Function-Oriented Synthesis: Design, Synthesis, and Evaluation of Highly Simplified Bryostatin Analogues

Abstract

Using a function-oriented synthesis strategy, we designed, synthesized, and evaluated the simplest bryostatin 1 analogues reported to date, in which bryostatin’s A- and B-rings are replaced by a glutarate linker. These analogues, one without and one with a C26-methyl group, exhibit remarkably different protein kinase C (PKC) isoform affinities. The former exhibited bryostatin-like binding to several PKC isoforms with K_i’s < 5 nM, while the latter exhibited PKC affinities that were up to ~180-fold less potent. The analogue with bryostatin-like PKC affinities also exhibited bryostatin-like PKC translocation kinetics in vitro, indicating rapid cell permeation and engagement of its PKC target. This study exemplifies the power of function-oriented synthesis in reducing structural complexity by activity-informed design, thus enhancing synthetic accessibility, while still maintaining function (biological activity), collectively providing new leads for addressing the growing list of therapeutic indications exhibited by PKC modulators.

Graphical Abstract

INTRODUCTION

Bryostatin 1 (1, henceforth bryostatin) has attracted intense multidisciplinary interest because of its unique molecular architecture, polypharmacology, and potential to address unmet clinical needs. It has been tested in over 40 cancer trials,¹ most conducted before its fascinating immunomodulatory activities were uncovered. As its promising modes of action unfolded, it has been used to increase antigen density for antigen-targeted cell and monoclonal therapies,² as an adjuvant for reversing latent HIV reservoirs directed at HIV/AIDS eradication³ and as a treatment for Alzheimer’s disease.⁴ It is also being investigated for the treatment of fragile X disorder,⁵ multiple sclerosis,⁶ Niemann–Pick disease,⁷ Char-cot–Marie–Tooth syndrome,⁸ and ischemic stroke.⁹ Its use as an adjuvant to enhance chimeric antigen receptor T cell therapy for children and young adults with acute lymphoblastic leukemia is being planned for clinical evaluation as part of a cooperative research and development agreement (CRADA) involving the National Cancer Institute (NCI) and Bryologyx.¹⁰ This range of therapeutic indications is thought to result from bryostatin’s potent modulation of conventional and novel protein kinase C (PKC) isoforms,^1–9,11 a family of eight C1-containing PKC isozymes (conventional: α, βI, βII, and γ; novel: δ, ε, η, and θ) that serve as central signaling kinases in a variety of therapeutically relevant biological processes.¹²

Despite the growing clinical significance of bryostatin, its natural supply, similar to that of many natural products, is scarce and variable. In 1991, the NCI sponsored the large-scale harvesting and extraction of 14 tons of Bugula neritina, the marine source organism, which afforded only 18 g of bryostatin.¹³ To circumvent the cost, time, and environmental issues associated with harvesting wild-type marine organisms, aquaculture had also been explored, but the production of bryostatin by a symbiotic bacterium (Endobugula sertula) in a non-natural ecosystem proved to be low-yielding, irreproducible, and expensive.¹⁴ Engineered biosynthesis has also been considered, but researchers are yet to find an appropriate expression system for the bryostatin polyketide synthase.¹⁵ Finally, while several research groups have reported impressive syntheses of various members of the bryostatin family,¹⁶ until recently, only the Keck group had produced a total synthesis of bryostatin 1, which, while noteworthy, at 57 steps would be a cost, time, and scalability challenge to meet supply needs.^16c In 2017, the Wender group provided a sustainable, step-economical, scalable, and cost-effective solution to this clinical supply problem in the form of a 29-step (19 steps in the longest linear sequence) gram-scale synthesis of bryostatin, reducing by half the step count of the only other synthesis.¹⁷ Based on current clinical dosing, this now completed multigram-scale GMP synthesis^10a,c addresses clinical needs as 1 g of bryostatin could treat upward of 1000 patients based on reported clinical trials.^2a,3a,4a

Notwithstanding its clinical potential, bryostatin, putatively used by its source organism as an antifeedant, is neither evolved nor optimized for clinical use. This is not unusual, as only 4% of all new drugs reported between 1981 and 2019 (n = 1881) were natural products,¹⁸ while >40% were natural product-inspired or close-in derivatives or precursors such as taxol-inspired taxotere and avermectin-inspired ivermectin.^19,20 In the case of bryostatin, however, access to immediate precursors or derivatives has been blocked by the lack of natural material, its sequestration for compassionate use, and the expected challenges in modifying its complex and densely functionalized structure. Illustrative of its scarcity, the cost of a milligram of bryostatin, the minimum needed for modification and characterization, is currently approximately $18,000. To address these problems, we proposed, based on computer analyses, the first pharmacophoric model for bryostatin’s interaction with its PKC target, leading to the suggestion that its association with PKC is a function of only part of its complex structure with its C ring contacting the C1 domain of PKC, while its A and B rings preorganize the PKC contacting units for binding and additionally influence transport and membrane association that drive downstream effects.^21a These studies have been more recently augmented by long timescale dynamics simulations of bryostatin bound to PKC in a membrane and the REDOR NMR structures of a bryostatin analogue bound to the PKC binding domain in a membrane microenvironment.²² Using this blueprint for design, we proposed a new class of bryostatin analogues in which the target recognition domain (largely the C ring) was kept intact, but the conformation controlling domain (A and B rings) was vastly simplified. This function-oriented synthesis (FOS) approach²¹ led to the first bryostatin analogues 2a and 2b (Figure 1) in 1998 and 2002, respectively,^21a with a simplified AB ring system. In line with our pharmacophore hypothesis, these analogues had PKC affinities comparable to bryostatin. Moreover, in functional studies conducted by the NCI, they were shown to significantly outperform bryostatin in cancer cell growth inhibition assays with some analogues being 2 orders of magnitude more effective.^21a As a nonmacrocyclic control, we also prepared 2c, which, in line with our pharmacophore model and the proposed role of the macro-cycle in controlling the conformation of the PKC binding elements, exhibited a binding affinity to rat brain PKC mixtures that was several orders of magnitude weaker than bryostatin and our designed analogues.^21a

Figure 1.

FOS leading to a library of bryostatin analogues with increasing simplification, some of which possess biological activity that is comparable to bryostatin 1.

Since then, our group and others have used this FOS strategy to design and synthesize several simplified bryostatin analogues (Figure 1). In agreement with our pharmacophore model, deletion of the bryostatin A- or B-ring systems (classes of analogues 3²³ and 4²⁴), or incorporation of readily available diacid linkers as spacer domains (classes of analogues 5²⁵ and 6²⁶), provided analogues that exhibited bryostatin-like biological activity, highlighting the clinical potential of bio-inspired, designed analogues that are considerably more synthetically accessible and several-fold more active than the natural product itself.^21a,23–26

Prompted by the above progress in which the spacer domain is progressively simplified, yielding analogues that are more synthetically accessible, we asked to what extent the complexity of the spacer domain could be further reduced and thus synthetic access increased while still maintaining relevant biological activity. Herein, we report an answer to that question in the form of a new class of highly simplified analogues 7a and 7b (without and with the C26 methyl group, respectively) whose spacer domains are derived simply from commercially available glutaric acid conjoined to two slightly different recognition domains without and with a C26 methyl group. These compounds represent the simplest bryostatin analogues reported to date, with one retaining biological activity comparable to the natural product.

RESULTS AND DISCUSSION

Synthesis of C26-Desmethyl Bryostatin Analogue 7a.

We hypothesized that if the spatial array of binding elements in the recognition domain of bryostatin, that is, the groups contacting the PKC protein, were retained using commercially available spacer domains, new bryostatin analogues could be realized with unprecedented step economy. This strategy would also furnish analogues needed to further advance our understanding of the pharmacophoric requirements for binding to PKC. While beyond the scope of this current study, this FOS approach would allow further investigation of selectivity, efficacy, and tolerability associated with disease-specific applications of PKC modulators. We first tested this hypothesis computationally with the des-C26-methyl recognition domain (Figure 2). We used Hoye’s protocol to perform conformational searches by Monte Carlo/molecular mechanics followed by further geometric optimizations and frequency calculations at the M06−2X/6−31+G(d,p) level of theory.²⁷ We then overlaid a conformer of the C20 truncated bryostatin 1′ (see the Supporting Information, page S26) and the lowest energy conformer of compound 7a′, the C20 truncated analogue of 7a (Figure 2).²⁸ Despite the highly simplified spacer domain of 7a′, it adopted a similar conformation to that portion of bryostatin proposed to contact PKC and a very similar spatial array of pharmacophoric elements at C1, C19, and C26 (rmsd = 0.13 c5). Prompted by this encouraging computational correspondence, we embarked on the synthesis of 7a, starting from the known intermediate 8, which can readily be prepared in 11 steps from neopentyl glycol in decagram quantities.^21a

Figure 2.

A conformer (gray) of C20-truncated bryostatin (1′) based on the X-ray crystal structure of bryostatin,¹⁷ the lowest energy conformer (green) of the equivalently truncated C26-desmethyl analogue with a glutarate linker (7a′) and the overlaid structures. *The rmsd was calculated for the oxygen at C19, the carbonyl oxygen at C1, and the C26–OH as they are the proposed PKC pharmacophoric elements of bryostatin.^21a

From compound 8 (Figure 3), the silyl ether was deprotected, immediately followed by Steglich esterification with monobenzyl glutarate to form ester 9. Ester 9 underwent Sharpless asymmetric dihydroxylation on the terminal alkene to afford, with modest diastereoselectivity, a mixture of diols 10.²⁹ This mixture was treated with aqueous hydrogen fluoride (HF) in acetonitrile to hydrolyze the mixed C19 ketal. Subsequently, the primary C26 hydroxyl group was selectively silylated using tert-butyldimethylsilyl chloride (TBSCl) in dichloromethane (DCM), and the mixture of diastereomers at C25 was separated to afford 11 as a single diastereomer in 38% yield over three steps without intermediate silica gel purification. Hydrogenolysis of the benzyl ester followed by Yamaguchi lactonization afforded macrocyclic lactone 12 in 20% yield over two steps.³⁰ Final deprotection with p-toluenesulfonic acid monohydrate (pTsOH·H₂O) in acetonitrile/water gave the desired product 7a in 75% yield. Overall, analogue 7a could be prepared in eight steps, with only four chromatographic purifications from 8 (total step count: 19 from neopentyl glycol) in an unoptimized 3.3% overall yield.

Figure 3.

Synthetic sequence for preparation of the C26-desmethyl highly simplified analogue. Reagents and conditions: (a) 3HF·Et₃N, THF, r.t.; (b) BnO₂C–(CH₂)₃–CO₂H, DCC, DCM, DMAP, r.t.; (c) K₂OsO₂(OH)₄, (DHQD)₂PYR, K₂CO₃, K₃Fe(CN)₆, ^tBuOH/H₂O (1:1 v/v), 4 °C; (d) HF (aq), MeCN, r.t.; (e) TBSCl, imidazole, DMAP, DCM, r.t.; (f) H₂, Pd/C, r.t.; (g) 2,4,6-trichlorobenzoyl chloride, Et₃N, DMAP, THF, toluene; (h) pTsOH·H₂O, MeCN/H₂O (4:1 v/v).

Preparation of the Simplified C26-Methyl Bryostatin Analogue 7b.

Having successfully prepared the highly simplified analogue 7a, we next addressed the synthesis of 7b bearing a C26 methyl group. Growth inhibition studies conducted by the NCI showed that for many cancer cell lines, analogues 2a and 2b both performed better than bryostatin (for some cell lines 2 orders of magnitude better) but 2b, lacking the C26 methyl group present in the natural product, was generally 10-fold more effective than 2a,^21a raising the question whether this difference would be observed in these more simplified analogues.

We started with ketone 13, a key intermediate in our bryostatin synthesis, which was prepared in five steps and three chromatographic purifications starting from 3,4-dihydro-2H-pyran.¹⁷ Ketone 13 (Figure 4A) was reduced using the Luche protocol,¹⁷ with subsequent esterification furnishing ester 14 in 90% yield. Ester 14 was then submitted to a Sharpless asymmetric dihydroxylation,²⁹ which proceeded chemoselectively at the more substituted, less hindered double bond, providing an intermediate diol in good yield and diaster-eoselectivity (14:1) in contrast to the lower selectivity found in the dihydroxylation leading to analogue 7a. Without purification, the C26–OH of the resulting diol was regioselectively silylated with TBSCl in DMF. The resulting alkene 15 was then subjected to our stoichiometric ozonolysis protocol.¹⁷ We initially envisioned the possibility of an in situ reduction of the ozonide 16 with sodium borohydride (NaBH₄) to make alcohol 17 in a single transformation from the terminal alkene.³¹ Interestingly, the ozonide group in 16 (generated in situ) was inert under these conditions and even survived column chromatography putatively because of its neopentyl environment. Fortunately, reduction of ozonide 16 with triphenylphosphine (PPh₃) provided aldehyde 18 in 96% yield.³² We then reduced aldehyde 18 under routine NaBH₄/methanol conditions, but the initially formed alcohol 17 readily underwent an irreversible trans-esterification to give the undesired ester 19. We attempted this reduction in a variety of different alcoholic solvents (e.g., isopropanol) as well as using other hydride sources (LiBH₄, ⁿBu₄NBH₄, NaBH-(OAc)₃, etc.), but all either led exclusively to trans-esterification or gave very slow conversion to mixtures of alcohols 17 and 19 (Figure 4B, and Supporting Information, page S2). We postulated that both the basicity of borohydride reducing conditions and the hydrogen bonding network in alcoholic solvents encouraged the trans-esterification process and that this could be minimized by the utilization of neutral borane–amine complexes in aprotic solvents.³³ Gratifyingly, treatment of aldehyde 18 with a tert-butylamine borane (^tBuNH₂·BH₃) complex,³³ followed by HCl workup, provided alcohol 17 cleanly without any sign of trans-esterification. At this point, we envisioned the use of glutaric anhydride, instead of monobenzyl glutarate, for esterification of neopentyl alcohol 17. Strategically, this would save an additional debenzylation step via hydrogenolysis that had been utilized in the synthesis of analogue 7a. Furthermore, the success of this strategy would demonstrate that the spacer domain for analogue 7b could be derived directly from a commercial starting material. To avoid subsequent trans-esterification upon standing, crude alcohol 17 was immediately subjected to monoesterification with glutaric anhydride, and to our pleasure seco-acid 20 was isolated in 65% yield. Subsequently, seco-acid 20 underwent macrolactonization using 2-methyl-6-nitrobenzoic anhydride in toluene at 50 °C, a condition that was developed by Shiina and co-workers,³⁴ to give bis-lactone 21 in 25% isolated yield, 41% yield based on recovered starting material 20.

Figure 4.

(A) Synthetic sequence for preparation of the highly simplified C26-methyl bryostatin analogue. Reagents and conditions: (a) NaBH₄, CeCl₃·7H₂O, MeOH, −50 °C; (b) octanoic anhydride, DMAP, DCM, r.t.; (c) K₂OsO₂(OH)₄, (DHQD)₂PHAL, K₂CO₃, K₃Fe(CN)₆, MeSO₂NH₂, tBuOH: H₂O (1:1 v/v), 4 °C; (d) TBSCl, imidazole, DMF, 0 °C to r.t.; (e) O₃, MeOH/DCM, −78 °C; (f) PPh₃, MeOH, −78 °C to r.t.; (g) tBuNH₂·BH₃, THF, r.t.; (h) glutaric anhydride, DMAP, DCM, r.t.; (i) 2-methyl-6-nitrobenzoic anhydride, DMAP, PhCH₃, 50 °C; and (j) camphorsulfonic acid (CSA), CH₃CN/H₂O (4:1 v/v). (B) Trans-esterification as a decomposition pathway upon treatment of aldehyde 18 with basic hydride sources.

With macrolactone 21 in hand, we attempted the final deprotection to yield analogue 7b. The previous conditions employing pTsOH·H₂O in acetonitrile/water that worked in the C26-desmethyl analogue series led principally to decomposition. Aqueous HF or Et₃N·3HF as acidic fluoride sources either led solely to ketal hydrolysis or decomposition. Basic fluoride sources such as tetrabutylammonium fluoride also led to decomposition, presumably because of hydrolysis of the macrolactone. We then turned to CSA as a milder acid source, and the desired product 7b was produced reliably in 65–70% yield.

Comparative Binding Affinities (K_is) of Bryostatin and Analogues 7a and 7b to Representative PKC Isozymes.

Having both the C26-methyl and C26-des-methyl analogues in hand, we evaluated their cell-free competitive binding to five different PKC isoforms representing conventional and novel PKCs. We included the parent compound, bryostatin, as a reference (Table 1). Significantly, the C26-des-methyl lactone 7a exhibited potent PKC binding affinities (K_i’s), comparable to bryostatin. Not unlike our first-generation analogues with and without a C26 methyl group (2a and 2b), the simplified des-methyl analogue 7a bound better than analogue 7b. The first-ever bryostatin analogue 2a containing a methyl group at C26 exhibited a K of 3.4 nM to rat brain PKC mixtures, then the only readily accessible PKC assay.^21a Analogue 2b lacking the C26 methyl group was a better binder (K_i = 0.25 nM) to rat brain PKC mixtures by an order of magnitude.^21a These sensitivities to minor structural changes are in line with studies on diacyl glycerol analogues,³⁵ the endogenous modulator of PKC, both studies suggesting that the reach of this hydroxyl group deep into the PKC binding pocket is subject to steric effects. This is still an ongoing subject of interest as other bryostatin analogue scaffolds prepared recently by Keck and co-workers showed no difference in biological response with respect to the presence or absence of the C26-methyl.³⁶ Compound 7a represents the most simplified bryostatin analogue reported to date with K_i values comparable to those of bryostatin. Achieving comparable affinity to bryostatin is more than sufficient for advancement as more potent analogues would encounter the formidable detection challenges arising from monitoring trace blood levels of bryostatin in clinical trials as it is administered at doses of ca. 20–40 μg/m² with correspondingly lower blood levels.^3a

Table 1.

Comparative Binding Affinity (K_i,, nM) of Highly Simplified Bryostatin Analogues across Various PKC Isoforms

	PKCα	PKCβI	PKCδ5	PKCε	PKCθ
bryostatin 1	0.81	1.6	2.4	3	1.5
7a	4.4	4	1.7	2.7	1.5
7b	240	184	241	498	283

While binding is necessary for function, it is not sufficient.^26,37 We next sought to determine the relationship of binding to intracellular PKC translocation, which requires cell entry and engagement of PKC, both being a prelude to all downstream PKC effects. The hallmark of ligand-induced PKC activation is translocation to the cell membrane, where PKC associates with related scaffolding proteins and phosphorylates its biological targets, which in turn induce relevant downstream signaling effects.¹² To test whether our more potent analogue 7a could induce PKC translocation, we treated cells that are transfected with a PKCδ-GFP fusion protein with 7a, and we measured the analogue’s ability to translocate the fluorescently tagged GFP-PKCδ fusion protein from the cytosol to the plasma membrane as a function of time. Significantly, compound 7a translocated PKCδ-GFP to a comparable extent and with comparable kinetics to that of bryostatin, both at 200 nM (Figure 5). The similarity in kinetics and extent of translocation is striking, considering the simplicity of this analogue.

Figure 5.

(A) Representative translocation images of CHO-K1 cells treated with the C26-desmethyl glutarate bislactone analogue (7a). (a,b) PKCδ-GFP translocation induced by analogue 7a at 200 nM. (c,d) PKCδ-GFP translocation induced by bryostatin 1 (1) at 200 nM and (B) normalized cytosolic fluorescence intensity readings from PKCδ-GFP in translocation assays in CHO-K1 cells at 200 nM.

CONCLUSIONS

Bryostatin has been and currently is in human clinical trials for treating significant and unsolved medical problems. As such, it represents the fastest path to the clinic for new treatment strategies. However, although bryostatin itself, the focus of over 4 decades of research, shows special promise for various clinical indications, it is not optimized for human therapy and unlikely to be the best candidate for the wide range of proposed clinical applications from neurology to oncology and virology. Bryostatin-inspired analogues, not unlike taxol-inspired taxotere and avermectin-inspired ivermectin,^19,20 could prove superior to the natural product in addressing clinical needs while offering the potential of disease-specific optimization. More accessible functional analogues also offer potentially more efficacious and better-tolerated candidates.³⁸ Guided by an FOS strategy exemplified by our computational analysis leading to the first designed simplified PKC modulators,^21a we designed, synthesized, and evaluated highly simplified bryostatin analogues to investigate the role of the bryostatin spacer domain in binding to cell free PKC and in translocating a PKC isoform fusion protein in real time in cells. We previously showed that a spacer domain is required as the C-ring recognition domain alone exhibits minimal binding to PKC (compound 2c).^21a Remarkably, analogue 7a containing a simple glutarate spacer exhibited PKC binding affinities comparable to bryostatin, meeting our potency goal as more potent analogues would, like bryostatin, pose in vivo detection problems in the clinic.^3a Significantly, the highly simplified analogue 7a translocated PKC with kinetics and efficiency comparable to those of bryostatin, indicating facile cell entry and productive engagement of PKC. We also found that the presence of a methyl group at C26 in 7b strongly attenuated binding affinity relative to the des-methyl compound 7a.

The use of FOS to identify structurally simplified agents with improved activity and synthetic accessibility is a general strategy for the design of tool compounds, preclinical leads, imaging agents, and even ligands for catalysis. Its focus on function first allows one to identify requirements for function and then use synthesis- and function-informed design or structure-generating algorithms to create diverse and more step economically accessible structures with superior activity. Informed and inspired by the structure and function of the clinical lead bryostatin, this study has produced the most simplified and most synthetically accessible bryostatin analogue (7a) reported to date, which exhibits potency and translocation activity comparable to that of bryostatin and can serve as a tool compound for advancing PKC research. More generally, this study shows how design-driven and synthesis-informed target identification, in this case inspired by bryostatin, can lead to simplified and more accessible agents with natural product-like function/activity and clinical potential.

EXPERIMENTAL SECTION

General Procedure for Chemical Synthesis.

Unless otherwise specified, all commercial reagents, solvents, and solutions were used without further purification with the following exceptions. THF and diethyl ether were purified via distillation from sodium benzophenone ketyl or by passage through an activated alumina drying column (Solv-Tek, Inc.). DCM was purified via distillation from calcium hydride or by passage through an activated alumina drying column (Solv-Tek, Inc.). Toluene was purified by passage through an activated alumina drying column (Solv-Tek, Inc.). Pyridine and triethylamine were distilled from calcium hydride under a positive pressure of nitrogen. Deuterated chloroform was dried and deacidified by storing over activated 4 c5 molecular sieves and solid potassium carbonate. Unless otherwise specified, all reactions were carried out in oven-dried (>120 °C) and/or flame-dried glassware equipped with a Teflon-coated magnetic stir bar and a rubber septum under a positive pressure of nitrogen or argon. Air- or moisture-sensitive reagents were transferred to the reaction vessel under positive pressure of nitrogen or argon via syringe or stainless steel cannula. Reactions were run at room temperature (20–25 °C) unless otherwise noted in the experimental procedure, and reported reaction temperatures refer to the external temperatures measured for the bath in which the reaction vessel was immersed. “Concentration” refers to the removal of solvent using a Buchi rotary evaporator equipped with a portable vacuum pump. Analytical thin-layer chromatography (TLC) was performed by using glass-backed silica plates coated with a 0.25 mm thickness of silica gel 60 F254 (EDM Millipore), visualized with an ultraviolet light, followed by exposure to p-anisaldehyde solution, potassium permanganate solution, or ceric ammonium molybdate solution and gentle heating. Products were purified by flash-chromatography on (230–400 mesh, grade 60, particle size 40–63 μm) purchased from Fisher Scientific or by preparative TLC with glass-backed silica plates coated with a 0.25 mm thickness of silica gel 60 F254 (EMD Millipore). pH 7 buffered silica gel was prepared by adding 10% weight pH 7 phosphate buffer to silica and rotating for ~12 h. Proton NMR spectra were recorded in CDCl₃ on a Varian 400 (400 MHz), Varian Inova 500 (500 MHz), and/or Varian Inova 600 (600 MHz). Carbon-13 NMR spectra were recorded on a Varian Inova 500 (125 MHz). The following format is used to report the proton NMR data: chemical shift in ppm [multiplicity, coupling constant(s) in Hz, integral, and assignment]. Coupling constant analysis was informed by methods Hoye and Zhao have previously reported.³⁹ Chemical shifts for proton spectra are referenced to residual solvent peak (δ = 7.26 ppm for chloroform, δ = 7.16 for benzene). First-order multiplicity is described as s (singlet), d (doublet), t (triplet), q (quartet), or a combination thereof. Non-first-order multiplets and complex overlapped peaks are identified as “m”. Chemical shifts for ¹³C NMR spectra in CDCl₃ are referenced to the carbon resonance in CDCl₃ (δ = 77.16 ppm) or C₆D₆ (δ = 128.06 ppm). Infrared spectra were acquired on a Nicolet iS 5 FT-IR Spectrometer (Thermo Fisher) and/or Nicolet 6700 FT-IR spectrometer (Thermo Fisher). Optical rotations were acquired on a P-2000 Digital Polarimeter (Jasco). High-resolution mass spectra (HRMS) were acquired at the Vincent Coates Foundation Mass Spectrometry Laboratory at Stanford.

2-((2S,3S,6R,E)-6-Allyl-2-methoxy-4-(2-methoxy-2-oxoethylidene)-3-(octanoyloxy)tetrahydro-2H-pyran-2-yl)-2 Methyl-propyl Benzyl Glutarate (9).

To a solution of 8^21a (128.7 mg, 0.232 mmol) in THF (2.3 mL, ca. 0.1 M) was added 3HF·Et₃N (ca. 389 μL, 2.32 mmol) at room temperature. The reaction mixture was stirred for 24 h and then diluted with diethyl ether (30 mL). The organic phase was washed with saturated aqueous sodium bicarbonate (2 × 20 mL), brine (2 × 20 mL), dried over Na₂SO₄, and filtered. The solution was concentrated in vacuo, and the resulting residue was subjected to the next reaction without purification. To a solution of the residue obtained above in CH₂Cl₂ (2 mL) was added a solution of BnO₂C(CH₂)₃CO₂H⁴⁰ (154.6 mg, 0.696 mmol) and 4-dimethylaminopyridine (DMAP) (85.0 mg, 0.696 mmol) in CH₂Cl₂ (1 mL). A solution of N,N′-dicyclohexylcarbodiimide (DCC) (143.6 mg, 0.696 mmol) was then added dropwise at room temperature. The reaction was stirred at room temperature for 4 h and then filtered through a plug of Celite. The filtrate was washed with H₂O (10 mL), brine (10 mL), and dried over Na₂SO₄. The solution after filtration was concentrated in vacuo. The residue was purified by chromatography (silica gel, with the gradient of 10–20% EtOAc/pentane) to give benzyl ester 9 (99.3 mg, 59% over 2 steps) as a colorless oil. Compound purity was established by TLC (one spot) analysis. R_f (pentane/EtOAc 6/1): 0.41. ¹H NMR (500 MHz, CDCl₃): δ 7.37–7.30 (m, 5H), 5.94–5.85 (m, 2H), 5.60 (s, 1H), 5.16–5.10 (m, 4H), 4.22 (d, J = 10.9 Hz, 1H), 4.10 (d, J = 10.9 Hz, 1H), 3.94 (dtd, J = 2.8, 5.8, and 11.7 Hz, 1H), 3.69 (s, 3H), 3.43 (dd, J = 1.9, and 16.5 Hz, 1H), 3.31 (s, 3H), 2.44–2.35 (m, 9H), 2.00–1.94 (m, 2H), 1.66–1.60 (m, 2H), 1.31–1.23 (m, 8H), 1.03 (s, 3H), 0.97 (s, 3H), and 0.87 (app t, J = 7.1, 3H). ¹³C{¹H} NMR (126 MHz, CDCl₃): δ 173.1, 172.9, 172.1, 166.6, 152.8, 136.0, 133.7, 128.7, 128.4, 128.3, 118.1, 116.6, 102.7, 71.7, 71.5, 69.3, 66.4, 51.3, 51.0, 45.5, 40.1, 34.5, 33.5, 33.5, 32.5, 31.8, 29.1, 29.0, 24.8, 22.7, 21.2, 20.6, 20.3, and 14.2. IR (thin film): 3070, 2952, 2931, 2958, 2360, 1738, 1733, 1665, 1456, 1435, 1392, 1362, 1312, 1224, 1154, 1044, 996, 918, 880, 827, 799, 751, and 698 cm⁻¹. HRMS (ESI+, m/z): [M + Na]⁺ calcd for C₃₆H₅₂O₁₀Na, 667.3453; found, 667.3448. [α]²²_D −15.4° (c = 1.0, CH₂Cl₂).

Benzyl (2-((2S,3S,6S,E)-6-((R)-3-((tert-Butyldimethylsilyl)oxy)-2-hydroxypropyl)-2-hydroxy-4-(2-methoxy-2-oxoethylidene)-3-(octanoyloxy)tetrahydro-2H-pyran-2-yl)-2-methyl-propyl) Glutarate (11).

K₂OsO₂(OH)₄ (1.0 mg, 2.7 μmol), (DHQD)₂PYR (6.0 mg, 6.8 μmol), K₃Fe(CN)₆ (670 mg, 2.03 mmol), and K₂CO₃ (282 mg, 2.04 mmol) were combined in a round-bottom flask, followed by the addition of H₂O (3.38 mL) and ^tBuOH (3.38 mL). The two-phase system was vigorously stirred at room temperature for 2 h. A flask containing compound 9 (47.5 mg, 0.0735 mmol) was cooled at 0 °C, and an aliquot (0.8 mL) of the stock solution was added. The yellow-orange colored reaction mixture was stirred at 0 °C for 4 days in a cold room. After dilution with H₂O, the aqueous phase was extracted with EtOAc. The combined organic phases were washed with brine, dried over Na₂SO₄, filtered, and concentrated in vacuo to afford 10 (1.5:1 dr by crude NMR). The crude 10 was dissolved in CH₃CN (2 mL) and H₂O (0.8 mL), followed by the addition of pTsOH·H₂O (98 mg, 0.516 mmol) at room temperature. The reaction was stirred at room temperature for 3 h and then quenched by the addition of saturated aqueous NaHCO₃ solution (10 mL). The resulting mixture was extracted with ethyl acetate (5 times × 10 mL). The combined organic layers were washed with brine, dried over Na₂SO₄, and filtered. The resulting solution was concentrated in vacuo to give crude hemiketal, which was immediately dissolved in DCM (7 mL). To the DCM solution of the crude hemiketal diol was added a solution of TBSCl (35 mg, 0.232 mmol) in DCM. The reaction mixture was stirred at room temperature for 3 h and then poured into a separatory funnel containing saturated aqueous NH₄Cl solution (20 mL). The organic layer was separated, and the aqueous layer was extracted with DCM (3 times × 20 mL). The combined organic layers were washed with brine, dried over Na₂SO₄, and filtered. The resulting solution was concentrated in vacuo, and the residue was purified by silica gel chromatography with the gradient 10–30% ethyl acetate in pentane to afford compound 11 (22 mg, 38% yield over three steps) as a colorless oil. Compound purity was established by TLC (one spot) analysis. Note: The intermediates in this sequence are prone to decomposition, so they were pushed forward through the next steps without purification. R_f (pentane/EtOAc 2/1): 0.64. ¹H NMR (400 MHz, CDCl₃): δ 7.37–7.31 (m, 5H), 6.01 (s, 1H), 5.16 (s, 1H), 5.11 (s, 2H), 4.27 (d, J = 11.3 Hz, 1H), 4.18 (m, 1H), 3.99 (d, J = 11.3 Hz, 1H), 3.89 (m, 1H), 3.67 (s, 3H), 3.70–3.62 (m, 2H), 3.45 (dd, J = 6.3, and 10.1 Hz, 1H), 2.87 (br s, 1H), 2.44–2.30 (m, 6H), 2.14–2.08 (m, 1H), 1.96 (app pent, J = 7.4 Hz, 2H), 1.68–1.55 (m, 4H), 1.33–1.21 (m, 10H), 1.03 (s, 3H), 1.00 (s, 3H), 0.89 (s, 9H), 0.86 (t, J = 7.0 Hz, 3H), and 0.06 (s, 6H). ¹³C{¹H} NMR (125 MHz, CDCl₃): δ 173.0, 172.8, 172.1, 166.6, 151.0, 135.9, 128.7, 128.4, 128.3, 120.2, 99.7, 73.9, 70.1, 68.3, 67.5, 67.2, 66.4, 51.3, 42.9, 39.0, 34.6, 33.4, 33.3, 31.7, 31.1, 29.1, 29.0, 26.0, 24.8, 22.7, 20.6, 20.2, 20.1, 18.5, 14.2, −5.2, and −5.3. IR (thin film): 3458, 2953, 2930, 2857, 2360, 1738, 1731, 1666, 1498, 1463, 1436, 1383, 1255, 1156, 1095, 1004, 939, 885, 837, 779, 751, 698, and 668 cm⁻¹. HRMS (ESI+, m/z): [M + Na]⁺ calcd for C₄₁H₆₆O₁₂SiNa, 801.4216; found, 801.4206. [α]_D²² −16.7° (c = 1.0, CH₂Cl₂).

(1S,3R,13S,14S,E)-3-(((tert-Butyldimethylsilyl)oxy)methyl)-13-hydroxy-15-(2-methoxy-2-oxoethylidene)-12,12-dimethyl-5,9-dioxo-4,10,17-trioxabicyclo[11.3.1]heptadecan-14-yl Octanoate (12).

To a solution of 11 (500 mg, 0.631 mmol) in EtOAc (6 mL) was added 10% w/w Pd/C (50 mg). The reaction was stirred under 1 atm of H₂ at room temperature for 1 h and then filtered through a plug of Celite. The filtrate was concentrated in vacuo to give crude carboxylic acid, which was used immediately in the next step without further purification. To a solution of crude carboxylic acid in THF (6 mL) were added 2,4,6-trichlorobenzoyl chloride (760 mg, 3.16 mmol) and Et₃N (637 mg, 6.31 mmol) at 0 °C. Once the addition was complete, the ice bath was removed and the reaction was allowed to stir at room temperature for 3 h and then diluted with toluene (14 mL, final concentration ca. 0.005 M). DMAP (1.92 g, 15.8 mmol) was added to the reaction mixture in one portion and then the reaction stirred at room temperature for 15 h. The reaction was quenched by the addition of saturated aqueous NaHCO₃ (20 mL). The resulting mixture was extracted with EtOAc (20 mL × 5 times). The combined organic layers were washed with brine, dried over Na₂SO₄, filtered, and concentrated in vacuo. The residue was purified by silica gel chromatography with gradient 5–20% EtOAc/pentane to give compound 11 (85 mg, 20% in 2 steps) as a colorless oil. Compound purity was established by TLC (one spot) analysis. R_f (pentane/EtOAc 4/1): 0.33. ¹H NMR (500 MHz, CDCl₃): δ 5.98 (d, J = 1.4 Hz, 1H), 5.26 (s, 1H), 5.18 (dddd, J = 3.8, 3.8, 3.8, and 11.6 Hz, 1H), 4.53 (s, 1H), 4.25 (tt, J = 2.5, and 11.3 Hz, 1H), 4.01 (d, J = 11.0 Hz, 1H), 3.88 (d, J = 11.0 Hz, 1H), 3.73–3.66 (m, 2H), 3.68 (s, 3H), 3.61 (dd, J = 3.7, and 11.0 Hz), 2.55–2.49 (m, 2H), 2.38–2.22 (m, 4H), 2.13–2.02 (m, 3H), 1.81–1.74 (m, 2H), 1.61–1.56 (m, 2H), 1.30–1.19 (m, 9H), 1.08 (s, 3H), 0.95 (s, 3H), 0.86 (s, 9H), 0.85 (t, 3H, J = 7.0 Hz), 0.02 (s, 3H), and 0.01 (s, 3H). ¹³C{¹H} NMR (125 MHz, CDCl₃): δ 172.2, 172.1, 172.5, 166.7, 151.9, 119.4, 100.1, 72.9, 72.6, 70.4, 65.8, 64.4, 51.3, 40.9, 35.8, 35.3, 34.7, 33.6, 31.7, 31.1, 29.1, 29.0, 26.0, 24.7, 22.7, 21.9, 21.2, 20.6, 18.5, 14.2, −5.28, and −5.30. IR (thin film): 3515, 2954, 2930, 2857, 2360, 2341, 1741, 1723, 1667, 1471, 1436, 1380, 1337, 1228, 1176, 1155, 1098, 1049, 1005, 934, 836, and 778 cm⁻¹. HRMS (ESI+, m/z): [M + Na]⁺ calcd for C₃₄H₅₈O₁₁SiNa, 693.3641; found, 693.3637. [α]_D²² −16.6° (c = 1.0, CH₂Cl₂).

(1S,3R,13S,14S,E)-13-Hydroxy-3-(hydroxymethyl)-15-(2-methoxy-2-oxoethylidene)-12,12-dimethyl-5,9-dioxo-4,10,17-trioxabicyclo[11.3.1]heptadecan-14-yl Octanoate (7a).

To a solution of 12 (42.3 mg, 0.063 mmol) in CH₃CN (1 mL) and H₂O (0.25 mL) was added pTsOH·H₂O (120 mg, 0.631 mmol) at room temperature. The reaction was stirred at rt for 30 min and then quenched by the addition of saturated aqueous NaHCO₃ solution (2 mL). The resulting mixture was extracted with EtOAc (5 times × 2 mL). The combined organic layers were washed with brine, dried over Na₂SO₄, filtered, and the filtrate was concentrated in vacuo. The residue was purified by silica gel chromatography with 50% EtOAc in pentane as the eluent to give compound 7a (26.3 mg, 75%) as an amorphous white solid. Compound purity was established by TLC (one spot) analysis. R_f (pentane/EtOAc 1/3) = 0.46. ¹H NMR (500 MHz, CDCl₃): δ 5.99 (d, J = 1.7 Hz, 1H), 5.29 (dddd, J = 3.0, 3.0, 5.6, and 11.9 Hz, 1H), 5.28 (s, 1H), 4.56 (s, 1H), 4.27 (tt, J = 2.7, and 11.3 Hz, 1H), 4.01 (d, J = 11.0 Hz, 1H), 3.90 (d, J = 10.9 Hz, 1H), 3.77 (tt, J = 3.5, and 11.6 Hz, 1H), 3.72 (dd, J = 2.1, and 14.0 Hz, 1H), 3.69 (s, 3H), 3.59 (app pent, J = 6.0 Hz, 1H), 2.63 (dd, J = 6.7, and 17.1 Hz, 1H), 2.53 (ddd, J = 2.1, 6.8, and 13.0 Hz, 1H), 2.39 (m, 1H), 2.33–2.22 (m, 3H), 2.14–2.06 (m, 2H), 1.99 (ddd, J = 3.0, 12.1, and 14.2 Hz, 1H), 1.91 (app t, J = 6.0 Hz, 1H), 1.85–1.78 (m, 2H), 1.60–1.58 (m, 2H), 1.31–1.21 (m, 9H), 1.01 (s, 3H), 0.96 (s, 3H), and 0.86 (t, J = 7.1, 3H). ¹³C{¹H} NMR (125 MHz, CDCl₃): δ 173.4, 172.2, 171.5, 166.6, 151.5, 119.6, 100.1, 72.8, 72.7, 71.6, 65.7, 65.4, 51.3, 41.0, 35.8, 35.2, 34.7, 33.5, 31.8, 30.9, 29.1, 29.0, 24.7, 22.7, 21.9, 21.2, 20.5, and 14.2. IR (thin film): 3519, 2931, 2858, 1739, 1722, 1666, 1469, 1437, 1380, 1337, 1274, 1231, 1175, 1136, 1107, 1083, 1049, 1004, 974, 937, 883, 847, and 815 cm⁻¹. HRMS (ESI+, m/z): [M + Na]⁺ calcd for C₂₈H₄₄O₁₁Na, 579.2776; found, 579.2779. [α]_D²² −49.8° (c = 1.0, CH₂Cl₂).

(2S,3S,6R,E)-6-((E)-But-2-en-1-yl)-2-methoxy-4-(2-methoxy-2-oxoethylidene)-2-(2-methylbut-3-en-2-yl)tetrahydro-2H-pyran-3-yl Octanoate (14).

A flask containing ketone 13¹⁶ (1.24 g, 3.85 mmol, 1 equiv) in dry MeOH (48 mL) was cooled to −40 °C and CeCl₃·7H₂O (716 mg, 1.92 mmol, 0.5 equiv) was added in a single portion. After 10 min, NaBH₄ (291 mg, 7.69 mmol, 2 equiv) was added in a single portion. After 35 min, the reaction had turned colorless and was deemed complete by TLC. After warming to room temperature, the reaction was partitioned between Et₂O and 3:2:1 NH₄Cl/brine/water. The aqueous layer was extracted with Et₂O (2×) and the combined organics were washed sequentially with water and brine and then dried over MgSO₄, filtered, and concentrated to a crude oil that was dissolved immediately in DCM (38 mL). Then, DMAP (2.8 g, 23.1 mmol, 6 equiv) followed by octanoic anhydride (3.4 mL, 11.55 mmol, 3 equiv) were added in single portions. After stirring at room temperature for 2 h, the reaction was partitioned between Et₂O and saturated NaHCO₃. The organics were washed with water and then dried over MgSO₄, filtered, and concentrated to a crude oil that was purified by flash chromatography (5% Et₂O/pentane) to provide ester 14 (1.56 g, 90% yield). Compound purity was established by TLC (one-spot) analysis. R_f (30% EtOAc/petroleum ether): 0.8 ¹H NMR (CDCl₃, 600 MHz): δ = 6.23 (dd, J = 10.9, and 17.7 Hz, 1H), 5.88 (td, J = 0.8, and 1.9 Hz, 1H), 5.60–5.51 (m, 2H), 5.35 (s, 1H), 4.85 (dd, J = 1.5, and 17.7 Hz, 1H), 4.82 (dd, J = 1.4, and 10.9 Hz, 1H), 3.74 (dddd, J = 2.8, 5.2, 6.6, and 11.6 Hz, 1H), 3.68 (s, 3H), 3.50 (ddd, J = 0.8, 2.8, and 15.1 Hz, 1H), 3.30 (s, 3H), 2.36–2.21 (m, 5H), 1.69–1.68 (m, 3H), 1.58 (pent, J = 7.4 Hz, 2H), 1.31–1.24 (m, 9H), 1.11 (s, 3H), 1.09 (s, 3H), and 0.88 (app t, J = 7.2 Hz, 3H). ¹³C{¹H} NMR (125 MHz, CDCl₃): δ 172.1, 166.8, 153.0, 146.3, 128.4, 126.3, 117.5, 108.7, 102.9, 72.2, 71.9, 51.6, 51.2, 46.8, 39.1, 34.7, 31.8, 31.4, 29.7, 29.0, 24.7, 24.2, 22.9, 22.7, 18.2, and 14.2. IR (thin film): 3054, 2929, 2954, 2873, 2857, 1740, 1722, 1668, 1461, 1436, 1414, 1378, 1361, 1259, 1220, 1157, 1106, 1080, 1056, 982, 942, 904, 880, 823, and 799 cm⁻¹. HRMS (ESI+, m/z): [M + Na]⁺ calcd for C₂₆H₄₂O₆Na, 473.2874; found, 473.2860. [α]_D²² −13.2° (c = 0.355, CH₂Cl₂).

(2S,3S,6S,E)-6-((2R,3R)-3-((tert-Butyldimethylsilyl)oxy)-2-hydroxybutyl)-2-methoxy-4-(2-methoxy-2-oxoethylidene)-2-(2-methylbut-3-en-2-yl)tetrahydro-2H-pyran-3-yl Octanoate (15).

To a vial charged with a stir bar were added K₂OsO₂(OH)₄ (2.2 mg, 0.006 mmol, 0.01 equiv), K₃Fe(CN)₆ (543 mg, 1.65 mmol, 3 equiv), K CO (228 mg, 1.65 mmol, 3 equiv), MeSO NH (52.3 mg, 0.55 mmol, 1 equiv), and (DHQD)₂PHAL (21.4 mg, 0.0275 mmol, 0.05 equiv). To the mixture were added ^tBuOH and H₂O (1:1 v/v, 11 mL) and it was stirred vigorously at room temperature for 30 min, at which time the biphasic solution was poured into a precooled (0 °C) flask containing olefin 14 (259 mg, 0.55 mmol, 1 equiv). After stirring at this temperature for 3 h, the reaction mixture was moved to a cold room and held at 4 °C and stirred vigorously for 4 h. The mixture was then partitioned between EtOAc and water. The aqueous layer was extracted with EtOAc (2×) and the combined organics were dried over MgSO₄, filtered, and concentrated to a crude oil, which was carried forward without further purification. A flame-dried vial was charged with crude diol (assume 0.55 mmol, 1 equiv) and a stir bar. Anhydrous DMF (2.7 mL) was added, followed by imidazole (93.6 mg, 1.38 mmol, 2.5 equiv) in a single portion. The reaction flask was cooled to 0 °C and TBSCl (124 mg, 0.825 mmol, 1.5 equiv) was added in a single portion. After stirring at this temperature for 30 min, the reaction flask was removed from the ice bath and stirred at room temperature for 1.5 h. Another 0.25 equiv of TBSCl was added in a single portion, and the reaction was stirred for an additional 1 h, at which time it was partitioned between EtOAc and water. The aqueous layer was extracted with EtOAc (2×) and the combined organic layers were dried over MgSO₄, filtered, concentrated, and the residue by flash chromatography (20% Et₂O/pentane) to provide silyl ether 15 (219 mg, 71% over 2 steps, 14:1 dr). Compound purity was established by TLC (one spot) analysis. R_f (30% EtOAc/hexanes): 0.7. ¹H NMR (500 MHz, CDCl₃): δ 6.22 (dd, J = 10.8, and 17.7 Hz, 1H), 5.88 (td, J = 0.9, and 1.9 Hz, 1H), 5.36 (s, 1H), 4.83 (dd, J = 1.4, and 17.7 Hz, 1H), 4.79 (dd, J = 1.4, and 10.8 Hz), 4.10 (tdd, J = 2.7, 9.8, and 12.0 Hz, 1H), 3.74–3.64 (m, 2H), 3.67 (s, 3H), 3.50 (dd, J = 2.1, and 14.9 Hz, 1H), 3.36 (s, 3H), 2.32–2.19 (m, 4H), 2.28–1.54 (m, 4H), 1.31–1.22 (m, 11H), 1.20 (d, J = 6.1 Hz, 3H), 1.11 (s, 3H), 1.08 (s, 3H), 0.90 (s, 9H), 0.86 (dt, J = 2.7, and 6.8 Hz, 3H), 0.10 (s, 3H), and 0.09 (s, 3H). ¹³C{¹H} NMR (125 MHz, CDCl₃): δ 172.0, 166.7, 152.7, 146.3, 117.6, 108.8, 102.7, 72.2, 72.1, 71.8, 68.4, 51.7, 51.2, 46.7, 40.5, 34.6, 34.2, 32.2, 31.8, 29.1, 29.0, 26.0, 24.6, 24.3, 22.9, 22.7, 22.5, 20.4, 18.2, 14.2, −4.1, and −4.7. IR (thin film): 3054, 2929, 2954, 2873, 2857, 1740, 1722, 1668, 1461, 1436, 1414, 1378, 1361, 1259, 1220, 1157, 1106, 1080, 1056, 982, 942, 904, 880, 823, and 799 cm⁻¹. HRMS (ESI+, m/z): [M + Na]⁺ calcd for C₃₂H₅₈O₈SiNa, 621.3793; found, 621.3774. [α]_D²² −1.6° (c = 0.475, CH₂Cl₂).

(2S,3S,6S,E)-6-((2R,3R)-3-((tert-Butyldimethylsilyl)oxy)-2-hydroxybutyl)-2-methoxy-4-(2-methoxy-2-oxoethylidene)-2-(2-methyl-1-oxopropan-2-yl)tetrahydro-2H-pyran-3-yl Octanoate (18).

A solution of olefin 15 (120 mg, 0.2 mmol, 1 equiv) in MeOH (2 mL) was cooled to −78 °C. A preformed saturated ozone solution (0.025 M, maintained at −78 °C) was added portionwise (19.5 mL, 0.48 mmol, 2.4 equiv) until the reaction was complete as monitored by TLC. Triphenylphosphine was added in a single portion (105 mg, 0.4 mmol, 2 equiv) and the reaction was removed from its ice bath. After 45 min of stirring, the reaction mixture was concentrated and directly loaded onto a silica column. Flash chromatography (35% Et₂O/pentane) provided aldehyde 18 as a foaming oil (91 mg, 81% yield). Compound purity was established by TLC (one spot) analysis. R_f (20% EtOAc/hexanes): 0.45. ¹H NMR (500 MHz, CDCl₃): δ 9.70 (s, 1H), 5.93 (d, J = 1.7 Hz), 5.19 (s, 1H), 4.14 (tdd, J = 2.8, 9.8, and 11.9 Hz, 1H), 3.38–3.63 (m, 2H), 3.68 (s, 3H), 3.60 (dd, J = 2.4, and 14.8 Hz, 1H), 3.47 (s, 3H), 2.37–2.29 (m, 1H), 2.18 (t, J = 7.5, 2H), 2.16 (m, 1H), 1.67 (m, 3H), 1.56–1.50 (m, 2H), 1.29–1.24 (m, 10H), 1.21 (d, J = 6.1 Hz, 3H), 1.16 (s, 3H), 1.00 (s, 3H), 0.90 (s, 9H), 0.86 (t, J = 7.1 Hz, 3H), 0.11 (s, 3H), and 0.10 (s, 3H). ¹³C{¹H} NMR (125 MHz, CDCl₃): δ 202.7, 171.8, 166.5, 150.5, 119.6, 102.2, 72.2, 71.8, 71.7, 68.8, 54.1, 51.7, 51.4, 40.6, 34.0, 31.8, 31.8, 29.0, 26.0, 24.5, 22.7, 20.5, 19.3, 18.2, 16.6, 14.2, −4.0, and −4.7. IR (thin film): 3555, 2954, 2930, 2857, 2723, 1751, 1723, 1667, 1463, 1437, 1393, 1378, 1362, 1258, 1211, 1157, 1108, 1083, 1007, 948, 882, 838, 810, 777, 733, and 667 cm⁻¹. HRMS (ESI+, m/z): [M + Na]⁺ calcd for C₃₁H₅₆O₉SiNa, 623.3586; found, 623.3568. [α]_D²² 6.4° (c = 0.04, CH₂Cl₂).

5-(2-((2S,3S,6S,E)-6-((2R,3R)-3-((tert-Butyldimethylsilyl)oxy)-2-hydroxybutyl)-2-methoxy-4-(2-methoxy-2-oxoethylidene)-3-(octanoyloxy)tetrahydro-2H-pyran-2-yl)-2-methylpropoxy)-5-oxopentanoic Acid (20).

To a solution of aldehyde 18 (49.8 mg, 0.083 mmol, 1 equiv) in dry THF (0.5 mL) was added a ^tBuNH₂·BH₃ complex (36.2 mg, 0.415 mmol). The reaction was stirred at room temperature for 1 h upon which time TLC indicated complete conversion. It was then partitioned between 5% HCl (3 mL) and Et₂O (3 mL). The aqueous layer was washed with Et₂O (5 mL × 4 times). The combined organic layers were dried over Na₂SO₄, filtered, and concentrated to provide crude neopentyl alcohol 17, which was carried forward without further purification because of instability upon standing. To a solution of crude alcohol 17 (assume 0.083 mmol, 1 equiv) in DCM (1 mL) were added glutaric anhydride (28.4 mg, 0.249 mmol, 3 equiv) and DMAP (45.6 mg, 0.373 mmol, 4.5 equiv) in DCM (1 mL). After stirring for 2.5 h, another 1.5 equiv (15.2 mg) of DMAP was added. After another 30 min, the reaction was partitioned between 5% HCl (5 mL) and Et₂O (5 mL). The organic layer was separated, and the aqueous layer was washed with EtOAc (5 mL × 5 times). The combined organic layers were dried over Na₂SO₄, filtered, and concentrated. Purification by flash chromatography with the gradient 15–45% v/v EtOAc/pentane provided seco-acid 20 as a pale yellow oil (38.5 mg, 65% yield over two steps). Compound purity was established by TLC (one spot) analysis. R_f (60% EtOAc/hexanes): 0.65. ¹H NMR (600 MHz, CDCl₃): δ 5.88 (s, 1H), 5.67 (s, 1H), 4.27 (m, 1H), 4.19 (d, J = 10.9 Hz, 1H), 4.11 (d, J = 11.1 Hz, 1H), 3.69 (s, 3H), 3.69–3.62 (m, 2H), 3.40–3.38 (app d, J = 12.3 Hz, 1H), 3.38 (s, 3H), 2.75 (t, J = 6.7 Hz, 3H), 2.45–2.38 (m, 8H), 2.02 (app pentet, J = 6.8 Hz, 1H), 1.98–1.93 (m, 2H), 1.67–1.58 (m, 4H), 1.34–1.23 (m, 8H), 1.19 (d, J = 6.00 Hz, 3H), 1.04 (s, 3H), 0.98 (s, 3H), 0.91 (s, 9H), 0.87 (t, J = 7.1, 3H), 0.10 (s, 3H), and 0.09 (s, 3H). ¹³C{¹H} NMR (125 MHz, CDCl₃): δ 178.0, 173.1, 172.1, 166.6, 153.0, 116.3, 102.0, 72.3, 71.7, 69.6, 68.3, 51.3, 51.0, 45.5, 40.6, 34.5, 33.4, 33.1, 31.8, 30.1, 29.1, 29.0, 26.0, 24.9, 22.7, 21.2, 20.7, 20.4, 20.1, 18.2, 16.5, 14.2, −4.1, and −4.7. IR (thin film): 2951, 2928, 2857, 1715, 1462, 1435, 1392, 1376, 1221, 1150, 1103, 1058, 941, 882, 835, and 776 cm⁻¹. HRMS (ESI+, m/z) [M + Na]+ calcd for C₃₆H₆₄O₁₂SiNa, 739.4059, found, 739.4058. [α]_D²² −6.5° (c = 1.33, CH₂Cl₂).

(1S,3R,13S,14S,E)-3-((R)-1-((tert-Butyldimethylsilyl)oxy)ethyl)-13-methoxy-15-(2-methoxy-2-oxoethylidene)-12,12-dimethyl-5,9-dioxo-4,10,17-trioxabicyclo[11.3.1]heptadecan-14-yl Octanoate (21).

To a two-neck 100-mL round-bottom flask charged with a stir bar was added 20 mL of toluene, followed by the addition of 2-methyl-6-nitrobenzoic anhydride (69.0 mg, 1.5 equiv) and DMAP (97.9 mg, 6 equiv). The reaction mixture was placed in an oil bath that was heated to 50 °C. Seco-acid 20 (95.8 mg, 0.043 mmol, 1 equiv) was added as a solution in PhCH₃ (46 mL) dropwise over 30 min. The final concentration was 0.002 M. The resulting reaction mixture was stirred at 50 °C for 16 h, at which point it was cooled to room temperature and partitioned between Et₂O (50 mL) and saturated aqueous NH₄Cl (100 mL). The organics were dried over Na₂SO₄, filtered, and concentrated. Purification by flash chromatography (15% EtOAc/pentane) provided protected macrocycle 21 as a colorless residue (12 mg, 29% yield over two steps, 41% BRSM). Compound purity was established by TLC (one spot) analysis. R_f (30% EtOAc/hexanes): 0.8. ¹H NMR (500 MHz, CDCl₃): δ 5.90 (s, 1H), 5.53 (s, 1H), 4.92 (dt, J = 3.6, and 7.6 Hz, 1H), 4.25 (d, J = 11.7 Hz, 1H), 4.08 (m, 1H), 4.01 (qd, J = 6.2, and 7.6 Hz, 1H), 3.70 (s, 3H), 3.65 (d, J = 12.1 Hz, 1H), 3.42 (s, 3H), 3.37 (app d, J = 15.8 Hz, 1H), 2.48–2.29 (m, 9H), 2.17 (dd, J = 4.0, and 7.4 Hz, 1H), 2.13–2.04 (m, 1H), 1.94–1.83 (m, 2H), 1.63 (m, 2H), 1.29–1.25 (m, 10H), 1.12 (d, J = 6.2, 3H), 1.05 (s, 3H), 0.88 (t, J = 7.2 Hz, 3H), 0.85 (s, 3H), 0.04 (s, 3H), and 0.00 (s, 3H). 13C{¹H} NMR (126 MHz, CDCl₃): δ 173.9, 172.5, 172.2, 166.5, 152.3, 116.9, 103.6, 76.0, 71.6, 68.3, 68.1, 51.3, 51.1, 44.1, 35.7, 34.6, 33.0, 32.3, 31.8, 29.8, 29.1, 29.0, 25.9, 25.8, 24.8, 22.7, 22.5, 22.1, 20.9, 20.1, 18.0, 14.2, −4.5, and −4.6. IR (thin film): 2952, 2929, 2857, 1721, 1667, 1463, 1436, 1379, 1244, 1193, 1146, 1099, 1068, 1033, 963, 884, and 834 cm⁻¹. HRMS (ESI+, m/z): [M + Na]+ calcd for C₃₆H₆₂O₁₁SiNa, 721.3954; found, 721.3955. [α]_D²² −26.74° (c = 0.500, CH₂Cl₂).

(1S,3R,13S,14S,E)-13-Hydroxy-3-((R)-1-hydroxyethyl)-15-(2-methoxy-2-oxoethylidene)-12,12-dimethyl-5,9-dioxo-4,10,17-trioxabicyclo[11.3.1]heptadecan-14-yl Octanoate (7b).

Protected macrocycle 21 (15.0 mg, 0.022 mmol, 1 equiv) was suspended in 4:1 v/v MeCN/H₂O (1.0 mL, final concentration: 0.02 M). Racemic CSA (49.8 mg, 0.22 mmol, 10 equiv) was added in a single portion and the solution rapidly went from cloudy to clear. Upon an hour of stirring, TLC indicated minimal starting material left, so the reaction mixture was partitioned between EtOAc (5 mL) and saturated aqueous NaHCO₃ (5 mL). The organic layer was separated, and the aqueous layer was washed with EtOAc (5 mL*5 times). The combined organic layers were dried over Na₂SO₄, filtered, and concentrated. Purification by flash chromatography (30–50% EtOAc/pentane) provided macrocycle 7a as a white solid (8.5 mg, 65% yield). Compound purity was established by TLC (one spot) analysis. R_f (50% EtOAc/hexanes): 0.2. ¹H NMR (CDCl₃, 600 MHz): δ = 6.00 (s, 1H), 5.28 (s, 1H), 5.15 (td, J = 3.7, and 11.7 Hz, 1H), 4.64 (s, 1H), 4.25 (app t, J = 11.4 Hz, 1H), 4.03 (d, J = 10.9 Hz, 1H), 3.92 (d, J = 10.8 Hz, 1H), 3.75–3.64 (m, 2H), 3.70 (s, 3H), 2.67 (dd, J = 6.8, and 17.8 Hz, 1H), 2.54 (dd, J = 2.2, and 13.3 Hz, 1H), 2.49–2.24 (m, 6H), 2.14–2.05 (m, 2H), 1.98 (app t, J = 12.5 Hz, 1H), 1.81 (m, 2H), 1.72–1.52 (m, 3H), 1.38–1.13 (m, 8H), 1.19 (d, J = 6.4 Hz, 3H), 1.11 (s, 3H), 0.97 (s, 3H), and 0.87 (m, 3H). ¹³C{¹H} NMR (CDCl₃, 125 MHz): δ 173.4, 172.2, 171.6, 166.7, 151.5, 119.6, 100.3, 77.4, 73.8, 72.9, 72.7, 69.0, 51.4, 40.9, 35.9, 35.5, 34.8, 33.2, 31.8, 31.1, 29.8, 29.2, 29.0, 24.7, 22.7, 21.9, 21.3, 20.6, and 14.0. IR (thin film): 3520, 2932, 2859, 1742, 1721, 1667, 1471, 1439, 1381, 1440, 1275, 1231, 1177, 1135, 1109, 1085, 1050, 1004, 974, 938, 885, 848, 814 cm⁻¹. HRMS (ESI+, m/z): [M + Na]⁺ calcd for C₂₉H₄₆O₁₁SiNa, 593.2938; found, 593.2932. [α]_D²² −13.44° (c = 0.100, CH₂Cl₂).

PKC Binding Assay.

The PKC affinity of bryostatin 1 and bryostatin analogues was studied via competition with ³H-phorbol-12,13-dibutyrate (³H-PDBu) as described below. This procedure entails a glass-fiber filtration method to determine the bound radioligand. First, to a 50 mL polypropylene tube were added Tris–HCl (pH 7.4, 1 M, 1 mL), KCl (1 M, 2 mL), CaCl₂ (0.1 M, 30 μL), and bovine serum albumin (40 mg, Sigma-Aldrich). This mixture was diluted to 20 mL with deionized H₂O and mixed gently. The buffer was stored on ice until use. For every two assays, 3.5 mg of phosphatidylserine (PS)(Avanti Polar Lipids, porcine, 25 mg/mL CHCl₃ solution) was concentrated by removing chloroform under a stream of nitrogen followed by reduced pressure. The solid PS was suspended as vesicles in freshly prepared PKC binding assay buffer (3.5 mL) by sonicating six times for 30 s, with a 30 s rest between sonications (Branson Sonifier 250, power = 2, 50% duty cycle). The resulting milky cloudy mixture (1 mg/mL) was stored on ice until use. Assay PKC was prepared by dissolving a 4 μg aliquot of the indicated recombinant human PKC isoform (Invitrogen) into 11.6 mL of PKC binding assay buffer (this amount is sufficient for two assays). The diluted PKC was stored on ice for immediate use. To prepare the solution of the radioligand, ³H-PDBu (American Radiolabeled Chemicals, Inc.; 1 mCi/mL acetone solution; specific activity: 20 Ci/mmol) was diluted 10-fold with DMSO. The resulting 500 nM stock solution was further diluted with DMSO to 30 nM. Compound dilutions were prepared by serially diluting from a chosen “high” concentration by factors of 3 or 4. For each analogue compound, seven concentrations were used to define the inhibition curve (i.e., for all compounds, the analogue concentrations used were 3000, 750, 188, 46.9, 11.7, 2.93, and 0.73 nM). To prepare the master mix solution, to a polypropylene tube were added 3.3 mL of 1 mg/mL PS vesicle solution, 11 mL of PKC isoform solution, and 1.1 mL of 30 nM ³H-PDBu solution. The resulting solution was vortexed to mix and stored on ice. Prior to performing the assay, glass-fiber filters (Whatman GF/B) were prepared by soaking them in a solution of aqueous polyethyleneimine (10% by vol, 18 mL) in deionized water (600 mL) for ≥1 h. Additionally, 500 mL of “rinsing buffer” of 20 mM Tris, pH 7.4, was cooled on ice for the duration of the incubation period and for the remainder of the assay. Triplicate data points were obtained for each analogue concentration. For each data point, 280 μL of the “Master Mix” solution and 20 μL of the analogue compound at a specified concentration were added to a polypropylene tube. Nonspecific ³H-PDBu binding was assessed in triplicate by substitution of the analogue compound with unlabeled PDBu (20 μL of a 75 μM stock, assay concentration: 5 μM). Maximal ³H-PDBu binding was assessed in triplicate by substitution of the analogue compound with 20 μL of DMSO. The solutions were vortexed to mix, incubated at 37 °C for 10 min, and incubated on ice for at least 30 min prior to filtration. Using a Brandel Harvester, the assay contents from each polypropylene tube were vacuum-filtered through polyethylenimine-soaked filters, washed with a rinsing buffer (3×), and dried first under vacuum for 5 min and then under ambient conditions for ≥2 h. The resulting filters had circular perforations for each data point, which were removed with forceps and placed in a scintillation vial. Scintillation vials were filled with Bio-Safe II scintillation fluid (5 mL) and measured for radioactivity using a Beckman LS 6000SC scintillation counter. Counts per minute (cpm) were averaged for each triplicate dilution. The data were plotted—cpm versus log(concentration)—using Prism by GraphPad Software and the IC₅₀ was determined using that program’s built-in one-site competition least squares regression function. $K_{\mathrm{i}}$ values were calculated using the equation: $K_{\mathrm{i}}={\mathrm{I}\mathrm{C}}_{50}/\left(1+\left(\left[{}^3\mathrm{H}-\mathrm{P}\mathrm{D}\mathrm{B}\mathrm{u}\right]/K_{\mathrm{d}}\right)\right)$. The $K_{\mathrm{d}}$ of ³H-PDBu for PKC isoforms was measured separately via saturation binding experiments under identical conditions ($\alpha =15.1\mathrm{nM}$, $\beta I=8.8\mathrm{nM}$, $\gamma =13.8\mathrm{nM}$, $\delta =4.5\mathrm{nM}$, $\epsilon =6.2\mathrm{nM}$, $\eta =18.4\mathrm{nM}$, $\theta =28.8\mathrm{nM}$).

PKCδ-GFP Translocation Protocols.

CHO-k1 cells (ATCC) were cultured in F-12 Kaighn’s media (HyClone, with 10% fetal bovine serum and 1% penicillin/streptomycin added, referred to as F-12 +/+) at 37 °C (5% CO₂). Cells were plated at 600,000 cells per well in a six-well plate for a final volume of 2.5 mL of F-12 +/+ media. The cells were then incubated for 24 h at 37 °C (5% CO₂) before transfection with lipofectamine 2000 reagent (Thermo Fisher) or CART DA 13:11.⁴¹ Before the transfections, the F-12 +/+ media was aspirated, and the cells were washed with F-12 −/− media. Fresh F-12 −/− media was then added to each of the wells, for a final volume of 2.5 mL of F-12 −/− media per well (2 mL for the Lipofectamine 2000 condition, 2.4 mL for the CART DA 13:11 conditions). Briefly, for each well of Lipofectamine 2000-treated CHO-k1 cells, 12.5 μL of Lipofectamine 2000 reagent (Thermo Fisher) was added to 250 μL of Opti-MEM media (Thermo Fisher) in a polypropylene tube and incubated for 20 min at r.t. Simultaneously, for each well, 4 μg of PKCδ-GFP pDNA and 250 μL of Opti-MEM media were added to a separate polypropylene tube. Lipofectamine 2000 suspension (250 μL) was added to the DNA suspension, and the solution was incubated for 30 min at room temperature. Then, 500 μL of the lipofectamine/DNA suspension was added to the respective wells of the six-well plate. For each well of CART DA 13:11-treated CHO-K1 cells, 4 μg of PKCδ-GFP pDNA was added to PBS (pH 5.5, final volume 100 μL). CART DA 13:11 (5.6 μL of 2 mM stock in DMSO) was then added to the DNA solution to achieve the final cation/anion ratio of 10:11. The solution was gently mixed (by flicking) for 20 s, at which point it was added directly to the respective wells of the six-well plate. After treatment with lipofectamine 2000 or CART DA 13:11, the cells were incubated at 37 °C (5% CO₂) for ~24 h. The media was then aspirated, and cells were washed with PBS (2.0 mL) and trypsinized (500 μL). The cell suspension was then diluted with 2.0 mL of F-12 +/+. 200 μL aliquots were added to three wells of a Lab-Tek II four-well-chambered coverglass slide (Fisher), producing four slides in total, each with three wells of cells. The cell suspension was directly diluted with 600 μL of additional F-12 +/+. The resulting samples were incubated for ~24 h prior to imaging. Fluorescent images were obtained using a Leica SP8 White Light Confocal microscope and the Leica AF software package. Prior to analysis, the media was aspirated, and 800 μL of PBS (HyClone, without Ca²⁺ or Mg²⁺) supplemented with glucose (10 mM) was added to each well of the chambered coverglass slide. Bryostatin and bryostatin analogues were diluted to the appropriate concentration in 200 μL of 10 mM glucose in PBS. Cells were located for imaging and data were recorded for three wells in parallel, imaging at predetermined positions in each well using adaptive focus control. Cells were imaged at 30 s intervals following the addition of the compound (set to t = 0) for 20–40 min. Data were recorded at room temperature. Images were exported as .lif files and fluorescence intensity was analyzed using FIJI (NIH) software. To monitor the translocation, small cytosolic regions of interest were selected in each cell, and fluorescence intensity values were plotted versus time following background subtraction and normalization. Graphed data represent the average of at least three replicates.

Supplementary Material

ASSOCIATED CONTENT

Supporting Information

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acs.joc.0c01988.

Condition screening for the reduction of aldehyde 18, summary of biological data, computational methodologies and data, as well as ¹H and ¹³C NMR spectra of new compounds (PDF)