Design, Synthesis, and Characterization of Novel sn-1 Heterocyclic DAG-Lactones as PKC Activators

Abstract

DAG-lactones represent useful templates for the design of potent and selective C1 domain ligands for PKC isozymes. The ester moiety at the sn-1 position, a common feature in this template, is relevant for C1 domain interactions, but it represents a labile group susceptible to endogenous esterases. An interesting challenge involves replacing the ester group of these ligands while still maintaining biological activity. Here, we present the synthesis and functional characterization of novel diacylglycerol-lactones containing heterocyclic ring substituents at the sn-1 position. Our results showed that the new compound 10B12, a DAG-lactone with an isoxazole ring, binds PKCα and PKCε with nanomolar affinity. Remarkably, 10B12 displays preferential selectivity for PKCε translocation in cells and induces a PKCε-dependent reorganization of the actin cytoskeleton into peripheral ruffles in lung cancer cells. We conclude that introducing a stable isoxazole ring as an ester surrogate in DAG-lactones emerges as a novel structural approach to achieve PKC isozyme selectivity.

Graphical Abstract

INTRODUCTION

Protein kinase C (PKC) isozymes comprise a family of signaling proteins that play essential roles in the control of cellular functions as diverse as cell cycle progression, differentiation, apoptosis, proliferation, motility, and gene expression.¹ The central role of PKC in cellular signal transduction has established it as a validated therapeutic target for various pathologies such as cancer, diabetes, Alzheimer’s, and cardiovascular diseases.^2–5

All members of the PKC family contain in their N-terminal regulatory domain one or two copies of C1 domains, a zinc-binding, cysteine-rich motif of 50–51 amino acid residues. In classical/conventional (cPKCs: α, β1, β2, and γ) and novel isoforms (nPKCs: δ, ε, η, and θ), the second messenger diacylglycerol (DAG) binds to the C1a and C1b domain present in their regulatory region, which promotes their translocation to the plasma membrane or other intracellular compartments and subsequent enzyme activation. The C1 domains in cPKCs and nPKCs are also high-affinity binding sites for the phorbol ester tumor promoters, natural compounds that bind to the same site as DAG. Unlike the C1 domains in cPKCs and nPKCs, a single C1 domain present in atypical isoforms (aPKCs: ζ and ι/λ) is unable to bind either phorbol esters or DAG.⁶ Individual members of the PKC family have been shown to confer distinctive patterns of cellular responses, including opposite responses in the context of tumor growth.¹ Although PKCs are the best-studied class of signaling proteins with C1 domains,⁷ other families of proteins containing DAG/phorbol ester-responsive C1 domains have been later identified. These include RasGRPs, which are guanyl exchange factors for Ras; chimaerins, which are GTPase activating proteins for Rac; DAG kinases (DGKs), enzymes responsible for DAG phosphorylation and the generation of phosphatidic acid; MRCK (myotonic dystrophy kinase-related cdc42-binding kinase), protein kinase D isoforms (PKDs), and Munc-13 isoforms.^8,9 Despite the similar overall structure of C1 domains in these proteins, subtle differences may confer differential ligand binding and membrane association properties to individual members of the family, thus encouraging the generation of agents capable of displaying binding selectivity among the different phorbol ester receptors and specificity for activation. This would also be useful to dissect PKC isozyme-specific responses in cellular models.

Among the multiple C1 domain ligands generated as agents capable of activating PKC isozymes, diacylglycerol lactones (DAG-lactones) represent a unique class of DAG mimetics in which the flexibility of the structure has been constrained to reduce the entropic loss due to binding. Indeed, the natural second messenger DAG displays a rather weak affinity for cPKCs and nPKCs, largely reflecting the flexibility of the glycerol backbone, with the consequence that the immobilization of the glycerol upon binding is entropically unfavorable. The design of DAG-lactones, pioneered by the Marquez and Blumberg laboratories, constitutes the greatest synthetic efforts to date, leading to the development of potent C1 domain ligands in which the glycerol backbone of DAG has been cyclized, leading to a rigid structure that eliminates the entropic penalty. Through appropriate manipulation of the hydrophobic side chains, it has been possible to obtain DAG-lactone analogues displaying higher affinity than the natural DAGs and, in many cases, achieving nanomolar binding affinities similar to those of phorbol esters or related compounds derived from natural sources.^10–12

Modeling and docking experiments have revealed two different modes of binding for diacylglycerol itself and the DAG-lactones, defined by the interaction between the C1 domain and either the sn-1 carbonyl or the sn-2 carbonyl in the DAG-lactone structure (Figure 1).¹³ These two comparable binding modes form identical networks of hydrogen bonds with receptor amino acids Thr 12, Leu21, and Gly23, as is observed in the crystal structure of PKCδ C1b with phorbol 13-O-acetate.¹⁴ For convenience in comparing C1 domains from different isozymes, we use residue numbering internal to the C1 domain, with residue 1 corresponding to the first zinc-binding histidine and residue 50 corresponding to the last zinc-binding cysteine.

Figure 1.

Schematic illustration of the C1 domain hydrogen bonding interactions in the alternative sn-1 (A) and sn-2 (B) binding modes for DAG-lactones.

The carbonyl that is not directly involved in C1 domain binding contributes to the binding energy through interactions with the C1 domain surface and the lipid bilayer in the C1 domain–ligand–membrane ternary complex. Thus, both carbonyl groups have been shown to be essential for a strong interaction with PKC. The distinct patterns of acyl (R1) and alkyl (R2) substitution are key elements that determine selectivity, providing significant differences in biological outcome patterns. When this approach was applied, combinatorial libraries of DAG-lactones were produced, leading to compounds with marked selectivity for the RasGRP family relative to their affinity for PKCα and other PKC isozymes.^11,12,15 A study by Ann et al.¹⁶ designed to generate isozyme-specific PKC ligands by the incorporation of linoleic acid derivatives as well as saturated and unsaturated alkyl chains into the side chains of DAG-lactones led to the identification of DAG-lactones with prominent isozyme selectivity. One of these compounds, the DAG-lactone (E)-(2-(hydroxymethyl)-4-(3-isobutyl-5-methylhexylidene)-5-oxotetrahydrofuran-2-yl)methyl pivalate (referred to as AJH-836), exhibits pronounced selectivity in vitro for novel PKCε and PKCδ relative to classical PKCα and PKCβ.¹⁷ Remarkably, AJH-836 displayed significant selectivity for the translocation of PKCε to the plasma membrane (a readout of activation) relative to PKCα. This selectivity could also be observed at a functional level in cellular models. For example, AJH-836 is capable of inducing major changes in the reorganization of the actin cytoskeleton, specifically inducing the formation of peripheral membrane ruffles, a response that is mediated primarily by PKCε.¹⁷ On the other hand, AJH-836 has limited ability to induce transcriptional activation of genes in lung cancer cells, a response that is mainly mediated by PKCα,¹⁸ ultimately reflecting its selectivity for the nPKCs relative to cPKCs.

A fundamental feature common to DAG-lactones is the presence of an ester moiety at the sn-1 position that, as discussed above, is important in mediating membrane and receptor interactions; however, this makes the DAG-lactones susceptible to hydrolysis by endogenous esterases, reducing their stability in cells.¹⁹ Previous attempts to replace this labile functional group led to a significant loss in biological activity. Interestingly, the isosteric replacement of the side-chain ester group at the sn-1 position with an amide or an N-hydroxyl amide group, developed by Marquez and colleagues, revealed a dramatic drop in binding affinity of more than 3 orders of magnitude, suggesting that the new functional groups were not able to adequately recapitulate the polar interactions that the sn-1 carbonyl makes in the C1 domain–membrane–DAG-lactone ternary complex.^19,20 In the present study, we designed and synthesized heterocycle-containing DAG-lactones in order to expand the chemical space covered by these ligands, leading to the development of novel activators of DAG-responsive C1 domain-containing proteins. Using AJH-836 as a lead DAG-lactone structure, we synthesized six new compounds in which we varied the nature of the heterocyclic ring at the sn-1 position, specifically replacing the ester moiety with isoxazole (1–2), triazole (3–4), and tetrazole (5–6) groups (Figure 2). Evaluation of their binding and functional properties in cellular models revealed that a stable isoxazole ring can function as an ester surrogate to allow binding to the C1 domain.

Figure 2.

Structure of new heterocyclic DAG-lactones 1−6 and AJH-836.

RESULTS AND DISCUSSION

Design of the Compounds.

Heterocycle rings play a central role in modern drug design. According to statistics, more than 85% of all biologically active chemical entities contain a heterocycle. The presence of heterocycles provides a useful tool for the modification of solubility, lipophilicity, polarity, and hydrogen bonding capacity of biologically active agents, which results in the optimization of the ADME/Tox properties of drugs or drug candidates.²¹ With the aim of increasing metabolic stability, we selected three popular heterocyclic ring moieties present in numerous biologically active compounds for the design of the new DAG-lactones derivatives. In this sense, isoxazole, triazole, and tetrazole heterocycles have provided improved pharmacological properties that have resulted in the development of wide synthetic strategies and methods for the synthesis of these valuable building blocks.²²

Isoxazole is an azole with an oxygen atom attached to nitrogen often found in natural products and in various pharmaceutical and therapeutic products such as insecticides, antibacterials, antibiotics, antifungals, among others. It is interesting to obtain the isoxazolol derivative since this group represents a bioisostere of the ester group present in our lead structure at the sn-1 position.^23,24

The 1,2,3-triazole unit is a fundamental building block found in different bioactive compounds. One of the most interesting features of triazoles is their ability to mimic the characteristics of different functional groups. This aromatic heterocycle is stable to metabolic and chemical degradation and is capable of acting as a hydrogen bond donor and acceptor, which can be convenient in the binding of biomolecular targets. This explains its wide use as a bioisostere for the synthesis of new active molecules, mainly amide bonds, although there are also some examples where it is used as an ester group isostere to reduce the susceptibility to enzymatic degradation in vivo.²⁵

Tetrazole is a five-member, planar heterocycle with four nitrogen atoms arranged in a regular fashion rarely present in nature. It has attracted significant attention, especially in medicinal chemistry, because it increases drug bioavailability and is resistant to biological degradation, extending their action.^26,27

Molecular Modeling.

After planning the synthesis, a series of heterocycle-containing DAG-lactones was docked into the crystal structure of the PKCδ C1b domain bound to phorbol ester.¹⁴ This domain has a 62% sequence identity with the PKCα and PKCε C1b domains, and the ligand-binding residues are strictly conserved. These docking experiments allowed us to ensure that the designed heterocyclic compounds fit into the binding site, to compare their binding mode to other known and previously studied DAG-lactones, and to analyze the interactions formed by the new heterocyclic rings and the C1 domain.

Docking predicted that the isoxazole-containing compounds, which retain an ether oxygen in the linkage between the heterocycle and the lactone ring, are capable of binding in both sn-1 and sn-2 orientations, like the parent DAG-lactone structures (Figure 3A,B). The triazole and tetrazole DAG-lactones, with a single methylene spacer between the two rings, bind exclusively sn-1 (Figure 3C,D). These compounds cannot bind sn-2 because the shorter linker does not give enough space to the heterocyclic ring, and instead of extending out of the back of the binding slot, it bumps against a side loop without making any productive interactions.

Figure 3.

Docked structures of compound 2 in sn-1 (A) and sn-2 orientations (B) and compounds 4 (C) and 6 (D) in sn-1 orientation. Hydrogen bonds are indicated with dashed orange lines. The receptor crystal structure is the PKCδ C1b domain¹⁴ (PDB ID: 1PTR).

In the sn-1 orientation, the hydrogen bonding interaction between the C1 domain backbone at Gly23 and the DAG-lactone is stronger in the tri- and tetrazole compounds because, with the shorter linker, the ring can adopt an orientation such that the hydrogen bond is in the plane of the ring. With the additional ether oxygen in the linker, in the isoxazole compounds, the NH donor is not aligned with the ring plane, and the hydrogen bonding is a weaker, out-of-plane interaction with an isoxazole oxygen atom.²⁸

Previous studies have shown that DAG-lactones prefer to bind in the sn-2 orientation, particularly when there is a large, branched alkyl chain at R₂, adjacent to the sn-2 carbonyl (Figure 1) as in compounds 2, 4, and 6. In this binding mode, the entropic advantage of constraining the DAG backbone into a lactone ring is realized, with a concomitant increase in binding affinity relative to an equivalent open-chain DAG.²⁹ This suggested that the isoxazole compounds would exhibit stronger PKC C1 domain binding overall than the other compounds, even though their sn-1 binding is predicted to be weaker.

Chemistry.

We started the synthesis of new analogues with orthogonally protected lactone 7 bearing a benzyl and a silyl ether group.³⁰ In order to explore the chemical compatibility between reaction conditions required to build the heterocycles and for hydroxyl group deprotection, we decided to synthesize simple derivatives bearing a propan-2-ylidene group at the sn-2 position. This alkyl substituent was introduced through a well-established procedure using an alkylation–elimination sequence to give lactone 8 according to published methods (Scheme 1).^13,31

Scheme 1. Synthesis of Isoxazole (1) and Triazole (3−4) Derivatives Employing Orthogonally Protected Lactones^a.

^aReagents and conditions: (a) [1] R₁R₂CO, LiHMDS, THF, −78 °C, 2−4 h; [2] (i) CH₃SO₂Cl, Et₃N, 0 °C, 5 h; (ii) DBU, CH₂Cl₂, room temp, 24 h, 62% for 8, 35% for 14 (E/Z regioisomers 69%); (b) BCl₃, CH₂Cl₂, −78 °C, 2 h, 97% for 9, 89% for 15; (c) 5-methy-3-isoxazolol, PPh₃, DIAD, THF, 24 h, 52%; (d) TBAF, THF, 0 °C, 1 h, 80% for 1, 80% for 3, 60% for 4; (e) p-TsCl, pyr, CH₂Cl₂, 24 h, (40 °C) 88% for 11, (90 °C) 90% for 16 (f) NaN₃, DMF, 100 °C, 48 h, 96% for 12, 72% for 17; (g) 4-ethynyltoluene, sodium ascorbate 1 M, CuSO₄·5H₂O 0.3 M, t-BuOH/H₂O, 48 h, 75% for 13, 77% for 18.

Considering that deprotection conditions used for silyl ethers cleavage were, in principle, compatible with the presence of the isoxazole ring, we decided first to selectively remove the benzyl ether group, with BCl₃ at low temperatures, to obtain the known lactone 9, already described in our previous work, in a good yield.³¹

The heterocyclic ring 5-methy-3-isoxazolol, a commercial product known as himexazole that is employed as a pesticide, was previously synthesized according to methods described in the literature.³² A convergent strategy based on the Mitsunobu reaction was used to couple this heterocyclic ring to the free hydroxyl group of lactone 9, using DIAD and PPh₃, to produce compound 10 in a moderate yield (52%). Selective desilylation of the tert-butyldiphenylsilyl ether (TBDPS) group under standard conditions (TBAF) was carried out in an excellent yield, obtaining the first heterocyclic-containing DAG-lactone 1 (Scheme 1).

In the past decade, the use of triazoles in medicinal chemistry has received marked attention, leading to the development of regioselective methodologies based on Huisgen 1,3-dipolar cycloaddition. We envisioned the synthesis of triazole analogues in the archetypical “click” reaction: the Huisgen [3+2] cycloaddition between alkynes and azides catalyzed by copper(I) salts (Cu-AAC) that conduce to the regioselective formation of 1,4-regioisomer.³³ With that idea in mind, the hydroxyl group in compound 9 was transformed to the corresponding tosylate with tosyl chloride in pyridine, which in turn was treated with sodium azide in DMF at elevated temperatures, yielding azide 12 in 85% yield, for both steps. Preparation of triazole 13 was carried out using the click reaction as a key step, between azide 12 and a commercial alkyne, 4-ethynyl toluene, in the presence of CuSO₄ and sodium ascorbate, a well-known catalytic system.^34,35 The reaction proceeded regioselectively at room temperature to give 13, as the 1,4-regioisomer exclusively, in 75% yield, although a long reaction time (48 h) was required to achieve full conversion (Scheme 1). Deprotection of the TBDPS group was accomplished under standard conditions (TBAF) to obtain triazole 3 in 80% yields.

The introduction of variations in the patterns of the substituents attached to the sn-1 and sn-2 positions on the DAG-lactone has resulted, over the years, in the development of very powerful agonists in the low nM range. One of the most potent analogues prepared to date, obtained by Márquez’s group, was the DAG-lactone AJH-836. Encouraged by this, we decided to synthesize heterocyclic DAG-lactones maintaining the same alkyl substituent present in AJH-836 at the sn-2 position. In this sense, aldehyde 5-methyl-3-(2-methylpropyl)-hexan-1-one was prepared according to published methods^13,19 and immediately employed in an aldol condensation with lactone 7, followed by olefination through the alkylation–elimination sequence already mentioned, yielding 69% of the mixture of E/Z isomers (Scheme 1). Separation of E/Z geometric isomers was achieved at this stage by column chromatography. The vinyl proton corresponding to the E-isomer (14) appears in its ¹H NMR spectrum as a characteristic multiplet that is further downfield compared to that of the corresponding Z-isomer. Following the same synthetic sequence described for the preparation of triazole 3, we obtained DAG-lactone 4 with similar yields for almost every step, except for TBDPS remotion with TBAF that renders a lower yield.

Attempts to obtain the branched isoxazole 2, starting with compound 15 and following the strategy based on Mitsunobu coupling employed to synthesize DAG-lactone 1, were unsuccessful. Consequently, we decided to change the starting material and prepared the known lactone 19³⁶ protected with benzyl and p-methoxyphenyl (PMP) groups. The sn-2 alkyl substituent was introduced through condensation with the branched aldehyde mentioned before,³⁷ and after elimination of the p-methoxyphenyl (PMP)-protecting group, Mitsunobu coupling with isoxazolol was performed to yield compound 23 in 37% yield. Finally, deprotection of the benzyl group with BCl₃ afforded DAG-lactone 2 in a good yield (Scheme 2).

Scheme 2. Different Strategies to the Synthesis of Isoxazole 2^a.

^aReagents and conditions: (a) [1] 3-isobutyl-5-methylhexanal, LiHMDS, THF, −78 °C, 2−4 h; [2] (i) CH₃SO₂Cl, Et₃N, 0 °C, 5 h; (ii) DBU, CH₂Cl₂, room temp, 24 h, 35% for 20 (69% for E/Z mixture); (b) CAN, H₂O/CH₃CN, 75%; (c) 5-methy-3-isoxazolol, PPh₃, DIAD, THF, 24 h, 37%; (d) BCl₃, CH₂Cl₂, −78 °C, 0.5 h, 67%; (e) I₂, PPh₃, imidazole, toluene, 90 °C, 1 h, 89%; (f) K₂CO₃, DMF, sealed tube, 140 °C, 4 h, 62%.

In view of the low yields obtained, it was decided to test an alternative synthetic route. The conversion of the hydroxyl group in compound 21 to an iodide group was carried out with iodine in the presence of PPh₃ and imidazole in excellent yield.³⁸ To favor the substitution reaction, compound 22 was treated with isoxazolol and a base (K₂CO₃) at an elevated temperature, in a sealed tube, obtaining the same intermediate 23 in higher yield (55% in two steps) than the Mitsunobu approach.³⁹

Both synthetic strategies used in the preparation of isoxazole derivatives could, in principle, be applied in obtaining tetrazole analogues. We decided to prepare derivatives from 5-phenyl-1H-tetrazole, a commercially available compound that is easily prepared according to literature procedures by a cycloaddition reaction between benzonitrile and sodium azide catalyzed by copper.⁴⁰ Considering that the Mitsunobu route was straightforward, we tried the coupling between tetrazole 28 and the known lactone 25⁴¹ (Scheme 3). Unfortunately, conversion was low even when adding a large excess of PPh₃ and DIAD. The reaction rendered a complex mixture, and several attempts to purify the desired compound were unsuccessful. Compound 27 was always obtained with variable amounts of byproducts. Deprotection under classical conditions for this type of lactones (BCl₃) allowed us to obtain 5 in 17% total yield from compound 25. Hence, the alternative strategy based on the nucleophilic substitution of a good leaving group, such as iodine, by the phenyltetrazole anion generated in situ in the presence of base was attempted. Indeed, compound 27 was obtained in 70% yield from 25 (two steps). Based on two-dimensional NMR experiments, it was concluded that the 2,5-disubstituted regioisomer was obtained as the only product. Finally, benzyl deprotection led to derivative 5. Unlike the previous case, when we tested the Mitsunobu reaction employing the branched lactone 21 and compound 28, we could introduce the tetrazole ring into compound 29, although in a moderate yield. The complete conversion was observed, and the byproducts were removed efficiently by column chromatography. Benzyl deprotection under standard conditions gave the tetrazolic DAG-lactone 6 (Scheme 3).

Scheme 3. Synthesis of Tetrazole Analogues DAG-Lactone 5 and DAG-Lactone 6^a.

^aReagents and conditions: (a) [1] acetone, LiHMDS, THF, −78 °C, 2−4 h; [2](i) CH₃SO₂Cl, Et₃N, 0 °C, 5 h; (ii) DBU, CH₂Cl₂, room temp, 24 h, 90% for 24; (b) CAN, H₂O/CH₃CN, 92%; (c) I₂, PPh₃, imidazole, toluene, 90 °C, 1 h, 95%; (d) 5-phenyl-1H-tetrazole, K₂CO₃, DMF, sealed tube, 120 °C, 20 h, 74%; (e) BCl₃, CH₂Cl₂, −78 °C, 0.5 h, 90% for 5, 78% for 6; (f) 5-phenyl-1H-tetrazole, PPh₃, DIAD, THF, 24 h, 53%.

Biological Results.

In Vitro Binding of Novel Heterocyclic DAG-Lactones to PKC Isozymes.

In order to determine the affinity of the newly synthesized DAG-lactones, we used an established in vitro competition binding assay, with [³H]phorbol 12,13-dibutyrate ([³H]PDBu) as a radioligand. The assay was carried out in the presence of 100 μg/mL phosphatidylserine, as previously described (for more details, see the General Procedures section).⁴² We first examined the affinity of the compounds for PKCα. Results revealed that DAG-lactone 2, from now on 10B12, could compete [³H]PDBu, displaying a K_i = 79.3 nM. The other DAG-lactones only marginally displaced the radioligand at a concentration of 30 μM. For comparison, the DAG-lactone AJH-836 bound to PKCα, with a K_i = 23.6 nM.¹⁷ We next determined the ability of 10B12 to bind to PKCε. Competition assays revealed that the K_i for PKCε was 2.4-fold lower than that for PKCα (33.6 nM), suggesting a marginal, yet significant preference for the nPKC, at least in vitro. For comparison, AJH-836 displayed a K_i = 1.89 nM for PKCε, therefore showing 12.4-fold lower K_i than that for PKCα.¹⁷ Two of the DAG-lactones with limited binding for PKCα (4 and 6) were also examined for their ability to bind PKCε, and they were found to display limited binding for this nPKC using a fixed concentration of 30 μM (Table 1).

Table 1.

DAG-Lactones Binding Affinities for PKCα and PKCε In Vitro^a

		KI (nM) (or % inhibition, if partial)
compound	ClogP	PKCα	PKCε
AJH-836	5.52	23.6 ± 2.017	1.89 ± 0.2017
1	0.78	31.3% at 30 μM	ND
2 (10B12)	4.75	79.3 ± 20.7	33.6 ± 6.8
3	2.61	11.7% at 30 μM	ND
4	6.58	21.3% at 30 μM	17% at 10 μM
5	2.08	23.5% at 30 μM	ND
6	6.05	37.5% at 30 μM	69.2% at 30 μM

^aValues represent the mean ± SEM of three independent experiments and were calculated from the ID₅₀ values determined from the competition curves. ND: nondetermined. ClogP calculated in ChemDraw.

Previous studies with chemical series of DAG-lactones have shown that there is a parabolic relationship between C1 domain binding affinity and logP, with a peak at a logP of approximately 5. A logP close to this value gives compounds adequate lipophilicity required for partitioning into membranes with the lowest possible amount of nonspecific membrane affinity.¹⁰ Of the heterocyclic DAG-lactones, compounds 10B12 and 6 are closest in logP value to the model compound AJH-836 and closest to this log P “sweet spot”, suggesting that these compounds have the ideal balance between the membrane and C1 domain binding ability.

Differential Translocation of PKCα and PKCε by Heterocyclic DAG-Lactones in Cells.

While the [³H]PDBu binding assay using reconstituted 100% phosphatidylserine vesicles is a well-accepted approach to determine the in vitro binding affinities of C1 domain ligands to PKC isozymes, assessing their effects in cellular models provides more reliable information about their ability to activate these kinases in their physiological environment. Indeed, small differences in K_i’s for individual PKCs, as determined in vitro, may translate into significant differences in isozyme specificity in cellular-based assays. We, therefore, examined the ability of heterocyclic DAG lactones to translocate PKC isozymes to membranes, a well-established hallmark of enzyme activation in cells. As an experimental strategy, we expressed Green Fluorescence Protein (GFP)-fused PKCα or PKCε in HeLa cervical adenocarcinoma cells and examined the ability of DAG lactones to induce kinase relocalization to the plasma membrane using fluorescence microscopy. Coupling this imaging analysis to the determination of membrane to cytosol fluorescent ratio allows accurate quantification of enzyme translocation and the precise determination of potencies in cellular models. Using this approach, we determined that AJH-836 displays 43-fold higher potency for translocation of PKCε relative to PKCα.¹⁷ As shown in the images from Figure 4A and subsequent quantification of the membrane/cytosol fluorescence ratio in Figure 4B, 10B12 (2) induced an evident redistribution of fluorescence in GFP-PKCε expressing HeLa cells. On the other hand, compounds 1, 3, 4, and 5 had essentially no effect on GFP-PKCε redistribution, and fluorescence remained cytoplasmic after treatment. Despite its weak in vitro binding affinity, compound 6 also displays some PKCε translocation effect, although smaller than that of AJH-836, possibly reflecting other mechanisms contributing to its efficacy in cells such as preferential partition into membranes or unique lipid associations. As controls, we found that the phorbol ester PMA (100 nM), which displays similar activity on cPKCs and nPKCs, fully translocated GFP-PKCα and GFP-PKCε to the plasma membrane, whereas the PKCε-selective DAG lactone AJH-836 (1 μM) only translocated GFP-PKCε, as we established in a previous study.¹⁷

Figure 4.

Translocation of PKC isozymes by DAG-lactones. HeLa cells expressing GFP (control), GFP-PKCα, or GFP-PKCε were stimulated for 30 min with the indicated compounds, fixed, counterstained with DAPI, and visualized by fluorescent microscopy. (A) Representative micrographs are shown. (B) Representative graphs obtained using Image G that depicts the quantification of membrane/cytosol ratio. (C) Membrane/cytosol ratio (n = 10−15 cells/group). Results are expressed as mean ± SD *, p < 0.05. ****, p < 0.001 vs control. Similar results were observed in two additional experiments.

10B12 Induces Major Changes in Cytoskeletal Morphology.

In order to pursue a deeper biological characterization of the heterocyclic DAG-lactones, we next examine a cellular function known to be specifically mediated by PKCε. Specifically, PKCε, but not PKCα, is the PKC implicated in the formation of membrane ruffles, actin-rich protrusions involved in cancer cell motility. Indeed, both PMA- and AJH-836-induced ruffle formation in A549 lung cancer cells was severely affected upon silencing PKCε by means of RNA interference (RNAi) or pretreatment with a selective PKCε pharmacological inhibitor.¹⁷ To address the effect of the newly synthesized DAG-lactones on ruffle formation, we treated A549 cells with the different compounds (1 μM, 30 min) and then subjected cells to polymerized actin staining using phalloidin. Upon microscopy examination, we observed a prominent stimulation of ruffle formation in response to 10B12 treatment. No effects were observed for the other synthesized DAG-lactones, other than a small response by DAG-lactone 6 (Figure 5A). Quantification of ruffle formation using a densitometric approach¹⁷ revealed that the effect of 10B12 was near the maximum response caused by 1 μM AJH-836 and 100 nM PMA (Figure 5B). We also tested a higher dose of the DAG lactones (10 μM), but at this concentration, those compounds that induce ruffles at 1 μM showed significant toxicity, as evidenced by cell detachment and death (data not shown). The rank in potency observed in these experiments essentially matched that observed for GFP-PKCε peripheral translocation.

Figure 5.

Stimulation of ruffle formation by DAG-lactones. A549 cells were stimulated for 30 min with DAG-lactones (1 μM) or PMA (0.1 μM). Cells were fixed and stained with phalloidin-rhodamine. (A) Representative micrographs. (B) Quantification of ruffle area/cell using ImageJ. ****, p < 0.001 vs vehicle.

Finally, in order to confirm the involvement of PKCε in the formation of ruffles induced by 10B12, we used PKCε RNAi to silence the expression of this kinase in A549 cells. In order to minimize misinterpretations due to the “off-target” effects of RNAi duplexes, we used two different RNAi duplexes (ε1 and ε2). We observed an 80–90% reduction in PKCε mRNA levels (as determined by Q-PCR) (Figure 6A) and ~70% reduction in PKCε protein expression (as determined by Western blot) 48 h after transfection of either RNAi duplex (Figure 6B). Notably, the formation of ruffles by 10B12 was markedly impaired in PKCε-depleted cells in a magnitude that was consistent with its depletion (Figure 6C,D), as was previously observed with AJH-836 using a similar approach.¹⁷ Together with the membrane translocation studies, these actin reorganization experiments establish 10B12 as a bona fide PKCε-activating DAG-lactone.

Figure 6.

PKCε mediates the ruffle formation response of compound 10B12. A549 cells were subjected to PKCε RNAi by delivery of specific siRNA duplexes (PKCε1 and PKCε2). As a control, cells were transfected with a nontarget control (NTC) siRNA duplex. Experiments were carried out 48 h later. (A) PKCε mRNA levels relative to NTC (dotted line). (B) PKCε protein levels. Upper panel, representative Western blot. Lower panel, densitometric analysis of PKCε protein levels, expressed as arbitrary units relative to β-actin. (C) A549 cells were stimulated for 30 min with 10B12 (1 μM), fixed, and stained with phalloidin-rhodamine. Upper panel. Representative micrographs. (D) Quantification of ruffle area/cell using ImageJ. ****, p < 0.001 vs 10B12-treated cells, NTC.

CONCLUSION

Over the past decades, a large number of DAG-lactones analogues with broad structural variation have been designed and synthesized, generating diversity at the sn-1 and sn-2 positions and covering an extensive chemical space. Most of these compounds have an ester moiety at the sn-1 position that is important for receptor interaction, and only a few attempts have been made to replace it. In this study, different types of heterocyclic groups were introduced as substituents for the DAG-lactone structure at the sn-1 position. Synthetic strategies were developed for the introduction of isoxazolol, triazole, and tetrazole rings as ester moiety replacements. The isoxazolol ring was introduced in DAG-lactone 1 by a Mitsunobu approach starting from lactone 7. A change in the protective groups of the starting material, following the Mitsunobu route, led the analogues with a saturated-branched alkyl chain at the sn-2 position (DAG-lactone 2, 10B12) in low yields. An improvement in the 2 yield was achieved by employing a synthetic alternative based on a substitution reaction. Triazole analogues 3 and 4 were obtained in good yields by methodologies based on the copper(I)-catalyzed variant of the Huisgen 1,3-dipolar cycloaddition between alkynes and azides catalyzed. The first attempts to introduce a tetrazole ring in DAG-lactones 5 and 6 were carried out in Mitsunobu conditions. Compound 6 was straightforwardly obtained by this route, but the strategy did not proceed well for compound 5, which was obtained using the alternative based on the nucleophilic substitution of a good leaving group by the phenyltetrazole anion.

Modeling predicted that compound 10B12, with its higher logP and its ability to adopt the favored sn-2 binding mode for DAG-lactones, would have a stronger C1 domain binding ability, and this was in agreement with the in vitro binding assay results. However, as has been shown previously with PKCα-binding and RasGRP-binding DAG-lactones,⁴³ the binding affinity to C1 domains is only one aspect of the biological effects of a DAG-lactone, and compound 6, with relatively weak C1 domain binding, was also able to induce GFP-PKCε translocation and a small effect on ruffle formation. Agents that belong to the different families of C1 domain ligands, such as phorbol esters, deoxyphorbol esters, and bryostatins, display similar affinities in vitro for individual PKC isoforms, yet these ligands have widely diverse biological functions.^44,45 This is epitomized by bryostatins and deoxyphorbol esters, compounds that display antitumor-promoting activities in mouse models rather than the typical tumor promoting effects caused by phorbol esters. Although not fully understood, it may be possible that additional mechanisms, namely, the interactions of ligands with lipid microdomains, the time-dependent concentration of ligand in different subcellular membranes, and/or the formation of specific ligand–protein–membrane ternary complexes, may greatly influence the biological responses of C1 domain-binding compounds. While these mechanisms still remain subject to experimental examination, it is clear that such rich diversity in the effects of C1 domain ligands underscores opportunities for the generation of PKC isozyme-specific ligands. In this regard, our data is in line with previous studies demonstrating that it is possible to achieve remarkable selectivity upon introduction of side chains in cyclized DAG structures, as has been previously demonstrated by means of combinatorial libraries for DAG-lactones with different alkyl-chain and acyl-chain substitutions.¹²

The present results highlight the feasibility of generating C1 domain ligands with preferential selectivity toward specific members of the PKC family, a highly complex task due to the remarkable similarity among C1 domains. In previous studies, we demonstrated significant selectivity toward PKCε for the DAG-lactone AJH-836. This derivative represents, to date, the best-characterized example for a cyclized DAG with preferential selectivity for an nPKC relative to cPKCs.^16–18 Similar to AJH-836, the DAG-lactone 10B12 synthesized in the present study, in which the ester moiety at the sn-1 position has been replaced by an isoxazole ring, displays a significant degree of specificity for PKCε relative to PKCα, particularly in a cellular context. Indeed, although the [³H]PDBu competition binding assays revealed a lower PKCε/PKCα selectivity index for 10B12 compared to AJH-836, it shows a preferred ability to translocate PKCε relative to PKCα. This selectivity is reflected in its ability to induce the reorganization of the cellular actin cytoskeleton into peripheral ruffles, a well-described PKCε-mediated effect. We carefully validated this conclusion by means of PKCε RNAi silencing, which was able to abrogate the activity of the new sn-1 derivative to form ruffles in lung cancer cells. Thus, a stable isoxazole ring can function as an ester surrogate to allow binding to C1 domain and dictate specificity toward PKCε as a receptor in a cellular environment. Since PKCε is very important for tumor growth and metastasis and has been implicated in cancer cell motility and invasiveness,^46,47 novel sn-1 derivatives may have significant utility for the study of this kinase in cancer progression. Considering the limited availability of isozyme-specific PKC activators, this class of C1 domain ligands may represent a novel tool that promotes PKC isozyme biological response patterns.

In summary, our study provides a proof of principle for the generation of DAG-lactones with replacements in the labile ester group at the sn-1 position having significant potency for PKC activation. DAG-lactones with a stable isoxazole ring that can function as an ester surrogate represent a novel class of PKC activators, thus expanding the diversity of C1 domain ligands with PKC isozyme selectivity. Given the simplicity of the DAG-lactone structure for chemical modifications, the search for additional substitutions could help in the design of additional potent nPKC selective agents for experimentally assessing the functional involvement of these kinases in human diseases.

EXPERIMENTAL SECTION

General Procedures.

All chemical compounds were purchased from commercial sources or synthesized. Reaction monitoring was performed on Merck silica gel 254F TLC plates. Column chromatography was performed on a Teledyne Isco CombiFlash Rf+ instrument under gradient elution conditions with RediSep disposable flash columns. All melting points were determined on an Electrothermal IA9000 series digital melting point apparatus and are uncorrected. ¹H NMR and ¹³C NMR spectra were carried out on a Bruker Fourier 300, 300 MHz or Bruker Avance DPX 400, 400 MHz spectrometer using CDCl₃, MeOD, or DMSO-d₆ as the solvent and TMS as the internal standard. HRMS was performed at CIBION (Bioanalytical Mass Spectrometry Group). High-resolution positive ion electrospray ionization MS was conducted on a Xevo G2S Q-TOF spectrometer (Waters Corporation, Manchester, UK) operated at both positive and negative modes. Data processing was carried out using MassLynx v4.1 software (Waters Corp.). Data was corrected during acquisition using a reference compound (LockSpray). The purity was determined by high-performance liquid chromatography (HPLC). The purity of all final compounds was 95% or higher. HPLC analysis was performed on a Shimadzu SPD-M20A UV–vis photo diode array (PDA) detector, HPLC system (Shimadzu, Tokyo, Japan). The column used was a Zorbax Eclipse XDB C8 (5 μm, 150 mm × 4.60 mm) at a temperature of 40 °C, using a linear gradient of 0–100% H₂O (0.1% HAcO)/acetonitrile at a flow rate of 1.0 mL/min over 20 min. All title compounds were examined for PAINS and passed.⁴⁸

Docking Studies.

Compounds were docked into the crystal structure of the C1B domain of PKC δ¹⁴ (PDB ID: 1PTR) using the program GOLD, version 5.2.2.^49,50 We used standard default genetic algorithm (GA) settings, with the binding site defined by atoms within a 10.0 Å radius of the Nε atom of residue Q257. The scoring function was GoldScore, and we added a template similarity constraint to the hydrogen bond acceptor atoms in bound phorbol-13-O-acetate from the crystal structure. Twenty GA runs were performed for each compound. After docking, the docked poses were refined with 500 steps of conjugate gradient energy minimization using the Embrace facility in MacroModel.⁵¹ The ligand and residues 240–242, 250–253, and 257 in the C1 domain were free to move, residues 237–239, 243, and 254–256 were restrained with a force constant of 100 kJ/mol Ų, and all other protein residues were fixed in place.

Chemistry.

General Procedure A: Aldol Condensation Followed by Olefination. Standard Alkylation Procedure.

A solution of lactone 7 or 19 (1 equiv) in THF (5 mL/mmol) at −78 °C was treated dropwise with LiHMDS (1.5–2.0 equiv, 1 M in THF). After the mixture was stirred at −78 °C for 1 h, a solution of R₂CHO (1.5 equiv) or acetone (5 equiv) in THF (1 mL/mmol) was added, and the stirring continued for 2–3 h at −78 °C. The reaction was quenched by the slow addition of a saturated aqueous solution of NH₄Cl and allowed to warm to room temperature. The layers were separated, and the aqueous layer was extracted with Et₂O (3×). The combined organic phases were washed with H₂O (1×) and brine (1×), dried over MgSO₄, and concentrated in vacuo. Purification by silica gel flash column chromatography (gradient 0–20% EtOAc/hexanes) gave a mixture of alcohol diastereomers, which were used directly in the next step.

Standard Mesylation–Olefination Procedure.

A solution of the alkylation product in dichloromethane (10 mL/mmol) at 0 °C was treated with methanesulfonyl chloride (2 equiv) and triethylamine (4 equiv) and then stirred at room temperature for 2–5 h. DBU (3–5 equiv) was added at 0 °C, and the resulting solution was stirred overnight at room temperature. The reaction mixture was treated with a saturated solution of NH₄Cl (10 mL/mmol) and extracted with CH₂Cl₂ (3×). The combined organics were washed with H₂O (2×) and brine (1×), dried (Na₂SO₄), and concentrated in vacuo. Purification by silica gel flash column chromatography (gradient 0–20% EtOAc/hexanes) gave 14, 20, and 24.

General Procedure B: BnO Deprotection.

A solution of 14, 22, 27, or 29 (1 equiv) in anhydrous dichloromethane (20 mL/mmol) at −78 °C was treated dropwise with boron trichloride (1 M in CH₂Cl₂, 2 equiv). The reaction was monitored by TLC and quenched upon completion (0.5–1 h) by the slow addition of NaHCO₃, and the aqueous phase was extracted with dichloromethane (2 × 15 mL). The organic phase was dried (Na₂SO₄) and concentrated in vacuo. Purification by silica gel flash column chromatography (gradient 0–20% EtOAc/hexanes) gave 15, 2, 5, and 6.

General Procedure C: PMP Deprotection.

Ceric ammonium nitrate (3 equiv) was added to a solution of 20 or 24 (1 equiv) in 80% CH₃CN/H₂O (v/v, 10 mL/mmol) at 0 °C. The reaction mixture was stirred for 15 min. A solution of 5% Na₂S₂O₃ was added at room temperature, and the aqueous phase was extracted with CH₂Cl₂ (3×). The organic phase was dried (Na₂SO₄) and concentrated in vacuo. Purification by silica gel flash column chromatography (gradient 0–30% EtOAc/hexanes) gave 21 and 25.

5-(((tert-Butyldiphenylsilyl)oxy)methyl)-5-(((5-methylisoxazol-3-yl)oxy)methyl)-3-(propan-2-ylidene)dihydrofuran-2(3H)-one (10).

A solution of triphenylphosphine (557 mg, 2.0 mmol), lactone 9³¹ (638 mg, 1.5 mmol), and 5-methylisoxazol-3-ol³² (119 mg, 1.2 mmol) in anhydrous tetrahydrofuran (6 mL) was treated with diisopropyl azodicarboxylate (410 μL, 2.0 mmol) and stirred at room temperature overnight. The reaction mixture was evaporated, and the residue was purified by flash chromatography using hexane/ethyl acetate (8:2) as an eluent to afford 10 as a white solid (314 mg, 52%). Mp: 93−94 °C. ¹H NMR (400 MHz, CDCl₃): δ 7.67−7.65 (m, 4H), 7.42−7.40 (m, 6H), 5.62 (s, 1H), 4.38−4.32 (m, 2H), 3.78 (s, 2H), 2.90 (AB d, J = 16.4 Hz, 1H), 2.78 (AB d, J = 16.4 Hz, 1H), 2.34 (s, 3H), 2.28 (s, 3H), 1.88 (s, 3H), 1.05 (s, 9H). ¹³C NMR (101 MHz, CDCl₃): δ 171.6, 170.6, 169.1, 150.2, 135.6, 135.6, 132.7, 132.5, 129.9, 127.8, 127.8, 119.5, 92.9, 81.2, 70.5, 66.0, 35.3, 26.6, 24.5, 19.8, 19.1, 12.9. HRMS (ESI): m/z calcd for C₂₉H₃₅NO₅Si [M + H]⁺, 506.2363; found, 506.2364.

5-(Hydroxymethyl)-5-(((5-methylisoxazol-3-yl)oxy)methyl)-3-(propan-2-ylidene)dihydrofuran-2(3H)-one (1).

To a stirred solution of 10 (160 mg, 0.32 mmol) in anhydrous THF (3 mL) was added tetrabutylammonium fluoride (1.0 M in THF, 320 μL, 0.32 mmol) at 0 °C. After the mixture was stirred for 1 h, the reaction was evaporated. The residue was purified by column chromatography over silica gel with hexane/ethyl acetate (6:4) as an eluent to afford 1 (68 mg, 80%) as a white solid. Mp: 139−140 °C. ¹H NMR (400 MHz, CDCl₃): δ 5.63 (s, 1H), 4.33−4.27 (AB q, J = 11.1 Hz, 2H), 3.69 (m, 2H), 3.15 (t, J = 6.6 Hz, 1H), 2.85 (AB d, J = 16.5 Hz, 1H), 2.73 (AB d, J = 16.5 Hz, 1H), 2.30 (s, 3H), 2.23 (s, 3H), 1.86 (s, 3H). ¹³C NMR (101 MHz, CDCl₃): δ 171.8, 170.9, 169.1, 151.9, 118.7, 92.9, 81.3, 70.5, 64.3, 31.9, 24.5, 19.9, 12.8. HRMS (ESI): m/z calcd for C₁₃H₁₈NO₅ [M + H]⁺, 268.1185; found, 268.1190.

(2-(((tert-Butyldiphenylsilyl)oxy)methyl)-5-oxo-4-(propan-2-ylidene)tetrahydrofuran-2-yl)methyl 4-methylbenzenesulfonate (11).

To a solution of compound 9 (127 mg, 0.30 mmol) in CH₂Cl₂ (2 mL) in a sealed vial were added triethylamine (606 mg, 6 mmol) and p-toluenesulfonyl chloride (114 mg, 0.6 mmol). The reaction mixture was heated at 40 °C overnight and treated with 1 N HCl (5 mL). The aqueous layer was extracted with CH₂Cl₂ (3 × 5 mL). The combined organic layers were washed with NaHCO₃ (sat. solution, 5 mL) and NaCl (sat. solution, 5 mL), dried over Na₂SO₄, and concentrated under a vacuum. The residue was purified by silica gel flash column chromatography hexane/ethyl acetate (9:1) to afford 11 as a white solid (148 mg, 88%). Mp: 44−46 °C. ¹H NMR (400 MHz, CDCl₃): δ 7.75 (d, J = 8.3 Hz, 2H), 7.63−7.56 (m, 4 H), 7.47−7.35 (m, 6H), 7.31 (d, J = 8.0 Hz, 1H), 4.12 (d, J = 10.1 Hz, 1H), 4.05 (d, J = 10.1 Hz, 1H), 3.60 (s, 2H), 2.79−2.65 (m, 2H), 2.43 (s, 3H), 2.21 (t, J = 2.0 Hz, 3H), 1.82 (s, 3H), 0.98 (s, 9H). ¹³C NMR (101 MHz, CDCl₃): δ 168.5, 151.1, 145.2, 135.5, 132.4, 132.3, 132.1, 129.9, 128.0, 127.8, 127.8, 118.7, 80.3, 70.2, 66.0, 32.0, 26.5, 24.5, 21.6, 19.8, 19.1. HRMS (ESI): m/z calcd for C₃₂H₃₉O₆SSi [M + H]⁺, 579.2237; found, 579.2231.

5-(Azidomethyl)-5-((tert-butyldiphenylsilyloxy)methyl)-3-(propan-2-ylidene)dihydrofuran-2(3H)-one (12).

To a solution of tosylate 11 (140 mg, 0.245 mmol) in DMF (3 mL) in a sealed vial was added sodium azide (319 mg, 4.9 mmol). The reaction mixture was heated at 120 °C for 4 h, then cooled to room temperature, and evaporated. The residue was dissolved in CH₂Cl₂ and washed with H₂O and brine, dried over anhydrous Na₂SO₄, and concentrated under a vacuum. The residue was purified by silica gel flash column chromatography hexane/ethyl acetate (95:5) to afford 12 (105 mg, 96%) as a colorless oil. ¹H NMR (400 MHz, CDCl₃): δ 7.67−7.62 (m, 4H), 7.47−7.36 (m, 6H), 3.65 (AB m, 2H), 3.52 (d, J = 12.8 Hz, 1H), 3.46 (d, J = 12.8 Hz, 1H), 2.73 (AB m, 2H), 2.25 (s, 3H), 1.85 (s, 3H), 1.04 (s, 9H). ¹³C NMR (101 MHz, CDCl₃): δ 168.7, 150.7, 135.6, 135.6, 132.6, 132.5, 129.9, 127.8, 119.3, 81.9, 66.8, 55.1, 32.8, 26.6, 24.5, 19.9, 19.2. HRMS (ESI): m/z calcd for C₂₅H₃₂N₃O₃Si [M + H]⁺, 450.2213; found, 450.2218.

5-((tert-Butyldiphenylsilyloxy)methyl)-3-(propan-2-ylidene)-5-((4-p-tolyl-1H-1,2,3-triazol-1-yl)methyl)dihydrofuran-2(3H)-one (13).

4-Ethynyltoluene (25 mg, 26 μL, 0.21 mmol) and compound 12 (95 mg, 0.21 mmol) were dissolved in 5 mL of a 1:1 tert-BuOH/H₂O mixture. While the mixture was being stirred, sodium ascorbate (1 M in H₂O, 189 μL, 0.189 mmol) was added, followed by copper(II) sulfate pentahydrate (0.3 M, 280 μL, 0.084 mmol). The resulting mixture was stirred at room temperature for 18 h before dilution with ethyl acetate (10 mL) and saturation with EDTA solution (0.4 mL) and water (10 mL). The aqueous layer was extracted with ethyl acetate (10 mL × 3). The combined organic layers were dried over Na₂SO₄, filtered through a short silica gel column, and concentrated in vacuo. The residue was purified by silica gel flash column chromatography hexane/ethyl acetate (9:1) to afford 13 (89 mg, 75%) as a white solid. Mp: 59−62 °C. ¹H NMR (400 MHz, CDCl₃): δ 7.88 (s, 1H), 7.68 (d, J = 8.3 Hz, 2H), 7.66−7.63 (m, 4H), 7.46−7.36 (m, 6H), 7.23 (d, J = 8.0 Hz, 2H), 4.72 (AB d, J = 14.5 Hz, 1H), 4.61 (AB d, J = 14.5 Hz, 1H), 3.72 (AB d, J = 10.9 Hz, 1H), 3.65 (AB d, J = 10.8 Hz, 1H), 2.75 (AB q, J = 16.6 Hz, 2H), 2.38 (s, 3H), 2.06 (s, 3H), 1.74 (s, 3H), 1.08 (s, 9H). ¹³C NMR (101 MHz, CDCl₃): δ 168.5, 151.8, 148.3, 138.1, 135.6, 135.5, 132.3, 132.2, 130.1, 129.5, 127.9, 127.4, 125.7, 121.0, 118.3, 81.5, 67.2, 54.0, 32.3, 26.7, 24.5, 21.3, 19.9, 19.2. HRMS (ESI): m/z calcd for C₃₄H₄₀N₃O₃Si [M + H]⁺, 566.2839; found, 566.2847.

5-(Hydroxymethyl)-3-(propan-2-ylidene)-5-((4-p-tolyl-1H-1,2,3-triazol-1-yl)methyl) dihydrofuran-2(3H)-one (3).

To a stirred solution of 13 (60 mg, 0.106 mmol) in anhydrous THF (3 mL) was added tetrabutylammonium fluoride (TBAF, 1.0 M solution in THF, 106 μL, 0.106 mmol) at 0 °C. After the mixture was stirred for 1 h, the reaction was evaporated. The residue was purified by column chromatography over silica gel with hexane/ethyl acetate (2:8) as an eluent to give the alcohol 3 (28 mg, 80%) as a white solid. Mp: 167−168 °C. ¹H NMR (400 MHz, CDCl₃): δ 7.88 (s, 1H), 7.69 (d, J = 8.0 Hz, 2H), 7.23 (d, J = 8.0 Hz, 2H), 4.69 (AB q, J = 14.6 Hz, 2H), 3.61 (dq, J = 12.2 Hz, 6.5 Hz, 2H), 3.02 (t, J = 6.6 Hz, 1H), 2.84 (AB q, J = 16.5 Hz, 2H), 2.37 (s, 3H), 2.10 (s, 3H), 1.80 (s, 3H). ¹³C NMR (101 MHz, CDCl₃): δ 168.5, 153.5, 148.3, 138.3, 129.5, 127.3, 125.7, 121.1, 117.6, 81.4, 65.3, 53.9, 32.5, 24.6, 21.3, 20.0. HRMS (ESI): m/z calcd for C₁₈H₂₂N₃O₃ [M + H]⁺, 328.1661; found, 328.1665.

(E)-5-((Benzyloxy)methyl)-5-(((tert-butyldiphenylsilyl)oxy)-methyl)-3-(3-isobutyl-5-methylhexylidene)dihydrofuran-2(3H)-one (14).

Starting from lactone 7³⁰ (1.654 g, 3.46 mmol) and 3-isobutyl-5-methylhexanal¹³ (940 mg, 5.45 mmol), and following general procedure A, the reaction product was isolated as a 1:1 mixture of E- and Z-isomers. Two fractions were isolated after purification by silica gel flash column chromatography with hexane/ethyl acetate (98:2) as an eluent. The first fraction corresponded to the Z-isomer (724 mg, 34%), and the second fraction was the E-isomer 14 (763 mg, 35%). ¹H NMR (400 MHz, CDCl₃): δ 7.64 (d, J = 6.9 Hz, 4H), 7.46−7.22 (m, 11H), 6.74 (m, 1H), 4.53 (s, 2H), 3.73 (AB q, J = 10.7 Hz, 2H), 3.59 (d, J = 10.1 Hz, 1H), 3.53 (d, J = 10.1 Hz, 1H), 2.82−2.69 (AB m, 2H), 2.10 (t, J = 6.2 Hz, 2H), 1.64−1.60 (m, 3H), 1.14−1.06 (m, 4H), 1.02 (s, 9H), 0.85 (d, J = 6.5 Hz, 12H). ¹³C NMR (101 MHz, CDCl₃): δ 170.3, 139.2, 137.7, 135.6, 135.6, 132.9, 132.6, 129.8, 129.8, 128.4, 127.9, 127.8, 127.7, 127.6, 84.4, 73.7, 71.9, 66.2, 43.9, 43.8, 34.7, 32.8, 30.3, 26.7, 25.2, 23.0, 22.9, 22.7, 22.6, 19.2. HRMS (ESI): m/z calcd for C₄₀H₅₅O₄Si [M + H]⁺, 627.3870; found, 627.3879.

(E)-5-((tert-Butyldiphenylsilyloxy)methyl)-5-(hydroxymethyl)-3-(3-isobutyl-5-methylhexylidene)dihydrofuran-2(3H)-one (15).

Starting from 14 (644 mg, 1.03 mmol) and following general procedure B, 15 (490 mg, 89%) was obtained as a colorless oil after purification by silica gel flash column chromatography with hexane/ethyl acetate (9:1). ¹H NMR (400 MHz, CDCl₃): δ 7.64−7.62 (m, 4H), 7.48−7.35 (m, 6H), 6.76 (m, 1H), 3.77−3.62 (m, 4H), 2.87−2.65 (AB m, 2H), 2.12 (t, J = 6.7 Hz, 2H), 1.93 (t, J = 6.8 Hz, 1H), 1.75−1.56 (m, 3H), 1.14−1.07 (m, 4H), 1.03 (s, 9H), 0.86 (d, J = 6.5 Hz, 12H). ¹³C NMR (101 MHz, CDCl₃): δ 170.2, 140.0, 135.6, 135.6, 132.7, 132.5, 130.0, 129.9, 127.8, 127.7, 85.1, 66.0, 65.3, 43.9, 43.8, 34.8, 32.8, 29.7, 26.7, 25.2, 23.0, 22.9, 22.7, 22.6, 19.2. HRMS (ESI): m/z calcd for C₃₃H₄₈O₄SiNa [M + Na]⁺, 559.3220; found, 559.3224.

(E)-(2-((tert-Butyldiphenylsilyloxy)methyl)-4-(3-isobutyl-5-methylhexylidene)-5-oxotetra hydro furan-2-yl)methyl 4-methylbenzenesulfonate (16).

To a solution of 15 (474 mg, 0.88 mmol) in pyridine (4 mL) was added p-toluenesulfonyl chloride (421 mg, 2.2 mmol). The reaction mixture was heated at 90 °C overnight and treated with 1 N HCl (8 mL). The aqueous layer was extracted with CH₂Cl₂ (3 × 10 mL). The combined organic layers were washed with NaHCO₃ (sat. solution, 10 mL) and NaCl (sat. solution, 10 mL), dried over Na₂SO₄, and concentrated under a vacuum. The residue was purified by silica gel flash column chromatography with hexane/ethyl acetate (98:2) to afford 16 (423 mg, 90%) as a white solid. Mp: 135−136 °C. ¹H NMR (400 MHz, CDCl₃): δ 7.74 (d, J = 8.1 Hz, 2H), 7.63−7.55 (m, 4 H), 7.48−7.34 (m, 6H), 7.31 (d, J = 8.1 Hz, 1H), 6.74 (m, 1H), 4.14 (d, J = 10.2 Hz, 1H), 4.06 (d, J = 10.2 Hz, 1H), 3.62 (s, 2H), 2.79−2.65 (m, 2H), 2.43 (s, 3H), 2.08 (t, J = 6.4 Hz, 2H), 1.73−1.55 (m, 3H), 1.15−1.02 (m, 4H), 0.98 (s, 9H), 0.86 (m, 12H). ¹³C NMR (101 MHz, CDCl₃): δ 169.2, 145.2, 140.9, 135.5, 132.3, 132.2, 130.0, 130.0, 128.0, 127.9, 126.3, 82.1, 70.1, 65.9, 43.9, 43.8, 34.9, 32.8, 30.0, 26.6, 25.2, 22.9, 22.9, 22.6, 22.6, 21.6, 19.2. HRMS (ESI): m/z calcd for C₄₀H₅₅O₆SSi [M + H]⁺, 691.3489; found, 691.3475.

(E)-5-(Azidomethyl)-5-((tert-butyldiphenylsilyloxy)methyl)-3-(3-isobutyl-5-methylhexylidene)dihydrofuran-2(3H)-one (17).

To a solution of tosylate 16 (400 mg, 0.58 mmol) in DMF (3 mL) was added sodium azide (754 mg, 11.6 mmol). The reaction mixture was heated at 90 °C for 72 h, then cooled to room temperature, and evaporated. The residue was dissolved in CH₂Cl₂, washed with H₂O and brine, dried over anhydrous Na₂SO₄, and concentrated under a vacuum. The residue was purified by silica gel flash column chromatography hexane/ethyl acetate (99:1) to afford 17 (235 mg, 72%) as a colorless syrup. ¹H NMR (300 MHz, CDCl₃): δ 7.66−7.61 (m, 4H), 7.54−7.40 (m, 6H), 6.78 (m, 1H), 3.75−3.67 (mAB, 2H), 3.54 (AB d, J = 12.8 Hz, 1H), 3.47 (AB d, J = 12.8 Hz, 1H), 2.70 (AB m, 2H), 2.11 (t, J = 6.7 Hz, 2H), 1.75−1.56 (m, 3H), 1.17−1.04 (m, 4H), 1.04 (s, 9H),0.86 (m, 12H). ¹³C NMR (75 MHz, CDCl₃): δ 169.5, 140.5, 135.6, 135.6, 132.5, 132.3, 130.0, 127.9, 126.9, 83.7, 66.5, 55.1, 43.9, 43.8, 34.8, 32.8, 30.7, 26.7, 25.2, 23.0, 22.9, 22.7, 22.6, 19.2. HRMS (ESI): m/z calcd for C₃₃H₄₇N₃O₃SiNa [M + Na]⁺, 584.3284; found, 584.3290.

(E)-5-((tert-Butyldiphenylsilyloxy)methyl)-3-(3-isobutyl-5-methylhexylidene)-5-((4-p-tolyl-1H-1,2,3-triazol-1-yl)methyl)-dihydrofuran-2(3H)-one (18).

4-Ethynyltoluene (23 mg, 25 μL, 0.20 mmol) and 17 (102 mg, 0.18 mmol) were dissolved in 4 mL of a 1:1 tert-BuOH/H₂O mixture. While the mixture was being stirred, sodium ascorbate (0.3 M in H₂O, 480 μL, 0.144 mmol) was added, followed by copper(II) sulfate pentahydrate (0.3 M, 240 μL, 0.072 mmol). The resulting mixture was stirred at room temperature for 2 h before dilution with ethyl acetate (10 mL) and saturation with EDTA solution (0.4 mL) and water (10 mL). The aqueous layer was extracted with ethyl acetate (3 × 10 mL). The combined organic layers were dried over Na₂SO₄, filtered through a short silica gel column, and concentrated in vacuo. The residue was purified by silica gel flash column chromatography hexane/ethyl acetate (93:7) to afford 18 (95 mg, 77%) as a colorless syrup. ¹H NMR (300 MHz, CDCl₃): δ 7.85 (s, 1H), 7.67 (d, J = 8.1 Hz, 2H), 7.66−7.61 (m, 4H), 7.49−7.34 (m, 6H), 7.22 (d, J = 8.0 Hz, 2H), 6.62 (m, 1H), 4.74 (AB d, J = 14.5 Hz, 1H), 4.64 (AB d, J = 14.5 Hz, 1H), 3.73 (AB d, J = 11 Hz, 1H), 3.65 (AB q, J = 11 Hz, 1H), 2.82 (AB q, J = 17.6 Hz, 1H), 2.71 (AB q, J = 17.6 Hz, 1H), 2.38 (s, 3H), 1.99 (t, J = 6.6 Hz, 2H), 1.67−1.46 (m, 3H), 1.13−0.93 (m, 4H), 1.08 (s, 9H),0.81 (d, J = 6.5 Hz, 12H). ¹³C NMR (75 MHz, CDCl₃): δ 169.2, 148.4, 141.5, 138.1, 135.6, 135.5, 132.2, 132.1, 130.1, 129.5, 127.9, 127.4, 125.8, 125.7, 120.8, 83.1, 66.9, 53.9, 43.8, 34.8, 32.7, 30.2, 29.7, 26.7, 25.1, 22.9, 22.6, 22.6, 21.3, 19.2. HRMS (ESI): m/z calcd for C₄₂H₅₆N₃O₃Si [M + H]⁺, 678.4091; found, 678.4121.

(E)-5-(Hydroxymethyl)-3-(3-isobutyl-5-methylhexylidene)-5-((4-p-tolyl-1H-1,2,3-triazol-1-yl)methyl)dihydrofuran-2(3H)-one (4).

To a stirred solution of 18 (79 mg, 0.12 mmol) in anhydrous THF (3 mL) was added tetrabutylammonium fluoride (1.0 M solution in THF, 120 μL, 0.12 mmol) at 0 °C. After the mixture was stirred for 1 h, the reaction was evaporated. The residue was purified by column chromatography over silica gel with hexane/ethyl acetate (75:25) as an eluent to afford 4 (31 mg, 60%) as a white solid. Mp: 155−156 °C. ¹H NMR (400 MHz, CDCl₃): δ 7.88 (s, 1H), 7.69 (d, J = 7.7 Hz, 2H), 7.23 (d, J = 7.7 Hz, 2H), 6.72 (m, 1H), 4.79 (AB d, J = 14.6 Hz, 1H), 4.63 (AB d, J = 14.6 Hz, 1H), 3.70−3.45 (m, 2H), 2.93 (t, J = 6.3 Hz, 1H), 2.83 (s, 2H), 2.37 (s, 3H), 2.07 (m, 2H), 1.71−1.49 (m, 3H), 1.15−0.93 (m, 4H), 0.83 (d, J = 5.9 Hz, 12H). ¹³C NMR (101 MHz, CDCl₃): δ 169.1, 148.4, 142.7, 138.3, 129.5, 127.2, 125.7, 125.3, 121.0, 83.10, 65.0, 53.7, 43.9, 34.9, 32.8, 30.6, 25.2, 22.9, 22.9, 22.6, 22.5, 21.3. HRMS (ESI): m/z calcd for C₂₆H₃₈N₃O₃ [M + H]⁺, 440.2913; found, 440.2926.

(E)-5-((Benzyloxy)methyl)-5-(hydroxymethyl)-3-(3-isobutyl-5-methylhexylidene)dihydrofuran-2(3H)-one (21).

Compound 21 was synthesized following a procedure described elsewhere.³⁷

Starting from lactone 19³⁶ (2.52 g, 7.4 mmol) and 3-isobutyl-5-methylhexanal¹³ (1.9 g, 11.41 mmol) and following general procedure A, the reaction product was isolated as a 1:1 mixture of E- and Z-isomers. Two fractions were isolated after purification by silica gel flash column chromatography with hexane/ethyl acetate (98:2) as an eluent. The first fraction corresponded to the Z-isomer (1.23 g, 34%) and the second fraction corresponded to the E-isomer 20 (1.27 g, 35%). Recorded spectra perfectly matched the published results. ¹H NMR (400 MHz, CDCl₃): δ 7.36−7.26 (m, 5 H), 6.81 (s, 4 H), 6.82−6.76 (m, 1 H), 4.58 (AB q, J = 12.3 Hz, 2H), 4.03 (AB q, J = 9.8, 26.6 Hz, 2H), 3.76 (s, 3H), 3.66 (AB q, J = 10.3, 23.7 Hz, 2H), 2.92−2.77 (AB m, 2H), 2.13 (t, J = 11.6 Hz, 2H), 1.76−1.58 (m, 3H), 1.14−1.06 (m, 4H), 0.85 (m, 12H).

Starting from E-isomer 20 (1.26 g, 2.55 mmol) and following general procedure C, 21 was obtained as a colorless oil (746 mg, 75% yield). Its spectrum data perfectly matched with published results. ¹H NMR (400 MHz, CDCl₃): δ 7.40−7.26 (m, 5H), 6.82−6.73 (m, 1H), 4.56 (m, 2H), 3.84−3.64 (m, 2H), 3.57 (AB q, J = 9.9, 28.0 Hz, 2H), 2.83−2.65 (m, 2H), 2.19−2.07 (m, 3H), 1.77−1.55 (m, 3H), 1.15−1.05 (m, 4H), 0.84−0.86 (m, 12H).

(E)-5-((Benzyloxy)methyl)-5-(iodomethyl)-3-(3-isobutyl-5-methylhexylidene)dihydrofuran-2(3H)-one (22).

A solution of 21³⁷ (128 mg, 0.33 mmol) in toluene (6 mL) was treated with triphenylphosphine (173 mg, 0.66 mmol), imidazole (67 mg, 0.99 mmol), and iodine (126 mg, 0.5 mmol). The mixture was heated at 90 °C for 1 h and, after reaching room temperature, was poured into a saturated solution of NaHCO₃. Excess of triphenylphosphine was destroyed by the addition of iodine until the iodine coloration persisted in the organic layer. The organic layer was washed with Na₂S₂O₃ (10%, 5 mL) and brine, dried over (MgSO₄), filtered, and concentrated in vacuo. The residue was purified on silica gel (hexanes:ethyl acetate, 98:2) to give 22 (146 mg, 89%) as a colorless syrup. ¹H NMR (400 MHz, CDCl₃): δ 7.38−7.27(m, 5H), 6.82−6.75 (m, 1H), 4.58 (s, 2H), 3.68 (AB d, J = 10.0 Hz, 1H), 3.63 (AB d, J = 10.0 Hz, 1H), 3.49 (AB d, J = 10.8 Hz, 1H), 3.43 (AB d, J = 10.8 Hz, 1H), 2.92−2.72 (AB m, 2H), 2.15−2.08 (m, 2H), 1.76−1.67 (m, 1H), 1.66−1.58 (m, 2H), 1.17−1.02 (m, 4H), 0.85 (m, 12H). ¹³C NMR (101 MHz, CDCl₃): δ 169.3, 140.8, 137.3, 128.5, 127.9, 127.7, 127.1, 81.5, 73.8, 73.1, 43.9, 43.9, 34.8, 34.1, 32.8, 25.2, 23.0, 22.9, 22.6, 11.0. HRMS (ESI): m/z calcd for C₂₄H₃₆IO₃ [M + H]⁺, 499.1709; found, 499.1703.

(E)-5-((Benzyloxy)methyl)-3-(3-isobutyl-5-methylhexylidene)-5-(((5-methylisoxazol-3-yl)oxy)methyl)dihydrofuran-2(3H)-one (23).

From 21, a solution of triphenylphosphine (325 mg, 1.24 mmol), 21 (160 mg, 0.41 mmol), and 5-methylisoxazol-3-ol (62 mg, 0.62 mmol) in anhydrous tetrahydrofuran (5 mL) was treated with diisopropyl azodicarboxylate (243 μL, 1.24 mmol) and stirred at room temperature overnight. The reaction mixture was evaporated, and the residue was purified by flash chromatography using hexane/ethyl acetate (95:5) as an eluent to afford 23 (70 mg, 37%) as a colorless syrup. ¹H NMR (400 MHz, CDCl₃): δ 7.37−7.25 (m, 5H), 6.83−6.73 (m, 1H), 5.59 (s, 1H), 4.57 (s, 2H), 4.41−4.32 (AB q, 2H), 3.65 (AB d, J = 13.4 Hz, 1H), 3.57 (AB d, J = 13.4 Hz, 1H), 2.92−2.70 (AB m, 2H), 2.32 (s, 3H), 2.11 (t, J = 8.0 Hz, 2H), 1.74−1.54 (m, 3H), 1.15−1.02 (m, 4H), 0.85 (d, J = 6.5 Hz, 12H). ¹³C NMR (101 MHz, CDCl₃): δ 171.6, 170.7, 169.7, 140.3, 137.4, 128.4, 127.8, 127.6, 126.8, 92.9, 82.2, 73.7, 71.5, 70.8, 43.7, 34.7, 32.7, 30.5, 25.2, 22.9, 22.6, 12.9. HRMS (ESI): m/z calcd for C₂₈H₃₉NO₅Na [M + Na]⁺, 492.2726; found, 492.2742.

From 22, a solution of iodine 22 (50 mg, 0.10 mmol), K₂CO₃ (140 mg, 1 mmol), and 5-methylisoxazol-3-ol (99 mg, 1 mmol) in DMF (0.5 mL) was stirred at 140 °C for 4 h in a sealed vial. Afterward, the reaction was cooled to room temperature, diluted with CH₂Cl₂ (3 mL), and washed with NaOH (5%) and brine. The organic phase was dried over anhydrous Na₂SO₄, filtered, and concentrated under a vacuum. The residue was purified by silica gel flash column chromatography hexane/ethyl acetate (95:5) to afford 23 (29 mg, 62%).

(E)-5-(Hydroxymethyl)-3-(3-isobutyl-5-methylhexylidene)-5-(((5-methylisoxazol-3-yl)oxy)methyl)dihydrofuran-2(3H)-one (2, 10B12).

Starting from 22 (26 mg, 0.055 mmol) and following general procedure B, compound 2 (10B12) was obtained as a white solid (14 mg, 67%). Mp: 81−82 °C. ¹H NMR (400 MHz, CDCl₃): δ 6.80 (m, 1H), 5.63 (s, 1H), 4.40 (AB d, J = 11.3 Hz, 1H), 4.32 (AB d, J = 11.3 Hz, 1H), 3.78−3.67 (m, 2H), 2.85 (AB d, J = 17.1 Hz, 1H), 2.73 (AB d, J = 17.1 Hz, 1H), 2.71 (t, J = 6.9 Hz, 1H), 2.32 (s, 3H), 2.13 (t, J = 6.5 Hz, 2H), 1.76−1.55 (m, 3H), 1.15−1.03 (m, 4H), 0.85 (d, J = 6.5 Hz, 12H). ¹³C NMR (101 MHz, CDCl₃): δ 171.8, 171.0, 169.7, 141.2, 126.5, 92.9, 83.1, 70.5, 64.4, 43.8, 34.8, 32.8, 30.0, 25.2, 25.2, 22.9, 22.6, 12.9. HRMS (ESI): m/z calcd for C₂₁H₃₄NO₅ [M + H]⁺, 380.2437; found, 380.2451.

5-((Benzyloxy)methyl)-5-(hydroxymethyl)-3-(propan-2-ylidene)-dihydrofuran-2(3H)-one (25).

Compound 25 was synthesized following a procedure described elsehwere.⁴¹

Starting from lactone 19 (3.51 g, 10.24 mmol) and acetone (1.78 g, 30.72 mmol) and following general procedure A, 24 (3.5 g, 90%) was obtained as a white solid. Mp: 77−79 °C. Its spectrum data perfectly matched with published results. ¹H NMR (300 MHz, CDCl₃): δ 7.37−7.26 (m, 5H), 6.87 (s, 4H), 4.65 (AB q, J = 12.3 Hz, 2H), 4.11 (AB q, J = 9.6, 2H), 3.82 (s, 3H),3.71 (AB q, J = 10.2, 2H), 2.97−2.84 (AB m, 2H), 2.33 (t, J = 2.1 Hz, 2H), 1.92 (s, 3H).

Starting from 24 (890 mg, 2.32 mmol) and following general procedure C, 25 was obtained as a colorless oil (586 mg, 92%). Its spectrum data perfectly matched with published results.¹H NMR (400 MHz, CDCl₃): δ 7.40−7.26 (m, 5H), 4.56 (AB q, J = 12.1, 2H), 3.70 (AB dq, J = 6.6, 12.1, 34.5 Hz, 2H), 3.55 (AB q, J = 9.9, 27.6 Hz, 2H), 2.82−2.67 (m, 2H), 2.25 (m, 4H), 1.86 (s, 3H).

5-((Benzyloxy)methyl)-5-(iodomethyl)-3-(propan-2-ylidene)-dihydrofuran-2(3H)-one (26).

A solution of 25 (200 mg, 0.73 mmol) in toluene (13 mL) was treated with triphenylphosphine (380 mg, 1.45 mmol), imidazole (148 mg, 2.18 mmol), and iodine (277 mg, 1.1 mmol). The mixture was heated at 90 °C for 2 h and, after reaching room temperature, was poured into a saturated solution of NaHCO₃. Excess triphenylphosphine was destroyed by the addition of iodine until the iodine coloration persisted in the organic layer. The organic layer was washed with Na₂S₂O₃ (10%, 5 mL) and brine, dried over (MgSO₄), filtered, and concentrated in vacuo. The residue was purified on silica gel (hexanes/ethyl acetate, 9:1) to give 26 (266 mg, 95%) as a colorless oil. ¹H NMR (400 MHz, CDCl₃): δ 7.38−7.27 (m, 5H), 4.64−4.53 (AB m, 2H), 3.68−3.60 (AB m, 2H), 3.49 (AB d, J = 14.2 Hz, 1H), 3.40 (AB d, J = 14.2 Hz, 1H), 2.91−2.73 (AB m, 2H), 2.26 (t, J = 2.8 Hz, 3H), 1.86 (s, 3H). ¹³C NMR (101 MHz, CDCl₃): δ 168.6, 151.6, 137.4, 128.4, 127.8, 127.7, 119.3, 79.7, 73.7, 73.2, 35.9, 24.6, 20.0, 11.4. HRMS (ESI): m/z calcd for C₁₆H₂₀IO₃ [M + H]⁺, 387.0457; found, 387.0468.

5-((Benzyloxy)methyl)-5-((5-phenyl-2H-tetrazol-2-yl)methyl)-3-(propan-2-ylidene)dihydrofuran-2(3H)-one (27).

A solution of 26 (111 mg, 0.28 mmol), K₂CO₃ (386 mg, 2.8 mmol), and 5-phenyl-1H-tetrazole⁴⁰ (420 mg, 2.8 mmol) in DMF (1 mL) was stirred at 120 °C overnight in a sealed vial, then cooled to room temperature, diluted with CH₂Cl₂ (3 mL), and filtered. The organic phase was washed with brine, dried over anhydrous Na₂SO₄, filtered, and concentrated under a vacuum. The residue was purified by silica gel flash column chromatography hexane/ethyl acetate (92:8) to afford 27 (84 mg, 74%) as a white solid. Mp: 87−88 °C. ¹H NMR (400 MHz, CDCl₃): δ 8.15−8.08 (m, 2H), 7.52−7.45 (m, 3H), 7.38−7.27 (m, 5H), 5.00 (AB d, J = 14.0 Hz, 1H), 4.92 (AB d, J = 14.0 Hz, 1H), 4.67 (AB d, J = 11.9 Hz, 2H), 4.61 (AB d, J = 11.9 Hz, 1H), 3.68 (AB d, J = 10.1 Hz, 1H), 3.60 (AB d, J = 10.1 Hz, 1H), 2.95 (d, J = 16.6 Hz, 1H), 2.75 (d, J = 16.6 Hz, 1H), 2.04 (s, 3H), 1.72 (s, 3H). ¹³C NMR (101 MHz, CDCl₃): δ 168.0, 165.4, 152.0, 137.2, 130.5, 128.9, 128.5, 128.0, 127.8, 127.0, 126.9, 117.6, 79.9, 73.8, 72.7, 56.9, 33.4, 24.4, 19.8.HRMS (ESI): m/z calcd for C₂₃H₂₅N₄O₃ [M + H]⁺, 405.1927; found, 405.1936.

5-(Hydroxymethyl)-5-((5-phenyl-2H-tetrazol-2-yl)methyl)-3-(propan-2-ylidene)dihydrofuran-2(3H)-one (5).

Starting from 27 (78 mg, 0.194 mmol) and following general procedure B, 5 was obtained as a white solid (55 mg, 90%). Mp: 154−155 °C. ¹H NMR (400 MHz, CDCl₃): δ 8.15−8.08 (m, 2H), 7.52−7.45 (m, 3H), 4.96 (s, 2H), 3.79 (d, J = 6.4 Hz, 2H), 2.95 (d, J = 16.4 Hz, 1H), 2.73 (d, J = 16.4 Hz, 1H), 2.74−2.68 (m, 1H), 2.07 (s, 3H), 1.76 (s, 3H). ¹³C NMR (101 MHz, CDCl₃): δ 168.1, 165.5, 152.9, 130.6, 128.9, 126.9, 126.8, 117.5, 81.0, 65.8, 56.3, 32.4, 24.5, 19.9. HRMS (ESI): m/z calcd for C₁₆H₁₉N₄O₃ [M + H]⁺, 315.1457; found, 315.1467.

(E)-5-((Benzyloxy)methyl)-3-(3-isobutyl-5-methylhexylidene)-5-((5-phenyl-2H-tetrazol-2-yl)methyl)dihydrofuran-2(3H)-one (29).

A solution of triphenylphosphine (226 mg, 0.85 mmol), lactone 21 (110 mg, 0.29 mmol), and 5-phenyl-1H-tetrazole (124 mg, 0.85 mmol) in anhydrous THF (6 mL) was treated with DIAD (168 μL, 0.85 mmol) and stirred at room temperature overnight. The reaction mixture was evaporated, and the residue was purified by flash chromatography using hexane/ethyl acetate (92:8) as an eluent to afford 29 as a white solid (78 mg, 53%). Mp: 85−86 °C.¹H NMR (400 MHz, CDCl₃): δ 8.14−8.09 (m, 2H), 7.50−7.45 (m, 3H), 7.38−7.28 (m, 5H), 6.68−6.61 (m, 1H), 5.03 (d, J = 14.1 Hz, 1H), 4.93 (d, J = 14.1 Hz, 1H), 4.66 (d, J = 11.9 Hz, 1H), 4.60 (d, J = 11.9 Hz, 1H), 3.68 (d, J = 10.1 Hz, 1H), 3.59 (d, J = 10.1 Hz, 1H), 2.93 (d, J = 17.0 Hz, 1H), 2.76 (d, J = 17.0 Hz, 1H),1.99 (t, J = 6.2 Hz, 2H), 1.66−1.58 (m, 1H), 1.58−1.47 (m, 2H), 1.08−0.91 (m, 4H), 0.82−0.78 (m, 12H). ¹³C NMR (101 MHz, CDCl₃): δ 168.8, 165.4, 141.4, 137.0, 130.5, 128.9, 128.5, 128.0, 127.8, 126.9, 125.3, 81.5, 73.8, 72.4, 56.6, 43.7, 43.7, 34.7, 32.6, 31.2, 25.1, 22.9, 22.8, 22.6, 22.5. HRMS (ESI): m/z calcd for C₃₁H₄₁N₄O₃ [M + H]⁺, 517.3179; found, 517.3182.

(E)-5-(Hydroxymethyl)-3-(3-isobutyl-5-methylhexylidene)-5-((5-phenyl-2H-tetrazol-2-yl)methyl)dihydrofuran-2(3H)-one (6).

Starting from 29 (50 mg, 0.097 mmol) and following general procedure B, 6 was obtained as a white solid (32 mg, 78%). Mp: 110−112 °C.¹H NMR (400 MHz, CDCl₃): δ 8.20−8.15 (m, 2H), 7.51−7.49 (m, 3H), 6.70 (m, 1H), 5.10−4.94 (AB m, 2H), 3.79−3.76 (m, 2H), 2.94 (d, J = 17.2 Hz, 1H), 2.73 (d, J = 17.2 Hz, 1H), 2.57 (t, J = 6.4 Hz, 1H), 2.03 (t, J = 6.3 Hz, 2H), 1.67−1.60 (m, 1H), 1.60−1.48 (m, 2H), 1.11−0.93 (m, 4H), 0.82 (d, J = 6.5 Hz, 12H). ¹³C NMR (101 MHz, CDCl₃): δ 168.8, 165.6, 142.2, 130.6, 128.9, 126.9, 126.8, 125.3, 82.5, 65.7, 56.0, 43.8, 43.8, 34.8, 32.7, 30.4, 25.1, 22.9, 22.8, 22.6, 22.5. HRMS (ESI): m/z calcd for C₂₄H₃₅N₄O₃[M + H]⁺, 427.2709; found, 427.2733.

Determination of In Vitro Binding Affinities for PKC Isozymes.

Recombinant human PKCα and PKCε isozymes were obtained from ThermoFisher Scientific (Waltham, MA). Compounds were assayed in vitro by competition for the binding of [³H]PDBu (9 Ci/mmol; prepared as a custom synthesis by Quotient Bioresearch, Cardiff, United Kingdom). Porcine brain l-α-phosphatidylserine (PS) was from Avanti Polar Lipids (Alabaster, AL). Immunoglobulin G was purchased from Sigma-Aldrich (St. Louis, MO). Nonradioactive phorbol 12,13-dibutyrate (PDBu) was purchased from LC Laboratories (Woburn, MA). Polyethylene glycol (PEG) 6000 was obtained from EMD Millipore Corporation (Billerica, MA). Briefly, the assay mixture had a final volume of 250 μL and contained the following components: 50 mM Tris-HCl, pH 7.4; 100 μg/mL l-α-phosphatidylserine; 0.1 mM Ca²⁺ for PKCα assay and 1 mM EGTA for PKCε assay; 5 mg/mL bovine immunoglobulin G and 0.003% Tx-100, 10 μM of nonradioactive phorbol 12,13-dibutyrate (PDBu), 2 nM of [³H]PDBu and increasing concentrations of the competing ligand. The assay tubes were incubated at 37 °C for 5 min and then chilled on ice for 10 min, after which 200 μL of 35% polyethylene glycol (PEG) was added. The tubes were vortexed and chilled for an additional 10 min. The tubes were then centrifuged at 12,200 rpm for 15 min at 4 °C. A 200 μL aliquot of each supernatant was transferred to a scintillation vial for the determination of the free concentration of [³H]PDBu. The assay pellets were dried, cut off, and placed in a separate scintillation vial for the determination of bound [³H]PDBu. The supernatant and pellet vials were immersed in 3 mL of scintillation cocktail (CytoScintEcoSafe, MP Biomedicals, Santa Ana, CA). In each experiment, triplicate measurements at each concentration of ligand were performed. MS Excel was used to process data, and graphing was done using GraphPad Prism software (GraphPad Software, Inc., San Diego, CA). Final concentrations for both PKC isozymes assays (fixed): 10 μM cold PDBu (nonspecific), 2 nM [³H]PDBu, 20 ng/tube enzyme, 100 μg/mL l-α-phosphatidylserine (porcine brain). In each experiment, at least triplicate measurements at each concentration of ligand were performed. K_i values were calculated from ID₅₀ values determined from the competition curves. MS Excel was used to process data, and graphing was done using GraphPad Prism software (GraphPad Software, Inc., San Diego, CA).

Translocation of GFP-Fused PKCs.

Experiments were carried out essentially as previously described.¹⁷ Briefly, HeLa cells (5 × 10⁴) were transfected with 1 μg of either pEGFP-N1-PKCα or pEGFP-N1-PKCε using Lipofectamine 3000 and plated on cover slides in 24-well plates. After 24 h, cells were serum-starved for 24 h and then stimulated with the different DAG-lactones or PMA for 30 min. Cells were then fixed in methanol, mounted on a glass slide, and visualized with a Nikon TE2000-U fluorescence microscope. Cells were stained with DAPI for nucleus visualization. Translocation was quantitated using ImageJ. A line was traced across the cytoplasm of individual cells (~10−15 cells/group), and the signal intensities and profiles were obtained using the Plot profile tool of the program. The operator was blinded to treatment assignment.

Western Blotting.

A549 cells were harvested in lysis buffer containing 50 mM Tris-HCl, pH 6.8, 10% glycerol, and 2% β-mercaptoethanol, and lysates were subjected to SDS-PAGE. After transferring to polyvinylidene difluoride membranes (Millipore Corp., Burlington, MA) and blockade for 1 h with 5% milk in TBS and 0.1% Tween 20, membranes were incubated overnight with either anti-PKCε (1:1000 dilution, Cell Signaling Technology, Danvers, MA) or β-actin (1:500 000, Sigma-Aldrich, St. Louis, MO) antibodies. After extensive washing, membranes were incubated for 1 h with either antimouse (1:1000 dilution) or antirabbit (1:3000 dilution) secondary antibodies conjugated to horseradish peroxidase (Bio-Rad, Hercules, CA). Bands were visualized and subjected to densitometric analysis using an Odyssey Fc system (LI-COR Biosciences, Lincoln, NE).

Quantitative PCR (Q-PCR).

Total RNA from cultured cells was extracted using the RNeasy kit as directed by the manufacturer (Qiagen, Valencia, CA). One μg of mRNA template was added to the reverse-transcription master mix (Taq-Man Reverse Transcription kit, Applied Biosystems). The cDNA samples were then diluted with 90 μL of RNase-free water and stored at −20 °C. Q-PCR primers for PKCε and UBC (housekeeping gene for normalization) were purchased from Applied Biosystems. PCR amplifications were performed using an ABI PRISM 7300 Detection System in a total volume of 20 μL containing Taqman Universal PCR Master Mix (Applied Biosystems). PCR product formation was continuously monitored using the Sequence Detection System software version 1.7 (Applied Biosystems).

Phalloidin Staining for Analysis of Cell Ruffle Formation.

A549 lung cancer cells growing on glass coverslides at low confluence were serum-starved for 24 h and stimulated with the different compounds at the concentrations indicated. Following fixation with 4% formaldehyde, F-actin was stained with phalloidin-rhodamine, and nuclei were counterstained with DAPI. Slides were visualized by fluorescence microscopy, and five random fields were scored for ruffle formation. Ruffle area was measured by thresholding for signal intensity using ImageJ.

RNAi Silencing.

A549 cells were transfected with validated siRNA duplexes for PKCε or a nontarget control (NTC) siRNA duplex (Dharmacon) using Lipofectamine RNAiMAX (Invitrogen), as previously described.¹⁷ siRNAi’s for PKCε were as follows: J-004653−06 (ε1) and J-004653−08 (ε2). Experiments were carried out 48 h after transfection.

Statistical Analysis.

We used GraphPad Prism software built-in analysis tools for ANOVA. A p value <0.05 was considered statistically significant.

Supplementary Material

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acs.jmedchem.1c00739.

Data for ¹H, ¹³C NMR spectra and HPLC traces for the title compounds (PDF)

Molecular formula strings (CSV)

Structure representation of isoxazole_sn1 (PDB)

Structure representation of isoxazole_sn2 (PDB)

Structure representation of tetrazole_sn1 (PDB)

Structure representation of triazole_sn1 (PDB)

ABBREVIATIONS USED

CAN

ceric ammonium nitrate

DAG

diacylglycerol

DBU

1,8-diazabicyclo[5.4.0]undec-7-ene

DIAD

diisopropyl azodicarboxylate

DMAP

4-dimethylaminopyridine

DMF

dimethylformamide

GEFs

guanine nucleotide exchange factors

LiHMDS

lithium hexamethyldisilazide

mp

melting point

MRCKs

myotonic dystrophy kinase-related Cdc42-binding kinases (MRCKs)

PDBu

[20−3H]phorbol 12,13-dibutyrate

PKC

protein kinase C

PKD

protein kinase D

PMP

p-methoxyphenyl

pyr

pyridine

RasGRP

Ras guanine nucleotide-releasing protein

TBAF

tetra-n-butylammonium fluoride

TBDPS

tert-butyldiphenylsilyl

t-BuOH

tert-butylalcohol

THF

tetrahydrofuran