Synthesis, characterization, and reactivity of complex tricyclic oxonium ions, proposed intermediates in natural product biosynthesis

Article abstract of DOI:10.1021/jacs.9b07438

Reactive intermediates frequently play significant roles in the biosynthesis of numerous classes of natural products although the direct observation of these biosynthetically relevant species is rare. We present here direct evidence for the existence of c

Full text of DOI:10.1021/jacs.9b07438

Article
Cite This: J. Am. Chem. Soc. XXXX, XXX, XXX−XXX
pubs.acs.org/JACS
Synthesis, Characterization, and Reactivity of Complex Tricyclic
Oxonium Ions, Proposed Intermediates in Natural Product
Biosynthesis
†
†
,†,‡
and Jonathan W. Burton*,†
Hau Sun Sam Chan, Q. Nhu N. Nguyen, Robert S. Paton,*
†
Chemistry Research Laboratory, University of Oxford, Mansfield Road, Oxford OX1 3TA, U.K.
‡
Department of Chemistry, Colorado State University, Fort Collins, Colorado 80523, United States
*
S Supporting Information
ABSTRACT: Reactive intermediates frequently play significant
roles in the biosynthesis of numerous classes of natural products
although the direct observation of these biosynthetically relevant
species is rare. We present here direct evidence for the existence of
complex, thermally unstable, tricyclic oxonium ions that have been
postulated as key reactive intermediates in the biosynthesis of
numerous halogenated natural products from Laurencia species.
Evidence for their existence comes from full characterization of
these oxonium ions by low-temperature NMR spectroscopy
supported by density functional theory (DFT) calculations,
coupled with the direct generation of 10 natural products on
exposure of the oxonium ions to various nucleophiles.
INTRODUCTION
stereochemically complex oxonium ions characterized to
■
31
date. Additionally, the elucidation of the in vitro reactivity
profile of the tricyclic oxonium ions has resulted in the total
synthesis of more than 10 complex halogenated natural
products. This work provides chemical evidence for the
existence of complex, halogenated, tricyclic oxonium ions that
have been postulated as key reactive intermediates in the
biosynthesis of numerous halogenated natural products.
Oxonium Ions in Laurencia Natural Product Biosyn-
thesis. Red algae of the genus Laurencia and organisms that
feed on them produce a vast array of structurally diverse C
The biosynthesis of secondary metabolites frequently proceeds
via unstable reactive intermediates. With terpenes, one of the
largest classes of natural products, olefin cyclizations via
carbocations (frequently tertiary carbocations) dominate
and account for the vast structural diversity exhibited by this
class of natural products. Evidence for carbocation inter-
mediates has come from many sources
the isolation and condensed phase characterization of tertiary
1
2
−4
5
−7
and is supported by
8
,9
(
and other) carbocations (carbenium ions), along with
numerous biomimetic total syntheses of terpenoid natural
1
5
22
halogenated cyclic ether acetogenin natural products which
have stimulated synthetic chemists to develop numerous new
methods to gain access to these complex secondary
products designed so as to proceed via cationic intermedi-
1
0
ates. Increasingly, the subtleties of these biosynthetic
4
cyclizations are being probed computationally. In a related
2
2,32
metabolites.
Since Masamune’s ground-breaking synthesis
manner, the biosynthesis of numerous halogenated acetogenin
natural products from Laurencia species have been proposed to
proceed via complex bi- and tricyclic oxonium ions (Figure
33−35
of (±)-laurencin ((±)-4, Figure 1) reported in the 1970s,
the syntheses of over 50 cyclic halogenated acetogenin natural
1
1−21
products from Laurencia spp. have been reported with many
1
).
The existence of these trialkyloxonium ions, as with
36−39
targets being synthesized by multiple independent routes.
carbocations in terpene biosynthesis, has been implied from
Hand-in-hand with the development of the synthetic method-
ology has been the development of biosynthetic postulates
toward these natural products with many of these biosynthetic
postulates being rich in oxonium ion chemistry including
the structural and stereochemical diversity of these natural
2
2
products and initial biosynthetic studies with isolated
1
1,13,23
enzymes
as well as from a number of elegant biomimetic
syntheses primarily from Kim, Snyder, and Braddock that
involve ring opening/expansions of in situ generated oxonium
ions to give a host of structurally diverse Laurencia natural
1
5,16
19
bicyclic oxonium ions derived from epoxides,
oxetanes,
17,18
21
14,25
tetrahydrofurans,
and oxocenes as well as fused
and
11−13,20,26
1
5−21,24−30
bridged tricyclic oxonium ions.
Initial proposals
products.
However, no direct evidence for the
regarding the biosynthesis of C15 halogenated acetogenin
existence of these trialkyloxonium ions has been forthcoming.
Herein we report the synthesis and full characterization (both
chemically and computationally) of one family of tricyclic
trialkyloxonium ionsamong the most structurally and
Received: July 12, 2019
©
XXXX American Chemical Society
A
DOI: 10.1021/jacs.9b07438
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
Figure 1. Proposed biosynthesis of some halogenated ethers from Laurencia species from laurediols 1. (a) Proposals regarding the biosynthesis of
deacetyllaurencin. (b) Proposed biosynthesis of natural products involving tricyclic oxonium ions. (c) Conventional representations of the
lauroxocane natural products the chlorofucins, bromofucins, prelaurefucins, acetyllaurefucins, and the laurefucins.
natural products from Laurencia species were put forward by
lization of the (3E,R,R)-laurediol (3E,R,R)-1 gives rise to
1
3,23,40
12,43
Murai
which involved bromocyclizations of the linear
deacetyllaurencin (3E)-3 via bromonium ion 2;
indeed,
laurediols 1 that exist in nature as mixtures of diastereomers
experiments with isolated enzymes yielded deacetyllaurencin
4
1,42
13,23,40,44
(Figure 1a).
Murai postulated that enzymatic bromocyc-
(3E)-3 in very low yield (0.015%).
More recently,
B
DOI: 10.1021/jacs.9b07438
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
trialkyloxonium ions have been proposed as intermediates in
the bromocyclization of linear precursors to give deacetyllaur-
encins 3. Braddock proposed that laurepoxides 5, the potential
precursor of the laurediols 1, undergoes bromocyclization to
must be non-nucleophilic to prevent premature reaction with
the target oxonium ions. In the event, successful oxonium ion
formation and characterization was achieved using the weakly
−
coordinating Krossing’s anion [Al(pftb) ] (pftb = perfluoro-
4
15
57
give the bicyclic oxonium ions 7 via the bromonium ions 6.
tert-butoxy)]. The synthesis of the oxonium ions required the
Opening of the oxonium ions 7 by water at C-6 gives 3. Such
epoxide bromonium ion cyclizations were first reported by
total synthesis of their potential precursors, namely, the
bromo- and chlorofucin natural products 19 and 24, as well as
the prelaurefucins 12 and the corresponding C-10 chlorides
the neoprelaurefucins 29 (Scheme 1); the neoprelaurefucins
have previously been postulated as natural products based on
4
5
Davies with cyclooctadiene epoxide and expanded by
15,16
Braddock with biosynthetically relevant model systems.
Snyder has proposed that deacetyllaurencin (3E)-3 could be
biosynthesized from (3E,R,R)-laurediols (3E,R,R)-1 by two
favorable 5-endo bromocyclizations via bromonium ions 8 and
12,26
the proposed biogenesis of the laurefucins.
Oxonium Ion Precursor Synthesis. Our route to these
compounds proceeded from the enantiopure bromomesylates
25 and 31 which were readily prepared from diacetone-D-
glucose (Scheme 1 and Supporting Information, sections 2 and
3, pp S4−S24). The bromomesylate 25 would provide the
enantiomer of the oxonium ions 13, namely, ent-13, and the
bromomesylate 31 would give the oxonium ions 20. The key
step in our synthesis of the oxonium ion precursors involved
the generation of the diastereomerically pure bromonium ions
26 and 32 from the bromomesylates 25 and 31. Thus,
9
to give the bicyclic oxonium ion 10 which undergoes formal
17,18
loss of positive bromine to give deacetyllaurencin (3E)-3.
Again, this biomimetic ring expansion of an oxonium ion
17,18,28
enjoys experimental support.
In both the Braddock and
Snyder proposals the respective oxonium ions can open in
other ways to form different ring systems found in other
Laurencia natural products.
The focus of the current work is the synthesis and
characterization of the tricyclic oxonium ions 13 and 20 that
are downstream from deacetyllaurencin on the proposed
58
exposure of the bromomesylate 25 to titanium(IV) chloride,
12
biosynthetic pathway (Figure 1b). Suzuki and Murai and
followed by addition of the chloride scavenger Ag[Al(pftb) ]·
4
1
1,13
57
Fukuzawa
proposed the intermediacy of the oxonium ions
CH Cl , and then excess tetrabutylammonium chloride
2
2
1
3 in the biosynthesis of notoryne and the laurefucins. The
delivered the corresponding dioxabicyclo[5.2.1]decane 28 in
68% yield. We propose that this complex transformation
proceeds by initial bromonium ion formation, using
oxonium ions 13 are proposed to arise by transannular
displacement of bromide ion from the prelaurefucins 12 which
themselves are derived from bromocyclization of deacetyllaur-
5
8
Braddock’s procedure, followed by removal of excess
chloride on addition of the silver(I) salt. This allows time for
the bromonium ion 26 to evolve into the corresponding
46
encins 3 (Figure 1b, right). Opening of the oxonium ions 13
with chloride ion at C-7 or with water at C-10 generates the
1
1,12,47
28
natural products (E/Z)-notoryne 14
and (E/Z)-
A closely related natural
9
oxonium ion 27 via 5-endo ring closure. Subsequent addition
13,23,48
laurefucins 16, respectively.
of excess chloride results in C-10 opening of the oxonium ion
4
59
product, acetyllaurefucin (E)-15, may arise from opening of
the oxonium ions 13 with acetate at C-10 or by acetylation of
2
7 to give dioxabicyclo[5.2.1]decane 28. In a similar manner,
exposure of the bromomesylate 31 to titanium(IV), followed
(
late
E)-laurefucin (E)-16. Extension of this biosynthetic postu-
by silver(I), and then excess chloride or bromide gave the
20,26 50
26
provides a route to the natural products the bromo-
51 52
known dioxabicyclo[5.2.1]decanes 34 and 38 resulting from
C-10 opening of the oxonium ion 33 by halide anion, along
with the corresponding known inseparable 2,2′-bifuranyls 35
and chlorofucins 19 and 24, the elatenynes 22, and
laurendecumenyne B 21, suggesting the participation of
oxonium ions 20 as the key biosynthetic intermediates (Figure
53
20
and 39 arising from competitive C-7 oxonium ion opening.
The dioxabicyclo[5.2.1]decanes 28, 34, and 38 were readily
converted into the corresponding oxonium ion precursors, the
1
b, left). Analogous to the formation of oxonium ions 13, the
oxonium ions 20 are proposed to originate from (3E/Z,S,S)-
laurediols (S,S)-1, via a series of enzymatic bromocyclizations
followed by transannular bromide displacement from the
bicyclic bromofucins 19. Opening of the oxonium ions 20 at
C-10 or C-7 with bromide or chloride ions would generate the
natural products 19, 21, 22, and 24. Alternatively, opening of
5
9
ent-neoprelaurefucins ent-(E/Z)-29, and the halofucin
59
59
natural products (E/Z)-19 and (E/Z)-24 using standard
transformations. Thus, ent-(E)-neoprelaurefucin ent-(E)-29
was synthesized from 28 via a cross metathesis with
crotonaldehyde and catalyst 30 followed by a Colvin−Ohira
reaction to install the (E)-enyne in 44% yield for the two
54
2
0 with bromide ion at C-13 would give the ocellenyes, the
24
full stereostructures of which have recently been tentatively
reassigned based on density functional theory (DFT)
steps. The (Z)-diastereomer, ent-(Z)-29, was synthesized
from 28 via a four-step route involving oxidative cleavage of
the terminal olefin in 28, (Z)-vinyl iodide formation using a
12,55
calculations and biogenetic considerations as 23.
6
0
Stork−Zhao reaction, Sonogashira cross-coupling, and
deprotection. The (E)- and (Z)-halofucins were prepared
from the bromochlorides 34 and 35, and the dibromides 38
and 39 using closely related procedures. Cross metathesis of an
inseparable 7:1 mixture of the bromochlorides 34 and 35 with
crotonaldehyde and Grubbs catalyst 30 in dichloromethane at
40 °C gave a 4:1 mixture of separable α,β-unsaturated
RESULTS AND DISCUSSION
Our strategy for the synthesis of the key oxonium ions 13 and
0 takes its cue from their proposed biosynthesis. This
biomimetic approach involves the transannular displacement of
halide from bicyclic halo ethers such as 12, 19, or 24 using a
silver(I)-mediated halide abstraction. Here, precipitation of the
silver halide would provide the necessary thermodynamic
driving force to compensate for the expected endothermic
nature of the oxonium ion formation. The choice of the
counteranion of the silver(I) salt was also essential to the
successful synthesis of our target oxonium ions. Since oxonium
■
2
20
aldehydes 36 and 37 in 78% overall yield. This change of
isomer ratio from 7:1 in the starting materials to 4.2:1 in the
products is due to the thermal rearrangement of either 34 or
36 or both (see later) most likely via the corresponding
tricyclic oxonium ions. Pure 36 has previously been reported
56
26
ions are known to alkylate heteroatoms, the counteranion
by Kim en route to (E)-chlorofucin (E)-24, and we used
C
DOI: 10.1021/jacs.9b07438
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
a
Scheme 1. Synthesis of the Oxonium Ion Precursors, the Bromo- and Chlorofucins, and the ent-Neoprelaurefucins
a
Reagents and conditions: (a) TiCl , AgAl(pftb) ·CH Cl , CH Cl −40 °C, 2 h, and then −78 °C add Bu NCl or Bu NBr, 28 68%; 38:39 3:1
4
4
2
2
2
2
4
4
inseparable mixture 50%; 34:35 7:1 inseparable mixture 70%; (b) crotonaldehyde, cat. 30, CH Cl , 40 °C, 37 15%, 36 63%; (c) Me SiCHN ,
2
2
3
2
nBuLi, −78 °C−RT, ent-(E)-29 44% (two steps); (E)-19 42%; (E)-24 43%; (d) OsO , NaIO , 2,6-lutidine, dioxane, water; (e) ICH PPh I,
4
4
2
3
NaHMDS, HMPA, THF, −78 °C−RT; (f) Me SiCCH, Pd(PPh ) , CuI, Et N; (g) K CO , MeOH, ent-(Z)-29 20% (four steps); (Z)-19 17%
3
3
4
3
2
3
(
four steps); (Z)-24 32% (four steps); (h) crotonaldehyde, cat. 30, Cu(I)I, Et O, RT, 41 13%, 40 60%.
2
2
4,26
Kim’s method to convert 36 into (E)-chlorofucin (E)-24.
42 and 43, and 44 and 45 followed by Sonogashira cross-
coupling and deprotection.
With a mixture of the dibromides 38 and 39, cross metathesis
using the above conditions resulted in the products 40 and 41
being formed as an inseparable mixture with a small amount of
the chlorides 36 and 37. The formation of the chlorides 36 and
Oxonium Ion Synthesis and Characterization. Having
secured the synthesis of the oxonium ion precursors, the
halofucins 19 and 24 and the enantiomers of the neo-
prelaurefucins ent-29, investigation into oxonium ion for-
mation and characterization commenced. Initial experiments
demonstrated that a complex tricyclic oxonium ion related to
ent-13 was thermally unstable above ca. −40 °C (see Figure 4
below). It was therefore necessary to generate and analyze the
oxonium ions at −40 °C or below. Ultimately, we found that
exposure of either of the chlorides ent-(E)-29 and ent-(Z)-29
to Ag[Al(pftb) ]·CH Cl in CD Cl at −78 °C allowed
3
7 most likely occurs by oxonium ion formation from 38 or 40
followed by quenching with the chloride from Grubbs second-
generation catalyst. Chloride formation could be avoided by
conducting the cross metathesis at room temperature in the
presence of copper(I) iodide which gave the separable
dibromides 40 and 41 in 73% overall yield. Following
Kim’s precedent, (E)-bromofucin (E)-19 was readily prepared
from 40 using a Colvin−Ohira homologation. The (Z)-
halofucins were synthesized from the bromochlorides 34 and
5, and the dibromides 38 and 39 using a Stork−Zhao
olefination to give mixtures of the inseparable vinyl iodides
61
2
6
20
4
2
2
2
2
26
generation of the oxonium ions, ent-13·Al(pftb) , which were
4
characterized by NMR at −78 °C after low-temperature
1
13
3
filtration of the precipitated silver chloride. The H and
C
6
0
NMR spectra of the generated intermediates compared with
D
DOI: 10.1021/jacs.9b07438
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
Figure 2. Formation and NMR spectra (CD Cl , −78 °C) of the oxonium ion ent-(E)-13·Al(pftb) along with NMR spectra (CD Cl , −78 °C) of
2
2
4
2
2
the starting material ent-(E)-29. The 1-D and 2-D NMR spectra provide evidence for the formation of the oxonium ion ent-(E)-13·Al(pftb) . In
4
1
13
13
particular, the H− C HMBC spectrum demonstrates the formation of trivalent oxygen. Additionally, on formation of the oxonium ion, the
C
NMR resonances of the carbon atoms directly attached to the trivalent oxygen atom (C-7 blue, C-10 green, and C-13 purple) move to significantly
higher chemical shift compared with the starting material, whereas nonadjacent carbons (e.g., C-9 red) undergo smaller changes in chemical shift.
The protons adjacent to the trivalent oxygen atom (H-7 blue, H-10 green, and H-13 purple) also move to significantly higher chemical shift on
oxonium ion formation. The β-protons (H-6 light blue, H-9 red, and H-8′ yellow) also move to higher chemical shift indicative of overlap of the
occupied σ
orbital with the σ*
orbital, whereas the gray β-proton (H-8) does not have the correct geometry for σ
orbital overlap with
C−H
C−O
C−H
the antibonding σ*C−O orbital and undergoes a significantly smaller change in chemical shift (see the Supporting Information for NBO analysis,
Figure S32, pp S102−S103). (a) Formation of the oxonium ion ent-(E)-13·Al(pftb) by chloride abstraction from the dioxabicyclo[5.2.1]decane
4
1
13
13
ent-(E)-29 at −78 °C. (b) H− C HMBC (heteronuclear multiple bond correlation) NMR spectrum of ent-(E)-13·Al(pftb) . (c) C NMR
4
1
3
spectrum of ent-(E)-13·Al(pftb) (top) and C NMR spectrum of ent-(E)-29 (bottom) with corresponding carbons joined by filled lines (carbons
4
1
1
α to trivalent oxygen) and dashed lines (carbon β to trivalent oxygen). (d) H NMR spectrum of ent-(E)-13·Al(pftb) (top) and H NMR
4
spectrum of ent-(E)-29 (bottom) with corresponding protons joined by filled lines (protons α to trivalent oxygen) and dashed lines (protons β to
trivalent oxygen).
those of the precursors, along with the 1H− C HMBC
13
Referring to a section of the H− C HMBC spectrum shown
1
13
spectra, provided good evidence for the formation of the
in Figure 2b, the trivalent nature of the oxygen atom within
oxonium ions, illustrated for ent-(E)-13·Al(pftb) in Figure 2.
ent-(E)-13·Al(pftb) is evident from the correlations of C-7
4
4
E
DOI: 10.1021/jacs.9b07438
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
Figure 3. Comparison of experimental and computed 13C and H NMR chemical shifts of the oxonium ions ent-(E/Z)-13 and (E/Z)-20. For each
individual histogram, the x-axis corresponds to atom number and the y-axis corresponds to δexpt − δcomp. The computed chemical shifts for each
oxonium ion are Boltzmann-weighted averages of the chemical shifts from their contributing conformers at −78 °C. (a) Comparison of the
1
1
3
experimental and computed C NMR chemical shifts of the oxonium ions ent-(E/Z)-13 and(E/Z)-20 (C-4 and C-12 were excluded from the
1
analysissee the Supporting Information). (b) Comparison of the experimental and computed H NMR chemical shifts of the oxonium ions ent-
(
E/Z)-13 and(E/Z)-20. The entries highlighted in blue correspond to the least deviation between computed and experimental chemical shifts,
1
3
1
having the lowest mean absolute deviation (MAD) as well as the lowest individual deviations in C and H chemical shifts (Tables S22−S28, pp
S89−S95), hence confirming the structural assignment of the oxonium ions.
(
blue) to H-10 (green), H-9 (red), and H-13 (purple) as well
overlap of the occupied σC−H orbitals of H-6, H-9, and H-8′
with the corresponding σ*C−O orbitals of the trivalent oxygen.
The H-8 (gray) proton does not have the correct geometry to
maximize this overlap, and hence, its chemical shift remains
virtually unchanged (see Figure S32, pp S102−S103 for NBO
analysis). In a similar manner, the oxonium ions (E)- and
(Z)-20·Al(pftb)4 could be readily generated from the
bromofucins 19 and were characterized as above; equivalent
figures to Figure 3 for the oxonium ions ent-(Z)-13·Al(pftb)4
as from C-10 (green) to H-7 (blue). Additional evidence for
the formation of ent-13·Al(pftb) comes from the change in H
and C chemical shifts in moving from ent-(E)-29 to the
oxonium ion ent-(E)-13·Al(pftb) . Figure 2c shows clearly that
the resonances for C-7 (blue), C-10 (green), C-13 (purple)
carbon atoms move to significantly higher chemical shift on
treatment of ent-(E)-29 with a silver(I) salt, in keeping with
the large C NMR chemical shift changes observed on
oxonium ion formation in other tricyclic systems,
1
4
1
3
4
59
1
3
62,63
whereas
and (E)- and (Z)-20·Al(pftb) may be found in the Supporting
4
the resonance for C-9 (red) moves to slightly lower chemical
shift. In a similar manner, Figure 2d shows the change in
chemical shifts of a number of protons on moving from ent-
Information (Figures S10−S12, pp S47−S49).
1
3
1
DFT calculated C and H NMR chemical shifts of the
computed structures of each oxonium ion provided further
supporting evidence for the formation and structural assign-
ment of the four synthetic oxonium ions ent-(E/Z)-13·
Al(pftb) and (E/Z)-20·Al(pftb) (Figure 3). Following a
(
E)-29 to ent-(E)-13·Al(pftb) . The protons α to the trivalent
4
oxygen atom, H-7 (blue), H-10 (green), and H-13 (purple) in
ent-(E)-13·Al(pftb) resonate at significantly higher chemical
4
4
4
shift compared with the corresponding protons in ent-(E)-29.
The resonances for the β-protons H-6 (light blue), H-9 (red),
H-8′ (yellow) also move to higher chemical shift, indicative of
conformational search, Boltzmann-weighted shielding tensors
were calculated at the mPW1PW91/6-311+G(2d,p)//B3LYP/
6-31+G(d,p) level of theory in dichloromethane, from which
F
DOI: 10.1021/jacs.9b07438
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
Figure 4. Rearrangement of the oxonium ion 46·Al(pftb) to give the oxocarbenium ion 47·Al(pftb) . The stability of the oxonium ion 46·
4
4
1
Al(pftb) was quantified by measuring the half-life (ca. 5100 s) for conversion into the oxocarbenium ion 47·Al(pftb) at 5 °C by H NMR. The
4
4
calculated transition state for this rearrangement is a highly asynchronous 1,2-hydride shift. (a) Conversion of 46·Al(pftb) into 47·Al(pftb) with
4
4
proposed mechanism. (b) DFT calculated transition structure for the conversion of 46 into 47, distances in angstroms. (c) Plot of the change of the
1
H NMR integral for H-9 of 46·Al(pftb) (blue) and H-9 of 47·Al(pftb) (red) vs time, along with half-life intervals on the x-axis.
4
4
1
3
1
C and H NMR chemical shifts were obtained using the
was shown to be a highly asynchronous 1,2-hydride shift with a
calculated Gibbs energy barrier of 20.6 kcal/mol, consistent
with experiment.
64
scaling factors reported by Tantillo. For each oxonium ion,
the smallest deviation between the experimental chemical shifts
and the pool of computed chemical shifts was found only when
the configuration of the computed structure matched the
configuration of the structure expected from synthesis
65,66
Oxonium Ion Reactivity with Nucleophiles. Having
characterized the complex halogenated tricyclic oxonium ions
both spectroscopically and computationally we sought to
investigate their reactivity with a variety of nucleophiles to
garner evidence to support their proposed intermediacy in the
biosynthesis of Laurencia natural products. Trapping of the
various oxonium ions with a range of nucleophiles directly
generated 10 natural products (Scheme 2). Trapping of the
(
diagonals of Figure 3, parts a and b). The off-diagonal
entries, which correspond to comparison between structures of
different configurations, gave larger mean absolute deviations
(
MADs) as well as larger deviations of individual 13C and/or
1
H NMR chemical shifts. Furthermore, the chemical shifts of
oxonium ions (E)- and (Z)-20·Al(pftb) with bromide anion
the carbon and proton atoms adjacent to the trivalent oxygen
were in keeping with simpler tricyclic oxonium ions.
4
62,63
gave rise to the (E)- and (Z)-bromofucins 19 in 37% and 51%
26
Characterization of a Decomposition Pathway. The
yields, respectively, along with small amounts of (E)- and
20
synthetic oxonium ions ent-(E)- and ent-(Z)-13·Al(pftb) and
(Z)-elatenyne 22. Using chloride in place of bromide anion
gave the corresponding (E)- and (Z)-chlorofucins 24 in 46%
and 63% yields, respectively, along with a trace amount of
4
(
E)- and (Z)-20·Al(pftb) were thermally unstable and readily
4
decomposed above −40 °C. Indeed, it was not possible to
synthesize the oxonium ion (E)-20·Al(pftb) from (E)-
2
0,26
laurendecumenyne B 21.
series, treatment of the oxonium ions ent-(E)- and ent-(Z)-13·
Al(pftb) with water gave the corresponding ent-laurefucins
ent-16 in 34% and 51% yields, respectively.
anion with ent-(E)-13·Al(pftb) gave ent-acetyllaurefucin ent-
15 in 30% yield. Both ent-(E)- and ent-(Z)-13·Al(pftb) were
In the other diastereomeric
4
chlorofucin (E)-24 as the halide abstraction reaction only
occurred above −40 °C (for energy profile for oxonium ion
4
29
24,28
formation, see Figures S33−S35, pp S104−S108). The
thermal instability of a closely related oxonium ion was studied
which resulted in the characterization of a clean decomposition
pathway (Figure 4). The oxonium ion 46·Al(pftb)4 was
synthesized at −78 °C by halide abstraction from the
corresponding chloride precursor using AgAl(pftb) ·CH Cl
Using acetate
4
4
also treated with chloride anion which gave the ent-(E)- and
ent-(Z)-neoprelaurefucins ent-29 in 59 and 73% yields,
12,26
respectively.
What is clear from these quenching experi-
4
2
2
and characterized by NMR spectroscopy at −78 °C (see the
ments is that all four oxonium ions undergo kinetic quenching
59
24,26,29,67
Supporting Information, p S57); the chloride precursor was
readily prepared from the alkene 28 by a hydroboration/
oxidation/esterification sequence (see the Supporting Infor-
mation, pp S56−S57). At higher temperature (>−40 °C) the
at C-10 in keeping with results in related systems.
Additionally, the oxonium ions (E)- and (Z)-20·Al(pftb)
4
undergo a small amount of C-7 quenching analogous to
26,29,67
previous work;
in keeping with related results, products
oxonium ion 46·Al(pftb) was readily and cleanly transformed
into the oxocarbenium ion 47·Al(pftb) which itself was fully
from quenching of the oxonium ions at C-13 were not
4
55
observed. The quenching experiments provided the 2,2′-
bifuranyl natural products 21 and 22 in low yields. However, it
is known that C-10 halogenated dioxabicyclic compounds
related to the halofucins can undergo slow rearrangement to
the corresponding 2,2-bifuranyls in the presence of activated
4
characterized (see the Supporting Information, pp S56−S59).
Incubation of the oxonium ion 46·Al(pftb) at 5 °C in an
4
NMR spectrometer allowed the rate of this isomerization to be
determined. The half-life for the conversion of 46·Al(pftb)4
12,20,27,48b,67
into 47·Al(pftb) at 5 °C was ca. 5100 s (average of three
silica gel.
Hence, exposure of the bromofucins 19
4
experiments) corresponding to a Gibbs energy of activation of
to activated silica gel in hexanes or petroleum ether for 36 h at
2
1.2 kcal/mol. The conversion of 46 into 47 computationally
ambient temperature gave (E)- and (Z)-elatenyne 22 in 61%
G
DOI: 10.1021/jacs.9b07438
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
Scheme 2. Natural Products from Oxonium Ions via Quenching with Nucleophilesa,b,c
a
Reagents and conditions: (a) Bu NBr, −78 °C−RT, (Z)-19 51%, (Z)-22 9%; (E)-19 37%, (E)-22 7%; (b) Bu NCl, −78 °C−RT, (Z)-24 63%, 21
4
4
trace; (E)-24 46%; ent-(Z)-29 73%; ent-(E)-29 59%; (c) water, NaHCO , −78 °C−RT, ent-(Z)-16 34%, ent-(E)-16 51%; (d) Bu NOAc, −78 °C−
RT, ent-15 30%. The conventional representations of the lauroxocane natural products, the chlorofucins and bromofucins, and the enantiomers of
the neoprelaurefucins, the laurefucins, and acetyllaurefucin are given at the bottom of the scheme. The neoprelaurefucins 29 have previously been
3
4
b
c
predicted to be natural products (refs 12, 26).
2
0
and 60% yields, respectively (Scheme 3). In a similar
manner, heating (Z)-chlorofucin (Z)-24 with activated silica
gel in chloroform at 80 °C (bath temperature) gave
laurendecumenyne B 21 in 73% yield, while (E)-chlorofucin
which indicated clean and quantitative conversion after 12
days.
DFT calculations of transition structures (TSs) helped to
explain the observed regioselectivity in the nucleophilic
25
(
E)-24 gave (E)-laurendecumenyne B (E)-21, which has yet to
be isolated from nature, in 50% yield on exposure to
activated silica gel in hexanes at 80 °C (bath temperature).
ent-(E/Z)-Neoprelaurefucins ent-29 as solutions in hexanes
when treated with silica gel at 80 °C (bath temperature) for 12
h gave ent-(E)- and ent-(Z)-notoryne ent-14 quantitatively.
The course of the rearrangement of ent-29 into ent-(Z)-14 at
0 °C (bath temperature) in CDCl in the absence of silica gel
opening of oxonium ions ent-(E)-13 and (E)-20. For ent-
(E)-13, the energy barrier for opening at C-10 with chloride is
lower than for the corresponding openings at C-7 and C-13
(Figure 5b, also see Figures S36−S48, pp S109−S130 for
20
29
68
analyses with different oxonium ions and nucleophiles). The
O−C-10 bond is longest in the ground state (Figure 5a),
indicative of greater carbocationic character at C-10.
Accordingly, attack at this position involves the least structural
reorganization in the TS; this can also be quantified through
47
8
3
1
was readily followed by H NMR (see Figure S59, p S206)
H
DOI: 10.1021/jacs.9b07438
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
Scheme 3. Natural Products from Natural Products ,
a b
barrier for attack at this position is much greater than at C-7
and C-10. This principally results from a much larger
distortion energy term required to open at this position, also
resulting in a later TS with respect to the breaking C−O bond.
Consistent with synthetic studies (Scheme 2), the
preferential site of nucleophilic attack by chloride is computed
to be C-10, followed by C-7 for both ent-(E)-13 and (E)-20.
With water as the nucleophile, the selectivity between C-10
and C-7 for (E)-20 is calculated to be much more finely
⧧
⧧
balanced with ΔE slightly favoring C-10 opening while ΔG
slightly favors C-7 opening (Figures S36−S48, pp S109−
29
S130). Regardless, the energetic preference for attack at C-10
over C-7 is relatively modest and could be overcome by the
intervention of enzymatic control. The reorganization required
for attack at C-13 is substantial for ent-(E)-13, although we
found this pathway to be more competitive for (E)-20; from
this oxonium ion, enzymatic promotion of the C-13 pathway is
55,70
more plausible.
CONCLUSION
We have synthesized and characterized four complex
halogenated tricyclic oxonium ions (E)- and (Z)-20 and ent-
■
(
E)- and ent-(Z)-13, proposed as key reactive intermediates in
the biosynthesis of numerous halogenated acetogenin natural
products from Laurencia spp. using low-temperature NMR
experiments and DFT calculations. Furthermore, we have
shown that, on exposure to a range of nucleophiles, these
oxonium ions directly generate 10 natural products. The above
work provides chemical evidence to support the proposed
intermediacy of such intermediates biosynthetically and sets
the scene for future biosynthetic studies on C-15 halogenated
acetogenin natural products.
a
Reagents and conditions: (a) activated silica gel, hexanes or
petroleum ether, RT, (E)-22 61%; (b) activated silica gel, hexanes
or petroleum ether, 80 °C (bath temperature), (Z)-22, 60%; (E)-21
5
0%; ent-(Z)-14 quant, ent-(E)-14 quant; (c) activated silica gel,
b
CHCl , 80 °C (bath temperature), 21 73%. (E)-Laurendecumenyne
is likely to be a natural product that is yet to be isolated.
3
ASSOCIATED CONTENT
Supporting Information
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/jacs.9b07438.
■
*
S
Experimental procedures, characterization data, NMR
spectra of reported compounds, computational details
(PDF)
The xyz coordinates of computed structures (ZIP)
AUTHOR INFORMATION
Corresponding Authors
robert.paton@colostate.edu
jonathan.burton@chem.ox.ac.uk
■
*
*
ORCID
Robert S. Paton: 0000-0002-0104-4166
Jonathan W. Burton: 0000-0002-5181-5301
Notes
Figure 5. DFT analysis of oxonium ion opening with chloride as
nucleophile. (a) The lowest energy conformer of ent-(E)-13 and
relevant bond lengths and angles; O−C-10 is the longest bond. (b)
TSs for opening at C-13, C-10, and C-7. (c) Distortion-interaction/
activation-strain analysis along the opening reaction coordinate. The
oxonium ion undergoes least distortion when opening at C-10.
Distances in angstroms.
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
We thank the Croucher Foundation for the award of a
University of Oxford Croucher Scholarship to H.S.S.C. and
Project 752491 from the European Union’s Horizon 2020
research and innovation program under the Marie Skłodowska-
Curie Grant Agreement No. 752491. We used the RMACC
Summit supercomputer, which is supported by the National
Science Foundation (ACI-1532235 and ACI-1532236), the
University of Colorado Boulder and Colorado State University,
distortion-interaction/activation-strain analysis along the re-
69
action coordinate (Figure 5c). Attack at C-7 and C-10
involves very similar interaction energy profiles, while the
distortion energy term at C-10 is smaller leading up to the TS,
giving rise to a smaller barrier for this position. Consistent with
the absence of C-13 opening in our experimental studies, the
I
DOI: 10.1021/jacs.9b07438
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
the Extreme Science and Engineering Discovery Environment
(20) Dyson, B. S.; Burton, J. W.; Sohn, T. I.; Kim, B.; Bae, H.; Kim,
D. Total synthesis and structure confirmation of elatenyne: success of
computational methods for NMR prediction with highly flexible
diastereomers. J. Am. Chem. Soc. 2012, 134, 11781−11790.
(21) Taylor, M. T.; Fox, J. M. Biosynthesis of the C15-acetogenin
laurepoxide may involve bromine-induced skeletal rearrangement of a
Δ4-oxocene precursor. Tetrahedron Lett. 2015, 56, 3560−3563.
(22) For recent reviews concerning Laurencia natural products see:
(a) Wang, B. G.; Gloer, J. B.; Ji, N. Y.; Zhao, J. C. Halogenated
organic molecules of Rhodomelaceae origin: Chemistry and Biology.
Chem. Rev. 2013, 113, 3632−3685. (b) Zhou, Z. F.; Menna, M.; Cai,
Y. S.; Guo, Y. W. Polyacetylenes of marine origin: chemistry and
bioactivity. Chem. Rev. 2015, 115, 1543−1596. (c) Wanke, T.;
Philippus, A. C.; Zatelli, G. A.; Vieira, L. F. O.; Lhullier, C.;
Falkenberg, M. C15 acetogenins from the Laurencia complex: 50 years
of research − an overview. Rev. Bras. Farmacogn. 2015, 25, 569−587.
(23) Fukuzawa, A.; Aye, M.; Takasugi, Y.; Nakamura, M.; Tamura,
M.; Murai, A. Enzymatic Bromo-Ether Cyclization of Laurediols with
Bromoperoxidase. Chem. Lett. 1994, 23, 2307−2310.
(24) Kim, B.; Lee, M.; Kim, M. J.; Lee, H.; Kim, S.; Kim, D.; Koh,
M.; Park, S. B.; Shin, K. J. Biomimetic asymmetric total synthesis of
(−)-laurefucin via an organoselenium-mediated intramolecular
hydroxyetherification. J. Am. Chem. Soc. 2008, 130, 16807−16811.
(25) Clarke, J.; Bonney, K. J.; Yaqoob, M.; Solanki, S.; Rzepa, H. S.;
White, A. J.; Millan, D. S.; Braddock, D. C. Epimeric face-selective
oxidations and diastereodivergent transannular oxonium ion for-
mation fragmentations: computational modeling and total syntheses
of 12-epoxyobtusallene IV, 12-epoxyobtusallene II, obtusallene X,
marilzabicycloallene C, and marilzabicycloallene D. J. Org. Chem.
2016, 81, 9539−9552.
(
XSEDE) through allocation TG-CHE180056.
REFERENCES
■
(
1) McMurry, J.; Begley, T. The Organic Chemistry of Biological
Pathways; Roberts and Company: Englewood, CO, 2005.
(
2) Tantillo, D. J. The carbocation continuum in terpene
biosynthesis−where are the secondary cations? Chem. Soc. Rev.
010, 39, 2847−2854.
3) Dewick, P. M. The biosynthesis of C5−C25 terpenoid
compounds. Nat. Prod. Rep. 2002, 19, 181−222.
4) Tantillo, D. J. Biosynthesis via carbocations: theoretical studies
on terpene formation. Nat. Prod. Rep. 2011, 28, 1035−1053.
5) Thulasiram, H. V.; Erickson, H. K.; Poulter, C. D. A common
2
(
(
(
mechanism for branching, cyclopropanation, and cyclobutanation
reactions in the isoprenoid biosynthetic pathway. J. Am. Chem. Soc.
2
(
008, 130, 1966−1971.
6) Wendt, K. U.; Schulz, G. E.; Corey, E. J.; Liu, D. R. Enzyme
Mechanisms for Polycyclic Triterpene Formation. Angew. Chem., Int.
Ed. 2000, 39, 2812−2833.
(
7) Christianson, D. W. Structural and Chemical Biology of
Terpenoid Cyclases. Chem. Rev. 2017, 117, 11570−11648.
(
8) Olah, G. A.; Tolgyesi, W. S.; Kuhn, S. J.; Moffatt, M. E.; Bastien,
I. J.; Baker, E. B. Stable Carbonium Ions. IV. Secondary and Tertiary
Alkyl and Aralkyl Oxocarbonium Hexafluoroantimonates. Formation
and Identification of the Trimethylcarbonium Ion by Decarbonylation
of the tert-Butyl Oxocarbonium Ion. J. Am. Chem. Soc. 1963, 85,
1
(
328−1334.
9) Kato, T.; Reed, C. A. Putting tert-butyl cation in a bottle. Angew.
Chem., Int. Ed. 2004, 43, 2908−2911.
10) Yoder, R. A.; Johnston, J. N. A case study in biomimetic total
(26) Kim, B.; Sohn, T. I.; Kim, D.; Paton, R. S. Asymmetric total
syntheses and structure confirmation of chlorofucins and bromofu-
cins. Chem. - Eur. J. 2018, 24, 2634−2642.
(27) Shepherd, E. D.; Dyson, B. S.; Hak, W. E.; Nguyen, Q. N. N.;
Lee, M.; Kim, M. J.; Sohn, T. I.; Kim, D.; Burton, J. W.; Paton, R. S.
Structure Determination of a Chloroenyne from Laurencia majuscula
Using Computational Methods and Total Synthesis. J. Org. Chem.
2019, 84, 4971−4991.
(
synthesis: polyolefin carbocyclizations to terpenes and steroids. Chem.
Rev. 2005, 105, 4730−4756.
(
11) Fukuzawa, A.; Aye, M.; Nakamura, M.; Tanura, M.; Murai, A.
Structure elucidation of laureoxanyne, a new nonisoprenoid C-15-
enyne, using lactoperoxidase. Tetrahedron Lett. 1990, 31, 4895−4898.
(
12) Kikuchi, H.; Suzuki, T.; Kurosawa, E.; Suzuki, M. The Structure
of Notoryne, a Halogenated C15 Nonterpenoid with a Novel Carbon
Skeleton from The Red Alga Laurencia Nipponica Yamada. Bull. Chem.
Soc. Jpn. 1991, 64, 1763−1775.
(28) Snyder, S. A.; Brucks, A. P.; Treitler, D. S.; Moga, I. Concise
synthetic approaches for the Laurencia family: formal total syntheses
of (±)-laurefucin and (±)-E- and (±)-Z-pinnatifidenyne. J. Am.
Chem. Soc. 2012, 134, 17714−17721.
(
13) Murai, A. Biosynthesis of Cyclic Bromoethers from Red Algae.
In Comprehensive Natural Products Chemistry; Barton, D. H. R., Meth-
Cohn, O., Nakinishi, K., Eds.; Elsevier: Oxford, U.K., 1999; Vol. 1, pp
(29) Kim, D.; Sohn, T.-i.; Kim, B.; Paton, R. S. Asymmetric Total
Synthesis and Structure Confirmation of (+)-(3E)-Isolaurefucin
Methyl Ether. Heterocycles 2018, 97, 179−191.
3
(
03−324.
14) Braddock, D. C. A hypothesis concerning the biosynthesis of
(30) Braddock, D. C.; Millan, D. S.; Perez-Fuertes, Y.; Pouwer, R.
H.; Sheppard, R. N.; Solanki, S.; White, A. J. P. Bromonium Ion
Induced Transannular Oxonium Ion Formation-Fragmentation in
Model Obtusallene Systems and Structural Reassignment of
Obtusallenes V-VII. J. Org. Chem. 2009, 74, 1835−1841.
the obtusallene family of marine natural products via electrophilic
bromination. Org. Lett. 2006, 8, 6055−6058.
(
15) Bonney, K. J.; Braddock, D. C. A unifying stereochemical
analysis for the formation of halogenated C15-acetogenin medium-
ring ethers from Laurencia species via intramolecular bromonium ion
assisted epoxide ring-opening and experimental corroboration with a
model epoxide. J. Org. Chem. 2012, 77, 9574−9584.
(31) Stoyanov, E. S.; Gunbas, G.; Hafezi, N.; Mascal, M.; Stoyanova,
+
+
I. V.; Tham, F. S.; Reed, C. A. The R O •H hydrogen bond: toward
3
a tetracoordinate oxadionium(2+) ion. J. Am. Chem. Soc. 2012, 134,
707−714.
(
16) Braddock, D. C.; Sbircea, D. T. Proof-of-principle direct double
cyclisation of a linear C15-precursor to a dibrominated bicyclic
medium-ring ether relevant to Laurencia species. Chem. Commun.
(32) Fujiwara, K. Total Synthesis of Medium-Ring Ethers from
Laurencia Red Algae. In Marine Natural Products; Kiyota, H., Ed.;
Springer-Verlag: Berlin Heildelberg, 2006; Vol. 5, pp 97−148.
(33) Murai, A.; Murase, H.; Matsue, H.; Masamune, T. The
synthesis of (±)-Laurencin. Tetrahedron Lett. 1977, 18, 2507−2510.
(34) Masamune, T.; Matsue, H.; Murase, H. Synthetic Studies of
Laurencin and Related Compounds. IV. Synthesis of cis-2-Ethyl-8-
formyl-3,4,7,8-dihydro-2H-oxocin-3-one 3-Ethylene Acetal and Re-
lated Compounds. Bull. Chem. Soc. Jpn. 1979, 52, 127−134.
(35) Masamune, T.; Murase, H.; Matsue, H.; Murai, A. Synthetic
Studies of Laurencin and Related Compounds. V. Transformation of
cis-2-Ethyl-8-formyl-3,4,7,8-dihydro-2H-oxocin-3-one 3-Ethylene Ace-
tal into (±)-Laurencin. Bull. Chem. Soc. Jpn. 1979, 52, 135−141.
(36) For reviews covering the synthesis of halogenated cyclic ether
acetogenin natural products from Laurencia spp. see refs 22a and 32.
2
(
014, 50, 12691−12693.
17) Snyder, S. A.; Treitler, D. S.; Brucks, A. P.; Sattler, W. A general
strategy for the stereocontrolled preparation of diverse 8- and 9-
membered Laurencia-type bromoethers. J. Am. Chem. Soc. 2011, 133,
1
(
5898−15901.
18) Zhang, Y. A.; Yaw, N.; Snyder, S. A. General Synthetic
Approach for the Laurencia Family of Natural Products Empowered
by a Potentially Biomimetic Ring Expansion. J. Am. Chem. Soc. 2019,
1
(
41, 7776−7788.
19) Keshipeddy, S.; Martinez, I.; Castillo, B. F., 2nd; Morton, M.
D.; Howell, A. R. Toward a formal synthesis of laureatin: unexpected
rearrangements involving cyclic ether nucleophiles. J. Org. Chem.
2012, 77, 7883−7890.
J
DOI: 10.1021/jacs.9b07438
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
(
(
37) For biomimetic syntheses see refs 15−21 and 24−30.
4582. Reisolation: (c) Wratten, S. J.; Faulkner, D. J. Metabolites of
the red alga Laurencia subopposita. J. Org. Chem. 1977, 42, 3343−
3349. Previous synthesis see refs 24 and 28.
(49) For isolation and structure determination of acetyllaurefucin
see refs 48a−c.
38) For selected recent non-biomimetic syntheses see: (a) Kim, B.;
Sohn, T. I.; Kim, S.; Kim, D.; Lee, J. Concise Substrate-Controlled
Asymmetric Total Synthesis of (+)-3-(Z)-Dihydrorhodophytin.
Heterocycles 2010, 82, 1113−1118. (b) Kim, M. J.; Sohn, T. I.;
Kim, D.; Paton, R. S. Concise substrate-controlled asymmetric total
syntheses of dioxabicyclic marine natural products with 2,10-
dioxabicyclo-[7.3.0]dodecene and 2,9-dioxabicyclo[6.3.0]undecene
skeletons. J. Am. Chem. Soc. 2012, 134, 20178−20188. (c) Sohn, T.
I.; Kim, D.; Paton, R. S. Substrate-Controlled Asymmetric Total
Syntheses of Microcladallenes A, B, and C Based on the Proposed
Structures. Chem. - Eur. J. 2015, 21, 15988−15997. (d) Yoshimura, F.;
Okada, T.; Tanino, K. Asymmetric Total Synthesis of Laurallene. Org.
Lett. 2019, 21, 559−562. Ref 27.
(50) Bromofucins. Isolation: (a) Coll, J. C.; Wright, A. D. Tropical
Marine-Algae. IV. Novel Metabolites from the Red Alga Laurencia
implicata (Rhodophyta, Rhodophyceae, Ceramiales, Rhodomelaceae).
Aust. J. Chem. 1989, 42, 1685−1693. (b) McPhail, K. L.; Davies-
Coleman, M. T. (3Z)-bromofucin from a South African sea hare. Nat.
Prod. Res. 2005, 19, 449−452. (c) Suzuki, M.; Takahashi, Y.; Matsuo,
Y.; Masuda, M. Pannosallene, a brominated C-15 nonterpenoid from
Laurencia pannosa. Phytochemistry 1996, 41, 1101−1103. Synthesis:
ref 25.
(51) Chlorofucins. Isolation: (a) Howard, B. M.; Schulte, G. R.;
Fenical, W.; Solheim, B.; Clardy, J. Three new vinyl acetylenes from
the marine red alga. Tetrahedron 1980, 36, 1747−1751. (b) Denys, R.;
Coll, J. C.; Carroll, A. R.; Bowden, B. F. Tropical marine-algae. X.
Isolaurefucin methyl-ether, a new lauroxocane derivative from the red
alga Dasyphila-Plumariodes. Aust. J. Chem. 1993, 46, 1073−1077.
(c) Suzuki, M.; Daitoh, M.; Vairappan, C. S.; Abe, T.; Masuda, M.
Novel halogenated metabolites from the Malaysian Laurencia pannosa.
J. Nat. Prod. 2001, 64, 597−602. Synthesis: ref 25.
(52) Elatenynes. Isolation and structure determination: (a) Hall, J.
G.; Reiss, J. A. Elatenyne − a Pyrano[3,2-b]pyranyl Vinyl Acetylene
from the Red Alga. Aust. J. Chem. 1986, 39, 1401−1409. (b) Kim, I.
K.; Brennan, M. R.; Erickson, K. L. Lauroxolanes from the marins alga
Laurencia Majuscula. Tetrahedron Lett. 1989, 30, 1757−1760.
(c) Sheldrake, H. M.; Jamieson, C.; Burton, J. W. The changing
faces of halogenated marine natural products: Total synthesis of the
reported structures of elatenyne and an enyne from Laurencia
majuscula. Angew. Chem., Int. Ed. 2006, 45, 7199−7202. (d) Smith, S.
G.; Paton, R. S.; Burton, J. W.; Goodman, J. M. Stereostructure
assignment of flexible five-membered rings by GIAO (13)C NMR
calculations: Prediction of the stereochemistry of elatenyne. J. Org.
Chem. 2008, 73, 4053−4062. (e) Dias, D. A.; Urban, S.
Phytochemical studies of the southern Australian marine alga,
Laurencia elata. Phytochemistry 2011, 72, 2081−2089. Ref 17.
(f) Urban, S.; Brkljaca, R.; Hoshino, M.; Lee, S.; Fujita, M.
Determination of the Absolute Configuration of the Pseudo-
Symmetric Natural Product Elatenyne by the Crystalline Sponge
Method. Angew. Chem., Int. Ed. 2016, 55, 2678−2682. Synthesis: ref
20.
(
39) For earlier notable contributions to the synthesis of
halogenated ether acetogenins from Laurencia spp. see: (a) Overman,
L. E.; Thompson, A. S. Total synthesis of (−)-laurenyne. Use of
acetal-initiated cyclizations to prepare functionalized eight-membered
cyclic ethers. J. Am. Chem. Soc. 1988, 110, 2248−2256. (b) Tsushima,
K.; Murai, A. Total synthesis of (+)-laurencin. Tetrahedron Lett. 1992,
3
3, 4345−4348. (c) Lee, E.; Park, C. M.; Yun, J. S. Total Synthesis of
Dactomelynes. J. Am. Chem. Soc. 1995, 117, 8017−8018. (d) Burton,
J. W.; Clark, J. S.; Derrer, S.; Stork, T. C.; Bendall, J. G.; Holmes, A. B.
Synthesis of medium ring ethers. 5. The synthesis of (+)-laurencin. J.
Am. Chem. Soc. 1997, 119, 7483−7498. (e) Crimmins, M. T.; Choy,
A. L. An Asymmetric Aldol-Ring-Closing Metathesis Strategy for the
Enantioselective Construction of Oxygen Heterocycles: An Efficient
Approach to the Enantioselective Synthesis of (+)-Laurencin. J. Am.
Chem. Soc. 1999, 121, 5653−5660. (f) Fujiwara, K.; Awakura, D.;
Tsunashima, M.; Nakamura, A.; Honma, T.; Murai, A. Total Synthesis
of (+)-Obtusenyne. J. Org. Chem. 1999, 64, 2616−2617. (g) Kim, H.;
Choi, W. J.; Jung, J.; Kim, S.; Kim, D. Construction of eight-
membered ether rings by olefin geometry-dependent internal
alkylation: First asymmetric total syntheses of (+)-3-(E)- and (+)-3-
(
E)-pinnatifidenyne. J. Am. Chem. Soc. 2003, 125, 10238−10240.
40) Fukuzawa, A.; Aye, M.; Murai, A. A Direct Enzymatic Synthesis
of Laurencin from Laurediol. Chem. Lett. 1990, 19, 1579−1580.
41) Kurosawa, E.; Irie, T.; Fukuzawa, A. trans-Laurediol and cis-
(
(
laurediol, unsaturated glycols from Laurencia Nipponica yamada.
Tetrahedron Lett. 1972, 13, 2121−2124.
(
42) Fukuzawa, A.; Honma, T.; Takasugi, Y.; Murai, A. Biogenetic
intermediates, (3E and 3Z,12Z)-laurediols and (3E and 3Z)-12,13-
dihydrolaurediols, isolated from Laurencia Nipponica. Phytochemistry
1
(
993, 32, 1435−1438.
(53) Laurendecumenyne B. Isolation and structure determination:
(a) Ji, N. Y.; Li, X. M.; Li, K.; Wang, B. G. Laurendecumallenes A-B
and laurendecumenynes A-B, halogenated nonterpenoid C-15-
Acetogenins from the marine red alga Laurencia decumbens. J. Nat.
Prod. 2007, 70, 1499−1502. (b) Ji, N. Y.; Li, X. M.; Li, K.; Wang, B.
G. J. Nat. Prod. 2007, 70, 1499; J. Nat. Prod. 2010, 73, 1192. (c) Ref
20. Synthesis ref 20.
43) For reviews of halogenating enzymes see: (a) Butler, A.; Carter-
Franklin, J. N. The role of vanadium bromoperoxidase in the
biosynthesis of halogenated marine natural products. Nat. Prod. Rep.
2
004, 21, 180−188. (b) Vaillancourt, F. H.; Yeh, E.; Vosburg, D. A.;
Garneau-Tsodikova, S.; Walsh, C. T. Nature’s inventory of
halogenation catalysts: Oxidative strategies predominate. Chem. Rev.
2
(
006, 106, 3364−3378.
(54) Schulte, G. R.; Chung, M. C. H.; Scheuer, P. J. Two bicyclic
C15 enynes from the sea hare Aplysia oculifera. J. Org. Chem. 1981, 46,
3870−3873.
44) Kaneko, K.; Washio, K.; Umezawa, T.; Matsuda, F.; Morikawa,
M.; Okino, T. cDNA cloning and characterization of vanadium-
dependent bromoperoxidases from the red alga Laurencia nipponica.
Biosci., Biotechnol., Biochem. 2014, 78, 1310−1319.
(55) Jeong, D.; Sohn, T. I.; Kim, J. Y.; Kim, G.; Kim, D.; Paton, R. S.
Construction of 6,10-syn- and -anti-2,5-dioxabicyclo[2.2.1]heptane
skeletons via oxonium ion formation/fragmentation: prediction of
structure of (E)-ocellenyne by NMR calculation. Org. Lett. 2017, 19,
6252−6255.
(
45) Davies, S. G.; Polywka, M. E. C.; Thomas, S. E. Stereoselective
synthesis of cyclic ethers via bromine assisted epoxide ring expansion.
Tetrahedron Lett. 1985, 26, 1461−1464.
(
46) Suzuki proposed a different route from laurediol to
(56) Olah, G. A.; Laali, K. K.; Wang, Q.; Prakash, G. K. S. Onium
Ions; Wiley: New York, 1998.
prelaurefucin; see ref 12.
47) For previous syntheses of notoryne see: Senapati, S.; Das, S.;
Ramana, C. V. Total Synthesis of Notoryne. J. Org. Chem. 2018, 83,
2863−12868 and ref 27 .
48) Laurefucin. Isolation and structure determination: (a) Fukuza-
(
(57) Krossing, I. The facile preparation of weakly coordinating
anions: structure and characterisation of silverpolyfluoroalkoxyalumi-
nates AgAl(OR ) , calculation of the alkoxide ion affinity. Chem. - Eur.
1
(
F
4
J. 2001, 7, 490−502.
wa, A.; Kurosawa, E.; Irie, T. Laurefucin and acetyllaurefucin, new
bromo compounds from Laurencia Nipponica yamada. Tetrahedron
Lett. 1972, 13, 3−6. (b) Furusaki, A.; Kurosawa, E.; Fukuzawa, A.;
Irie, T. The revised structure and absolute configuration of Laurefucin
from Laurencia Nipponica Yamada. Tetrahedron Lett. 1973, 14, 4579−
(58) Braddock, D. C.; Hermitage, S. A.; Kwok, L.; Pouwer, R.;
Redmond, J. M.; White, A. J. The generation and trapping of
enantiopure bromonium ions. Chem. Commun. 2009, 1082−1084.
(59) The relative configuration of the dioxabicyclo[5.2.1]decanes
28, ent-(E/Z)-29, ent-15, ent-(Z)-16, (E/Z)-19, (E/Z)-24 and the
K
DOI: 10.1021/jacs.9b07438
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
oxonium ions ent-(E/Z)-13·Al(pftb) and (E/Z)-20·Al(pftb) , 46·
4
4
1
Al(pftb) were assigned by H NMR NOE experiments including
4
ROSEY where necessary (Supporting Information section 14, pp
S77−S84). For the natural products (E/Z)-19, (E/Z)-24, ent-(E)-16,
ent-(E/Z)-14, (Z)-21, (E/Z)-22, and α,β-unsaturated aldehydes 36
and 40 comparison with literature data confirmed their structures
(
Supporting Information, pp S26−S28 and sections 10 and 11, pp
S60−S76).
60) Stork, G.; Zhao, K. A stereoselective synthesis of (Z)-1-iodo-1-
alkenes. Tetrahedron Lett. 1989, 30, 2173−2174.
61) Voigtritter, K.; Ghorai, S.; Lipshutz, B. H. Rate enhanced olefin
cross-metathesis reactions: the copper iodide effect. J. Org. Chem.
011, 76, 4697−4702.
62) Etzkorn, M.; Aniszfeld, R.; Li, T.; Buchholz, H.; Rasul, G.;
Prakash, G. K. S.; Olah, G. A. 1-Oxoniaadamantane. Eur. J. Org. Chem.
008, 2008, 4555−4558.
63) Mascal, M.; Hafezi, N.; Meher, N. K.; Fettinger, J. C.
(
(
2
(
2
(
Oxatriquinane and oxatriquinacene: extraordinary oxonium ions. J.
Am. Chem. Soc. 2008, 130, 13532−13533.
(
64) Molecular mechanics conformational searches were performed
with Spartan16; Wavefunction, Inc.: Irvine, CA. DFT optimizations
with several functionals were performed with Gaussian09, rev. D.01;
Gaussian, Inc.: Wallingford, CT, 2016. Computational methods and
references are described in full in the Supporting Information.
(
65) For related 1,2-hydride shifts involving oxonium ions see:
Sugimoto, M.; Suzuki, T.; Hagiwara, H.; Hoshi, T. The first total
synthesis of (+)-(Z)-laureatin. Tetrahedron Lett. 2007, 48, 1109−1112
and refs 17 and 24 .
(
66) A referee suggested the oxocarbenium ion may be formed by
elimination to give an enol ether followed by C-protonation. We
cannot rule out this mechanism as it would have the same kinetic
profile as the proposed 1,2-hydride shift if enol ether protonation was
fast with respect to elimination. We favor the illustrated 1,2-hydride
shift due to precedent in related systems (ref 65) and the
computational analysis (Figure 4).
(
67) Sohn, T.; Kim, M. J.; Kim, D. Asymmetric Total Synthesis of
Trilobacin via Organoselenium-Mediated Oxonium Ion Formation/
SiO -Promoted Fragmentation. J. Am. Chem. Soc. 2010, 132, 12226−
2
1
2227.
(
68) The DFT calculations on the opening of oxonium ions ent-(E)-
3 and (E)-20 expand upon the recently reported studies of the
opening of closely related oxonium ions (truncated versions of 13 and
0) with water (ref 29). The current study (Figures 5 and S36−S48)
1
2
uses water and chloride as nucleophiles, on the complete substrates, at
several levels of theory, with distortion-interaction/activation-strain
analysis to analyze the preference for opening at C-7, C-10, and C-13
with both ent-(E)-13 and (E)-20.
(
69) Bickelhaupt, F. M.; Houk, K. N. Analyzing Reaction Rates with
the Distortion/Interaction-Activation Strain Model. Angew. Chem., Int.
Ed. 2017, 56, 10070−10086.
(
70) Tantillo, D. J. Importance of Inherent Substrate Reactivity in
Enzyme-Promoted Carbocation Cyclization/Rearrangements. Angew.
Chem., Int. Ed. 2017, 56, 10040−10045.
L
DOI: 10.1021/jacs.9b07438
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX