Rifamycin Biosynthetic Congeners: Isolation and Total Synthesis of Rifsaliniketal and Total Synthesis of Salinisporamycin and Saliniketals A and B

Article

pubs.acs.org/JACS

Rifamycin Biosynthetic Congeners: Isolation and Total Synthesis of

Rifsaliniketal and Total Synthesis of Salinisporamycin and

Saliniketals A and B

Yu Feng,^†Jun Liu,^†Yazmin P. Carrasco,^†John B. MacMillan,*^,†and Jef K. De Brabander*^,†,‡

‡

^†Department of Biochemistry and Harold C. Simmons Comprehensive Cancer Center, The University of Texas Southwestern

Medical Center at Dallas, 5323 Harry Hines Boulevard, Dallas, Texas 75390-9038, United States

S

* Supporting Information

ABSTRACT: We describe the isolation, structure elucidation, and

total synthesis of the novel marine natural product rifsaliniketal and

the total synthesis of the structurally related variants salinispor-

amycin and saliniketals A and B. Rifsaliniketal was previously

proposed, but not observed, as a diverted metabolite from a

biosynthetic precursor to rifamycin S. Decarboxylation of rifamycin

provides salinisporamycin, which upon truncation with loss of the

naphthoquinone ring leads to saliniketals. Our synthetic strategy

hinged upon a Pt(II)-catalyzed cycloisomerization of an alkynediol

to set the dioxabicyclo[3.2.1]octane ring system and a fragmenta-

tion of an intermediate dihydropyranone to forge a stereochemically

deﬁned (E,Z)-dienamide unit. Multiple routes were explored to

assemble fragments with high stereocontrol, an exercise that

provided additional insights into acyclic stereocontrol during

stereochemically complex fragment-assembly processes. The resulting 11−14 step synthesis of saliniketals then enabled us to

explore strategies for the synthesis and coupling of highly substituted naphthoquinones or the corresponding naphthalene

fragments. Whereas direct coupling with naphthoquinone fragments proved unsuccessful, both amidation and C−N bond

formation tactics with the more electron-rich naphthalene congeners provided an eﬃcient means to complete the ﬁrst total

synthesis of rifsaliniketal and salinisporamycin.

Chart 1. Structures of Saliniketals A (1a) and B (1b),

Salinisporamycin (2), and Rifamycin S (3)

INTRODUCTION

■

Since the discovery of the antimalarial activity of quinine,

isolated from the bark of the cinchona tree, natural products

have played a fundamental role as the starting point for the

discovery of drugs to combat infection, inﬂammation, and

cancer in humans.¹Over the past 10 years or so, marine-derived

microorganisms, more speciﬁcally actinomycetes, have provided

a new source of natural products. The genus Salinispora in

particular has been a proliﬁc producer of secondary metabolites.

Currently, three species that belong to this genus have been

identiﬁed: Salinispora tropica, Salinispora arenicola, and

Salinispora paciﬁca.²The potent proteasome inhibitor salino-

sporamide A represents the ﬁrst natural product isolated from

the obligate marine bacteria S. tropica,³which is currently in

clinical trials for the treatment of cancer. In 2007 the genome of

S. tropica was sequenced to provide insight into its biosynthetic

potential. It was shown that a large percentage of its genome is

devoted to the assembly of natural products,⁴suggesting that

there is still a large chemical diversity to explore within this

genus. In this article, we document our studies of novel

metabolites with a common rifamycin biosynthetic origin

isolated from the marine actinomycete S. arenicola.

were isolated on the basis of their ability to prevent phorbol

ester-mediated induction of ornithine decarboxylase (ODC),

In 2007, Fenical and co-workers isolated saliniketals A (1a)

Received: March 29, 2016

and B (1b) (Chart 1) from S. arenicola.⁵These metabolites

A

DOI: 10.1021/jacs.6b03248

J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society

Article

a

Scheme 1. Biosynthetic Relationship between Rifamycins, Rifsaliniketal, Salinisporamycin, and Saliniketals A and B

a

Adapted from ref 14.

which is an enzyme responsible for polyamine biosynthesis.⁶

Thus, by the use of T24 cells stimulated with the potent tumor

promotor 12-O-tetradecanoylphorbol-13-acetate (TPA), salini-

ketals A and B were shown to inhibit TPA-induced induction of

ODC with IC₅₀values of 1.95 and 7.83 μg/mL, respectively.

Amino acid-derived polyamines are important for developmen-

tal processes in mice.⁷In addition, ODC is a direct target of the

MYC oncogene, and it has been shown to be overexpressed in

diﬀerent tumor cells.⁸Because of their inhibitory properties

against ODC induction, saliniketals may be useful as chemo-

preventive agents.^6,9To date, one formal and two total

syntheses of saliniketals have been disclosed, including

ours.^10,11Later, in 2009, the Matsuda group isolated a

metabolite from a culture of the same species from which

saliniketals were isolated (S. arenicola) on the basis of its ability

to inhibit the growth of A549 human lung adenocarcinoma cells

(IC₅₀value of 3 μg/mL).¹²Moreover, this metabolite was

found to have signiﬁcant activity against the bacterial strains

Staphylococcus aureus IFO 12732 and Bacillus subtilis IFO 3134

with MIC values of 0.46 and 4.1 μg/mL, respectively. This

compound, which they named salinisporamycin (2), is in

essence the dihydroxynaphthoquinone amide of saliniketal A,

which inter alia co-occurred in the fermentation broth. In spite

of the unique 1,4-dimethyl-2,8-dioxabicyclo[3.2.1]octane ring

system, salinisporamycin and saliniketals structurally appear as

acyclic truncated variants of the rifamycin ansa-macrolide family

of compounds, suggesting that they might share common

biosynthetic origins. Indeed, rifamycin S (3), a metabolite

previously observed exclusively in terrestrial soil actinobacteria

(Amycolatopsis mediterranei), was coisolated with salinispor-

amycin.¹²Previously, around 2006, the groups of Fenical in

California and Fuerst in Australia independently demonstrated

the widespread production of rifamycins in marine Salinispora

species.¹³

Finally, elegant studies documented by Moore and co-

workers provided the ﬁrst solid evidence that saliniketals and

salinisporamycin are diverted products of rifamycin biosyn-

thesis.¹⁴Using PCR-directed mutagenesis, chemical comple-

mentation studies, and isotope feeding experiments, they

formulated a biosynthetic pathway wherein saliniketal produc-

tion diverges from rifamycin production at the stage of 34a-

deoxyrifamycin W (4) (Scheme 1). An S. arenicola-speciﬁc

cytochrome P450 (CYP) encoded by the biosynthetic gene

sare1259 doubly oxidizes C34a to form carboxylic acid 5,

temporally regulated decarboxylation of which speciﬁes

saliniketal or rifamycin synthesis. In accordance with the

proposed biosynthesis of terrestrial rifamycin, C12/C29

double-bond cleavage in intermediate 5 to form ketone

aldehyde 6 prior to decarboxylation (by a CYP encoded by

sare1259 in marine S. arenicola or rif-orf5 in terrestrial A.

mediterranei) leads to rifamycin via a decarboxylative ketaliza-

tion. When decarboxylation occurs prior to oxidative double-

bond cleavage, the resulting C11/C12 enol 7 is then

subsequently cleaved to form enone 8 upon action of the

sare1259 gene product. Subsequent C28/C29 oleﬁn reduction

followed by formation of the dioxabicyclo[3.2.1]octane ring

system then provides putative intermediate 9 (which we shall

term rifsaliniketal) that upon decarboxylation yields salinispor-

amycin (2), the metabolite isolated from S. arenicola by

Matsuda and co-workers. Moore and co-workers were not able

B

DOI: 10.1021/jacs.6b03248

J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society

Article

to identify a rif pathway gene in S. arenicola associated with the

putative conversion of 2 to saliniketal A (1a), suggesting that an

enzyme outside the rif gene cluster might be responsible.

Alternatively, one can envision a nonenzymatic naphthoqui-

none hydration (at C2 to yield 10) followed by amide

(saliniketal) elimination.

The above demonstrates the biosynthetic potential of marine

microorganisms to enzymatically tailor rifamycin biosynthetic

pathway intermediates for the generation of marine-speciﬁc

metabolites.¹⁴In view of their biological activity, biosynthetic

relationship to rifamycin, and novel structural features, we

initiated a synthetic program directed toward saliniketals and

salinisporamycin. In addition to our initial disclosure, only one

total synthesis and one formal synthesis of saliniketals have

been published.^10,11No total synthesis of salinisporamycin has

been described to date. Herein we describe in detail our eﬀorts

leading to the total synthesis of saliniketals A and B and the ﬁrst

total synthesis of salinisporamycin. In addition, we disclose for

the ﬁrst time the isolation (from a novel strain of S. arenicola),

full characterization, and total synthesis of rifsaliniketal. Our

studies thus lend credence to the intermediacy of this

previously proposed but heretofore uncharacterized biosyn-

thetic intermediate en route to saliniketals and salinispor-

amycin.¹⁴

Figure 1. Selected COSY and HMBC assignments for the structural

elucidation of rifsaliniketal (9).

H25 and H34, and H25/H26 provide the assignment of the

bicyclic ketal found in 9. In addition, by means of a powerful

technique that allows homonuclear decoupling of multiple

neighboring protons simultaneously, we were able to obtain

coupling constants of H18.¹⁵The proton on C18 is coupled to

three distinct sets of protons, giving rise to a doublet of

doublets of quartets. Application of multiple homonuclear

decoupling to H17 and H30 caused H18 to collapse to a

doublet with coupling to H19 (9.1 Hz). Additional decoupling

of H19 and H30 collapsed H18 to a doublet coupled to H17

(8.0 Hz). Finally, decoupling of H17 and H19 collapsed H18 to

a quartet coupled to H30 (6.9 Hz).

A challenge in the determination of the structure of 9 was the

lack of protons in the naphthoquinone ring, making it diﬃcult

to conﬁdently assess the presence of a carboxylic acid. To

circumvent this problem, we converted the carboxylic acid into

its methyl ester derivative 9a using TMSCHN₂. The presence

of a methyl peak in the proton NMR spectrum (3.87 ppm) and

a mass indicative of the methyl ester derivative ([M + Na]⁺, m/

z 678.3) together with an HMBC correlation of the resulting

methyl ester to the carbonyl carbon at C11 (δ_C172.6) validated

our suspicion that in fact there was a carboxylic acid present.

The remainder of the naphthoquinone core of 9 was assigned

by a combination of ¹³C chemical shift assignments and HMBC

correlations from H3 (δ_H7.66) to C2 (δ_C141.1), C4 (δ_C

183.8), and C10 (δ_C128.6) and from H12 (δ_H2.16) to C6 (δ_C

162.6) and C5 (δ_C117.5).

RESULTS AND DISCUSSION

■

Isolation and Characterization of Rifsaliniketal, the

Proposed Biosynthetic Precursor to Salinisporamycin

and Saliniketals. During our evaluation of natural product

fractions for cytotoxicity toward a panel of cancer cell lines, we

identiﬁed a fraction from S. arenicola with potent cytotoxicity

toward the human glioblastoma cell line T98G. To follow up

on the activity, a 20 L scale fermentation of SNB-003 was

carried out, and the secondary metabolites were extracted using

XAD-7 resin. Following solvent/solvent partitioning and

reversed-phase column chromatography, the fraction with

activity against T98G was further puriﬁed by C18 HPLC to

yield an active compound and an inactive compound. The

active compound was determined to be staurosporine, whereas

the inactive molecule displayed NMR signals similar to those of

salinisporamycin. Further characterization led to the assignment

of this compound as rifsaliniketal (9), validating the Moore

biosynthetic proposal (Scheme 1).

Overall, the NMR data for 9 were nearly identical to those of

2.¹⁶All of the ansa-chain H chemical shifts and coupling

1

Rifsaliniketal was isolated as a yellow powder with a

molecular formula established as C₃₄H₄₃NO₁₁based on the

molecular ion peak at 642.2901 [M + H]⁺, with 14 degrees of

unsaturation calculated. On the basis of correlation spectros-

copy (COSY) correlations, we were able to build the polyketide

side chain of 9 and demonstrate that it is identical to that found

in 1a (Figure 1). Brieﬂy, the C13−C18 fragment was

established by COSY correlations of H15/H16, H16/H17,

and H17/H18 along with key heteronuclear multiple bond

correlation (HMBC) correlations from CH₃-29 (δ_H2.08) to

C13 (δ_C169.7), C14 (δ_C129.3), and C15 (δ_C138.0). The

C16−C17 double bond was assigned the E conﬁguration on the

basis of the H16/H17 proton−proton coupling constant of

15.2 Hz. COSY correlations of H18/H19 and H30, H19/H20,

H20/H21 and H31, H21/H22, and H22/H23 and H32

established the C18−C23 fragment. HMBC correlations from

H27 (δ_H1.78−1.83) to C26 (δ_C24.6), from H25 (δ_H4.22) to

C28 (δ_C106.2), and from CH₃-33 (δ_H1.39) to a quaternary

carbon at δ_C106.2 indicative of a ketal functionality and to C27

(δ_C34.9) along with COSY correlations of H23/H24, H24/

constants were nearly identical, while the ¹³C chemical shifts

were within 1.0 ppm. Besides the additional C11 carbonyl

carbon in 9 that led to changes in the ¹³C chemical shifts of C4

and C10, the other data for the naphthoquinone rings of 9 and

2 were very similar.

Synthesis of Saliniketals: Synthetic Strategy, Explora-

tion of Aldol-Based Couplings, and Final Dihydropyr-

anone Fragmentation. Structurally, saliniketals A (1a) and B

(1b) are endowed with a unique 1,4-dimethyl-2,8-

dioxabicyclo[3.2.1]octane ring system connected to a 2-

substituted conjugated (E,Z)-dienoic carboxamide via a

polypropionate-derived C5-stereopentad. Saliniketal B diﬀers

from saliniketal A by a hydroxylation at the C2-Me substituent.

Whereas Paterson utilized a late-stage Stille cross-coupling

approach to install the dienoic carboxamide,^10awe opted to

explore a base-mediated fragmentation (E2 elimination) of a

dihydropyranone (see 12 in Scheme 2).¹⁷This approach

imparts a higher degree of convergence by providing

saliniketal’s entire carbon skeleton during a late-stage aldol

coupling of two enantiomerically pure C1−C7 and C8−C17

C

DOI: 10.1021/jacs.6b03248

J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society

Article

As shown in Scheme 3, triethylsilyl-protected aldehyde 13

was readily available from commercial pent-4-ynol (16) in four

Scheme 2. Synthetic Strategy and Stereochemical Analysis

Scheme 3. Synthesis of Aldehyde 13 and Aldol Reaction with

a

3-Pentanone

a

Reagents and conditions: (a) Dess−Martin periodinane, CH₂Cl₂, rt,

1 h, 94%; (b) (+)-Ipc₂B(OMe), ⁿBuLi, KO^tBu, cis-2-butene, Et₂O, −78

°C, 87%, >10:1 dr; (c) TESCl, 2,6-lutidine, CH₂Cl₂; (d) NaIO₄, OsO₄,

2,6-lutidine, CH₂Cl₂, 81% (two steps); (e) 3-pentanone (1 equiv),

TiCl₄(1.1 equiv), Bu₃N (1.2 equiv), CH₂Cl₂, −78 °C, then 13 (1.2

equiv), −78 °C, 2 h, 43−53%; or the same conditions with 3-

pentanone (2 equiv), TiCl₄(2.2 equiv), Bu₃N (2.4 equiv), 13 (1

equiv), 37−46%. Abbreviations: DMP, Dess−Martin periodinane; Ipc,

isopinocampheyl; TES, triethylsilyl.

fragments (14 and 15, respectively). Our tactic to obtain the

dioxabicyclooctane ring would rely on a Pt(II)-catalyzed

cycloisomerization of a dihydroxyalkyne as previously devel-

oped by our group.¹⁸Paterson, on the other hand, forged this

ring system using an eﬃcient intramolecular Wacker-type

cyclization of the corresponding dihydroxy-substituted terminal

oleﬁn.^10a

steps and excellent overall yields. Dess−Martin periodinane

(DMP)-mediated oxidation²²of alkynol 16 was followed by a

reagent-controlled crotylation with in situ-prepared homochiral

((Z)-2-butenyl)diisopinocampheylborane to yield homoallylic

alcohol 17.²³Following silylation and oxidative double-bond

cleavage, the target aldehyde 13 was obtained in excellent yield.

This aldehyde proved to be a frustrated partner for the ensuing

aldol reaction with 3-pentanone. Reaction with lithium or

boron Z(O)-enolates were sluggish and provided complex

mixtures from which no pure identiﬁable products were

isolated.²⁴The Z(O)-titanium enolate proved more reactive

and selective, providing a single aldol product in variable

isolated yields (37−53%) but favoring the undesired (and

unpredicted!) diastereomer 18-syn (dr > 15:1) and varying

amounts of desilylated cyclodehydrated compound 19.^25,26

Z(O)-titanium enolates of achiral ethyl ketones, including 3-

pentanone, have been documented to provide anti-Felkin

products with syn-α-Me,β-alkoxyaldehydes (cf. 13).^21a,bWe

therefore do not understand our results in the context of

current models for asymmetric induction.²⁷

In an attempt to override the inherent Felkin bias of

aldehyde 13, we explored its engagement in an auxiliary-

controlled Evans aldol reaction with the Z(O)-titanium enolate

derived from (S)-4-benzyl-3-propionyloxazolidin-2-one (20)

(Scheme 4).²⁸Under these conditions, a 73% yield of aldol

products was obtained with a dramatically improved albeit still

unsatisfactory 1.2:1 ratio of the separable isomers 21-syn and

21-anti. Removal of the silyl protecting group followed by

formation of the corresponding acetonides (23-syn and 23-anti)

enabled assignment of the indicated relative conﬁguration by

¹³C NMR analysis.²⁹

Placing the above-described tactics into the proper stereo-

chemical contextsaliniketals contain eight contiguous stereo-

centersis best achieved visually when focusing on structure

12 (Scheme 2). Within structure 12, one can identify a C5−

C13 stereononad with apparent C_ssymmetry (symmetry plane

through C9) as it relates to the nature and conﬁguration of the

branching methyl and oxygen substituents. Although the C5

stereocenter will be destroyed during our planned dihydropyr-

anone fragmentation, one should note our judicious choice of

conﬁguration at C5 in this context. This stereochemical

framework readily reveals a bidirectional aldol-based fragment

coupling between 3-pentanone and aldehydes 13 and 14.¹⁹

According to currently accepted stereochemical models for

aldol reactions of Z(O)-enolates via cyclic transition states, the

aldehyde α-methyl and β-alkoxy substituents impart maximal

reinforcing stereochemical bias when in a syn relationship by

minimizing both syn-pentane and dipole−dipole interactions

(see the box in Scheme 2).^20,21The thus-anticipated absolute

stereochemistry at the newly formed stereocenters (C7/C8 for

aldol with 14 or C10/C11 for aldol with 13) would be as

shown in structure 12. However, one cannot ignore an

additional layer of complexity resulting from enolate stereo-

chemical bias during the second of these two aldol reactions

(i.e., after a C7−C8 or C10−C11 bond has been formed during

the ﬁrst aldol reaction), a point to which we shall return later.

Suﬃce it for now that we ultimately settled on constructing the

C10−C11 bond ﬁrst (→15) in order to minimize protecting

group manipulations (i.e., installation of the dioxabicyclooctane

prior to the second aldol) and bringing in the potentially more

labile dihydropyranone aldehyde 14 last during the ﬁnal of

these two aldol reactions.

Unable to override the inherent facial bias of aldehyde 13, we

took a step back and explored two alternative aldol routes

starting from achiral pentynals and chiral enolates. As shown in

D

DOI: 10.1021/jacs.6b03248

J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society

Article

with high selectivity (>10:1 dr) in 83% yield. An anti-selective

sodium triacetoxyborohydride reduction delivered the desired

stereotetrad 27 in 80% yield.³⁰Alternatively, aldol reaction of

the stannyl enolate derived from known oxazolidinone 29 with

aldehyde 28 (obtained from DMP oxidation of the

corresponding commercially available alkynol in 92% yield)²²

provided aldol product 30 in 82% yield with >16:1 dr.³¹Anti-

selective reduction with Na(OAc)₃BH (>20:1 dr)³⁰followed by

desilylation aﬀorded stereotetrad 22-anti (74% yield over two

steps).³²

Scheme 4. Chiral-Auxiliary-Controlled Aldol Reaction with

Aldehyde 13 and Stereochemical Assignment

a

As discussed previously (Scheme 2), we planned to construct

the saliniketal dioxabicyclooctane ring system by means of a

Pt(II)-catalyzed cycloisomerization of a dihydroxyalkyne as

previously developed by us.¹⁸The availability of a set of

diﬀerent alkynediols provided a solid basis to further investigate

the scope and limitations of this methodology (Table 1). As

Table 1. Pt(II)-Catalyzed Cycloisomerization of

a

Dihydroxyalkynes

a

Reagents and conditions: (a) 20, TiCl₄, (−)-sparteine, CH₂Cl₂, 0 °C,

5 min, then 13, 0 °C, 1 h, 73%, 1.2:1 dr; (b) PPTS, EtOH, rt; (c)

PPTS (6 mol %), acetone/(MeO)₂CMe₂(4:1, 0.03 M), rt, 30 min,

94% 23-syn (two steps), 90% 23-anti (two steps). Abbreviations: Bn,

benzyl; PPTS, pyridinium p-toluenesulfonate.

Scheme 5A, reaction of the Z(O)-boron enolate derived from

chiral ethyl ketone 25 with aldehyde 24 (prepared from

commercial alcohol 16 via DMP oxidation in 94% yield; see

Scheme 3)²²using a protocol pioneered by Paterson and co-

workers^10aaﬀorded the desired anti-dimethyl aldol product 26

Scheme 5. Two Successful Routes to Correctly Conﬁgured

a

Stereotetrads 27 and 22-anti

a

Reaction conditions: substrate (0.1 mmol), [PtCl₂(CH₂CH₂)]₂(5

b

mol %), THF, rt. Yields of isolated puriﬁed products.

shown in entries 1 and 2, we were pleased to observe that

substrates with stereochemistry reminiscent of saliniketals (27

and 22-anti) were smoothly converted to 1,4-dimethyl-2,8-

dioxabicyclo[3.2.1]octanes 32 and 33, respectively, in >95%

yield upon stirring with a catalytic amount of Ziese’s dimer

([PtCl₂(CH₂CH₂)]₂, 5 mol %) in THF at ambient temper-

ature. Interestingly, the diastereomeric dihydroxyalkyne 22-syn

was recalcitrant toward cycloisomerization and instead was

converted to terminal methyl ketone 34 in 81% yield via

competing hydration of the alkyne (entry 3), whereas the

reduced triol substrate 31-syn (entry 4) did proceed to yield the

diastereomeric dioxabicyclooctane 35, but only after a

substantially prolonged reaction time of 1.5 h (vs 5 min for

entries 1 and 2).

The above results are perhaps most easily explained in the

context of concepts formulated in Scheme 6. Upon treatment

with the π acid ([PtCl₂(CH₂CH₂)]₂, the terminal alkyne,

generically represented by structure 36, is activated for a

kinetically facile 5-exo cyclization to provide, after proton

transfer, the putative oxonium intermediate 37 (Scheme 6A),

which is trapped via 6-endo spiroketalization to deliver, after

a

i

Reagents and conditions: (a) (+)-Ipc₂BOTf, Pr₂NEt, Et₂O, −78 °C,

83%, >10:1 dr; (b) NaBH₄, HOAc, 0 °C to rt; (c) 29, Sn(OTf)₂, Et₃N,

CH₂Cl₂, −20 °C, then −78 °C, 28, 82%, 16:1 dr; (d) TBAF, THF, 3

min, 94%. Abbreviations: TBAF, tetrabutylammonium ﬂuoride; Tf,

triﬂuoromethanesulfonyl.

E

DOI: 10.1021/jacs.6b03248

J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society

Article

sequence) from commercially available materials via two

alternative and equally eﬃcient routes (45−55% overall yield).

The synthesis of the aldehyde coupling partners 14a and 14b

is based on the conceptually simple and practical approach

depicted in Scheme 8. According to a sequence adapted from

Scheme 6. Rationalization of the Results Presented in Table

1

a

Scheme 8. Synthesis of Coupling Fragments 14a and 14b

protodemetalation, the dioxabicyclo[3.2.1]octane ring system.¹⁸

As shown in Scheme 6B, low-energy chair conformations of

anti-conﬁgured substrates that avoid syn-pentane interactions,

as exempliﬁed by 22-anti and its in situ-generated oxomium

intermediate 37-anti, are readily accessible and allow facile,

rapid spiroketalization, as documented in Table 1, entries 1 and

2. Such low-energy conformations are not accessible for 22-syn

and 31-syn. Although the formation of intermediate oxonium

37 occurs unimpeded, potential transition states for intra-

molecular addition of the second alcohol (6-endo) do appear to

suﬀer steric consequences, as represented by 37-syn and 37-

diol. Potential alternate conformations (not shown) also suﬀer

steric penalties compared with cyclization intermediate 37-anti.

For intermediate 37-syn, this leads to competitive hydrolysis to

give ketone 34 (entry 3), whereas 37-diol did provide bicyclic

ketal 35 but at a greatly reduced rate (entry 4).

a

Reagents and conditions: (a) 4-MeOBnOC(NH)CCl₃, PPTS,

CH₂Cl₂, rt, 18 h, 87%; (b) DIBAL-H, CH₂Cl₂, −78 °C, 2 h, 93%;

(c) (+)-MeOB(Ipc)₂, allylMgBr, 0 °C, add 40, −98 °C, then NaOH,

30% H₂O₂, Et₂O, reﬂux, 90%; (d) paraformaldehyde, DABCO,

dioxane/H₂O (1:1), 72 h; (e) TIPSCl, imidazole, DMAP, CH₂Cl₂, 0

°C to rt, 1 h, 79% (two steps); (f) LiOH, THF/H₂O (1:1), rt, 36 h,

92%; (g) DCC, DMAP, CH₂Cl₂, 0 °C to rt, 12 h, 90% from 42a, 82%

from 42b; (h) Grubbs-II (10 mol %), CH₂Cl₂, reﬂux, 14 h, 64% 43a

(+12% recovered SM), 67% 43b (+15% recovered SM); (i) DDQ,

CH₂Cl₂/H₂O (20:1), rt, 1 h; (j) DMP, NaHCO₃, CH₂Cl₂, rt, 30 min,

90% 14a or 14b (two steps). Abbrevations: DABCO, 1,4-

diazabicyclo[2.2.2]octane; DDQ, 2,3-dichloro-5,6-dicyano-1,4-benzo-

quinone; DCC, dicyclohexylcarbodiimide; DIBAL-H, diisobutylalumi-

num hydride; DMAP, 4-dimethylaminopyridine; PMB-TCA, p-

methoxybenzyl trichloroacetimidate; TIPS, triisopropylsilyl.

With the cycloisomerization accomplished, oxazolidinone 33

was processed to obtain ethyl ketone 15 via Weinreb amide

formation and Grignard reaction with ethylmagnesium bromide

(87% for two steps; Scheme 7). Alternatively, hydrogenolysis of

Nicolaou and co-workers, protection and semireduction of

Roche ester 39 was followed by a highly selective reagent-

controlled Brown allylation of aldehyde 40 to provide syn

homoallylic alcohol 41 in 73% overall yield for the three

steps.^33,34Esteriﬁcation of this material with methacrylic acid

(42a) or its oxygenated congener 42b (prepared from methyl

acrylate via Baylis−Hillman reaction,³⁵silylation, and saponi-

ﬁcation in 73% overall yield) set the stage for a ring-closing

a

Scheme 7. Two Routes for the Synthesis of Ethyl Ketone 15

metathesis that provided dihydropyranones 43a and 43b.³⁶

A

ﬁnal oxidative deprotection (DDQ, 90%) and oxidation with

DMP (quantitative) delivered aldehydes 14a and 14b in seven

steps and 35−36% overall yield from commercial starting

material.

a

We next investigated the key aldol coupling reaction

(Scheme 9). As discussed during the retrosynthetic analysis,

the predicted anti-Felkin bias of aldehyde 14 in aldol reactions

with Z(O)-enolates would provide the stereochemical outcome

for saliniketals.^20,21In regard to coupling partner 15, the facial

bias of chiral α-Me,β-alkoxy-substituted Z(O)-enolates is

primarily dominated by the α-stereocenter and strongly

inﬂuenced by the enolate counterion.^21,37,38Whereas titaniu-

m^21a,band boron^37c,denolates provide syn-dimethyl aldol

products (referring to the product ketone α- and α′-Me

substituents) with high selectivity, the corresponding lithium

enolates are generally less selective,^21a,c,37a,bbut importantly,

Reagents and conditions: (a) H₂, Pd/C, EtOH, 2 h, rt, 96%; (b)

DMP, NaHCO₃, CH₂Cl₂; (c) EtMgBr, THF, 0 °C, 1 h; (d)

MeONHMe·HCl, AlMe₃, THF, 0 °C.

benzyl ether 32 provided an alcohol that was identical to the

one reported by Paterson and co-workers.^10a,32This alcohol

was oxidized to aldehyde 38, which was followed by Grignard

addition (EtMgBr) and another Dess−Martin oxidation to

form ketone 15. Thus, key coupling fragment 15 was

reproducibly available in eight or nine steps (longest linear

F

DOI: 10.1021/jacs.6b03248

J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society

Article

Scheme 9. Highly Selective Aldol Fragment Coupling

they exhibit a preference for the opposite anti-dimethyl aldol

product. These observations are conveniently rationalized via

consideration of chairlike transition states TTS-A and TTS-B

(see the box in Scheme 9, in which circles indicate potentially

relevant nonbonding interactions). Minimization of non-

bonding interactions between the enolate chiral α-substituent

and metal ligands dominate for the more compact transition

states with titanium (also boron) enolates, which is best

accomplished via TTS-A. Minimization of allylic A^1,3strain as in

TTS-B, on the other hand,³⁹is potentially a more important

contributor for reactions with lithium enolates, which proceed

via a looser transition state because of the shorter Ti−O bond

length versus Li−O.⁴⁰

Figure 2. Conformational analysis of lithium enolates derived from

ethyl ketone 15.

additional oﬀending syn-pentane interaction nevertheless

provide the opposite stereochemical outcome.⁴²

Having achieved a highly selective aldol coupling, we turned

our attention to completing the synthesis of saliniketals as

shown in Scheme 10. A stereoselective reduction of aldol

The above analysis predicted a fully matched situation for a

titanium enolate reaction with aldehyde partners 14a and 14b

to yield the desired aldol products 44a and 44b (TTS-A, anti-

Felkin aldehyde face).^21bUnfortunately, all attempts to engage

titanium enolates derived from 15 proved fruitless, leading to

decomposition of ketone 15 and recovery of aldehydes 14a and

14b. The lithium enolate of 15, on the other hand, was

predicted to provide low selectivity due to a developing syn-

pentane interaction between the enolate and aldehyde methyl

substituents, as shown in TTS-B (mismatch with Felkin

aldehyde face).^20a,21c,41Surprisingly, however, aldol products

44a and 44b were obtained in high yields and selectivity

(>10:1), indicating that the reaction had proceeded via anti-

Felkin attack of the enolate Si face to aldehydes 14a and 14b, as

if it had behaved as a Z(O)-titanium enolate (see TTS-A).

This unexpected yet fortuitous outcome deserves some

comment. Inspection of the four staggered (C−C bond

between the α- and β-stereocenters) lithium enolate con-

formations I−IV depicted in Figure 2 (two each with H,H and

H,Me A^1,3enolate conformation) reveals a potential deﬁning

role for the γ-Me substituent. In the absence of this substituent

(light-gray circle), conformations I and IV best minimize steric

strain and dipole−dipole interactions, of which A^1,3-minimized

conformation I is expected to engage in Si-face aldol reactions,

as shown in TTS-B (Scheme 9). The γ-Me substituent,

however, raises the energy of conformations I and III by virtue

of a syn-pentane interaction with the enolate carbon (dark-gray

circle), leading to the observed Re-face attack via conformation

IV (see TTS-A (Li) in the right box). Although we cannot rule

out a chelation model via inside (Re-face) attack on

conformation II, which would lead to the observed stereo-

chemical outcome, we disfavor this possibility because of the

developing syn-pentane interactions with the ethylene bridge

(TTS-chelate, left box) and because other lithium enolates

derived from α-Me,β-alkoxy ethyl ketones but lacking this

Scheme 10. Completion of the Saliniketal A and B

a

Synthesis

a

Reagents and conditions: (a) Me₄N(AcO)₃BH, MeCN/HOAc (1:1),

−20 °C; (b) (MeO)₂CMe₂, PPTS, acetone, rt; (c) TBAF (10 equiv),

THF, rt; then NH₃(g), HOBt (2 equiv), EDC (2 equiv), rt, 72%; (d)

LiHMDS (10 equiv), −78 °C, 1 h, THF, 0 °C, 95%; (e) NH₃(1.0 M

in dioxane), HOBt (2 equiv), EDC (2 equiv), THF, rt; (f) Dowex,

MeOH, 1 h, rt, 90% (two steps). Abbreviations: EDC, 1-ethyl-3-(3-

(dimethylamino)propyl)carbodiimide; Bt, benzotriazolyl; LiHMDS,

lithium hexamethyldisilazide.

products 44a and 44b provided the corresponding anti-diols

(>20:1),³⁰the stereochemistry of which was conﬁrmed via H

1

and ¹³C NMR analysis of their acetonide derivatives 45a and

45b.²⁹During our attempts to remove the silyl ether of the

anti-diol derived from 44b, we noted that tetrabutylammonium

ﬂuoride (TBAF) was basic enough to induce the subsequent

dihydropyranone fragmentation.⁴³This led to the development

of a one-pot protocol consisting of treatment of the anti-diol

derived from 44b with TBAF (10 equiv) for 24 h, saturation of

G

DOI: 10.1021/jacs.6b03248

J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society

Article

the resulting THF solution with ammonia, and addition of the

coupling reagents HOBt and EDC. After an additional 10 h of

stirring at ambient temperature, saliniketal B (1b) was obtained

in 64% overall yield for this three-step process. The

spectroscopic data and optical rotation of the thus-obtained

material are in full agreement with those reported by Fenical

and Paterson for natural and synthetic 1b, respectively.^5,10a

For the synthesis of saliniketal A, we opted to explore the

dihydropyranone fragmentation on acetonide 45a, as we would

need to engage a protected saliniketal A en route to the total

synthesis of salinisporamycin (2) and rifsaliniketal (9) to be

described later. Interestingly, the TBAF-mediated fragmenta-

tion conditions utilized for saliniketal B led to only partial

conversion to the dienoic acid (∼40% and 55% recovered 45a),

which led us to speculate that the corresponding fragmentation

for saliniketal B was at least in part mediated by in situ-

generated alkoxide (from ﬂuoride-mediated desilylation and/or

deprotonation of the C7-hydroxyl).⁴⁴Since ﬂuoride (TBAF)

was insuﬃcient for full conversion of 45a, we explored other

Scheme 11. Strategies toward Naphthoquinone and

Naphthalene Fragments for Rifsaliniketal and

Salinisporamycin Total Synthesis

17a,b

t

bases, among which BuOK (THF, 0 °C, 24 h)

and DBU

(THF, rt, 48 h) also led to low conversion (13−20%, 60−65%

recovered 45a). In the end, we found a satisfactory high-

yielding solution in which base-mediated fragmentation of 45a

with LiHMDS (10 equiv, THF, 0 °C) provided the desired

dienoic acid in 95% yield. Amide formation as described for

saliniketal B (NH₃, HOBt, EDC, THF, rt) followed by

methanolysis of the acetonide protecting group yielded

saliniketal A (1a) in 90% (two steps). Again, the spectral

data were fully congruent with those reported for natural and

synthetic 1a.^5,10aThus, we have achieved the total synthesis of

saliniketals A and B in 11−14 steps (longest linear sequence)

and 27−28% overall yield from commercially available materials

using two alternate but equally eﬃcient routes to the key

intermediate ethyl ketone 15.

Scheme 12. Synthesis of Naphthoquinone Coupling Partners

for Salinisporamycin Synthesis

a

Salinisporamycin and Rifsaliniketal Total Synthesis:

Synthesis of the Naphthoquinone and Naphthalene

Fragments and Coupling with the Saliniketal A Frag-

ment. Synthetically, we envisioned rifsaliniketal and salinispor-

amycinwhich diﬀer only by the presence or absence of a

carboxylic acid at C4to be derived from combining a

common saliniketal A fragment with an appropriately function-

alized naphthoquinone fragment via two potential routes: (1) a

condensation between saliniketal acid and aminonaphthoqui-

none Q or aminonaphthalene N (Scheme 11, X = NH₂) or (2)

transition-metal-catalyzed C−N bond formation between

saliniketal A and bromonaphthoquinone Q or bromonaph-

thalene N (Scheme 11, X = Br).⁴⁵Because of the prevalence of

highly substituted naphthoquinones in ansamycins such as

rifamycin,⁴⁶several approaches for their synthesis have been

reported.^47,48Of those, we considered two approaches most

pertinent for our goals. A Diels−Alder reaction between 46 and

47, according to protocols explored by Trost^48cand Roush,^48g,h

would provide facile access to bromonaphthoquinone Q (X =

Br, R = H), whereas a cycloaddition between in situ-generated

benzyne 50 (from C₂-symmetric dibromide 49) and furan as

described by Kinoshita^48e,fwould deliver oxygen-bridged

cycloadduct 48. Further elaboration of this substance would

provide an alternative entry into fragments Q/N wherein R ≠

H.

a

Reagents and conditions: (a) Et₃N, TMSCl, hexanes, rt, 24 h; (b)

ⁿBuLi, Pr₂NH, −78 to 0 °C, then −78 °C, TMSCl, THF, 2 h, 86%

(two steps); (c) benzene, rt, 6 h, then silica gel, rt, 8 h, 72%; (d)

MOMCl, Pr₂NEt, CH₂Cl₂, rt, 12 h, 76%; (e) NaN₃, THF/MeOH/

H₂O (9:2:2), rt, 12 h; (f) PPh₃, THF/H₂O (4:1); rt, 30 min, 71% (two

steps). Abbreviation: MOM, methoxymethyl.

i

Brassard-like diene 46⁴⁹with known dibromobenzoquinone

47⁵⁰in benzene in the presence of silica gel aﬀorded product

53 in 72% yield. Interestingly, when the reaction was performed

in the absence of silica gel,⁵¹instead exploiting an aqueous HCl

workup to decompose intermediate cycloadduct 52 as reported

previously,^48c,g,hno identiﬁable product could be isolated.

Bis(methoxymethyl) protection of dihydroxynaphthoquinone

53 yielded bromonaphthoquinone coupling partner 54 in 76%

yield. According to a modiﬁed protocol contributed by Boger,⁵²

The Diels−Alder approach to bromo- and aminonaphtho-

quinone coupling partners relevant to salinisporamycin is

depicted in Scheme 12. According to a modiﬁed protocol

adapted from Trost^48cand Roush,^48g,hDiels−Alder reaction of

H

DOI: 10.1021/jacs.6b03248

J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society

Article

carefully controlled addition of sodium azide (the reaction is

sensitive to the presence of excess azide) followed by azide

reduction with triphenylphosphine in THF/H₂O aﬀorded

aminonaphthoquinone coupling partner 55 in 71% yield over

two steps.

As shown in Scheme 13, reduction of bromonaphthoquinone

54 (aq. Na₂S₂O₄) followed by methylation of the crude reduced

decomposed upon dissolution in chloroform or standing in air.

Therefore, reduction of the stable nitro precursor 61 was best

performed immediately before use.

Signiﬁcant eﬀortsnone of them successfulwere devoted

to the C4 functionalization of intermediates of the Diels−Alder

route to salinisporamycin (Schemes 12 and 13) as an entry to

coupling partners useful for rifsaliniketal synthesis (bearing an

additional C4-carboxylic acid group).⁵⁶We therefore were

motivated to explore the Kinoshita benzyne cycloaddition route

to naphthalenes functionalized at C4 (Scheme 14). Kinoshita

Scheme 13. Synthesis of Naphthalene Coupling Partners for

a

Salinisporamycin Synthesis

Scheme 14. Initial Benzyne Cycloaddition Route to

a

Rifsaliniketal Naphthalene Coupling Partners

a

Reagents and conditions: (a) Na₂S₂O₄, ether/H₂O (7:1), rt, 30 min;

(b) NaH, MeI, DMF, rt, 12 h, 78% (two steps); (c) ⁿBuLi, THF, −78

°C, 30 min, then MgBr₂·OEt₂, 0 °C, PhSCH₂N₃, −78 °C, 2 h, 24% 56

n

(unstable) and 62% 60; or BuLi, THF, −78 °C, 30 min, then aq.

a

Reagents and conditions: (a) Br₂_i, CH₂Cl₂; (b) BnBr, K₂CO₃,

NH₄Cl, 72% 60; (d) Cu(NO₃)₂·H₂O_x, CaCl₂, Ac₂O, −40 °C, 71%; (e)

H₂, Pd/C, EtOAc, >95% (unstable product; used immediately in the

next step).

n

acetone, 93% (two steps); (c) BuLi, Pr₂NH, −78 to 0 °C, then −78

°C, furan, warm to rt, 12 h, 89%; (d) TMSOTf, 2,6-^tBu₂-pyridine, 0 °C

for 5 min, then rt, 3 h, then TBAF, 0 °C, 15 min, 77%; (e) 71 (10 mol

%), O₂(1 atm), DMF, 50 °C, 12 h, 73%; (f) Na₂S₂O₄, ether/H₂O

(3:1), 15 min, rt; (g) NaH, MeI, DMF, rt, 12 h, 80% (two steps); (h)

product with MeI (NaH, DMF) aﬀorded the more electron-

rich bromonaphthalene coupling partner 57 (78% yield, two

steps). Unfortunately, all attempts to reduce the corresponding

aminonaphthoquinone 55 met with failure.⁵³We therefore

explored an amination protocol developed by Trost using

azidomethylphenyl sulﬁde as a synthon for electrophilic

i

H₂, Pd/C, EtOAc, 2 h; (i) MOMCl, Pr₂NEt, CH₂Cl₂, rt, 12 h, 78%

(two steps); (j) BuLi, ClCO₂Me, THF, −78 °C, 1 h, 95%; (k) H₂,

Pd/C, EtOAc, 99%. Abbreviations: LDA, lithium diisopropylamide.

n

described the synthesis of 64 in 97% yield from the

cycloaddition of furan with the benzyne generated in situ

from 63 and sodium amide in THF at 57 °C.^48fIn our hands,

these conditions provided 64 in unreliable yields. Instead, we

found that slow addition of a solution of furan and dibromide

63 in THF over a 1 h period to a −78 °C solution of LDA in

THF provided after 12 h the desired cycloadduct 64 in more

reproducible yields of ∼85−90%.

The subsequent ring-opening isomerization of cycloadduct

64 proved to be challenging and demanded extensive

experimentation (see Table 2). Indeed, the ring-opening

isomerization of a cycloadduct congener of 64 reported by

Kinoshita (diﬀering only by virtue of a C4-propanoyl

substituent instead of the C4-bromide as in 64) calls for 60%

48c,54

+

NH₂.

Accordingly, treatment of the Grignard derived

from bromonaphthalene 57 with azidomethylphenyl sulﬁde

followed by a basic hydrolytic workup to decompose triazene

intermediate 59 provided unstable (air-sensitive) amine 56 in

low yield (∼24%), with the remainder of the mass balance

being reduced naphthalene 60 (62%). Quenching lithium

intermediate 58a with aqueous ammonium chloride yielded 60

in 72% yield. This material proved useful for conversion to

nitronaphthalene 61 using copper(II) nitrate hydrate and

calcium chloride in acetic anhydride at −40 °C (71% yield).⁵⁵

Hydrogenation of 61 (Pd/C, H₂) in ethyl acetate provided

target aminonaphthalene 56 in essentially quantitative yield.

However, this compound proved to be very unstable and

I

DOI: 10.1021/jacs.6b03248

J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society

Article

Table 2. Exploration of Conditions for the Ring-Opening

Isomerization of Cycloadduct 64 to Naphthol 65a

Scheme 15. Synthesis of Rifsaliniketal Aminonaphthalene

Coupling Partner 78

a

b

c

entry

conditions

HClO₄(5%), THF, rt, 4 h

yield

1

2

3

4

5

6

7

34%

NR

17%

23%

PPTS (20%), THF/MeOH (10:1), 40 °C, 20 h

TsOH (5%), DCM, 40 °C, 6 h

H₂SO₄(5%), MeOH/H₂O (5:1), 60 °C, 6 h

aq. HCl (12 M, 20%), MeOH, rt, 12 h

CF₃CO₂H (5%), DCM, 40 °C, 6 h

Me₂C(CH₂OH)₂, (MeO)₃CH, TsOH (5%), DCM, 40 °C,

12 h

8

9

TMSI (2 equiv), DCM, 0 °C, 1 h

TMSOTf (1 equiv), DCM, rt, 1 h

DC

10 TMSOTf (1 equiv), pyridine (1 equiv), DCM, rt, 1 h

11 TMSOTf (5 equiv), 2,6-lutidine (6 equiv), THF, rt, 4 h

d

35%

12 TMSOTf (5 equiv), 2,6-t-Bu₂-pyridine (6 equiv), THF, rt, 77%

a

n

3 h

Reagents and conditions: (a) BuLi, ClCO₂Me, THF, −78 °C, 1 h,

95%; (b) TMSOTf, 2,6-^t-Bu₂-pyridine, 0 °C for 5 min, then rt, 3 h; (c)

H₂, Pd/C, EtOAc; (d) MOMCl, Pr₂NEt, CH₂Cl₂, rt, 12 h, 18% 74

and 36% 75; (e) BCl₃, −78 °C, 79%; (f) 71 (10 mol %), O₂(1 atm),

DMF, 50 °C, 12 h, 68%; (g) Na₂S₂O₄, ether/H₂O (5:1), rt, 15 min;

(h) NaH, MeI, DMF, rt, 12 h, 68% (two steps); (i) Cu(NO₃)₂·H₂O_x,

CaCl₂, Ac₂O, −40 °C, 69%.

13 TBSOTf (6 equiv), 2,6-lutidine (6 equiv), DCM, rt, 4 h

NR

i

a

b

All reactions were performed on a 0.1 mmol scale. For entries 8−13,

TBAF (1 M in THF, 3 equiv) was added prior to workup to deprotect

c

partially silylated products. Isolated yields. NR = no reaction; DC =

d

decomposition. 30% of the starting material was recovered.

aq. perchloric acid in THF for 35 h.^48fThese conditions

completely destroyed our cycloadduct 64. Switching to 5% aq.

perchloric acid in THF provided milder conditions and partial

conversion to the desired naphthol 65a (Table 2, entry 1).

Other mildly acidic conditions led to recovered starting

material (entries 2−5) or low conversion (entries 6 and 7).⁵⁷

Lewis acids such as TMSI or TMSOTf (in CH₂Cl₂) were too

reactive and led to decomposition (entries 8 and 9), even when

buﬀered with pyridine (entry 10). Interestingly, the addition of

2,6-^tBu₂-pyridine led to the desired naphthol 65a in 77% yield

(entry 12), whereas 2,6-lutidine proved to be less eﬀective

(entry 11) and the use of TBSOTf led to recovered starting

material (entry 13).⁵⁸The regiochemistry of naphthol 65a was

validated by nuclear Overhauser eﬀect correlations obtained

from derivative 65b, congruent with results reported by

Kinoshita.

Continuing with the synthesis (Scheme 14), oxidation of

naphthol 65a to quinone 66 was best performed with oxygen in

the presence of the Co−salen catalyst 71 (salcomine).^59,60

Dithionite reduction of 66 followed by methylation (NaH,

MeI) provided the corresponding naphthalene 67 in 80% yield

(two steps). Envisioning protecting group issues later in the

synthesis, we decided to switch the benzyl ether protecting

groups to methoxymethyl ether protecting groups. Hydro-

genolytic removal of the benzyl ethers proceeded smoothly, but

the corresponding dihydroxynaphthalene 69 proved to be

unstable under the conditions for methoxymethyl ether

formation. We therefore decided to implement a protecting

group switch earlier in the synthesis. As an aside, hydrogenation

of 67 followed by MOM protection provided an alternative

means of producing intermediate 60 en route to amino-

naphthalene 56 (Scheme 13).

isomerization using the optimized conditions from Table 2,

entry 12 followed by hydrogenolysis of the crude naphthol 73

then enabled installation of the methoxymethyl ether protecting

groups. The desired bis(MOM ether) 75 was obtained in 36%

overall yield for the three-step sequence along with tris(MOM

ether) 74 (18%). The latter could be recycled to the 75 in 79%

yield upon treatment with boron trichloride (for an overall

yield of 75 from 72 of 52%). Salcomine-catalyzed oxidation (→

76, 68%),⁵⁹dithionite reduction, and methyl ether formation

proceeded smoothly as before to deliver naphthalene 70.

Nitration of 70 (using conditions optimized for nitration of 60;

Scheme 13) yielded stable nitronaphthalene 77, which could be

eﬃciently reduced to aminonaphthalene 78. As before (amine

56; Scheme 13), this substance was unstable and best prepared

immediately prior to the coupling step with the saliniketal acid

fragment (vide infra).

With various naphthoquinone/naphthalene fragments at

hand, we explored their coupling with a suitable saliniketal

surrogate (Scheme 16). Our ﬁrst attempts involving a direct

condensation strategy of acetonide-protected saliniketal acid

79⁶¹with aminonaphthoquinone 55 invariably met with failure.

A representative example includes the use of EDC and

hydroxybenzotriazole (HOBt), a reagent combination that

provided only the activated benzotriazolyl ester intermediate 81

(78%). Irrespective of prolonged reaction times, heating, and

concentration, active ester 81 was recalcitrant to engaging the

aminonaphthoquinone coupling partner 55. Given the reduced

nucleophilicity of aminonaphthoquinones,⁶²we were not really

surprised by these results and moved to explore a more

contemporary C−N bond formation between bromonaphtho-

quinone 54 and protected saliniketal amide 80.⁶¹However, this

approach also failed, as we did not identify suitable palladium-

or copper-catalyzed conditions to eﬀect this transforma-

tion.⁶³⁻⁶⁵

As shown in Scheme 15, transmetalation of bromide 64

(ⁿBuLi) followed by trapping with methyl chloroformate

yielded methyl ester 72 in 95% yield.^48fRing-opening

J

DOI: 10.1021/jacs.6b03248

J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society

Article

a

Scheme 16. Completion of Rifsaliniketal (9) and Salinisporamycin (2) Total Synthesis

a

Reagents and conditions: (a) 79, 55 or 56 or 78, HOBt, EDCI, Et₃N, CH₂Cl₂, rt, 2 h, 76% 81 (with 55) or 67% 84 (with 56) or 71% 85 (with 78);

(b) 80, 57, CuI (10 mol %), K₃PO₄, DMEDA, toluene, 100 °C, 12 h, 57%; (c) CAN, H₂O, MeCN, 0 °C, 15 min; (d) HCl, NaI, MeOH, H₂O, THF,

rt, 24 h, 62% 2 (two steps from 84); (e) crude 83, LiOH, MeOH, H₂O, 0 °C, 12 h, 47% (three steps from 85). Abbreviations: CAN, ceric

ammonium nitrate; DMEDA, N,N′-dimethylethylenediamine.

Moving to the more electron-rich naphthalene coupling

partners, we were pleased to observe a smooth Cu(I)-catalyzed

cross-coupling between acetonide-protected saliniketal 80 and

bromonaphthalene 57 to yield salinisporamycin precursor 84 in

57% yield according to a protocol described by Altman and

Buchwald.^65dEqually satisfying, the corresponding EDC/

HOBt-mediated condensation between acetonide-protected

saliniketal acid 79 and the more electron-rich aminonaph-

thalene 56 now also proceeded without impediment,⁶²

providing the same amide 84 in a slightly better isolated yield

of 67%. By means of the latter procedure, rifsaliniketal amide

precursor 85 was obtained in 71% yield via condensation of

aminonaphthalene 78 and saliniketal acid 79. With a viable

solution at hand, the only remaining steps entailed an oxidation

back to the naphthoquinoid oxidation state and global

deprotection. In the event, treatment of naphthalene amides

84 and 85 with ceric ammonium nitrate in aqueous acetonitrile

readily provided naphthoquinones 82 and 83, respectively. The

crude products were not puriﬁed but rather were immediately

engaged in a global deprotection. Extensive experimentation

was required to identify conditions for simultaneous removal of

the acetonide and methoxymethyl ether protecting groups.

Using substrate 82, we found that Dowex (in MeOH) removed

only the acetonide, whereas a variety of Lewis acidic conditions

led to decomposition. The acetonide group was similarly

removed with aqueous HCl (varying acid strength, solvents,

temperature, concentration), but the phenolic methoxymethyl

ethers were left largely intact. In the end, stirring an acidic

solution (HCl, 0.2 N ﬁnal concentration) of crude 82 and the

crucial additive NaI (∼1 equiv) in THF/MeOH/H₂O

(10:2.5:1) for 24 h at ambient temperature cleanly delivered

fully deprotected salinisporamycin (2) in 62% overall yield

from naphthalene 84.⁶⁶By means of the same procedure, but

with an additional saponiﬁcation (aq. LiOH/MeOH),

rifsaliniketal (9) was obtained in 47% overall yield for the

three-step procedure from precursor 85. The spectral data for

synthetic 2 and 9 were in full agreement with those reported for

natural 2 in ref 12 and natural 9 disclosed herein.

CONCLUSION

■

We have described the isolation, structure elucidation, and total

synthesis of the novel marine natural product rifsaliniketal and

the total synthesis of the structurally related variants

salinisporamycin and saliniketals A and B. Rifsaliniketal was

previously proposed, but not observed, as a diverted metabolite

from a biosynthetic precursor to rifamycin S.¹⁴Decarboxylation

of rifamycin provides salinisporamycin, which upon truncation

with loss of the naphthoquinone ring leads to saliniketals. Our

synthetic strategy hinged upon a Pt(II)-catalyzed cyclo-

isomerization of an alkynediol, methodology developed in our

lab,¹⁸to set the unique dioxabicyclo[3.2.1]octane ring system

and a fragmentation of an intermediate dihydropyranone to

forge a stereochemically deﬁned (E,Z)-dienamide unit. Multiple

routes were explored to assemble fragments with high

stereocontrol, an exercise that provided additional insights

into acyclic stereocontrol during stereochemically complex

fragment-assembly processes. The resulting 11−14 step

synthesis of saliniketals then enabled us to explore strategies

for the synthesis and coupling of highly substituted

naphthoquinones or the corresponding naphthalene fragments.

A Roush/Trost Diels−Alder approach eﬃciently provided the

heterocyclic fragments for salinisporamycin synthesis in 6−9

total steps (18−37% overall yield), whereas the corresponding

carboxylic acid-containing congeners were best derived using a

benzyne cycloaddition approach ﬁrst explored by Kinoshita (12

total steps, 12% overall yield). Whereas direct coupling with

K

DOI: 10.1021/jacs.6b03248

J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society

Article

(15) Espindola, A. P.; Crouch, R.; DeBergh, J. R.; Ready, J. M.;

MacMillan, J. B. J. Am. Chem. Soc. 2009, 131, 15994−15995.

(16) See Table S2 in the Supporting Information for a comparison of

naphthoquinone fragments proved unsuccessful, both amida-

tion and C−N bond formation tactics with the more electron-

rich naphthalene congeners provided an eﬃcient means to

complete the ﬁrst total synthesis of rifsaliniketal and

salinisporamycin in 15−16 steps (longest linear sequence)

and 13−17% overall yield from commercially available

materials.

1

the H and ¹³C NMR data of salinisporamycin and rifsaliniketal.

(17) This interesting fragmentation is rarely used in total synthesis.

For selected examples, see: (a) Corey, E. J.; Schmidt, G. Tetrahedron

Lett. 1979, 20, 2317−2320. (b) Roush, W. R.; Spada, A. P. Tetrahedron

Lett. 1982, 23, 3773−3776. (c) Masamune, S.; Imperiali, B.; Garvey,

D. S. J. Am. Chem. Soc. 1982, 104, 5528−5531.

(18) Liu, B.; De Brabander, J. K. Org. Lett. 2006, 8, 4907−4910.

(19) For a related bidirectional approach in the context of rifamycin,

see ref 17c.

(20) For selected reviews, see: (a) Roush, W. J. Org. Chem. 1991, 56,

4151−4157. (b) Franklin, A. S.; Paterson, I. Contemp. Org. Synth.

1994, 1, 317−338. For a computational study, see: (c) Gennari, C.;

Vieth, S.; Comotti, A.; Vulpetti, A.; Goodman, J. M.; Paterson, I.

Tetrahedron 1992, 48, 4439−4458.

ASSOCIATED CONTENT

* Supporting Information

■

S

The Supporting Information is available free of charge on the

ACS Publications website at DOI: 10.1021/jacs.6b03248.

Rifsaliniketal isolation procedures, experimental proce-

dures, compound characterization data, and copies of

NMR spectra (PDF)

(21) For selected examples, see: (a) Martin, S. F.; Lee, W.-C.

Tetrahedron Lett. 1993, 34, 2711−2714. (b) Evans, D. A.; Dart, M. J.;

Duffy, J. L.; Rieger, D. L. J. Am. Chem. Soc. 1995, 117, 9073−9074.

(c) Evans, D. A.; Yang, M. G.; Dart, M. J.; Duffy, J. L. Tetrahedron Lett.

1996, 37, 1957−1960.

AUTHOR INFORMATION

Corresponding Authors

*Jef.DeBrabander@UTSouthwestern.edu

*John.MacMillan@UTSouthwestern.edu

Notes

■

(22) Dess, D. B.; Martin, J. C. J. Org. Chem. 1983, 48, 4155−4156.

(23) Brown, H. C.; Bhat, K. S. J. Am. Chem. Soc. 1986, 108, 5919−

5923.

(24) ¹H NMR analysis of the crude mixture indicated the presence of

all four possible diastereomers in addition to the starting material and

unidentiﬁed material.

(25) The stereochemistry of the aldol product 18-syn was assigned by

comparison to the same ethyl ketone product obtained from 21-syn via

treatment of 21-syn with ethylmagnesium bromide (CH₂Cl₂, −78 °C,

33% yield). See the Supporting Information for details.

(26) Increasing the reaction time and/or the temperature (to −50

°C) led to the formation of signiﬁcant amounts of desilylated,

dehydrated cyclic dienol ether byproduct 19. See the Supporting

Information for details.

The authors declare no competing ﬁnancial interest.

ACKNOWLEDGMENTS

■

We thank Dr. Matsuda for providing NMR spectra of natural

salinisporamycin and Karen MacMillan for proofreading the

manuscript. Financial support was provided by the Robert A.

Welch Foundation (Grants I-1422 and I-1689). J.K.D. holds

the Julie and Louis Beecherl, Jr., Chair in Medical Science, and

J.B.M. is a Chilton/Bell Scholar in Biochemistry.

REFERENCES

(27) Unlike other syn-α-Me,β-alkoxyaldehydes, aldehyde 13 contains

a terminal alkyne, which might be responsible for the observed Felkin

aldol selectivity. In this context, we note that Danishefsky and

coworkers have observed an eﬀect of π-electron substituents (oleﬁn,

aryl) on aldehyde facial bias through long-range nonbonding

interactions between the aldehyde and substituent π systems. See:

Lee, C. B.; Wu, Z.; Zhang, F.; Chappell, M. D.; Stachel, S. J.; Chou, T.-

C.; Guan, Y.; Danishefsky, S. J. J. Am. Chem. Soc. 2001, 123, 5249−

5259.

■

(1) Prudhomme, J.; McDaniel, E.; Ponts, N.; Bertani, S.; Fenical, W.;

Jensen, P.; Le Roch, K. PLoS One 2008, 3, e2335.

(2) Jensen, P. R.; Mafnas, C. Environ. Microbiol. 2006, 8, 1881−1888.

(3) Feling, R. H.; Buchanan, G. O.; Mincer, T. J.; Kauffman, C. A.;

Jensen, P. R.; Fenical, W. Angew. Chem., Int. Ed. 2003, 42, 355−357.

(4) Udwary, D. W.; Zeigler, L.; Asolkar, R. N.; Singan, V.; Lapidus,

A.; Fenical, W.; Jensen, P. R.; Moore, B. S. Proc. Natl. Acad. Sci. U. S. A.

2007, 104, 10376−10381.

(5) Williams, P. G.; Asolkar, R. N.; Kondratyuk, T.; Pezzuto, J. M.;

Jensen, P. R.; Fenical, W. J. Nat. Prod. 2007, 70, 83−88.

(6) Gerner, E. W.; Meyskens, F. L., Jr. Nat. Rev. Cancer 2004, 4, 781−

792.

(7) Pendeville, H.; Carpino, N.; Marine, J. C.; Takahashi, Y.; Muller,

M.; Martial, J. A.; Cleveland, J. L. Mol. Cell. Biol. 2001, 21, 6549−6558.

(8) Saunders, L. R.; Verdin, E. Mol. Cancer Ther. 2006, 5, 2777−

2785.

(9) Basuroy, U. K.; Gerner, E. W. J. Biochem. 2006, 139, 27−33.

(10) For total syntheses, see: (a) Paterson, I.; Razzak, M.; Anderson,

E. A. Org. Lett. 2008, 10, 3295−3298. (b) Liu, J.; De Brabander, J. K. J.

Am. Chem. Soc. 2009, 131, 12562−12563.

(11) For a formal synthesis, see: Yadav, J. S.; Hossain, S. S.; Madhu,

M.; Mohapatra, D. K. J. Org. Chem. 2009, 74, 8822−8825.

(12) Matsuda, S.; Adachi, K.; Matsuo, Y.; Nukina, M.; Shizuri, Y. J.

Antibiot. 2009, 62, 519−526.

(13) (a) Jensen, P. R.; Williams, P. G.; Oh, D.-C.; Zeigler, L.; Fenical,

W. Appl. Environ. Microbiol. 2007, 73, 1146−1152. (b) Kim, T. K.;

Hewavitharana, A. K.; Shaw, P. N.; Fuerst, J. A. Appl. Environ.

Microbiol. 2006, 72, 2118−2125. (c) Bose, U.; Hewavitharana, A. K.;

Vidgen, M. E.; Ng, Y. K.; Shaw, P. N.; Fuerst, J. A.; Hodson, M. P.

PLoS One 2014, 9, e91488.

(28) Crimmins, M. T.; King, B. W.; Tabet, E. A.; Chaudhary, K. J.

Org. Chem. 2001, 66, 894−902.

(29) (a) Rychnovsky, S. D.; Skalitzky, D. J. Tetrahedron Lett. 1990,

31, 945−948. (b) Evans, D. A.; Rieger, D. L.; Gage, J. R. Tetrahedron

Lett. 1990, 31, 7099−7100.

(30) Evans, D. A.; Chapman, K. T.; Carreira, E. M. J. Am. Chem. Soc.

1988, 110, 3560−3578.

(31) (a) Evans, D. A.; Clark, J. S.; Metternich, R.; Novack, V. J.;

Sheppard, G. S. J. Am. Chem. Soc. 1990, 112, 866−868. (b) Evans, D.

A.; Ng, H. P.; Clark, S.; Rieger, D. L. Tetrahedron 1992, 48, 2127−

2142.

(32) The 22-anti diol thus prepared was identical to that prepared

according to Scheme 4. The structures of 27 and 22-anti were

conﬁrmed after transformation to the same primary alcohol obtained

from hydrogenolysis of 32 or LiBH₄reduction of 33 and comparison

to the same alcohol prepared independently by Paterson et al.^10aSee

the Supporting Information.

(33) Nicolaou, K. C.; Patron, A. P.; Ajito, K.; Richter, P. K.; Khatuya,

H.; Bertinato, P.; Miller, R. A.; Tomaszewski, M. J. Chem. - Eur. J.

1996, 2, 847−868.

(34) It should be noted that the stereochemical outcome is not

important, as the resulting stereocenter will be destroyed during the

ﬁnal dihydropyranone fragmentation reaction (Scheme 10). However,

homogeneous material facilitates characterization, and targeting the

(14) Wilson, M. C.; Gulder, T. A.; Mahmud, T.; Moore, B. S. J. Am.

Chem. Soc. 2010, 132, 12757−12765.

L

DOI: 10.1021/jacs.6b03248

J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society

Article

syn stereoisomer was predicted to provide maximal stereocontrol

during the aldol fragment coupling in this double diastereodiﬀerentiat-

ing process.

(35) Yu, C.; Liu, B.; Hu, L. J. Org. Chem. 2001, 66, 5413−5418.

(36) Scholl, M.; Ding, S.; Lee, C. W.; Grubbs, R. H. Org. Lett. 1999,

1, 953−956.

(54) Trost, B. M.; Runge, T. A. J. Am. Chem. Soc. 1981, 103, 2485−

2487.

(55) The conditions reported by Kinoshita failed, which we

attributed to the presence of crystal water in commercially available

Cu(NO₃)₂. The addition of dry CaCl₂provided an eﬃcient solution.

See: Gandy, M. N.; Piggott, M. J. J. Nat. Prod. 2008, 71, 866−868.

(56) We explored various Friedel−Crafts acylation conditions and

halogenations with intermediates from Schemes 12 and 13 as well as

congeners with various alternative protecting groups instead of

MOM.^48gAttempts to access more elaborate dienes for early C4

introduction during the Diels−Alder reaction were equally frustrating.

See, for example, ref 48d,j and: Okabayashi, T.; Iida, A.; Takai, K.;

Nawate, Y.; Misaki, T.; Tanabe, Y. J. Org. Chem. 2007, 72, 8142−8145.

(37) (a) Gustin, D. J.; VanNieuwenhze, M. S.; Roush, W. R.

Tetrahedron Lett. 1995, 36, 3447−3450. (b) McCarthy, P. A.;

Kageyama, M. J. Org. Chem. 1987, 52, 4681−4686. (c) Paterson, I.;

McClure, C. K. Tetrahedron Lett. 1987, 28, 1229−1232. (d) Duffy, J.

L.; Yoon, T. P.; Evans, D. A. Tetrahedron Lett. 1995, 36, 9245−9248.

(38) The selectivity tends to be largely independent of the nature and

relative conﬁguration of the enolate β-alkoxy substituent.

(39) For a review of allylic 1,3-strain, see: Hoffmann, R. W. Chem.

Rev. 1989, 89, 1841−1860.

(57) (a) Sorgel, S.; Azap, C.; Reißig, H.-U. Eur. J. Org. Chem. 2006,

̈

2006, 4405−4418. (b) Blanchot, M.; Candito, D. A.; Larnaud, F.;

Lautens, M. Org. Lett. 2011, 13, 1486−1489. (c) Kaelin, D. E.; Lopez,

O. D.; Martin, S. F. J. Am. Chem. Soc. 2001, 123, 6937−6938. (d) Batt,

D.; Jones, D. G.; La Greca, S. J. Org. Chem. 1991, 56, 6704−6708.

(e) Morton, G. E.; Barrett, A. G. M. J. Org. Chem. 2005, 70, 3525−

3529. (f) Etomi, N.; Kumamoto, T.; Nakanishi, W.; Ishikawa, T.

Beilstein J. Org. Chem. 2008, 4, 15.

(40) For an insightful analysis, see ref 37a.

(41) This particular stereochemical combination of lithium enolate

and aldehyde was classiﬁed as a fully mismatched case.^20cSimilar

literature examples of fully mismatched reactions provided aldol

products in a nonselective manner.^21a,c,37a

(42) Because lithium enolates are typically not monomeric, a

chelated transition state akin to II (Figure 2) could engage the

aldehyde via a second distinct lithium atom. To avoid developing syn-

pentane interactions with the ethylene bridge (as in TTS-chelate), it

would be anticipated that the aldehyde would engage the less-hindered

enolate Si face in such a situation.

(43) Nakata, T.; Hata, N.; Oishi, T. Heterocycles 1990, 30, 333−334.

(44) TBAF can contain signiﬁcant amounts of hydroxide, which is

potentially the base in the reaction. The poor reactivity could be due

to the need for a primary hydroxymethyl and/or C7-hydroxyl to act as

an internal base, both of which are absent in acetonide 45a.

(45) Given a potential facile redox interconversion, we envisioned

that both naphthoquinone Q and naphthalene N intermediates would

be viable coupling partners.

(46) Floss, H. G.; Yu, T.-W. Chem. Rev. 2005, 105, 621−632.

(47) For selected total syntheses, see: (a) Rifamycin S: Kishi, Y. Pure

Appl. Chem. 1981, 53, 1163−1180. (b) Rifamycin W: Nakata, M.;

Akiyama, N.; Kamata, J.-i.; Kojima, K.; Masuda, H.; Kinoshita, M.;

Tatsuta, K. Tetrahedron 1990, 46, 4629−4652. (c) Damavaricin D:

Roush, W. R.; Coffey, D. S.; Madar, D. J. J. Am. Chem. Soc. 1997, 119,

11331−11332.

(58) Magnus, P.; Eisenbeis, S. A.; Fairhurst, R. A.; Iliadis, T.; Magnus,

N. A.; Parry, D. J. Am. Chem. Soc. 1997, 119, 5591−5605.

(59) De Jonge, C. R. H. I.; Hageman, H. J.; Hoentjen, G.; Mijs, W.

Org. Synth. 1977, 57, 78.

(60) Oxidation with PhI(OAc)₂, Ag₂O (dioxane/HNO₃), NBS

(HOAc/H₂O), or iron(III) salts did not produce the desired 1,4-

naphthoquinone. See ref 58 and: (a) Kitani, Y.; Morita, A.;

Kumamoto, T.; Ishikawa, T. Helv. Chim. Acta 2002, 85, 1186−1195.

(b) Moller, K.; Wienhofer, G.; Schroder, K.; Join, B.; Junge, K.; Beller,

̈

M. Chem. - Eur. J. 2010, 16, 10300−10303.

(61) Acetonide-protected acid 79 and amide 80 are intermediates en

route to saliniketal A. See the Supporting Information.

(62) For similar observations regarding a recalcitrant intramolecular

macrolactamization of a rifamycin-derived aminoquinoid carboxylic

acid and a solution via the use of the more electron-rich

aminodihydroquinoid carboxylic acid, see: Corey, E. J.; Clark, D. A.

Tetrahedron Lett. 1980, 21, 2045−2048.

(63) To the best of our knowledge, there are no reported examples of

amidation of halo-substituted naphthoquinones similar to bromide 55

via C−N bond formation.

(64) For a user’s guide to palladium-catalyzed aminations and

amidations, see: Surry, D. S.; Buchwald, S. L. Chem. Sci. 2011, 2, 27−

50.

(65) For a review, see: (a) Surry, D. S.; Buchwald, S. L. Chem. Sci.

2010, 1, 13−31. For selected examples, see: (b) Klapars, A.; Huang,

X.; Buchwald, S. L. J. Am. Chem. Soc. 2002, 124, 7421−7428. (c) Jiang,

L.; Job, G. E.; Klapars, A.; Buchwald, S. L. Org. Lett. 2003, 5, 3667−

3669. (d) Altman, R. A.; Buchwald, S. L. Nat. Protoc. 2007, 2, 2474−

2479.

(66) For an example of the beneﬁcial use of NaI as an additive during

methoxymethyl ether deprotection, see: Williams, D. R.; Barner, B. A.;

Nishitani, K.; Phillips, J. G. J. Am. Chem. Soc. 1982, 104, 4708−4710.

(48) For selected approaches relevant to the work described here,

see: (a) Kozikowski, A. P.; Sugiyama, K.; Springer, J. P. Tetrahedron

Lett. 1980, 21, 3257−3260. (b) Parker, K. A.; Petraitis, J. J. Tetrahedron

Lett. 1981, 22, 397−400. (c) Trost, B. M.; Pearson, W. H. Tetrahedron

Lett. 1983, 24, 269−272. (d) Kelly, R. T.; Behforouz, M.; Echavarren,

A.; Vaya, J. Tetrahedron Lett. 1983, 24, 2331−2334. (e) Nakata, M.;

Kinoshita, M.; Ohba, S.; Saito, Y. Tetrahedron Lett. 1984, 25, 1373−

1376. (f) Nakata, M.; Wada, S.; Tatsuta, K.; Kinoshita, M. Bull. Chem.

Soc. Jpn. 1985, 58, 1801−1806. (g) Roush, W. R.; Madar, D. J.

Tetrahedron Lett. 1993, 34, 1553−1556. (h) Roush, W. R.; Coffey, D.

S. J. Org. Chem. 1995, 60, 4412−4418. (i) Kuttruff, C. A.; Geiger, S.;

Cakmak, M.; Mayer, P.; Trauner, D. Org. Lett. 2012, 14, 1070−1073.

(j) Nawrat, C. C.; Palmer, L. I.; Blake, A. J.; Moody, C. J. J. Org. Chem.

2013, 78, 5587−5603. (k) Nawrat, C. C.; Lewis, W.; Moody, C. J. J.

Org. Chem. 2011, 76, 7872−7881.

(49) Langer, P.; Schneider, T.; Stoll, M. Chem. - Eur. J. 2000, 6,

3204−3214.

(50) For the preparation of 47, see: Bringmann, G.; Gotz, R.; Keller,

̈

P. A.; Walter, R.; Boyd, M. R.; Lang, F.; Garcia, A.; Walsh, J. J.; Tellitu,

I.; Bhaskar, K. V.; Kelly, T. R. J. Org. Chem. 1998, 63, 1090−1097.

(51) For a similar beneﬁcial eﬀect of silica gel on aromatization of

Diels−Alder intermediates en route to naphthomycin, see ref 48i.

(52) Boger, D. L.; Panek, J. S.; Duff, S. R.; Yasuda, M. J. Org. Chem.

1985, 50, 5790−5795.

(53) In addition to dithionite reduction, Luche reduction (see:

Gemal, A. L.; Luche, J. L. J. Am. Chem. Soc. 1981, 103, 5454−5459 )

and various hydrogenation conditions were explored..

M

DOI: 10.1021/jacs.6b03248

J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Article Doi

Rifamycin Biosynthetic Congeners: Isolation and Total Synthesis of Rifsaliniketal and Total Synthesis of Salinisporamycin and Saliniketals A and B

DOI: 10.1021/jacs.6b03248

Source and publish data:

Authors:

Article abstract of DOI:10.1021/jacs.6b03248

Full text of DOI:10.1021/jacs.6b03248

Products guided by the article

R&D Labs maybe for 1187192-23-9

Relevant to this article

Hot Product