Biological Investigations of (+)-Danicalipin A Enabled Through Synthesis

Article abstract of DOI:10.1002/anie.201509082

A total synthesis of the chlorosulfolipid (+)-danicalipin A has been accomplished in 12 steps and 4.4 % overall yield. The efficient and scalable synthesis enabled in-depth investigations of the lipid's biological properties, in particular cytotoxicity towards various mammalian cell lines. Furthermore, the ability of (+)-danicalipin A to increase the uptake of fluorophores into bacteria and mammalian cells was demonstrated, indicating it may enhance membrane permeability. By comparing (+)-danicalipin A with racemic 1,14-docosane disulfate, and the diol precursor of (+)-danicalipin A, we have shown that both chlorine and sulfate functionalities are necessary for biological activity.

Full text of DOI:10.1002/anie.201509082

Angewandte
Communications
Chemie
International Edition: DOI: 10.1002/anie.201509082
German Edition: DOI: 10.1002/ange.201509082
Chlorosulfolipids
Biological Investigations of (+)-Danicalipin A Enabled Through
Synthesis
Adrian M. Bailey, Susanne Wolfrum, and Erick M. Carreira*
Abstract: A total synthesis of the chlorosulfolipid (+)-dan-
icalipin A has been accomplished in 12 steps and 4.4% overall
yield. The efficient and scalable synthesis enabled in-depth
investigations of the lipidꢀs biological properties, in particular
cytotoxicity towards various mammalian cell lines. Further-
more, the ability of (+)-danicalipin A to increase the uptake of
fluorophores into bacteria and mammalian cells was demon-
strated, indicating it may enhance membrane permeability. By
comparing ( ++ )-danicalipin A with racemic 1,14-docosane
disulfate, and the diol precursor of (+)-danicalipin A, we
have shown that both chlorine and sulfate functionalities are
necessary for biological activity.
C
hlorosulfolipids are a unique family of natural lipids, the
first of which were isolated in the 1960ꢀs independently by the
groups of Haines and Elovson from the fresh water alga
[
1]
Ochromonas danica. Since then, several additional chloro-
sulfolipids have been isolated and characterized, namely
[2]
malhamensilipin A from Poterioochromonas malhamensis
and mytilipin A–C (Figure 1a) from unspecified, harmful
microalgae, which accumulate in Mediterranean sea mussels
[
3]
Mytilus galloprovincialis rendering them toxic. These chiral
lipids are characterized by polychlorinated, sulfated aliphatic
[
4]
chains, structural features uncommon in nature. Despite
a number of total syntheses of these structurally complex
Figure 1. a) The structures of the most complex chlorosulfolipids, and
b) the polar and apolar regions of 1 and phosphatidylcholine.
[5]
chlorosulfolipids, little is known about their biological
properties. Herein, we disclose a scalable total synthesis of
+)-danicalipin A (1) and evidence that this lipid has
a significant effect on the membranes of mammalian cells
and the walls of Gram-negative bacteria.
Chlorosulfolipids are most abundant in O. danica, con-
tributing to 14.4% of the total lipid weight, whereas they
account for only 2.3% of the total lipid weight of P.
malhamensis. Furthermore, in the case of O. danica, these
lipids comprise more than 80 mol% of the total polar lipid
content of the entire cell. This observation has led to the
speculation that the larger contribution of the chlorinated
lipids is related to an unspecified important role in membrane
[
8]
(
through self-assembly. A comparison of the conventional
phospholipid structure to that of 1 reveals a number of
differences. Firstly, in contrast to typical phospholipids, 1 is
[
1f, 9]
water soluble and highly polar.
This is presumably
a consequence of the pair of anionic O-sulfates. Secondly,
the charged domains of phospholipids are localized at one
extremity of the lipid chain, which facilitates interaction
between the extra- and/or intracellular space of the cell with
[
6]
[
7]
[8]
the interior of the membrane. Chlorosulfolipid 1 has two
charged O-sulfates separated by an expanse of fourteen
methylenes. It has been proposed that the O-sulfate posi-
tioned towards the center of the lipid chain, would find itself
inside the hydrophobic region of a biomembrane, and
[1f]
structure and function in O. danica.
Most membrane lipids have well defined domains includ-
ing hydrophilic heads and hydrophobic tails (Figure 1b);
these structural attributes enable formation of lipid bilayers
[1f]
presumably lead to its destabilization.
Lastly, the fatty
acid components of phospholipids incorporate long carbon
chains, with palmitic (C ) and oleic (C ) acids being most
16
18
[
10]
abundant in mammalian cells. By contrast, the hydrophobic
domains in 1 are much shorter. This presents a problem in
understanding the structural integrity of membranes largely
[
*] A. M. Bailey, Dr. S. Wolfrum, Prof. Dr. E. M. Carreira
Laboratorium für Organische Chemie, Eidgençssische Technische
Hochschule Zürich, HCI H335
[11]
comprised of 1. Collectively, these structural features of this
chlorosulfolipid are at odds with the generally accepted
characteristics and associated functions of lipids in biomem-
branes.
Vladimir-Prelog-Weg 3, 8093 Zürich (Switzerland)
E-mail: carreira@org.chem.ethz.ch
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201509082.
Angew. Chem. Int. Ed. 2016, 55, 639 –643
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
639

Angewandte
Communications
Chemie
The biological data for 1 is limited. It is associated with
other chlorosulfolipids (3–5) for which there are anecdotal
Hydroboration/oxidation of this electron deficient olefin
proved problematic and intriguing. Surprisingly, the use of
[5e, 12]
reports of their causing seafood poisoning,
its toxicity
and that it
BH ·THF furnished dehalogenated material in addition to
3
[5b, 13]
[20]
towards an unspecified genus of brine shrimp,
small quantities (15–30%) of desired product. A screening
is one of the dominant polar components of the membrane of
of various hydroboration reagents and conditions led to the
use of dicyclohexylborane followed by work up with sodium
[
7b,14]
O. danica.
The lack of any kind of understanding of their
[21]
role and function stimulated us to investigate these rare lipids.
Specifically for 1, we sought to gain insight into its effect on
cell membranes, to generate an array of biological assays, and
to compare biological data with related non-chlorinated
lipids. To achieve these goals, we required the development
of a route towards appreciable quantities of this synthetically
challenging, intriguing target.
perborate
to provide primary alcohol 10 in 75% yield.
After benzoylation, the enantiomeric excess of 10 was
determined to be 91% by chiral HPLC analysis. Despite the
undesired propensity for elimination of b-chloroaldehydes to
[22]
a,b-enals, oxidation with Dess–Martin periodinane to the
unpurified b-chloroaldehyde proceeded smoothly with no
1
elimination evident as determined by H NMR analysis.
[5a–c,h]
[23]
To date, 1 has been synthesized four times
with the
Subsequent Brown allylation
of the crude b-chloroalde-
most elegant approach in 14 total steps and 4.6% overall
hyde yielded homoallylic alcohol 11 in 8.3:1 d.r. and 71%
yield over two steps. Conversion of 11 to the desired
secondary chloride with complete inversion was achieved in
[
5h]
yield. Despite its conciseness, midway through the route is
a step that requires the use of a catalyst that is not
commercially available and must be prepared in 3 steps to
effect a kinetic resolution reaction. This is followed by a low
yielding cross metathesis reaction (29% yield) with 30 mol%
of a catalyst (single turnover) that has only recently become
[24]
93% yield, utilizing Ghosezꢀs reagent (12). One-pot cross-
[25]
metathesis of 13 with 14 and subsequent hydrogenation
afforded 15 in 77% yield. To complete the synthesis, 15 was
deprotected with hot methanolic HCl, and the free diol was
sulfated with ClSO H in CH Cl yielding 1 in 82% yield over
[15]
commercially available. Consequently, we looked to for-
mulate a synthesis utilizing robust, scalable chemistry to
facilitate the production of large quantities of material
3
2
2
2 steps. For the longest linear sequence, the route required 12
steps and afforded 1.45 g of 15, with an overall yield of 4.4%.
With sufficient quantities of 1 in hand, it was then
examined in the sole biological assay reported in which its
toxicity towards brine shrimp (unspecified genera) was
(
> 1 g).
Synthesis of 1 commenced with MnO -mediated oxidation
2
of commercially available (Z)-non-2-en-1-ol (6) to the
corresponding (Z)-enal (Z/E = 24:1) isolated without purifi-
cation in 97% yield (Scheme 1). Oehlschlager–Brown halo-
[5b,13]
investigated.
Artemia salina brine shrimp were hatched
and allowed to grow for 24–96 h. Following distribution into
tissue culture plates (well volume 2 mL), yeast extract (aq.)
along with 1, 16, or 17 were added. After 24 h, live-dead
shrimps were counted, and the results were plotted to provide
LC50 data (see Supporting Information, SI). The value
[
16]
allylation selectively yielded syn-halohydrin 7 (a/g = 5:1;
syn/anti > 20:1) utilizing (À)-Ipc BOMe, which was prepared
2
[17]
following the procedure of Lautens.
Difficulties were
encountered in attempts to effect separation of the desired
product from the co-product, isopinocampheol. However,
subjecting the unpurified mixture to directed epoxidation
À1
measured for the toxicity of 1 (LC = 5.3 mm = 3.8 mgmL ;
50
Table 1) was in agreement to that previously reported (LC =
5
0
[18]
À1 [5b,13]
(mCPBA, d.r. = 11:1 syn/anti as determined by NMR)
3.3 mgmL ).
Unfortunately, in the previous studies, non-
followed by TBS protection afforded the all syn protected
halohydrin 8 in 30% yield over 3 steps. Subsequent applica-
tion of Yoshimitsuꢀs conditions for dichlorination of epox-
chlorinated sulfolipids were not examined. Interestingly,
[26,27]
docosane disulfate (17)
has been proposed to be a bio-
synthetic precursor to 1; consequently, we decided to inves-
tigate it and assess the effect of chloride substitution. For
[19]
ides to 8 provided trichloride 9 as a single diastereomer.
Scheme 1. Reagents and conditions: a) MnO (25 equiv), CH Cl , RT, Z:E=24:1; b) (À)-Ipc BOMe (1.0 equiv), allyl chloride (1.3 equiv),
2
2
2
2
(
C H ) NLi (1.6 equiv), BF ·OEt (2.6 equiv), Et O:THF (1.8:1), À788C to RT, then ethanolamine (1.0 equiv), Et O; c) mCPBA (1.1 equiv), CH Cl ,
6
11
2
3
2
2
2
2
2
0
8C to RT, d.r.=11:1; d) TBSCl (1.3 equiv), imidazole (2.0 equiv), DMAP (0.1 equiv), CH Cl , 08C to 408C, 30% (3 steps); e) NCS (3.6 equiv),
2 2
ClPPh (3.0 equiv), CH Cl , 08C to RT, 44%; f) (C H ) BH (2.4 equiv), THF, 08C, then NaBO (30 equiv), THF:H O (1.4:1), RT, 75%; g) DMP
2
2
2
6
11
2
3
2
(
1.2 equiv), CH Cl , 08C to RT; h) (+)-Ipc BCl (2.0 equiv), allylMgBr (1.0m in Et O, 1.5 equiv), THF, À788C to À1008C to RT, d.r. =8.3:1, 71% (2
2
2
2
2
steps); i) 12 (3.0 equiv), CHCl , 08C to RT, 93%; j) 14 (3.0 equiv), Grubbs II (10 mol%), PtO (10 mol%), H , 408C to RT, 77%; k) AcCl
3
2
2
(
50 equiv), MeOH, 08C to 808C, 89%; l) ClSO H (30 equiv), CH Cl , 08C to RT, 92%. Ipc=isopinocampheyl, mCPBA=m-ClC H CO H,
3
2
2
6
4
3
TBS=tBuMe Si, DMAP=p-(Me N)C H N, NCS=N-chlorosuccinimide, DMP=Dess–Martin periodinane, Ac=acyl.
2
2
5
5
6
40
www.angewandte.org
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2016, 55, 639 –643

Angewandte
Communications
Chemie
Table 1: Toxicity of 1, 16, and 17 towards brine shrimp and various cell
lines.
[
a]
Target
Compound toxicity
1
16
17
[
b]
Brine shrimp LC50
A549 cells EC50
HT-29 cells EC50
Hepa 1-6 cells EC50
5.3
>141
63.8
26.5Æ0.9
15.5Æ1.2
14.3Æ0.7
41.4Æ0.9
69.3Æ1.2
84.4Æ0.7
39.1Æ0.2
>166
17.3Æ0.1
[
a] LC50 and EC50 data are reported in units of mm. [b] The exact LC50 of
this compound could not be determined due to its limited water
À1
solubility above 100 mgmL
.
comparison purposes, we included the non-sulfated chloro-
[
13]
lipid 16, which was the penultimate intermediate in our
synthesis. In the assay, danicalipin A (1) is an order of
magnitude more toxic than 17 and at least 2 orders of
magnitude more toxic than 16.
We next decided to expand the panoply of assays to
include exploration of cytotoxicity towards various cell lines:
A549 and HT-29 human adenocarcinomic cell lines, as well as
Hepa 1–6 murine liver cell line (Table 1). Our aim was to
establish robust cell culture conditions for more than one cell
line that could ultimately be used to study the effect of 1, 16,
1
7, and any future derivatives on membrane permeability.
The cell lines were chosen based on high and consistent
growth capacity as well as batch to batch homogeneity to
assure high reproducibility vis-à-vis cell numbers and minimal
phenotypic variation. Similar criteria were applied in the
selection of Gram-negative bacteria (see below). Before
beginning experiments on cell cultures, toxicity thresholds
were determined towards the three cell lines (Table 1; see SI
for details). The EC50 was determined for each compound,
wherein 1 was shown to be most toxic. With this information,
we began to investigate the effect of these lipids on cells at or
below this threshold value.
A staining method was developed to distinguish cells with
compromised membranes from healthy cells. Two DNA
staining dyes were selected: 1) Hoechst 33342 stain, capable
of crossing healthy membranes and subsequently staining the
[
28]
DNA of the cell,
and 2) Sytox Green, known to only
[29]
penetrate cells with compromised plasma membranes.
Upon exposure of HT-29 cells to 1 (Figure 2, entries 1–3)
a vast majority of the cells stained positive for both dyes,
a result that is particularly evident at 10 mm (Figure 2,
entry 2). A qualitative positive correlation between the
concentration of 1 and the number of cells stained with
Sytox Green was observed. In contrast, exposure of HT-29
cells, under equivalent conditions, to 16 or 17 resulted in
minimal DNA staining by Sytox Green at concentrations
below their EC50 values (Figure 2, entries 4–9), consistent
with results from negative control experiments (1% DMSO).
Alternatively, positive control experiments (20% EtOH)
furnished results in alignment with those produced by 1 (see
SI). The data strongly implicate, for the first time, that the
natural product danicalipin A affects the integrity of the
cellular membrane. Similar observations were made with
Hepa 1–6 cells (see SI), demonstrating that the effect of 1 is
not cell line specific.
Figure 2. Fluorescent images of HT-29 cells following exposures to 1,
16, or 17 at various concentrations.
To further probe the role of these lipids, we targeted
a structurally distinct organism, namely, E. coli DH5a, a strain
of Gram-negative bacteria. These bacteria were chosen as
they have two protective layers, namely an outer membrane
comprised largely of lipopolysaccharides and proteins and
[30]
[31]
a peptidoglycan cell wall. Utilization of an MTT assay
demonstrated that viability of bacteria was reduced (70–
90%) compared to untreated cells only upon exposure to 1 at
ꢀ 250 mm. Non-chlorinated lipid 17 was shown to be slightly
antibacterial in a dose independent manner at all concen-
trations ꢁ 250 mm. Bacterial viability was not affected upon
treatment with 16, even at 250 mm (see SI). Thus, the presence
Angew. Chem. Int. Ed. 2016, 55, 639 –643
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
641

Angewandte
Communications
Chemie
of chlorines in danicalipin A renders it somewhat less toxic
towards bacteria (> 90% viability) when compared to the
dechlorinated alkyl disulfate 17 (70–90% viability).
1 but this study has shown that this unique property is
dependent on the chlorines and the sulfates, indicating that
both are essential. These structural features in particular,
given the effect on membrane fluidity, may be pertinent to the
We then moved on to examine whether 1 would have
similar effects on membrane integrity, in the more challenging
bacterial setting, as was observed with mammalian cell lines.
Hoechst 33342 staining of E. coli DH5a bacterial cultures in
the presence of 1% DMSO was taken as baseline, while dye
uptake following heat inactivation of the bacteria served as
[
7]
presence of danicalipin A in the flagella of O. danica.
In conclusion, we have accomplished a concise and
scalable synthesis of (+)-danicalipin A (1), utilizing haloally-
lation, chlorination, and one-pot metathesis/hydrogenation
reactions providing significant quantities of material. Its
purported casual association with seafood poisoning renders
its study essential. We have also documented a comparative
study of 1 with its non-chlorinated biosynthetic precursor 17.
Through the application of DNA staining experiments, we
have shown that both the chlorines and the sulfates are
necessary to compromise the cellular membranes of mamma-
lian (e.g. human colorectal) cells as well as Gram-negative
bacteria. In this respect, the study described with danicali-
pin A provides a crucial benchmark for any future inves-
tigations. Further efforts toward establishing the biophysical
role of 1 in the membrane of O. danica and a more detailed
analysis of the biological properties of 1, its stereoisomers,
and its congeners are currently ongoing and will be reported
in due course.
[32]
a positive control.
Incubation of bacteria in the presence of Hoechst 33342
and 1 elicited a dramatic dose-dependent positive correlation
between the fluorescence measured and the concentration of
1
(Figure 3). Compounds 16 and 17 had minimal effect on the
Acknowledgements
We thank the ERC for support of this work through grant
3
20666_CHLIP.
Keywords: biological function · cell membranes ·
Figure 3. The fluorescence response due to nuclear staining of bacteria
E. coli DH5a by Hoechst 33342 as a function of the concentration of
chlorosulfolipids · natural products · total synthesis
1
, 16, or 17 as well as positive and negative control experiments. The
How to cite: Angew. Chem. Int. Ed. 2016, 55, 639–643
Angew. Chem. 2016, 128, 649–653
data is normalized to the untreated results. The significance of each
result vs. DMSO is shown above the bar: p<0.01=**;
p<0.001=***. For a color image, see SI.
[
1] a) T. H. Haines, M. Pousada, B. Stern, G. L. Mayers, Biochem. J.
969, 113, 565; b) J. Elovson, P. R. Vagelos, Proc. Natl. Acad. Sci.
1
amount of dye entering the cell, even at 250 mm, and produced
results similar to the negative control. Notably, at the 125 mm
concentration of 1, a 5-times greater fluorescence response
was observed and bacterial viability remained > 90%. In
analogy to the mammalian cell experiments, the bacterial
analysis shows that the chlorides and sulfates of danicalipin A
are necessary for effective uptake of Hoechst 33342 into E.
coli DH5a cells.
USA 1969, 62, 957; c) J. Elovson, P. R. Vagelos, Biochemistry
1970, 9, 3110; d) C. L. Mooney, E. M. Mahoney, M. Pousada,
T. H. Haines, Biochemistry 1972, 11, 4839; e) C. L. Mooney,
T. H. Haines, Biochemistry 1973, 12, 4469; f) T. H. Haines, Annu.
Rev. Microbiol. 1973, 27, 403; g) J. Elovson, Biochemistry 1974,
1
3, 403.
[
2] a) J. L. Chen, P. J. Proteau, M. A. Roberts, W. H. Gerwick, J.
Nat. Prod. 1994, 57, 524; b) A. R. Pereira, T. Byrum, G. M.
Shibuya, C. D. Vanderwal, W. H. Gerwick, J. Nat. Prod. 2010, 73,
2
79.
It is generally accepted that sterols, particularly choles-
terol and ergosterol, govern the fluidity and mechanical
[
[
3] P. Ciminiello, C. DellꢀAversano, E. Fattorusso, M. Forino, S.
[33,34]
Magno, M. D. Rosa, A. Ianaro, R. Poletti, J. Am. Chem. Soc.
strength of membranes.
The effect of sterols on the
2
002, 124, 13114.
structure of biomembranes is conserved among a large variety
of living organisms including vertebrates, fungi, protozoa,
higher plants, and algae. The optimal membrane composition
defined at one end by high permeability and at the other by
intransigency, has come about through evolutionary pressures
and is unique to each organismꢀs particular set of membrane
sterols. The remarkable ability of 1 to increase the amount of
dye to pass through biological membranes without leading to
4] For recent reviews on chlorosulfolipids, see: a) D. K. Bedke,
C. D. Vanderwal, Nat. Prod. Rep. 2011, 28, 15; b) C. Nilewski,
E. M. Carreira, Eur. J. Org. Chem. 2012, 1685; c) T. Umezawa, F.
Matsuda, Tetrahedron Lett. 2014, 55, 3003.
5] For syntheses of (+)-danicalipin A, see: a) D. K. Bedke, G. M.
Shibuya, A. Pereira, W. H. Gerwick, T. H. Haines, C. D.
Vanderwal, J. Am. Chem. Soc. 2009, 131, 7570; b) T. Umezawa,
M. Shibata, K. Kaneko, T. Okino, F. Matsuda, Org. Lett. 2011, 13,
[
9
04; c) T. Yoshimitsu, R. Nakatani, A. Kobayashi, T. Tanaka,
[35]
cell death is rare for a lipid outside of sterols. At present
there is much uncertainty about the mechanism of action of
Org. Lett. 2011, 13, 908; For a synthesis of malhamensilipin A,
see: d) D. K. Bedke, G. M. Shibuya, A. Pereira, W. H. Gerwick,
6
42
www.angewandte.org
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2016, 55, 639 –643

Angewandte
Communications
Chemie
C. D. Vanderwal, J. Am. Chem. Soc. 2010, 132, 2542; For
in: M. Lautens, M. L. Maddess, E. L. O. Sauer, S. G. Ouellet,
syntheses of mytilipin A, see: e) C. Nilewski, R. W. Geisser,
E. M. Carreira, Nature 2009, 457, 573; f) T. Yoshimitsu, N.
Fukumoto, R. Nakatami, N. Kojima, T. Tanaka, J. Org. Chem.
Org. Lett. 2002, 4, 83.
[18] a) H. B. Henbest, R. A. L. Wilson, J. Chem. Soc. 1957, 1958;
b) B. E. Rossiter, T. R. Verhoeven, K. B. Sharpless, Tetrahedron
Lett. 1979, 20, 4733.
2
010, 75, 5425; g) W. Chung, J. S. Carlson, D. K. Bedke, C. D.
Vanderwal, Angew. Chem. Int. Ed. 2013, 52, 10052; Angew.
Chem. 2013, 125, 10236; For a general approach towards
[19] T. Yoshimitsu, N. Fukumoto, T. Tanaka, J. Org. Chem. 2009, 74,
696.
(
+)-danicalipin A, malhamensilipin A, and mytilipin A, see:
h) W. Chung, J. S. Carlson, C. D. Vanderwal, J. Org. Chem.
014, 79, 2226; For a synthesis of mytilipin B, see: i) C. Nilewski,
[20] It is unclear at this point in time how dehalogenation occurs. It
may very well be that trace metal contamination leads to metal
borides that effect CÀCl reduction, see: B. Ganem, J. O. Osby,
2
N. R. Deprez, T. C. Fessard, D. B. Li, R. W. Geisser, E. M.
Chem. Rev. 1986, 86, 763.
Carreira, Angew. Chem. Int. Ed. 2011, 50, 7940; Angew. Chem.
[21] G. W. Kabalka, T. M. Shoup, N. G. Goudgaun, J. Org. Chem.
1989, 54, 5930.
2
011, 123, 8087.
[
[
6] E. I. Mercer, C. L. Davies, Phytochemistry 1979, 18, 457.
7] a) L. L. Chen, M. Pousada, T. H. Haines, J. Biol. Chem. 1976,
[22] D. B. Dess, J. C. Martin, J. Org. Chem. 1983, 48, 4155.
[23] a) H. C. Brown, P. K. Jadhav, J. Am. Chem. Soc. 1983, 105, 2092;
b) H. C. Brown, K. S. Bhat, R. S. Randad, J. Org. Chem. 1987, 52,
319.
2
51, 1835; b) L. L. Chen, T. H. Haines, J. Biol. Chem. 1976, 251,
1
828.
[
[
8] B. Alberts, A. Johnson, J. Lewis, D. Morgan, M. Raff, K. Roberts,
P. Walter in Molecular Biology of the Cell, Garland Science
Taylor & Francis, New York, 2015, pp. 565 – 640.
9] A statement corroborated by the requirement for very polar
conditions during chromatographic purification. See referen-
ces [5a–c,g], or the Supporting Information of this paper.
[24] a) B. Haveaux, A. Dekoker, M. Rens, A. R. Sidani, J. Toye, L.
Ghosez, Org. Synth. 1980, 59, 26; b) F. Munyemana, A. Frisque-
Hesbain, A. Devos, L. Ghosez, Tetrahedron Lett. 1989, 30, 3077.
[25] a) J. Louie, C. W. Bielawski, R. H. Grubbs, J. Am. Chem. Soc.
2001, 123, 11312; b) J. Cossy, F. Bargiggia, S. BouzBouz, Org.
Lett. 2003, 5, 459.
[
[
10] G. van Meer, D. R. Voelker, G. W. Feigenson, Nat. Rev. Mol. Cell
Biol. 2008, 9, 112.
[26] J. Elovson, Biochemistry 1974, 13, 3483.
[27] For a previous synthesis of 17, see: G. L. Mayers, M. Pousada,
T. H. Haines, Biochemistry 1969, 8, 2981.
[28] M. J. Lydon, K. D. Keeler, D. B. Thomas, J. Cell. Physiol. 1980,
102, 175.
[29] M. P. Bova, D. Tam, G. MeMahon, M. N. Mattson, Toxicol. Lett.
2005, 155, 41.
11] a) W. Dowhan, M. Bogdanov, E. Mileykovskaya in Biochemistry
of Lipids, Lipoproteins and Membranes (Eds.: D. E. Vance, J. E.
Vance), Elsevier, San Diego, 2008, pp. 1 – 37; b) D. R. Voelker in
Biochemistry of Lipids, Lipoproteins and Membranes (Eds.:
D. E. Vance, J. E. Vance), Elsevier, San Diego, 2008, pp. 441 –
4
84.
[30] B. Alberts, A. Johnson, J. Lewis, D. Morgan, M. Raff, K. Roberts,
P. Walter in Molecular Biology of the Cell, Garland Science
Taylor & Francis Group, New York, 2015, pp. 1263 – 1296.
[31] T. Mosmann, J. Immunol. Methods 1983, 65, 55.
[32] Heat led to bacterial death and, in doing so, exposed genetic
material enabling the Hoechst 33342 stain to bind to it.
[33] E. J. Dufourc, J. Chem. Biol. 2008, 1, 63.
[34] M. R. Vist, J. H. Davis, Biochemistry 1990, 29, 451.
[35] Increased membrane permeability has been observed with
anionic surfactants, specifically, alkyl sulfates, and fatty acids.
The effects of these compounds on human cellular membranes
have been studied. For examples, see: a) F. Yamaguchi, S.
Watanabe, F. Harada, M. Miyake, M. Yoshida, T. Okano, J. Oleo
Sci. 2014, 63, 995; b) A. C. Chao, J. V. Nguyen, M. Broughall, A.
Griffin, J. A. Fix, P. E. Daddona, Int. J. Pharm. 1999, 191, 15;
c) E. R. Cooper, J. Pharm. Sci. 1984, 73, 1153.
[
12] For reviews, see: a) P. Ciminiello, E. Fattorusso, Eur. J. Org.
Chem. 2004, 2533; b) T. Yasumoto, M. Murata, Chem. Rev. 1993,
9
3, 1897. For additional information, see references [2,5b,7,13].
[
[
13] T. Kawahara, Y. Kumaki, T. Kamada, T. Ishii, T. Okino, J. Org.
Chem. 2009, 74, 6016.
14] Although other lipids are proposed to be active in the
membrane, their presence in low quantities has led to the
suggestion that they have little effect on overall characteristics of
membrane structure, see: a) A. E. Brown, J. Elovson, Biochem-
istry 1974, 13, 3476; b) M. C. Gershengorn, A. R. H. Smith, G.
Goulston, L. J. Goad, T. W. Goodwin, T. H. Haines, Biochem-
istry 1968, 7, 1698.
15] B. K. Keitz, K. Endo, P. R. Patel, M. B. Herbert, R. H. Grubbs, J.
Am. Chem. Soc. 2012, 134, 693.
16] a) S. Hu, S. Jayaraman, A. C. Oehlschlager, J. Org. Chem. 1996,
[
[
6
1, 7513; b) S. Hu, S. Jayaraman, A. C. Oehlschlager, J. Org.
Chem. 1998, 63, 8843.
[
17] The Oeschlager– Brown reagent is commercially available. We
found optimal yield and selectivity are obtained with freshly
prepared material. In our hands, the best procedure can be found
Received: September 28, 2015
Published online: November 27, 2015
Angew. Chem. Int. Ed. 2016, 55, 639 –643
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
643