Chiral aggregates formed from methylated tetraenoic fatty acids: Formation of both antipodes of chiral aggregates from a single enantiomer and time-dependent stereomutation

Article abstract of DOI:10.1021/ol062834d

(Chemical Equation Presented) Formation and behaviors of chiral aggregates of methylated tetraenoic fatty acids (TE-FAs) were reported. In the C-24 TE-FA series, the aggregate prepared by dispersing an ethanol stock solution of C-24 TE-FA 1b into sodium phosphate buffer gave a strong positive Cotton effect, whereas the aggregate prepared from a methanol stock solution gave a negative Cotton effect. In the C-20 TE-FA series, the aggregate prepared from an ethanol stock solution of C-20 TE-FA 2b exhibited time-dependent chirality inversion.

Full text of DOI:10.1021/ol062834d

ORGANIC
LETTERS
2007
Vol. 9, No. 2
319-322
Chiral Aggregates Formed from
Methylated Tetraenoic Fatty Acids:
Formation of Both Antipodes of Chiral
Aggregates from a Single Enantiomer
and Time-Dependent Stereomutation
Jianguo Ma, Hwan-Sung Cheon, and Yoshito Kishi*
Department of Chemistry and Chemical Biology, HarVard UniVersity,
12 Oxford Street, Cambridge, Massachusetts 02138
kishi@chemistry.harVard.edu
Received November 21, 2006
ABSTRACT
Formation and behaviors of chiral aggregates of methylated tetraenoic fatty acids (TE-FAs) were reported. In the C-24 TE-FA series, the
aggregate prepared by dispersing an ethanol stock solution of C-24 TE-FA 1b into sodium phosphate buffer gave a strong positive Cotton
effect, whereas the aggregate prepared from a methanol stock solution gave a negative Cotton effect. In the C-20 TE-FA series, the aggregate
prepared from an ethanol stock solution of C-20 TE-FA 2b exhibited time-dependent chirality inversion.
We recently reported the aggregate-forming process of
tetraenoic fatty acids (TE-FAs) and the properties of resultant
aggregates.1 The aggregate formation of C-24 TE-FA 1a
proceeds through the states of dynamic K-aggregate, static
T1-aggregate, and static T2-aggregate. The step of dynamic
K-aggregate formation is reversible, whereas the following
steps, i.e., dynamic K-aggregatefstatic T1-aggregate and
static T1-aggregatefstatic T2-aggregate, are (almost) irrevers-
ible (Figure 1).2 We suggested that these steps may be
universally involved in aggregation of TE-FAs. On the
balance of enthalpy and entropy, TE-FAs exhibit a unique
property, depending on the FA backbone length. Among a
number of TE-FAs studied, C-24 TE-FA 1a was found to
be an ideal substrate to study the static T1-aggregate, whereas
(1) Wang, Y.; Ma, J.; Cheon, H.-S.; Kishi, Y. Angew. Chem., Int. Ed. In
press.
(2) On the basis of the doping experiment of C-24 TE-FA 1a with
saturated FAs, we suggested that the process of K- to T1-aggregates was
virtually irreversible. However, on the basis of the titration experiment with
sMMPs/sMGPs, we noticed a tendency for the T1-aggregate to slowly
deaggregate. Interestingly, in the C-26 TE-FA series, both experiments
demonstrated that the T1-aggregate is irreversible.1
Figure 1. Schematic representation of the aggregate-forming
process and the characteristics of resultant aggregates.1,2
10.1021/ol062834d CCC: $37.00
© 2007 American Chemical Society
Published on Web 12/20/2006

C-20 TE-FA 2a was found to be an ideal substrate to study
the dynamic K-aggregate. With this insight, we have noticed
the possibility that TE-FAs with a chiral center incorporated
in the FA backbone might form chiral aggregates. Chiral
supramolecular assemblies are known for certain classes of
compounds, most notably π-conjugated compounds.3 To the
best of our knowledge, however, monomeric fatty acids (FAs)
have not been studied for chiral aggregation.4 In this
communication, we report our observations on the chiral
aggregates formed from methylated TE-FAs 1b,c and 2b,c
(Figure 2).
Figure 2. Structure of TE-FAs.
Figure 3. UV and CD spectra of the aggregates formed from 1b,c
in sodium phosphate buffer (pH ) 7.0, 1 × 10-5 M) at room
temperature (23 ( 2 °C). Panels A-1 and B-1: UV spectra of 1b
and 1c, respectively. Panels A-2 and B-2: CD spectra of the
aggregate prepared by dispersing an ethanol stock solution of 1b
and 1c into buffer, respectively. Panels A-3 and B-3: CD spectra
of the aggregate prepared by dispersing a methanol stock solution
of 1b and 1c into buffer, respectively.
Our previous study demonstrated that TE-FA 1a in the
C-24 series forms a “static” aggregate with the stability of
at least 1 day at room temperature, and therefore, we felt
that a methylated TE-FA is a suitable substrate to test the
possibility of forming a chiral aggregate and arbitrarily chose
two TE-FAs, 1b,c.5 Thus, we synthesized TE-FAs 1b,c and
their antipodes. The optical purity of these substrates was
confirmed to be better than 95%. The UV spectrum (Figure
3, panel A-1) clearly demonstrated that 1b in aqueous
solution (sodium phosphate buffer, pH ) 7.0) exists as an
aggregate with stability comparable to that observed for TE-
FA 1a. However, unlike TE-FA 1a, the shape of the UV
spectrum is not symmetrical, thereby suggesting that an
H-aggregate is indeed formed, but it is twisted.
of the Cotton curve was found to be consistent with the
aggregate stability estimated from the UV experiment. The
antipode of TE-FA 1b exhibited the mirror image of the CD
spectrum shown in panel A-2 in Figure 3.
We then observed the surprising CD behavior of the
aggregate derived from TE-FA 1b. Namely, the CD spectrum
(panel A-3, Figure 3) of the aggregate prepared by dispersing
a methanol stock solution of TE-FA 1b into sodium
phosphate buffer gave a negatiVe Cotton effect, which has
the opposite sign of the Cotton effect shown in panel A-2.6
The CD behaviors of the aggregates prepared from ethanol
and methanol stock solutions were found to be highly
reproducible. In addition, the antipode of TE-FA 1b exhibited
the mirror images of behaviors.
Knowing that TE-FA 1a forms a static aggregate, we
assume that, once formed, the resultant aggregates remain
as formed. Then why did dispersing an ethanol or methanol
stock solution into sodium phosphate buffer result in the left-
or right-handed chiral aggregates? We speculate that it might
Encouraged by this UV experiment, we then recorded the
CD spectrum of the aggregate prepared by dispersing an
ethanol stock solution of TE-FA 1b into sodium phosphate
buffer (pH ) 7.0). To our delight, we detected a strong
positiVe Cotton effect (panel A-2, Figure 3). The stability
(3) For recent reviews, see: (a) Hoeben, F. J. M.; Jonkheijin, P.; Meijer,
E. W.; Schenning, A. P. H. J. Chem. ReV. 2005, 105, 1491. (b) Brunsveld,
L.; Folmer, B. J. B.; Meijer, E. W.; Sijbesma, R. P. Chem. ReV. 2001, 101,
4071. (c) Hill, D. J.; Mio, M. J.; Prince, R. B.; Hughes, T. S.; Moore, J. S.
Chem. ReV. 2001, 101, 3893. (d) Stone, M. T.; Heemstra, J. M.; Morre, J.
S. Acc. Chem. Res. 2006, 39, 11. (e) Pu, L. Acta Polym. 1997, 48, 116.
(4) Supramolecular aggregates of azobenzene phospholipids and related
compounds were studied by Whitten and co-workers: Song, X.; Perlstein,
J.; Whitten, D. G. J. Am. Chem. Soc. 1997, 119, 9144.
(5) In preliminary experiments, we observed that i and ii and their anti-
diastereomers exhibited the monomer-like UV absorption, thereby suggest-
ing that a methyl group at the adjacent carbon to the tetraene chromophore
interferes with the intermolecular electronic interaction of the chromophore
within a H-aggregate.
(6) Formation of both left- and right-handed aggregates from a single
enantiomer is known. See: (a) Boiadjiev, S. E.; Lightner, D. A. J. Am.
Chem. Soc. 2000, 122, 378. (b) See ref 4. (c) Langeveld-Voss, B. M. W.;
Christiaans, M. P. T.; Janseen, R. A. J.; Meijer, E. W. Macromolecules
1998, 31, 6702. (d) Sinkeldam, R. W.; van Houtem, M. H. C. J.; Pieterse,
K.; Vekemans, J. A. J. M.; Meijer, E. W. Chem.-Eur. J. 2006, 12, 6129.
(e) Bouman, M. M.; Meijer, E. W. AdV. Mater. 1995, 7, 385. (f) Satrijo,
A.; Meskers, S. C. J.; Swager, T. M. J. Am. Chem. Soc. 2006, 128, 9030.
320
Org. Lett., Vol. 9, No. 2, 2007

be attributed to the difference in solvent polarity between
ethanol and methanol; the solution structure of 1b in
methanol and ethanol might be different, and these structural
characteristics might be conserved on dispersion into sodium
phosphate buffer.7 Clearly, further work is required to shed
light on the intriguing behaviors.
The UV spectrum of TE-FA 1c was found to resemble
that of TE-FA 1b, except for one aspect: the time-course
study of UV behaviors showed that the aggregate formed
from TE-FA 1c is less stable than the corresponding
aggregate formed from TE-FA 1b (panel B-1, Figure 3). The
observed difference in stability may be attributed to the fact
that the packing near the head of the TE-FA is tighter than
the packing near the tail, and therefore the chiral methyl
group at C-3 causes more severe disruption on aggregate
formation/stability than that at C-16. The CD spectrum of
TE-FA 1c was found to resemble that of TE-FA 1b, except
for two aspects (panel B-2, Figure 3). First, consistent with
the time-course study of the UV spectrum, the chiral
aggregate in the 1c series was found to be less stable than
the chiral aggregate in the 1b series. Second, both aggregates
prepared from ethanol and methanol stock solutions gave
Cotton curves with the same sign (panel B-2 vs panel B-3,
Figure 3).
Figure 4. UV and CD spectra of the aggregates formed from 2b
in sodium phosphate buffer (pH ) 7.0, 1 × 10-5 M) at room
temperature (23 ( 2 °C). Panel A-1: UV spectrum of 2b. Panel
A-2: CD spectrum of the aggregate prepared by dispersing an
ethanol stock solution of 2b into buffer. Panel A-3: CD spectrum
of the aggregate prepared by dispersing a methanol stock solution
into buffer.
Our previous study showed that TE-FA 2a in the C-20
series forms a kinetic aggregate, which is stable for several
hours up to overnight at room temperature. Through the
doping experiment with a saturated FA and a complexation
experiment with sMMP and sMGP,1,8 we demonstrated that
the aggregate in the C-20 series exhibits a vastly different
property from the aggregate in the C-24 series. For example,
upon addition of sMMP or sMGP, the former aggregate was
deaggregated, whereas the latter aggregate was not. We refer
to these properties as “dynamic” and “static”.1,2 Thus, we
anticipated that TE-FAs 2b,c (methylated analogues of 2a)
exhibit properties different from those of TE-FAs 1b,c
(methylated analogues of 1a). With this anticipation, we
synthesized both antipodes of TE-FAs 2b,c and tested their
UV behaviors, thereby demonstrating: (1) that TE-FA 2b
exists as an aggregate in sodium phosphate buffer and (2)
that this aggregate exhibits the stability anticipated from the
aggregate behavior observed for TE-FA 2a (panel A-1,
Figure 4).
The aggregate prepared by dispersing an ethanol stock
solution of TE-FA 2b into phosphate buffer exhibited a
surprising CD behavior (panel A-2, Figure 4). Extrapolating
from the TE-FA 1b case, we anticipated that the chiral
aggregate thus prepared should give a positive Cotton effect.
Indeed, the CD spectrum recorded immediately after the
sample preparation did give a positive Cotton curve. How-
ever, within 10 min, the positive Cotton curve inverted to
the negative Cotton curve, and its intensity reached a
maximum at around 20 min and then gradually decreased.10
The observed phenomenon is an example of so-called
stereomutation. Stereomutation is known for a number of
biomacromolecules11,12 but is rarely recognized for small
organic molecules.13 It is worth noting that most of the
reported examples for stereomutation are due to a change
of the surrounding environments such as solvents14-17 and
temperature16 or due to a method of sample preparation.18,19
Perhaps, the case of helical merocyanine dye nanorods
The aggregate prepared by dispersing a methanol stock
solution of TE-FA 2b into sodium phosphate buffer exhibited
the anticipated CD behavior (panel A-3, Figure 4).9 This CD
behavior compares well with that observed for TE-FA 1b,
including the handedness of the resultant chiral aggregate.
However, as anticipated from the UV experiment, the chiral
aggregate in the 2b series had a shorter lifetime than that in
the 1b series.
(10) The aggregate prepared from TE-FA 2c exhibited UV and CD
behaviors parallel with those observed for the aggregate prepared from TE-
FA 1c, except aggregate stability. For details, see Supporting Information.
(11) Stereomutation was observed for polypeptides. For example, see:
(a) Steinberg, I. Z.; Harrington, W. F.; Berger, A.; Sela, M.; Katchalski, E.
J. Am. Chem. Soc. 1960, 82, 5263. (b) Gratzer, W. B.; Rhodes, W.; Fasman,
G. D.; Biopolymers 1963, 1, 319. (c) Bidan, G.; Guilerez, S.; Sorokin, V.
AdV. Mater. 1996, 8, 157.
(12) Stereomutation was observed for nucleotides. For example, see: (a)
Pohl, F. M.; Jovin, T. M. J. Mol. Biol. 1972, 67, 375. (b) McIntosh, L. P.;
Zielinski, W. S.; Kalisch, B. W.; Pfeifer, G. P.; Sprinzl, M.; Drahovsky,
D.; van de Sande, J. H.; Jovin, T. M. Biochemistry 1985, 24, 4806.
(13) For stereomutation of small organic compounds, see, for example:
(a) ref 6e and (b) ref 4.
(14) Boiadjiev and Lightner observed a solvent-dependent chirality
inversion for an optically active dimethylmesobilirubin-XIIIR: see ref 6a.
(15) Meijer and co-workers reported at least two cases: (a) solvent
dependency of optical activity of chiral polythiophene aggregates (see ref
6c); (b) solvent dependency of chiral poly(ureidophthalimide) foldamers
(see ref 6d).
(7) An i-PrOH stock solution gave a result similar to an EtOH stock
solution.
(8) sMMP and sMGP are abbreviations of synthetic 3-O-methylmannose-
and 6-O-methylglucose-containing polysaccharides, respectively. Design and
synthesis of sMMP and sMGP. (a) sMMP: Hsu, M. C.; Lee, J.; Kishi, Y.
J. Org. Chem., submitted for publication. (b) sMGP: Meppen, M.; Wang,
Y.; Cheon, H.-S.; Kishi, Y. J. Org. Chem., submitted for publication.
(9) The antipode of 2b gave the identical CD results, but in the mirror
image.
Org. Lett., Vol. 9, No. 2, 2007
321

reported by Wu¨rthner and co-workers20 is closest to the
current case. We should specifically note that time-dependent
stereomutation, i.e., stereomutation via a metastable state,
is a very rare phenomenon and that, in this context alone,
the example reported here should warrant further studies.
As suggested, the aggregate formed from TE-FA 2b is
dynamic, and therefore an aggregate reorganization within
the H-aggregate should be feasible in the C-20 TE-FA series.
In contrary, the aggregate formed from TE-FA 1b is static,
and therefore an aggregate reorganization within the ag-
gregate is not feasible in the C-24 TE-FA series. It is worth
pointing out again that the handedness of the chiral aggregate
formed from an ethanol stock solution of 1b matches the
handedness of the chiral aggregate initially observed in the
2b series.
Structurally, we have suggested that both K- and T1-
aggregates are lamellar types in a local sense but are different
in the packing mode of the FA backbone; the K-aggregate
is a melted type, whereas the T1-aggregate is a crystalline
type (Figure 1).1 A chiral methyl group introduced on the
FA backbone would enforce the aggregate formation in a
twisted manner, to furnish the chiral lamellar-type aggregates
(Figure 5). In this context, it is worthwhile adding that
bismethylated TE-FAs, i.e., R1 ) R2 ) Me in 1 and 2 and
their anti-diastereomers, showed only very weak or no Cotton
effect,21 thereby suggesting that the two methyl groups are
Figure 5. Stereostructure for two possible modes, A and B, of
chiral lamellar-type aggregates. A shadowed plate and a solid or
broken line attached to it represent a TE-FA and a chiral methyl
group, respectively.
counteracting the enforcement of aggregate formation in one
twisted direction.
In summary, we reported the formation and properties of
chiral aggregates from monomeric methylated TE-FAs for
the first time. In a broad sense, these aggregates may be
relevant to the supramolecular assemblies of π-conjugated
and other systems.3 Interestingly, the chiral aggregates
formed from methylated TE-FAs appear to share the unique
properties/behaviors with some of these supramolecular
assemblies.
(16) Fujiki and co-workers reported solvent and temperature effects on
the chiral aggregates of poly(alkylarylsilane)s bearing remote chiral
groups: Nakashima, H.; Fujiki, M.; Koe, J. R.; Motonaga, M. J. Am. Chem.
Soc. 2001, 123, 1963.
(17) Swager and co-workers reported solvent dependency for transforma-
tion of a chiral poly(p-phenylenevinylene) derivative from solutions to
films: see ref 6f.
(18) Whitten and co-workers showed the cooling rate dependency for
chiral aggregate formation of azobenzene phospholipids: see ref 4.
(19) Bouman and Meijer reported the sample preparation dependency,
i.e., the fast- and slow-cooled polymer films prepared from optically active
regioregular polythiophenes (see ref 6e).
Acknowledgment. Financial support from Amgen, Eisai
Research Institute, and Eli Lilly is gratefully acknowledged.
Supporting Information Available: Experimental de-
tails. This material is available free of charge via the Internet
at http://pubs.acs.org.
OL062834D
(20) Lohr, A.; Lysetska, M.; Wu¨rthner, F. Angew. Chem., Int. Ed. 2005,
44, 5071.
(21) For details, see Supporting Information.
322
Org. Lett., Vol. 9, No. 2, 2007