Total Synthesis of the Alleged Structure of Crenarchaeol Enables Structure Revision**

Article abstract of DOI:10.1002/anie.202105384

Crenarchaeol is a glycerol dialkyl glycerol tetraether lipid produced exclusively in Archaea of the phylum Thaumarchaeota. This membrane-spanning lipid is undoubtedly the structurally most sophisticated of all known archaeal lipids and an iconic molecule in organic geochemistry. The 66-membered macrocycle possesses a unique chemical structure featuring 22 mostly remote stereocenters, and a cyclohexane ring connected by a single bond to a cyclopentane ring. Herein we report the first total synthesis of the proposed structure of crenarchaeol. Comparison with natural crenarchaeol allowed us to propose a revised structure of crenarchaeol, wherein one of the 22 stereocenters is inverted.

Full text of DOI:10.1002/anie.202105384

A Journal of the Gesellschaft Deutscher Chemiker
Angewandte
Chemie
International Edition
www.angewandte.org
Accepted Article
Title: Total Synthesis of the Alleged Structure of Crenarchaeol enables
Structure Revision
Authors: Mira Holzheimer, Jaap S. Sinninghe Damsté, Stefan
Schouten, Remco W. A. Havenith, Ana V. Cunha, and
Adriaan J. Minnaard
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: Angew. Chem. Int. Ed. 10.1002/anie.202105384
Link to VoR: https://doi.org/10.1002/anie.202105384

10.1002/anie.202105384
Angewandte Chemie International Edition
RESEARCH ARTICLE
Total Synthesis of the Alleged Structure of Crenarchaeol enables
Structure Revision
Mira Holzheimer,[a] Jaap S. Sinninghe Damsté,[e],[f] Stefan Schouten,[e],[f] Remco W. A. Havenith,[a],[b],[c]
Ana V. Cunha,[d] and Adriaan J. Minnaard[a]*
Dedication ((optional))
[a]
[b]
[c]
[d]
[e]
[f]
Dr. Mira Holzheimer, Prof. Dr. Remco W. A. Havenith, Prof. Dr. Adriaan J. Minnaard
Stratingh Institute for Chemistry
University of Groningen
Nijenborgh 7, 9747 AG, Groningen, The Netherlands
Prof. Dr. Remco W. A. Havenith
Zernike Institute for Advanced Materials
University of Groningen
Nijenborgh 4, 9747 AG, Groningen, The Netherlands
Prof. Dr. Remco W. A. Havenith
Ghent Quantum Chemistry Group, Department of Chemistry
Ghent University
Krijgslaan 281 (S3), B-9000 Gent, Belgium
Dr. Ana V. Cunha
Eenheid Algemene Chemie (ALGC)
Vrije Universiteit Brussel (VUB)
Pleinlaan 2, 1050 Brussels, Belgium
Prof. Dr. Jaap S. Sinninghe Damsté, Prof. Dr. Stefan Schouten
NIOZ Royal Netherlands Institute for Sea Research
Department of Marine Microbiology and Biogeochemistry
PO Box 59, 1790 AB Den Burg, The Netherlands
Prof. Dr. Jaap S. Sinninghe Damsté, Prof. Dr. Stefan Schouten
Faculty of Geosciences, Department of Earth Sciences
Utrecht University
PO Box 80.021, 3508 TA Utrecht, The Netherlands
Supporting information for this article is given via a link at the end of the document.
Abstract: Crenarchaeol is a glycerol dialkyl glycerol tetraether lipid
produced exclusively in Archaea of the phylum Thaumarchaeota. This
membrane-spanning lipid is undoubtedly the structurally most
sophisticated of all known archaeal lipids and an iconic molecule in
organic geochemistry. The 66-membered macrocycle possesses a
unique chemical structure featuring 22 mostly remote stereocenters,
and a cyclohexane ring connected by a single bond to a cyclopentane
ring. Herein we report the first total synthesis of the proposed structure
of crenarchaeol. Comparison with natural crenarchaeol allowed us to
propose a revised structure of crenarchaeol, wherein one of the 22
stereocenters is inverted.
chains, contrary to the straight chain fatty acid glycerol ester lipids
found in Bacteria and Eukarya.[5] Apart from the difference in lipid
linkage, the stereochemistry of the glycerol backbone in archaeal
Introduction
Figure 1 The alleged structure of crenarchaeol.
In 1990, Woese proposed to classify all living organisms in three
domains of life: Archaea, Bacteria and Eukarya.[1] Before that,
‘archaeabacteria’ were considered to belong to the Bacteria.
Based on differences in their genome and lipidome, Archaea were
ultimately recognized as separate, third domain.[2] For a long time,
Archaea were primarily associated with extreme habitats such as
high temperature, extreme pH, and hypersaline environments.[3]
Growing interest over the years, however, led to the discovery of
meso- and extremophilic Archaea in virtually any habitat on
Earth.[4] The cell membrane of Archaea is built up of diether or
membrane-spanning tetraether lipids containing isoprenoid
isoprenoidal glycerol dialkyl glycerol tetraether lipids (GDGTs) is
opposite to bacterial or eukaryotic glycerolipids, raising questions
on the evolution of archaeal and bacterial/eukaryotic lipid
membranes.[6] The lipid composition of Archaea varies,
depending on the species and environmental factors, and this is
considered an adaptation to their habitat.[7] The ether-linkages
provide chemical stability against hydrolysis, and the presence of
methyl-branches and cyclopentane moieties, which are formed by
1
This article is protected by copyright. All rights reserved.

10.1002/anie.202105384
Angewandte Chemie International Edition
RESEARCH ARTICLE
Scheme 1 Retrosynthetic analysis of crenarchaeol.
internal cyclization of the biphytanol chain,[8] leads to decreased
membrane permeability, allowing growth at extreme pH, salinity,
and temperature.[9] One archaeal GDGT – named crenarchaeol –
stands out from all other archaeal membrane lipids due to its
unique chemical structure (Figure 1). Crenarchaeol is produced
by a specific lineage of Archaea, the Thaumarchaeota,[10] and was
first isolated from surface sediments of the Arabian Sea. After
extensive GC-MS and NMR analysis, the structure and
Results and Discussion
Our retrosynthetic analysis of crenarchaeol made use of the
inherent symmetry of the bicyclic biphytanyl chain of the molecule
(Scheme 1). It started with the disconnection of the central C–C-
bond of the bicyclic biphytanyl moiety by intramolecular alkene
metathesis and ether bond disconnection of 1. This led to two key
intermediates, termed Fragment A and B, and protected glycerol
building block 2. Fragment A can be further simplified via dithiane
disconnections to arrive at building blocks 4 and 6, both carrying
a methyl-branched stereocenter, and cyclopentane building block
5. In turn, 5 can be traced back to hydroxyketone 7, which is
accessed from commercially available (S)-carvone via ring
contraction. Syntheses of archaeal cis-[15] and trans-substituted[14]
cyclopentane containing lipids have been previously reported. As
we planned to build the macrocycle by alkylation of a suitably
functionalized glycerol building block and ring-closing metathesis,
we required differentially protected lipid chains containing the
trans-substituted cyclopentane and the methyl-branches. Based
on the stereochemical assessment of the bicyclic biphytane
moiety in crenarchaeol17 and its subsequent confirmation
provided by Helmchen et al.22, we planned the synthesis of the
desired stereoisomer.
stereochemistry of this unique GDGT was proposed,
a
considerable achievement given the fact that the molecular
complexity originates merely from its unusual hydrocarbon
framework.[11] It contains four 1,3-trans-substituted cyclopentane
moieties. One of these is connected by a single bond to a
cyclohexane ring, a structural feature rarely found in natural
products.[12] This feature of crenarchaeol is likely formed by
further internal cyclization of the bicyclic biphytanyl moiety.[5b]
Crenarchaeol contains a total of 22 stereocenters, most of which
are remote, including an all-carbon quaternary stereocenter.
Recently, a parallel
glycerol configuration of sedimentary
crenarchaeol was inferred from chemical derivatization
experiments.[13] Montenegro et al. confirmed the structure of the
bicyclic biphytanyl moiety in archaeal GDGTs by total
synthesis,[14] yet to date, there is no proof of structure of the
tricyclic biphytanyl moiety of crenarchaeol and no total synthesis.
The 5-6 ring motif of crenarchaeol is particularly interesting due to
its complexity and uniqueness in nature. In order to ultimately
confirm the structure and stereochemistry of crenarchaeol, we
embarked on its total synthesis.
Retrosynthesis of Fragment B commenced with the C–C-bond
disconnection of 8 arriving at dithiane 10 and iodide 9, the latter
originating from Fragment A. Further simplification of 10 by
asymmetric Cu-catalyzed Grignard alkylations and a Wittig
olefination delivered diacetate 11. The 5-6 ring motif of 11 was
disconnected at the C–C-bond joining the two carbocycles.[16]
2
This article is protected by copyright. All rights reserved.

10.1002/anie.202105384
Angewandte Chemie International Edition
RESEARCH ARTICLE
1. NaH
CS2
CH3I
H
nOe
Hd
H
ethylene glycol
CH(OMe)3
PPTS
O
O
OH
O
O
OH
O
O
O
Hc
OH
ref. 27 (see SI)
THF, 0 ºC
H
Ha
Recycling
1. PPh3
I2
43%
over 4 steps
2. AIBN
n-Bu3SnH
benzene, 75 ºC
96% over 2 steps
PhCH3, 80 ºC
Hb
H
92%
H
7
15
16
imidazole
CH2Cl2, 0 ºC to rt
2. DBU
H
(S)-carvone
BH3·SMe2
THF, 0 ºC
then NaOH, H2O2
THF, 0 ºC to rt
dr 1:1
CH3CN, 80 ºC
77%
over 2 steps
O
1. PhNTf2
KHMDS
O
O
O
O
O
O
TBDPSCl
imidazole
DMAP
FeCl3·SiO2
THF, -78 ºC
+
OH
OH
OTBDPS
OTBDPS
OTBDPS
H
H
H
H
H
acetone, rt, 12 min
CH2Cl2, 0 ºC to rt
2. Pd(PPh3)4
LiCl
n-Bu3SnH
THF, rt
quant.
quant.
18
19
20
17
43%
epi-17
80% over 2 steps
9-BBN
dithiane 4
t-BuLi
THF/HMPA, -78 ºC
THF, 0 ºC
then NaOH
H2O2
HO
CBr4
PPh3
Br
S
S
BnO
TBAF
CH2Cl2, 0 ºC
then bromide 5
THF/HMPA, -78 ºC
87%
THF, rt
quant.
OTBDPS
THF, 0 ºC to rt
OTBDPS
OTBDPS
H
H
H
97%
87%
22
21
5
dithiane 6
I2
n-BuLi
THF, 0 ºC
imidazole
PPh3
S
S
S
S
S
S
BnO
BnO
BnO
OTBDPS
Raney Ni
H2
THF, 65 ºC
86%
then iodide 9
THF, 0 ºC
68%
CH2Cl2, 0 ºC
OH
I
H
H
H
97%
S
S
3
23
9
HO
S
OTBDPS
S
S
H
BnO
S
OTBDPS
H
4
6
Fragment A
Scheme 2 Synthesis of Fragment A.
We realized that for this challenging transformation an advanced
intermolecular Pd-catalyzed asymmetric allylic alkylation could be
instrumental, inspired by the work of Trost.[17] By this, we arrived
at building blocks 13 and 14, readily accessible from pimelic acid
and cyclopentadiene, respectively.
Table 1. Optimization of the Pd-catalyzed allylic alkylation.
Synthesis of Fragment A
The synthesis of Fragment A was initiated by the preparation of
Entry[a]
Ligand
Base
Solvent[c]
Conversion[b]
(yield) [c]
dr[d]
known -hydroxyketone 7 from (S)-carvone (Scheme 2). Via a
four-step sequence involving a hydrolytic ring contraction,[18]
7
was obtained as single diastereomer, as confirmed by NOESY
Notably, this sequence proved robust and scalable and allowed
multigram synthesis of 7 (see SI). After acetal protection of 7, the
hydroxyl group of 15 was removed by Barton-McCombie
deoxygenation, providing 16 in excellent yield. Notably, acetal
protection was necessary to avoid elimination of the -hydroxyl
group in the synthesis of the xanthate intermediate. Initially, we
envisioned to stereoselectively install the methyl stereocenter
adjacent to the 5-membered ring by means of Cu- or Co-catalyzed
asymmetric hydroboration.[19] No published method to perform the
asymmetric hydroboration of the 1,1-disubstituted terminal alkene
of 16 delivered 17 in acceptable yield and stereoselectivity,
however. Thus, we resorted to non-stereoselective
1
L1
L2
L3
L4
L4
L4
L4
L4
L4
LHMDS
LHMDS
LHMDS
LHMDS
LHMDS
LHMDS
NaHMDS
LDA
THF
80%
25:75
51:49
81:19
86:14
85:15
94:6
2
THF
40%
3
THF
40%
4
THF
40% (27%)
40%
5
PhCH3
DME
DME
DME
DME
6
42%
7
10-15%
41%
92:8
8
92:8
9[e]
LHMDS
full (53%)
93:7
hydroboration-oxidation of 16 followed by diastereomer
separation, giving 17 in 43% yield as single stereoisomer. The
stereochemistry of the methyl-branched center in 17 was
determined by amidation of its corresponding acid with
phenylglycine methyl ester, followed by 1H NMR analysis
[a] See SI for details. [b] Determined by 1H NMR. [c] Isolated yield.
[d] Determined by 13C NMR of the crude product. [e] 1.6 eq. of LHMDS and 3 eq.
LiCl were used.
3
This article is protected by copyright. All rights reserved.

10.1002/anie.202105384
Angewandte Chemie International Edition
RESEARCH ARTICLE
TBDPSCl
imidazole
DMAP
Building block synthesis
PPh2
O
Ph2P
O
Amano lipase
OAc
OH
OTBDPS
phosphate buffer, rt
94%, 99% ee
CH2Cl2, 0 ºC to rt
Ligands for
Pd-catalyzed
intermolecular
allylic alkylation
AcO
AcO
AcO
O
HN
24
83%
14
25
NH
2 steps from
cyclopentadiene
PPh2
N
1. ethylene glycol
PPTS
benzene, reflux
1. NaH
BnBr
THF, 0 ºC to rt
L1
L2
O
BnO
O
HO
O
O
O
O
O
O
2. 9-BBN
THF, 0 ºC to rt
then NaOH, H2O2
THF, 0 ºC to rt
2. p-TsOH
H2O, acetone, 50 ºC
92% over 2 steps
NH HN
NH HN
26
27
13, 99% ee
5 steps from
pimelic acid
PPh2 Ph2P
PPh2 Ph2P
88% over 2 steps
L4
L3
Fragment B synthesis
BnO
O
Pd2(dba)3CHCl3 (5 mol%)
nOe
L4 (11 mol%)
LHMDS
LiCl
1.TBAF/AcOH
THF, rt
1. DIBAL-H
CH2Cl2, -78 ºC
BnO
O
BnO
AcO
H
BnO
AcO
H
H
OTBDPS
H
OTBDPS
OAc
+
27
OTBDPS
H
H
2. Ac2O
pyridine
DMAP
2. Ac2O
DME, 0 ºC
64%
dr >20:1
pyridine, DMAP
CH2Cl2, 0 ºC to rt
83% over two steps
30
31
11
AcO
CH2Cl2, 0 ºC to rt
95% over two steps
14
1. KHMDS
CS2
THF, -78 ºC,
then CH3I, 0 ºC
Grignard 32
CuI (15 mol%)
THF, 0 ºC
1. Flavin-cat. 34
H2N-NH2·H2O
EtOH, rt
BnO
AcO
BnO
BnO
HO
H
H
H
MOMO
MOMO
MOMO
then 11
THF, -20 ºC
75%
2.LiAlH4
THF, 0 ºC to rt
98% over two steps
2. n-Bu3SnH
AIBN
PhCH3, 90 ºC
H
H
H
33
36
35
93% over two steps
Pd/C
Pd(OH)2
H2 (1 atm.)
THF, rt
1. (COCl)2
DMSO
96%
L5·CuBr (8 mol%)
Et3N
CH2Cl2, -60 ºC
H
H
HO
MeMgBr
MOMO
MOMO
H
EtS
EtS
MOMO
H
H
MTBE, -78 ºC
87%
dr >20:1
2. 38
CH2Cl2, 40 ºC
85% over two steps
H
O
O
39
1. HCl, THF, rt
2. DIBAL-H
37
40
CH2Cl2, -78 ºC
3. BF3·OEt2
1,3-propanedithiol
CH2Cl2, -78 ºC
82% over three steps
OH
S
S
S
S
BnO
H
t-BuLi
BnO
HO
+
H
H
S
H
THF/HMPA, -78 ºC
I
H
H
67%
10
S
9
S
S
8
O
O
O
1. Raney Nickel
H2 (1 atm.)
THF, 60 ºC
BrMg
OH
MOMO
32
Cy
O
H
HO
P
H
H
N
N
N
O
NH
Cy
P
2. Pd/C
Pd(OH)2
H2 (1 atm.)
THF, rt
Fe
O
H
PPh3
O
EtS
Fragment B
42%
Flavin-cat. 34
(R,S)-Josiphos L5
38
Scheme 3 Synthesis of Fragment B.
Synthesis of Fragment B
stage was set for the first dithiane alkylation.[22] After optimization
of the lithiation conditions of 4 (prepared using known methods,
see SI), the alkylation proceeded in high yield (87%) giving 22.
Desilylation followed by Appel iodination delivered iodide 9, which
serves as intermediate in the synthesis of both Fragment A and
B. In turn, after identification of the optimal lithiation conditions,
deprotonation of dithiane 6 with n-BuLi at 0 ºC followed by addition
of 9 produced bis-dithiane 3 in 68% yield. With the carbon
skeleton of Fragment A constructed, the dithiane moieties and the
benzyl ether of 3 were removed by Raney-nickel reduction in good
yield, thus concluding the synthesis of Fragment A.
(see SI).[20] In addition, the efficiency of the synthesis was further
increased by ‘recycling’ of the undesired epi-17 by iodination and
elimination, giving alkene 16 in 77% yield over two steps. After
silyl protection of 17, the acetal moiety of 18 was removed.
Optimization of the reaction conditions, to minimize epimerization,
resulted in treatment of 17 in acetone with FeCl3 adsorbed to
silica,[21] giving 19 in quantitative yield with 3% epimerization.
Ketone 19 was converted to the corresponding terminal alkene by
enol-triflation and Pd-catalyzed triflate reduction, delivering 20 in
80% yield over two steps. Hydroboration-oxidation of 20 gave
alcohol 21 in 87% yield, which was converted to the
corresponding bromide 5 in excellent yield. With 5 in hand, the
Next, the considerably more complex Fragment B was to be
constructed. The synthesis started with the preparation of two
4
This article is protected by copyright. All rights reserved.

10.1002/anie.202105384
Angewandte Chemie International Edition
RESEARCH ARTICLE
Scheme 4 Completion of the synthesis of the proposed structure of crenarchaeol.
building blocks 14 and 27 (Scheme 3). The synthesis of
cyclopentene 14 started from meso-diacetate 24, accessible in
two steps from cyclopentadiene.[23] Diacetate 24 was subjected to
enzymatic desymmetrization[24] in excellent yield and ee, followed
by silyl protection giving 14. Cyclohexanone 27 was prepared
according to the method developed by the Stoltz laboratory from
allyl cyclohexanone 13,[25] which was protected and subjected to
hydroboration/oxidation to deliver 26. Omission of the protection
of the ketone in 13 led to the formation of the
corresponding hemiacetal. Benzylation and acetal hydrolysis
provided the desired cyclohexanone 27 in 92% yield over two
steps.
increasing the equivalents of LHMDS to 1.6 and using LiCl as
additive resulted in full conversion (Table 1, entry 9). The product
was isolated in 53% yield with an excellent dr of 93:7.
We decided to apply these conditions to acetate 14 and
cyclohexanone 27, and found this system to be superior to the
model reaction. Product 30 was obtained in 67% yield with a dr
>20:1, and no undesired diastereomer detected (Scheme 3). This
variant of the intermolecular Pd-catalyzed asymmetric allylic
alkylation further expands the toolbox of this type of reaction and
we expect it to open up new avenues for future asymmetric
construction of joint ring systems in a convergent manner.
Progressing the synthesis of Fragment B, the ketone moiety was
reduced and acetylated, giving 31 as single diastereomer.
Subsequent desilylation and acetylation delivered diacetate 11 in
excellent yield. Notably, attempts to shorten this sequence by
performing reduction, desilylation, and double acetylation led to
significantly lower yields. This was due to the formation of a
tricyclic product arising from SN2’ addition of the non-allylic
hydroxy group to the double bond (see SI). With diacetate 11 in
hand, a regioselective copper-catalyzed Grignard alkylation with
32 (prepared from (R)-citronellol, see SI) was performed providing
a crude dr of 4:1 and, after separation of the isomers, alkylation
product 33 in 75% yield as single stereoisomer. The double bond
of 33 was reduced by a flavin-catalyzed diimide reduction[26]
followed by deacetylation providing 35 in 98% yield over two steps.
The hydroxyl moiety of 35 was then removed by a Barton-
McCombie deoxygenation reaction in excellent yield. After Pd-
catalyzed hydrogenolysis of the benzyl ether in 36, alcohol 37 was
oxidized to the corresponding aldehyde and subjected to a Wittig
olefination delivering ,-unsaturated thioester 39. The last
methyl-branched stereocenter of Fragment B was then introduced
in an excellent dr of 20:1 (see SI for details) by copper-catalyzed
With acetate 14 in hand, we chose to investigate the key step –
the intermolecular Pd-catalyzed Tsuji-Trost alkylation – with 2,2-
dimethylcyclohexanone 28 as model substrate (Table 1).
We started by screening ligands L1-L4 (Scheme 3) in
combination with Pd2(dba)3CHCl3 in order to achieve good chiral
induction. In presence of LHMDS as base and THF as solvent at
0 ºC, (R,R)-ANDEN-Phenyl Trost L1 gave good conversion to the
alkylation product 29, albeit with a dr of 75:25 favoring the
undesired diastereomer (Table 1, entry 1). Under the same
conditions, (R)-t-ButylPHOX L2 failed to give chiral induction
(Table 1, entry 2). When using DACH ligands L3 and L4, good
diastereoselectivities of 81:19 and 86:14 were achieved (entries
3 and 4), yet with a low conversion of around 40% and in the case
of L4 only 27% isolated yield. Since acceptable stereo-induction
was achieved, we continued the optimization with L4. Changing
the solvent to toluene or DME (Table 1, entry 5 and 6) did not
result in higher conversion, but the latter gave the product with
improved dr of 94:6. When using NaHMDS the conversion
dropped significantly to around 10-15% (Table 1, entry 7), while
LDA performed comparable to LHMDS (entry 8). Ultimately,
5
This article is protected by copyright. All rights reserved.

10.1002/anie.202105384
Angewandte Chemie International Edition
RESEARCH ARTICLE
Figure 2 A: Conversion of natural crenarchaeol and Fragment B into biphytanes. B-D: Partial gas chromatograms of the formed biphytane(s). B: biphytanes I-III
from the GDGTs in the Bligh Dyer extract of “Ca. Nitrosotenuis uzonensis”.[29] C: biphytane IV from Fragment B. D: Co-injection of IV with the biphytane mixture of
“Ca. N. uzonensis”. * Indicates an isomeric bicyclic biphytane, most likely originating from GDGT-4.
asymmetric conjugate addition of methylmagnesium
bromide[27] producing 40 in 87% yield. With the last
stereocenter of the biphytane core of crenarchaeol set, the
dithiane moiety of 10 was installed, after MOM deprotection of
40, through thioester reduction and treatment with 1,3-
propanedithiol in the presence of BF3OEt2. Dithiane 10 was
obtained in 82% over the three steps. Notably, dithiane
synthesis in presence of the MOM ether resulted in a complex
mixture of 10 and various trans-acetalization products. With 10
in hand, the last dithiane alkylation was performed, in presence
of the free hydroxyl group. After optimization of the lithiation
conditions, the reaction of lithiated 10 with iodide 9 smoothly
provided the coupling product 8 in 67% yield, containing the
entire carbon-skeleton of Fragment B. The synthesis of
Fragment B was concluded by a two-step sequence, involving
removal of the dithianes with Raney-nickel, followed by Pd-
catalyzed hydrogenolysis of the remaining benzyl ether.
amounts of elimination products were observed. Consequently,
given the difficulty of this step, we continued with the synthesis.
In order to perform the final ring closure of the macrocycle, 45
was converted to bis-alkene 1 by oxidation and Wittig reaction.
The 66-membered macrocycle was closed by means of ring-
closing metathesis with Grubbs 2nd generation catalyst,
a method often used for the construction of large rings.[28] This
provided 46 in 65% yield, given the size of the produced
macrocycle a more than satisfactory result. In the final step,
the double bond as well as the benzyl ethers were removed by
hydrogenolysis with palladium on carbon in low yield of 34%,
which could be partially attributed to the scale of the reaction.
This concluded the synthesis of this structurally complex lipid
and provided 1.2 mg of synthetic crenarchaeol. With both
synthetic crenarchaeol and the tricyclic intermediate Fragment
B in hand we sought to investigate the chemical structure of
natural crenarchaeol. For this purpose, we re-isolated natural
crenarchaeol in a laborious procedure (see SI) and made a
comparison of their NMR spectra. Furthermore, we performed
chemical derivatization in combination with GC-MS analysis.
Endgame – Completion of the total synthesis of the
proposed structure of crenarchaeol
After the successful stereoselective synthesis of both
Fragment A and B, the macrocycle of crenarchaeol was
assembled (Scheme 4). The endgame of the synthesis started
with the O-alkylation of protected glycerol 2 with mesylate 41
prepared from Fragment A. During the reaction using sodium
hydride in DMF, partial cleavage of the TBDPS ether was
observed. Therefore, after O-alkylation, the silyl ether was
reintroduced, giving alkylation product 42 in 62% yield. The
trityl ether was removed delivering 43, the substrate for the
next ether synthesis, in 94% yield. The double O-alkylation of
43 with bis-mesylate 44 came about after considerable
experimentation, by reaction with KOtBu as the base in toluene
in the presence of TBAB as phase-transfer catalyst. After
desilylation of the crude double alkylation product, the desired
diol 45 was obtained in a poor yield of 27% over the two steps.
There are multiple factors complicating this reaction. It is a
double O-alkylation of a bis-mesylate. The sheer size and
flexibility of this electrophile plays a role in the reaction rate as
we expect that the site of alkylation is not always exposed for
reaction with the weak alkoxide nucleophile. In addition, small
Comparison of natural crenarchaeol and Fragment B by
GC-MS
The Bligh Dyer extract of the thermophilic Thaumarchaeota
“Ca. Nitrosotenuis uzonensis” (dominated by crenarchaeol and
its cis-cyclopentyl isomer[29], see Fig. 2A) has previously been
treated with HI. This cleaves the ether bonds to produce a
mixture of biphytane diiodides.[29] Reduction of the iodides with
H2/PtO2 led to the corresponding hydrocarbons I-III, which
were analyzed by GC-MS.[29] This showed a ratio of bi- and
tricyclic biphytanes of approximately 1:1 (Fig. 2B). As a direct
comparison of the configuration of the tricyclic biphytane unit
within synthetic and natural crenarchaeol was considered
complicated, we subjected also Fragment
derivatization (Fig. 2A).[29-30] This enabled
B
a
to this
precise
comparison by GC-MS. Treatment of fragment B with HI
followed by reduction yielded biphytane IV which appeared, as
expected, as a single peak in the gas chromatogram (Fig. 2C),
but much to our surprise with a significantly different retention
time than the supposedly identical II derived from natural
6
This article is protected by copyright. All rights reserved.

10.1002/anie.202105384
Angewandte Chemie International Edition
RESEARCH ARTICLE
crenarchaeol. The mismatch in chemical structure was
confirmed by co-injection, showing retention time differences
of IV and II or III of approximately 1.5 and 2 min, respectively
(Fig. 2D).
Next, we turned our attention to the mass spectra of II-IV (see
SI). The fragmentation patterns of natural II and III were
equivalent to their previously reported mass spectra,[29, 31] and
featured the characteristic fragment m/z 262, originating from
bond cleavage adjacent to the quaternary stereocenter. This
fragment was also clearly visible in the mass spectrum of
synthetic IV.
Large differences ( >1 ppm) in chemical shift at the
(sub)terminal carbons (A1, A1’, A2 and A2’) of Fragment B
originate from the presence of primary hydroxyl moieties
contrary to the ether linkages in crenarchaeol. More
importantly, however, three carbon atoms around the all-
carbon quaternary stereocenter elicit large differences in
chemical shift at positions A14’ ( -5.96 ppm), A16’ ( -4.29
ppm) and A20’ ( +7.57 ppm), indicating a difference in
structure around these positions. It is noteworthy that the 13C
signals of the remaining stereocenters of the 5-6-ring system
(A10’ and A7’) in Fragment B show no significant difference. In
particular the good agreement of A10’ is indicative for the
ascribed stereochemistry of the single bond connecting the 5-
and 6-membered ring. It is expected that a difference in
stereochemistry on A11’ would translate to a significant 13C
chemical shift difference in A10’. This indicates that, on these
positions, the chemical structure of natural crenarchaeol
matches that of Fragment B. The chemical shifts of synthetic
nominal crenarchaeol (chemical shifts in brackets in Table 2)
show the same pattern of chemical shift differences. It should
be highlighted that there is no significant 13C chemical shift
difference between Fragment B and the tricyclic biphytane of
synthetic crenarchaeol (except for the terminal carbons A1/A1’
and A2/A2’) excluding an influence of the macrocyclic structure
on the chemical shifts.
Furthermore, the remaining fragmentation patterns of II/III and
IV are also virtually identical, providing strong evidence that the
overall chemical connectivity of II/III and synthetic IV is
identical. Thus, we concluded that II and IV are stereoisomers.
Comparison of Fragment
crenarchaeol by NMR
B
with isolated natural
In order to elucidate the exact structural difference between
synthetic Fragment B and the tricyclic biphytanyl moiety of
natural crenarchaeol, we compared their NMR spectra. The 1H
and 13C signals of natural crenarchaeol[11] and Fragment B
were assigned by thorough 2D NMR analysis. In addition, the
13C signals of synthetic crenarchaeol were assigned based on
the NMR analysis of Fragment B.
Table 2. Comparison of 13C NMR values of natural crenarchaeol with
Fragment B and synthetic nominal crenarchaeol.
Carbon
number[a]
13C shift natural
crenarchaeol (ppm)
13C shift
Fragment
(ppm)b]
 (ppm) [c]
B
A1, A1’
A2, A2’
70.23, 70.26
36.72, 36.75
61.42
(70.28, 70.25)
-8.81, -8.84
40.14, 40.15
(36.74)
+3.42,
+3.40
A11’
A12’
A13’
A14’
A15’
A16
39.38
32.27
22.40
44.13
33.20
30.12
37.80
22.55
39.01 (39.10)
31.80 (31.81)
22.10 (22.12)
38.17 (38.30)
32.92 (32.93)
30.50 (30.51)
33.51 (33.46)
30.12 (30.13)
-0.37
-0.47
-0.30
-5.96
-0.28
+0.38
-4.29
+7.57
Figure 3 Alleged structure of crenarchaeol with carbon numbering and the
synthetic Fragment B. Carbons with moderate ( 0.25–1 ppm) and larger
13C chemical shift differences are marked in orange and red, respectively.
The comparison of selected 13C NMR signals of Fragment B
and synthetic crenarchaeol with those of natural crenarchaeol
is shown in Table 2 (see SI for a table with all signal
assignments).
The carbon numbering is shown in Figure 3, and significant
differences in 13C NMR shifts between Fragment B and natural
crenarchaeol are marked in orange ( 0.25–1 ppm) and red
( >1 ppm). Upon comparison of the 13C NMR signals of
Fragment B with those of natural crenarchaeol,[11] it becomes
clear that the majority of the chemical shifts of Fragment B are
in very good agreement ( < 0.25 ppm) with those of the
tricyclic biphytane of
A16’
A20’
[a] Assignments of 13C NMR chemical shifts of crenarchaeol[11] and
Fragment B. Signals are reported relative to the solvent residual signal
(CDCl3  77.16 ppm). [b] Corresponding signals of synthetic nominal
crenarchaeol are shown in brackets.
crenarchaeol. In particular, the 13C chemical shifts of the three
cyclopentane rings (which are not connected to the
cyclohexane ring) and their alkyl substituents are virtually
identical (see SI).
Moderate chemical shift differences ( 0.25–1 ppm) were
ascribed to the cyclohexane ring carbons (A11’, A12’, A13’ and
A15’) and the alkyl chain adjacent to the cyclohexyl ring (A16).
Besides the good agreement of most of the 13C NMR chemical
shifts of crenarchaeol and Fragment B, the H NMR chemical
shifts of A7’, A10’ and A11’ correlate well (Table 3, see SI for
all assignments). At position A19’ (axial) and A20’, only minor
1H shift differences were observed. Only three positions show
significant chemical shift differences: the equatorial proton of
A14’ ( 0.27 ppm), A16’ ( 0.47 ppm) and the equatorial
1
7
This article is protected by copyright. All rights reserved.

10.1002/anie.202105384
Angewandte Chemie International Edition
RESEARCH ARTICLE
proton of A19’ ( 0.15 ppm). This provides further evidence
that the difference in structure of natural and synthetic
crenarchaeol is located around these positions.
calculated (See SI for the protocol and the calculated shifts).
The DFT prediction is in good agreement with the upfield shift
of methyl group A20‘ in natural crenarchaeol and the expected
downfield shift of methylene A16‘.
All in all this combined data provides overwhelming evidence
for an inverted stereochemistry of crenarchaeol at A15’
compared to Fragment B. On the basis of the evidence from
chemical derivatization, NMR studies, and computation, we
therefore propose a revised structure of crenarchaeol (Fig. 4),
in which the stereochemistry of the all-carbon quaternary
stereocenter is inverted compared to the original proposal.
Table 3. Comparison of 1H NMR values of natural crenarchaeol and
Fragment B.
Carbon
number[a]
1H shift
(ppm)
crenarchaeol
1H shift Fragment B (ppm)]
A7’
1.79
1.47
1.17
1.15
1.31
1.78
A10’
A11’
A14’
A16’
A19’
A20’
1.46
1.12
Conclusion
1.42
The first total synthesis of the originally proposed structure of
the thaumarchaeotal GDGT crenarchaeol has been achieved.
The synthesis involved the stereoselective construction of a
unique 5-6 ring motif as well as a late-stage 66-membered
macrocyclization by means of RCM. The structure
determination of crenarchaeol has a considerable history.[11]
Due to the very complex structure, including 22 stereocenters,
as well as the highly aliphatic character and its lack of rigidity,
NMR-based structural studies have been heavily complicated.
Furthermore, since this lipidic molecule does not have the
tendency to crystallize, X-ray diffraction was not possible. The
synthesis of the proposed structure of crenarchaeol and the
key intermediate Fragment B enabled direct comparison with
natural crenarchaeol by chemical derivatization and GC-MS
analysis. This revealed a mismatch of the chemical structure
of the tricyclic biphytane chain. Subsequently, detailed NMR
analysis including computational simulation of 13C chemical
shifts, of Fragment B and synthetic crenarchaeol, and
comparison with natural crenarchaeol isolated from sea
surface sediments was performed. Ultimately, from the
spectroscopic data of fragment B, synthetic and natural
crenarchaeol, we were able to revise the originally proposed
structure beyond reasonable doubt. Through this extensive
analysis we identified the inversion of just one out of the 22
stereocenters of crenarchaeol, namely the quaternary
stereocenter embedded in the cyclohexane ring.
1.78
ax.: 0.70; eq.: 1.39
0.84
ax.: 0.64; eq.: 1.52
0.79
[a] Assignments of 1H NMR chemical shifts of crenarchaeol[11] and Fragment
B. Signals are reported relative to the solvent residual signal (CDCl3  77.16
ppm).
Since the relative and absolute stereochemistry of Fragment B
is known, the methyl substituent A20’ of Fragment B is
assigned to be equatorial due to the 1,3-cis relationship of the
methyl and cyclopentyl substituents on the cyclohexane ring.
As a result of the deshielding -gauche effect, the 13C NMR
chemical shift of axial substituents in cyclohexanes is more
upfield relative to equatorial substituents.[32] In Fragment B the
13C signal of methyl group A20’ resonates at 30.12 ppm, while
the methyl group A20’ of natural crenarchaeol is shifted more
upfield at 22.55 ppm. This strongly suggests that the methyl
group A20’ in natural crenarchaeol is in axial position in
contrast to the initially proposed structure. To further support
this, we considered the 13C chemical shifts of A16’. In
Fragment B, the carbon atom A16’ of the alkyl side-chain of the
cyclohexyl ring is axially oriented. The 13C signal resonates at
33.51 ppm, whereas in crenarchaeol the 13C signal of A16’ is
shifted downfield to 37.80 ppm. Thus, the downfield shift of A16’
in natural crenarchaeol strongly suggests equatorial
Total synthesis not only comprises the access to complex
molecules, but serves also as a breeding ground for new
synthetic methodology as well as probing current synthetic
methods. Mistakes in the proposed structure of a natural
product are by no means
a
rare occurrence.[33] The
architectural and stereochemical complexity of a new unknown
structure, in combination with very small amounts of isolated
material often make assignments extremely difficult, in
particular in a case such as crenarchaeol, which features
almost no heteroatom functionalities and is highly flexible. By
using the information gathered from the synthetic epimer of
natural crenarchaeol, we were able to reassign the structure
without the need to repeat the entire, very complex, synthesis.
The correction of the structure of crenarchaeol has important
implications for the study of its role in archaeal membranes.
The current hypothesis is that the presence of crenarchaeol
regulates membrane fluidity and packing, an important
adaptation to temperature and pressure changes in the
environment. As the stereochemistry of the quaternary center
in crenarchaeol has a significant influence on its conformation,
Figure 4 Revised chemical structure of natural crenarchaeol.
substitution of the alkyl chain substituent on the cyclohexyl ring.
Further support comes from the computationally calculated 13
C
shift values for A16’ and A20’. First, MD simulations in
chloroform were carried out on fragment B and its isomer to
determine the lowest energy conformations. Subsequently, the
energies of the conformers from the MD trajectory were
evaluated using DFT calculations and the chemical shifts
8
This article is protected by copyright. All rights reserved.

10.1002/anie.202105384
Angewandte Chemie International Edition
RESEARCH ARTICLE
Richter, T. Benvegnu, D. Plusquellec, J. Am. Chem. Soc. 2004, 126,
10003-10012.
and thus membrane packing, we expect that an explanation
(supported by for instance molecular dynamics simulations) for
its role in membrane behavior is now within reach.
[16] W. Kinouchi, R. Saeki, H. Kawashima, Y. Kobayashi, Tet. Lett. 2015,
56, 2265-2268.
[17] B. M. Trost, D. L. Van Vranken, Chem. Rev. 1996, 96, 395-422.
[18] W. B. Kover, J. Jones Jr., J. Braz. Chem. Soc. 1996, 7, 257-263.
[19] a) J. Chen, T. Xi, X. Ren, B. Cheng, J. Guo, Z. Lu, Org. Chem. Front.
2014, 1, 1306-1309; b) L. Zhang, Z. Zuo, X. Wan, Z. Huang, J. Am.
Chem. Soc. 2014, 136, 15501-15504; c) W. J. Jang, S. M. Song, J.
H. Moon, J. Y. Lee, J. Yun, J. Am. Chem. Soc. 2017, 139, 13660-
13663.
Acknowledgements
The authors would like to thank Dr. J. Buter and E. Jonkheim
(University of Groningen) for their contribution to the isolation
of crenarchaeol. P. van der Meulen and Dr. J. Kemmink
(University of Groningen) are acknowledged for their
assistance in NMR measurements. This work was sponsored
by NWO Exact and Natural Sciences for the use of
supercomputer facilities and RWAH and AV thank S. Dolas
(SURF, NL) for allowing to perform experiments on the
experimental AMD platform kleurplaat, maintained and
operated by SURF Open Innovation Lab. This project received
funding to JSSD from the European Research Council (ERC)
under the European Union’s Horizon 2020 research and
innovation program (grant agreement no. 694569,
MICROLIPIDS).
[20] a) Y. Nagai, T. Kusumi, Tet. Lett. 1995, 36, 1853-1856; b) T.
Yabuuchi, T. Ooi, T. Kusumi, Chirality 1997, 9, 550-555; c) T.
Yabuuchi, T. Kusumi, J. Org. Chem. 2000, 65, 397-404.
[21] K. S. Kim, Y. H. Song, B. H. Lee, C. S. Hahn, J. Org. Chem. 1986,
51, 404-407.
[22] A. B. I. Smith, C. M. Adams, Acc. Chem. Res. 2004, 37, 365-377.
[23] D. R. Deardorff, D. C. Myles, K. D. MacFerrin, Tet. Lett. 1985, 26,
5615-5618.
[24] L. Tietze, C. Stadler, N. Böhnke, G. Brasche, A. Grube, Synlett 2007,
2007, 485-487.
[25] J. T. Mohr, M. R. Krout, B. M. Stoltz, Org. Synth. 2009, 86, 194-211.
[26] C. Smit, M. W. Fraaije, A. J. Minnaard, J. Org. Chem. 2008, 73, 9482-
9485.
[27] R. Des Mazery, M. Pullez, F. López, S. R. Harutyunyan, A. J.
Minnaard, B. L. Feringa, J. Am. Chem. Soc. 2005, 127, 9966-9967.
[28] a) K. Arakawa, T. Eguchi, K. Kakinuma, J. Org. Chem. 1998, 63,
4741-4745; b) L. Yet, in Organic Reactions, Vol. 89 (Ed.: S. E.
Denmark), John Wiley & Sons, Inc., 2016.
Keywords: crenarchaeol • tetraether lipid • structure revision
• archaea • total synthesis
[29] J. S. Sinninghe Damsté, W. I. C. Rijpstra, E. C. Hopmans, M. J. den
Uijl, J. W. H. Weijers, S. Schouten, Org. Geochem. 2018, 124, 22-28.
[30] a) O. Gräther, D. Arigoni, J. Chem. Soc., Chem. Commun. 1995, 405-
406; b) M. Kaneko, F. Kitajima, H. Naraoka, Org. Geochem. 2011, 42,
166-172; c) S. K. Lengger, Y. A. Lipsewers, H. de Haas, J. S.
Sinninghe Damsté, S. Schouten, Biogeosciences 2014, 11, 201-216.
[31] S. Schouten, M. J. L. Hoefs, M. P. Koopmans, H.-J. Bosch, J. S.
Sinnighe Damsté, Org. Geochem. 1998, 29, 1305-1319.
[32] a) G. W. Buchanan, Can. J. Chem. 1982, 60, 2908-2913; b) E.
Breitmaier, W. Voelter, Carbon-13 NMR Spectroscopy: High-
Resolution Methods and Applications in Organic Chemistry and
[1]
[2]
C. R. Woese, O. Kandler, M. L. Wheels, Proc. Natl. Acad. Sci. USA
1990, 87, 4576-4579.
a) C. R. Woese, G. E. Fox, Proc. Natl. Acad. Sci. USA 1977, 74,
5088-5090; b) Y. Koga, H. Morii, Microbiol. Mol. Biol. Rev. 2007, 71,
97-120; c) R. Cavicchioli, Nat. Rev. Microbiol. 2011, 9, 51-61.
C. R. Woese, L. J. Magrum, G. E. Fox, J. Mol. Evol. 1978, 11, 245-
252.
[3]
[4]
a) E. F. DeLong, Curr. Op. Genet. Dev. 1998, 8, 649-654; b) S.
Schouten, E. C. Hopmans, R. D. Pancost, J. S. Sinninghe Damsté,
Proc. Natl. Acad. Sci. USA 2000, 97, 14421-14426.
a) S. Jain, A. Caforio, A. J. Driessen, Front. Microbiol. 2014, 5, 641;
b) L. Villanueva, J. S. Damste, S. Schouten, Nat. Rev. Microbiol. 2014,
12, 438-448.
[5]
[6]
Biochemistry,
3 ed., VCH Verlagsgesellschaft mbH, Weinheim,
Germany, 1987; c) M. E. Squillacote, J. M. Neth, Magn. Reson. Chem.
1987, 25, 53-56; d) J. Xiong, J. Wan, J. Ding, P.-P. Wang, G.-L. Ma,
J. Li, J.-F. Hu, J. Nat. Prod. 2017, 80, 2874-2882.
a) J. Lombard, P. Lopez-Garcia, D. Moreira, Nat. Rev. Microbiol.
2012, 10, 507-515; b) Y. Koga, J. Mol. Evol. 2014, 78, 234-242.
P. M. Oger, A. Cario, Biophys. Chem. 2013, 183, 42-56.
Z. Zeng, X. L. Liu, K. R. Farley, J. H. Wei, W. W. Metcalf, R. E.
Summons, P. V. Welander, Proc. Natl. Acad. Sci. USA 2019, 116,
22505-22511.
[33] K. C. Nicolaou, S. A. Snyder, Angew. Chem. 2005, 44, 1012-1044.
[7]
[8]
[9]
Y. Koga, Archaea 2012, 2012, 789652.
[10] a) J. S. Sinninghe Damsté, W. I. Rijpstra, E. C. Hopmans, M. Y. Jung,
J. G. Kim, S. K. Rhee, M. Stieglmeier, C. Schleper, Appl. Environ.
Microbiol. 2012, 78, 6866-6874; b) N. J. Bale, M. Palatinszky, W. I. C.
Rijpstra, C. W. Herbold, M. Wagner, J. S. Sinnighe Damsté, Appl.
Environ. Microbiol. 2019, 85, e01332-01319.
[11] J. S. Sinninghe Damsté, S. Schouten, E. C. Hopmans, A. C. T. van
Duin, J. A. J. Geenevasen, J. Lipid. Res. 2002, 43, 1641-1651.
[12] a) A. Matsuo, I. Terada, M. Nakayama, S. Hayashi, Tet. Lett. 1977,
43, 3821-3824; b) A. V. Tkachev, M. M. Shakirov, V. A. Raldugin, J.
Nat. Prod. 1991, 54, 849-853; c) F. Nagashima, M. Suzuki, S.
Takaota, Y. Asakawa, J. Nat. Prod. 2001, 64, 1309-1317.
[13] X.-L. Liu, D. A. Russell, C. Bonfio, R. E. Summons, Org. Geochem.
2019, 128, 57-62.
[14] E. Montenegro, B. Gabler, G. Paradies, M. Seemann, G. Helmchen,
Angew. Chem. Int. Ed. 2003, 42, 2419-2421.
[15] a) G. Lecollinet, R. Auzély-Velty, M. Danel, T. Benvegnu, G.
Mackenzie, J. W. Goodby, D. Plusquellec, J. Org. Chem. 1999, 64,
3139-3150; b) G. Lecollinet, A. Gulik, G. Mackenzie, J. W. Goodby,
T. Benvegnu, D. Plusquellec, Chem. Eur. J. 2002, 8; c) M. Brard, W.
9
This article is protected by copyright. All rights reserved.

10.1002/anie.202105384
Angewandte Chemie International Edition
RESEARCH ARTICLE
Entry for the Table of Contents
Total synthesis of the proposed structure of crenarchaeol and comparison with the natural isolate has led to a revised structure of
crenarchaeol, wherein one of the 22 stereocenters is inverted. The synthesis featured an asymmetric intermolecular Tsuji-Trost-type
alkylation to access the unusual 5-6-ring motif as well as a late stage ring-closing metathesis to access the 66-membered macrocycle
of crenarchaeol.
Institute and/or researcher Twitter usernames: @mira_holzheimer @AJMinnaard @StratinghInst @NIOZnews
10
This article is protected by copyright. All rights reserved.