made it possible to test and measure the activities of these compounds, especially in membrane-associated proteins. The most popular method used to synthesize menaquinones and derivatives has been BF OEt -catalyzed Friedel-Crafts reaction. This reaction 3· 2 typically occurs over two steps from commercially available starting material, but it generally produces low yields and a mixture of isomers. Although Friedel-Crafts alkylation is the conventional, concise method, in this review we aimed to compile and evaluate syntheses of menaquinones from all the literature to summarize the advantages and disadvantages of each route based on overall yield, regioselectivity, and stereoretention of the α-isoprene double bond, as shown in Tables4 and6–9. We further evaluated the reactions to consider the number of steps in the synthesis, minimization of side reactions, and overall safety to highlight one representative method in each category, as outlined in Table 10. With examination of Table 10, it is possible to compare the syntheses across diﬀerent approaches. As a representative of the nucleophilic ring methods, Swenton and coworkers were chosen for the clean use of electrolysis to protect the quinone as bismethyl ketals. Simple acid hydrolysis deprotection of the ketals reformed the carbonyls of the quinone without oxidizing agents. As a representative of the metal-mediated and radical reactions, Liebeskind et al. were chosen for the easily synthesized cobalt complex and internal alkynes. The updated cobalt complex was found to be more tolerant toward a wide scope of Lewis acid catalysts, and, as a result decreased the amount of catalyst required to synthesize the product. Rüttimann and Büchi designed an auxiliary-directed Diels-Alder reaction using commercially available materials, cyclopentadiene and menadione. Alkylation of the adduct occurred chemoselectively, and then facile retro-Diels-Alder formed the desired product in high yields and excellent regioselectivity. Lipshutz et al. used homologation and Negishi carboalumination and cross-coupling as side chain extension methods to attach the prenyl side chain. The stereochemistry of the α-isoprene double bond was easily installed using Negishi carboalumination of the corresponding terminal alkyne. Finally, the representative electrophilic ring strategy was done in one synthesis of menaquinones using Michael additions was presented. The method developed by Naruta and Maruyuma retained the stereochemistry of α-isoprene double bond but was not regioselective due to competing 1,2- and 1,4-addition mechanisms. Molecules 2020, 25, 4477 31 of 37

Table 10. Summary of the best reactions within each strategy to be compared to each other.

Strategy- Advantages Disadvantages -Unique use of electrolysis as a protection method -Lithium organocuprate nucleophile -High yields for all reported steps ( 85%) Nucleophilic Ring 2.2. ≥ only used one of two bisketal -Regiocontrol through lithium bromide Transmetalation rings—poor atom economy exchange Swenton and coworkers -Difficult purification because of -Stereoretention of α-isoprene double bond Electrolysis and unreacted starting materials (<5% Z alkene estimated) Lithium Organocuprate [50,51] -5 step synthesis (not including starting -Deprotection of bisketals to menaquinone material) ring structure via hydrolysis, no oxidation required -High yields (>86%) Metal-Mediated -Simple coordination complex synthesis 3.2. Coordination Complex required using commercially available -The authors did not address α-isoprene Liebeskind et al. materials double bonds isomerization Cobalt Complex Cycloaddition -The regiochemistry of the system is [75] controlled by the alkynes 43a and 43b -1 step synthesis (not including catalyst) Electrophilic Ring 4.1. 1,2- vs. 1,4-Addition -Low yields for both formations Naruta and Maruyuma -Stereoretention of the α-isoprene double (30–48%) Organostannane bond -Prominent competition between C2 and Michael Addition C3 alkylation [81,83] -Uses commercially available starting Pericyclic materials (menadione and cyclopentadiene) 5.1. Diels-Adler -High regiocontrol through adduct 66 Rüttimann et al. -Slight competition between -Stereoretention of α-isoprene double bond Auxiliary-Directed C-alkylation and O-alkylation -Cyclopentadiene can be recycled Diels-Alder -High yields throughout the synthesis [43] ( 90%) ≥ -High yields throughout the synthesis (87–93%) -Method is applicable to a wide scope of Homologation & Side Chain benzo- and naphthoquinones Extensions -Stereochemistry of the α-isoprene double 6.1. Homologation -Requires the use of hydrogen chloride bond is defined by the configuration of the Lipshutz et al. gas organoalane Homologation to Negishi -Regiocontrolled by the installation of the Cross-Coupling [90] chloromethyl group at the C3 position -No extraneous coordination complex synthesis required

Although the methods described in Table 10 are short, high yielding, and selective, there are a few more methods that should be considered if a new synthetic target can accommodate a longer synthesis or the processes in Table 10 are incompatible for the particular target. The homologation method developed by Mehta et al. uses two one-carbon installations at the beginning to synthesize provide a functional handle to stereoselectively install the α-isoprene double bond, and thus can be a strong competitor to the other reactions. Even though many of the methods described have great success using prenyl halides, the reagents are expensive, particularly in comparison to synthesizing MK-derivatives from cheap prenyl alcohols. Saá & coworkers were successful in using commercially available prenyl aldehydes as electrophiles (no distillation necessary). Reduction of the resultant benzyl alcohol was achieved using Birch hydrogenolysis conditions to maintain the stereochemistry of the α-isoprene double bond. Despite the notorious use of Friedel-Crafts alkylations, Schmid et al. discovered a unique intramolecular Friedel-Crafts alkylation beginning with the O-alkylated product using BF OEt . This route adds at two steps, not including synthesis of the starting material and 3· 2 one protecting group, but the additional steps improved the overall yield, as well as the regio- and stereochemistry. Lastly, even though Jacobsen and Torssell only synthesized MK-1, their method Molecules 2020, 25, 4477 32 of 37 would most likely tolerate polyprenylcarboxylic acids based on the stereoretentive results of radical coupling with organotellurides reported by Yamago et al. With respect to Michael additions, the method developed by Naruta and Maruyuma was plagued by competing mechanisms (1,2- vs. 1,4-addition). It appears that alkylstannanes are not chemoselective enough to produce the Michael product for C-C bonds; however, many advancements have been made regarding Michael additions in general, suggesting more regioselective Michael additions may now be available for the synthesis of menaquinone and derivatives. Recently, Michael additions have been successfully used to synthesize C-N bonds using primary amines to form new analogues for biological testing. Recently, Salunke-Gawali and coworkers have used n-alkylamines and aminophenols to synthesize naphthoquinone derivatives which have been reported to assess their antiproliferative and antibacterial activities [98,99]. Salunke-Gawali and coworkers used single crystal x-ray crystallography to determine the structures of these compounds. The incorporation of C-N bonds introduces more hydrogen bonding sites for both intra- and intermolecular interactions. Similarly, Zacconi et al. synthesized naphthoquinone derivatives using benzylamine and 2-phenylethylamine to evaluate their isothermal solubility in supercritical carbon dioxide [100]. All these MK-derived compounds were solids, compared to the typical oil produced for the all-C-atoms MK-derivatives. It would be of great interest to synthesize truncated menaquinones with N-containing isoprenyl side chains and study their redox potentials, solubility, and potential solid-state structure compared to the compounds already reported. However, as demonstrated by this review, many attractive strategies are available that could be used to synthesize potential targets even if speciﬁc geometries are required.

Author Contributions: The contributions to the work are as follows conceptualization, D.C.C. and M.B.-T.; methodology, M.B.-T. and D.C.C.; investigation, M.B.-T. and D.C.C.; writing—original draft preparation, M.B.-T.; writing—review and editing, M.B.-T. and D.C.C.; supervision, D.C.C.; project administration, D.C.C.; funding acquisition, D.C.C. All authors have read and agreed to the published version of the manuscript. Funding: This research was partially funded by National Science Foundation (CHE-1709564) and Colorado State University, particularly for literature access. The APC was funded by MDPI. Acknowledgments: We thank MDPI for the invitation to write this review. We thank Ryan Cerbone for the substantial feedback provided on early drafts of this review. Conﬂicts of Interest: The authors declare no conﬂict of interest.

Abbreviations β-CD: β-cyclodextrin. BIHY: Birch hydrogenolysis. CAN: ceric ammonium nitrate. DCM: dichloromethane, methylene chloride. DIBAL-H: diisobutylaluminum hydride. EtOH: ethanol. HMDS: hexamethyldisilazane. HMDSZ: outdated term, hexamethyldisilazane. HMPA: hexamethylphosphoramide. HPLC: high performance liquid chromatography. Hν: light. Imid: imidazole. LiOt-Bu: lithium tert-butoxide. MeCN: acetonitrile. MeOH: methanol. MK: menaquinone, methylnaphthoquinone. NaHMDS: sodium hexamethylsilazide, sodium hexamethyldisilazane. n-BuLi: n-butyl lithium. PCC: pyridinium chlorochromate. PPh3: triphenylphosphine. Pyr: pyridine. SN2: substitution nucleophilic bimolecular. TBSCl: tert-butyldimethylsilyl chloride. t-BuLi: t-butyl lithium. t-BuOH: tert-butyl alcohol. t-BuOK: potassium tert-butoxide. THF: tetrahydrofuran. TMSCl: trimethylsilyl chloride. Tol: tolyl. U: ubiquinone, benzoquinone.

References

1. Kroppenstedt, R.M.; Mannheim, W. Lipoquinones in Members of the Family Pasteurellaceae. Int. J. Syst. Bacteriol. 1989, 39, 304–308. [CrossRef] 2. Collins, M.D.; Jones, D. Distribution of Isoprenoid Quinone Structural Types in Bacteria and Their Taxonomic Implications. Microbiol. Rev. 1981, 45, 316–354. [CrossRef] 3. Bergdoll, L.; Ten Brink, F.; Nitschke, W.; Picot, D.; Baymann, F. From low- to high-potential bioenergetic chains: Thermodynamic constraints of Q-cycle function. Biochim. Biophys. Acta 2016, 1857, 1569–1579. [CrossRef][PubMed] 4. Epand, R.M.; Walker, C.; Epand, R.F.; Magarvey, N.A. Molecular mechanisms of membrane targeting antibiotics. Biochim. Biophys. Acta 2016, 1858, 980–987. [CrossRef][PubMed] Molecules 2020, 25, 4477 33 of 37

5. Vilcheze, C.; Hartman, T.; Weinrick, B.; Jain, P.; Weisbrod, T.R.; Leung, L.W.; Freundlich, J.S.; Jacobs, W.R., Jr. Enhanced respiration prevents drug tolerance and drug resistance in Mycobacterium tuberculosis. Proc. Natl. Acad. Sci. USA 2017, 114, 4495–4500. [CrossRef][PubMed] 6. Upadhyay, A.; Fontes, F.L.; Gonzalez-Juarrero, M.; McNeil, M.R.; Crans, D.C.; Jackson, M.; Crick, D.C. Partial Saturation of Menaquinone in Mycobacterium tuberculosis: Function and Essentiality of a Novel Reductase, MenJ. ACS Cent. Sci. 2015, 1, 292–302. [CrossRef] 7. Nakagawa, K.; Hirota, Y.; Sawada, N.; Yuge, N.; Watanabe, M.; Uchino, Y.; Okuda, N.; Shimomura, Y.; Suhara, Y.; Okano, T. Identification of UBIAD1 as a novel human menaquinone-4 biosynthetic enzyme. Nature 2010, 468, 117–121. [CrossRef] 8. Mahdinia, E.; Demirci, A.; Berenjian, A. Production and application of menaquinone-7 (vitamin K2): A new perspective. World J. Microbiol. Biotechnol. 2017, 33, 1–7. [CrossRef] 9. Tarento, T.D.C.; McClure, D.D.; Talbot,A.M.; Regtop, H.L.; Biffin, J.R.; Valtchev,P.;Dehghani, F.; Kavanagh, J.M. A potential biotechnological process for the sustainable production of vitamin K1. Crit. Rev. Biotechnol. 2019, 39, 1–19. [CrossRef] 10. Zhang, C.; Wu, D.; Ren, H. Economical production of vitamin K2 using crude glycerol from the by-product of biodiesel. Sci. Rep. 2020, 10, 1–8. [CrossRef] 11. Dunphy, P.J.; Brodie, A.F. [233] The structure and function of quinones in respiratory metabolism. In Methods Enzymology; Elsevier: Amsterdam, The Netherlands, 1971; Volume 18, pp. 407–461. 12. Koehn, J.T.; Magallanes, E.S.; Peters, B.J.; Beuning, C.N.; Haase, A.A.; Zhu, M.J.; Rithner, C.D.; Crick, D.C.; Crans, D.C. A Synthetic Isoprenoid Lipoquinone, Menaquinone-2, Adopts a Folded Conformation in Solution and at a Model Membrane Interface. J. Org. Chem. 2018, 83, 275–288. [CrossRef] 13. Tetali, S.D. Terpenes and isoprenoids: A wealth of compounds for global use. Planta 2019, 249, 1–8. [CrossRef] [PubMed] 14. Ferreira, V.F.; Ferreira, V.; Nicoletti, C.; Ferreira, P.; Futuro, D.; da Silva, F. Strategies for Increasing the Solubility and Bioavailability of Anticancer Compounds: β-Lapachone and Other Naphthoquinones. Curr. Pharm. Des. 2016, 22, 5899–5914. [CrossRef][PubMed] 15. Ingles, D.P.; Cruz Rodriguez, J.B.; Garcia, H. Supplemental Vitamins and Minerals for Cardiovascular Disease Prevention and Treatment. Curr. Cardiol. Rep. 2020, 22, 1–8. [CrossRef] 16. Gengenbacher, M.; Kaufmann, S.H. Mycobacterium tuberculosis: Success through dormancy. FEMS Microbiol. Rev. 2012, 36, 514–532. [CrossRef] 17. Bahuguna, A.; Rawat, D.S. An overview of new antitubercular drugs, drug candidates, and their targets. Med. Res. Rev. 2020, 40, 263–292. [CrossRef][PubMed] 18. Karamzad, N.; Maleki, V.; Carson-Chahhoud, K.; Azizi, S.; Sahebkar, A.; Gargari, B.P. A systematic review on the mechanisms of vitamin K effects on the complications of diabetes and pre-diabetes. Biofactors 2020, 46, 21–37. [CrossRef] 19. Turner, M.E.; Adams, M.A.; Holden, R.M. The Vitamin K Metabolome in Chronic Kidney Disease. Nutrients 2018, 10, 1076. [CrossRef] 20. Simes, D.C.; Viegas, C.S.B.; Araujo, N.; Marreiros, C. Vitamin K as a Diet Supplement with Impact in Human Health: Current Evidence in Age-Related Diseases. Nutrients 2020, 12, 138. [CrossRef] 21. Schurgers, L.J.; Teunissen, K.J.; Hamulyak, K.; Knapen, M.H.; Vik, H.; Vermeer, C. Vitamin K-containing dietary supplements: Comparison of synthetic vitamin K1 and natto-derived menaquinone-7. Blood 2007, 109, 3279–3283. [CrossRef] 22. Bentley, R.; Meganathan, R. Biosynthesis of Vitamin K (Menaquinone) in Bacteria. Microbiol. Rev. 1982, 46, 241–280. [CrossRef] 23. Conly, J.; Stein, K. The production of menaquinones (vitamin K2) by intestinal bacteria and their role in maintaining coagulation homeostasis. Prog. Food Nutr. Sci. 1992, 16, 307–343. [PubMed] 24. Kurosu, M.; Begari, E. Vitamin K2 in electron transport system: Are enzymes involved in vitamin K2 biosynthesis promising drug targets? Molecules 2010, 15, 1531–1553. [CrossRef][PubMed] 25. Cockayne, S.; Adamson, J.; Lanham-New, S.; Shearer, M.J.; Gilbody, S.; Torgerson, D.J. Vitamin K and the Prevention of Fractures. Arch. Intern. Med. 2006, 166, 1256–1261. [CrossRef][PubMed] 26. Kumar, S.; Koehn, J.T.; Gonzalez-Juarrero, M.; Crans, D.C.; Crick, D.C. Mycobacterium tuberculosis survival in J774A.1 cells is dependent on MenJ moonlighting activity, not its enzymatic activity. ACS Infect. Dis. 2020. Accepted. [CrossRef][PubMed] Molecules 2020, 25, 4477 34 of 37

27. Debnath, J.; Siricilla, S.; Wan, B.; Crick, D.C.; Lenaerts, A.J.; Franzblau, S.G.; Kurosu, M. Discovery of selective menaquinone biosynthesis inhibitors against Mycobacterium tuberculosis. J. Med. Chem. 2012, 55, 3739–3755. [CrossRef] 28. Abe, T.; Ozaki, S.; Ueda, D.; Sato, T. Insight into Isoprenoid Biosynthesis by Functional Analysis of Isoprenyl Diphosphate Synthases from Mycobacterium vanbaalenii and Mycobacterium tuberculosis. ChemBioChem 2020, 21, 1–9. [CrossRef] 29. Prince, R.C.; Dutton, P.L.; Bruce, J.M. Menaquinones and plastoquinones in aprotic solvents. FEBS Lett. 1983, 160, 273–276. [CrossRef] 30. Guin, P.S.; Das, S.; Mandal, P.C. Electrochemical Reduction of Quinones in Different Media: A Review. Int. J. Electrochem. 2011, 2011, 816202. [CrossRef] 31. Dryhurst, G.; Kadish, K.M.; Scheller, F.; Renneberg, R. Biological Electrochemistry; Elsevier: Amsterdam, The Netherlands, 1982. 32. Jaworski, J.S.; Leniewska, E.; Kalinowski, M.K. Solvent Effect on the Redox Potential of Quinone-Semiquinone Systems. J. Electroanal. Chem. Interfacial Electrochem 1979, 105, 329–334. [CrossRef] 33. Gómez-Murcia, V.; Torrecillas, A.; de Godos, A.M.; Corbalán-García, S.; Gómez-Fernández, J.C. Both idebenone and idebenol are localized near the lipid–water interface of the membrane and increase its fluidity. Biochim. Biophys. Acta Biomembr. 2016, 1858, 1071–1081. [CrossRef][PubMed] 34. Kaurola, P.; Sharma, V.; Vonk, A.; Vattulainen, I.; Rog, T. Distribution and dynamics of quinones in the lipid bilayer mimicking the inner membrane of mitochondria. Biochim. Biophys. Acta 2016, 1858, 2116–2122. [CrossRef][PubMed] 35. Quinn, P.J.; Esfahani, M.A. Ubiquinones have Surface-Active Properties Suited to Transport Electrons and Protons across Membranes. Biochem. J. 1980, 185, 715–722. [CrossRef][PubMed] 36. Roche, Y.; Peretti, P.; Bernard, S. The Redox State Influences the Interactions of Ubiquinones with Phospholipid Bilayers: A DSC Study. J. Therm. Anal. Calorim. 2007, 89, 867–873. [CrossRef] 37. Koehn, J.T.; Beuning, C.N.; Peters, B.J.; Dellinger, S.K.; Van Cleave, C.; Crick, D.C.; Crans, D.C. Investigating Substrate Analogues for Mycobacterial MenJ: Truncated and Partially Saturated Menaquinones. Biochemistry 2019, 58, 1596–1615. [CrossRef] 38. Koehn, J.T.; Crick, D.C.; Crans, D.C. Synthesis and Characterization of Partially and Fully Saturated Menaquinone Derivatives. ACS Omega 2018, 3, 14889–14901. [CrossRef] 39. Upadhyay, A.; Kumar, S.; Rooker, S.A.; Koehn, J.T.; Crans, D.C.; McNeil, M.R.; Lott, J.S.; Crick, D.C. Mycobacterial MenJ: An Oxidoreductase Involved in Menaquinone Biosynthesis. ACS Chem. Biol. 2018, 13, 2498–2507. [CrossRef] 40. Ishihara, M.; Sakagami, H. QSAR of Molecular Structure and Cytotoxic Activity of Vitamin K2 Derivatives with Concept of Absolute Hardness. Anticancer Res. 2007, 27, 4059–4064. 41. Siricilla, S.; Mitachi, K.; Skorupinska-Tudek, K.; Swiezewska, E.; Kurosu, M. Biosynthesis of a water-soluble lipid I analogue and a convenient assay for translocase I. Anal. Biochem. 2014, 461, 36–45. [CrossRef] 42. Daines, A.M.; Payne, R.J.; Humphries, M.E.; Abell, A.D. The Synthesis of Naturally Occurring Vitamin K and Vitamin K Analogues. Curr. Org. Chem. 2003, 7, 1–15. [CrossRef] 43. Rüttimann, A. Recent Advances in the Synthesis of K-Vitamins. Chimia 1986, 40, 290–306. 44. Snyder, C.D.; Rapoport, H. Synthesis of menaquinones. J. Am. Chem. Soc. 1974, 96, 8046–8054. [CrossRef] [PubMed] 45. Tabushi, I.; Fujita, K.; Kawakubo, H. One-step preparation of vitamin K1 or K2 analogs by cyclodextrin inclusion catalysis. J. Am. Chem. Soc. 1977, 99, 6456–6457. [CrossRef][PubMed] 46. Tabushi, I.; Yamamura, K.; Fujita, K.; Kawakubo, H. Specific inclusion catalysis by beta.-cyclodextrin in the one-step preparation of vitamin K1 or K2 analogs. J. Am. Chem. Soc. 1979, 101, 1019–1026. [CrossRef] 47. Garcias, X.; Ballester, P.; Capo, M.; Saá, J.M. 2.DELTA.-Stereocontrolled Entry to (E)- or (Z)-Prenyl Aromatics and Quinones. Synthesis of Menaquinone-4. J. Org. Chem. 1994, 59, 5093–5096. [CrossRef] 48. Ballester, P.; Capo, M.; Garcias, X.; Saá, J.M. Synthesis of prenylated quinones by the oxidative degradation approach. Birch vs vinylogous Birch hydrogenolysis (BIHY vs VIBIHY) in controlling 2.DELTA. stereochemistry of the prenyl chain. J. Org. Chem. 1993, 58, 328–334. [CrossRef] 49. Ballester, P.; Saá, J.; Capo, M. Prenylation of Aromatics. Lithium-Ammonia Reduction of 0-Silylated Tertiary Cinnamyl Alcohols. Tetrehedron Lett. 1990, 31, 1339–1342. [CrossRef] Molecules 2020, 25, 4477 35 of 37

50. Raynolds, P.W.; Manning, M.J.; Swenton, J.S. Utilization of quinone bisacetal organocopper compounds in the synthesis of isoprenoid systems. ChemComm 1977, 14, 499–500. [CrossRef] 51. Chenard, B.L.; Manning, M.J.; Raynolds, P.W.; Swenton, J.S. Organocopper chemistry of quinone bisketals. Application to the synthesis of isoprenoid quinone systems. J. Org. Chem. 1980, 45, 378–384. [CrossRef] 52. Lindlar, H. Verfahren zur Herstellung von Kondensationsprodukten. U.S. Patent 320,582, 12 August 1953.

53. Isler, O.; Doebel, K. Synthesen in der Vitamin-K-Reihe I. Über totalsynthetisches Vitamin K 1. Helv. Chim. Acta 1954, 37, 225–233. [CrossRef] 54. Hirschmann, R.; Miller, R.; Wendler, N.L. The Synthesis of Vitamin K1. J. Am. Chem. Soc. 1954, 76, 4592–4594. [CrossRef]

55. Fieser, L.F. Synthesis of Vitamin K 1. J. Am. Chem. Soc. 1939, 61, 3467–3475. [CrossRef] 56. Fieser, L.F.; Campbell, W.P.; Fry, E.M. Synthesis of Quinones Related to Vitamins K 1 and K 2. J. Am. Chem. Soc. 1939, 61, 2206–2218. [CrossRef] 57. Fieser, L.F.; Campbell, W.P.; Fry, E.M.; Gates, M.D. Synthetic Approach to Vitamin K1. J. Am. Chem. Soc. 1939, 61, 2559. [CrossRef] 58. Almquist, H.J.; Klose, A.A. Synthetic and Natural Antihemorrhagic Compounds. J. Am. Chem. Soc. 1939, 61, 2557–2558. [CrossRef] 59. Binkley, S.B.; Cheney, L.C.; Holcomb, W.F.; McKee, R.W.; Thayer, S.A.; MacCorquodale, D.W.; Doisy, E.A.

The Constitution and Synthesis of Vitamin K1. J. Am. Chem. Soc. 1939, 61, 2558–2559. [CrossRef] 60. Klose, A.A.; Almquist, H.J. Synthesis of Vitamink K1. J. Biol. Chem. 1940, 132, 469–471. 61. Isler, O. Process for Producing Condensation Products of 1,4-Naphthohydroquinones. U.S. Patent 2325681, 3 August 1943.

62. Tishler, M.; Fieser, L.F.; Wendler, N.L. Nature of the By-product in the Synthesis of Vitamin K 1. J. Am. Chem. Soc. 1940, 62, 1982–1991. [CrossRef] 63. Schmid, R.; Antoulas, S.; Rüttimann, A.; Schmid, M.; Vecchi, M.; Weiserb, H. Synthesis of All Four Stereoisomers of (E)-Vitamin KT (Phylloquinone), Analysis of Their Diastereoisomeric and Enantiomeric Purities and Determination of Their Biopotencies. Helv. Chim. Acta 1990, 73, 1276–1299. [CrossRef] 64. Yoshizawa, T.; Toyofuku, H.; Tachibana, K.; Kuroda, T. Regioselective Polyprenyl Rearrangement of Polyprenyl 2,3,4,5-Tetrasubstituted Phenyl Ethers Promoted by Boron Triﬂuoride. Chem. Lett. 1982, 11, 1131–1134. [CrossRef] 65. Min, J.-H.; Lee, J.-S.; Yang, J.-D.; Koo, S. The Friedel Crafts Allylation of a Prenyl Group Stabilized by a − Sulfone Moiety: Expeditious Syntheses of Ubiquinones and Menaquinones. J. Org. Chem. 2003, 68, 7925–7927. [CrossRef][PubMed]

66. Coman, S.M.; Parvulescu, V.I.; Wuttke, S.; Kemnitz, E. Synthesis of Vitamin K 1 and K 1 -Chromanol by Friedel–Crafts Alkylation in Heterogeneous Catalysis. ChemCatChem 2010, 2, 92–97. [CrossRef] 67. Wuttke, S.; Coman, S.M.; Scholz, G.; Kirmse, H.; Vimont, A.; Daturi, M.; Schroeder, S.L.M.; Kemnitz, E. Novel Sol-Gel Synthesis of Acidic MgF2 x(OH)x Materials. Chem. Eur. J. 2008, 14, 11488–11499. [CrossRef] − [PubMed] 68. Scholz, G.; Stosiek, C.; Feist, M.; Kemnitz, E. Magnesium Hydroxide Fluorides-New Materials with Adjustable Composition and Properties. Eur. J. Inorg. Chem. 2012, 2012, 2337–2340. [CrossRef] 69. Sato, K.; Inoue, S.; Saito, K. A New Synthesis of Vitamin K via π-Allylnickel Intermediates. J. Chem. Soc. Perkin Trans. I 1973, 2289–2293. [CrossRef] 70. Godschalx, J.P.; Stille, J.K. Cross-coupling reaction of allyl bromides with organotin reagents catalyzed by zinc chloride. Tetrahedron Lett. 1983, 24, 1905–1908. [CrossRef] 71. Liebeskind, L.S.; Foster, B.S. Stannylquinones. Synthesis and utilization as quinone carbanion synthetic equivalents. J. Am. Chem. Soc. 1990, 112, 8612–8613. [CrossRef] 72. Dötz, K.H. Carbenkomplexe in der Organischen Synthese. Angew. Chem. 1984, 96, 573–594. [CrossRef] 73. Dötz, K.H.; Pruskil, I.; Muhlemeier, I. Vitamin-Synthesen mit Carben-Komplexen, II. Carbonyl(carben)-Komplex-induzierte Synthese von Vitaminen der K1- und K2-Reihe. Chem. Ber. 1982, 115, 1278. 74. Liebeskind, L.S.; Baysdon, S.L.; South, M.S. An organotransition-metal synthesis of naphthoquinones. J. Am. Chem. Soc. 1980, 102, 7397–7398. [CrossRef] Molecules 2020, 25, 4477 36 of 37

75. Liebeskind, L.S.; Baysdon, S.L.; Goedken, V.; Chidambaram, R. Phthaloylcobalt complexes in synthesis. Ligand modifications leading to a practical naphthoquinone synthesis. Organometallics 1986, 5, 1086–1092. [CrossRef] 76. Jacobsen, N.; Torssell, K. Radikalische Alkylierung von Chinonen: Erzeugung von Radikalen in Redoxreaktionen. Liebigs Ann. 1972, 763, 135–147. [CrossRef] 77. Jacobsen, N.; Torssell, K. Synthesis of Naturally Occuring Quinones: Alkylation with the Silver Ion-Peroxydisulphite-Carboxylic Acid System. Acta Chem. Scand. 1973, 27, 3211–3216. [CrossRef] 78. Yamago, S.; Hashidume, M.; Yoshida, J.-I. Radical-Mediated Synthesis of Substituted Quinones with Organotellurium Compounds. Chem. Lett. 2000, 29, 1234–1235. [CrossRef] 79. Yamago, S.; Hashidume, M.; Yoshida, J.-I. A new synthetic route to substituted quinones by radical-mediated coupling of organotellurium compounds with quinones. Tetrahedron 2002, 58, 6805–6813. [CrossRef] 80. Coppa, F.; Fontana, F.; Minisci, F.; Nogueira, M.C.; Vismara, E. Homolytic Alkylation of Naphthoquinone and Methyl-Naphthoquinone. Enthalpic, Steric, and Polar Effects. Tetrahedron 1991, 47, 7347–7352. [CrossRef] 81. Naruta, Y. Regio- and stereoselective synthesis of coenzymes Qn (n = 2–10), vitamin K, and related polyprenylquinones. J. Org. Chem. 1980, 45, 4097–4104. [CrossRef] 82. Naruta, Y. Allylation of quinones with allyltin reagents. J. Am. Chem. Soc. 1980, 102, 3774–3783. [CrossRef] 83. Maruyama, K.; Naruta, Y. Synthesis of naturally occurring quinones. Part 3. Allylation of quinones with allyltin reagents. New synthesis of coenzyme Q1 and plastoquinone-1. J. Org. Chem. 1978, 43, 3796–3798. [CrossRef] 84. Araki, S.; Katsumura, N.; Butsugan, Y. Allylation of quinones by allylic indium reagents. J. Organomet. Chem. 1991, 415, 7–24. [CrossRef] 85. Evans, D.A.; Hoffman, J.M. Regiospecific quinone isoprenylation. Examples of remarkably facile [3,3] sigmatropic processes. J. Am. Chem. Soc. 1976, 98, 1983–1984. [CrossRef] 86. Troll, T.; Schmid, K. Darstellung und reaktionen von 1,3-bis-(trimethylsiloxy)-isobenzofuranen. Tetrahedron Lett. 1984, 25, 2981–2984. [CrossRef] 87. Büchi, G.H.; Rüttimann, A. Process for the Manufacture of Quinone Derivatives. U.S. Patent 4,603,223, 29 July 1986. 88. Tso, H.-H.; Chen, Y.-J. A convenient one-flask synthesis of vitamin K. J. Chem. Res. Synop. 1995, 3, 104–105. 89. Mal, D.; Ghosh, K.; Jana, S. Synthesis of Vitamin K and Related Naphthoquinones via Demethoxycarbonylative Annulations and a Retro-Wittig Rearrangement. Org. Lett. 2015, 17, 5800–5803. [CrossRef][PubMed] 90. Lipshutz, B.H.; Kim, S.-K.; Mollard, P. An Expeditious Route to COQn, Vitamins K1 and K2, and Related Allylated para-Quinones Utilizing Ni(0) Catalysis. Tetrahedron 1998, 54, 1241–1253. [CrossRef] 91. Van Horn, D.E.; Negishi, E. Selective carbon-carbon bond formation via transition metal catalysts. 8. Controlled carbometalation. Reaction of acetylenes with organoalane-zirconocene dichloride complexes as a route to stereo- and regio-defined trisubstituted olefins. J. Am. Chem. Soc. 1978, 100, 2252–2254. [CrossRef] 92. Negishi, E.; Van Horn, D.E.; Yoshida, T. Controlled carbometalation. 20. Carbometalation reaction of alkynes with organoalene-zirconocene derivatives as a route to stereo-and regiodefined trisubstituted alkenes. J. Am. Chem. Soc. 1985, 107, 6639–6647. [CrossRef] 93. Matsushita, H.; Negishi, E. Palladium-catalyzed Synthesis of 1,4-Diene by Allylation of Alkenylalanes: α-Farnesene. Org. Synth. 1984, 62, 31–37. 94. Negishi, E. Bimetallic catalytic systems containing Ti, Zr, Ni, and Pd. Their applications to selective organic syntheses. Pure Appl. Chem. 1981, 53, 2333–2356. [CrossRef] 95. Mehta, D.S.; Mohan, P.; Shastri, M.; Reid, T. Method of Preparation of Stereospecific Quinone Derivatives. U.S. Patent 9,464,021, 11 October 2016. 96. Masaki, Y.; Hasimoto, K.; Sakuma, K.; Kaji, K. Synthetic Studies on Isoprenoidquinones. I. A Facile Regio- and Stereocontrolled Synthesis. Chem. Pharm. Bull. 1984, 32, 3952–3958. [CrossRef] 97. Masaki, Y.; Hasimoto, K.; Kaji, K. Synthetic Studies of Isoprenoidquinones II. Syntheses of Ubiquinone-10, Phylloquinone, and Menaquinone-4 by a Chain-Extending Method Utilizing Terminally Functionalized Isoprenoidhydroquinones. Chem. Pharm. Bull. 1984, 32, 3959–3967. [CrossRef] 98. Chadar, D.; Camilles, M.; Patil, R.; Khan, A.; Weyhermüller, T.; Salunke-Gawali, S. Synthesis and characterization of n-alkylamino derivatives of vitamin K3: Molecular structure of 2-propylamino-3-methyl-1,4-naphthoquinone and antibacterial activities. J. Mol. Struct. 2015, 1086, 179–189. [CrossRef] Molecules 2020, 25, 4477 37 of 37

99. Kathawate, L.; Joshi, P.V.; Dash, T.K.; Pal, S.; Nikalje, M.; Weyhermüller, T.; Puranik, V.G.; Konkimalla, V.B.; Salunke-Gawali, S. Reaction between lawsone and aminophenol derivatives: Synthesis, characterization, molecular structures and antiproliferative activity. J. Mol. Struct. 2014, 1075, 397–405. [CrossRef] 100. Zacconi, F.C.; Nuñez, O.N.; Cabrera, A.L.; Valenzuela, L.M.; del Valle, J.M.; de la Fuente, J.C. Synthesis and solubility measureme11nt in supercritical carbon dioxide of two solid derivatives of 2-methylnaphthalene-1,4-dione (menadione): 2-(Benzylamino)-3-methylnaphthalene-1,4-dione and 3-(phenethylamino)-2-methylnaphthalene-1,4-dione. J. Chem. Thermodyn. 2016, 103, 325–332. [CrossRef]

© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).