Aggregation behavior of tetraenoic fatty acids in aqueous solution

Article abstract of DOI:10.1002/anie.200603979

(Chemical Equation Presented) Why so blue? Unique blue-shifted UV absorptions were used to follow the aggregation of C₂₄ tetraene fatty acids in aqueous solution. Aggregation takes place through three distinct states (i.e. K→T¹→T²; see picture). It is suggested that all of these aggregates are lamellar-type in a local sense but differ in the packing mode of the fatty acid backbone.

Full text of DOI:10.1002/anie.200603979

Angewandte
Chemie
DOI: 10.1002/anie.200603979
Fatty Acid Aggregation
Aggregation Behavior of Tetraenoic Fatty Acids in Aqueous
Solution**
Yonghui Wang, Jianguo Ma, Hwan-Sung Cheon, and Yoshito Kishi*
In connection with an ongoing program in this laboratory, we
synthesized tetraenoic fatty acids (TE-FAs)[1] as a probe[2] for
the study of association/dissociation events between fatty
acids (FAs) and synthetic 3-O-methyl-d-mannose-containing
polysaccharides (sMMP)[3,4] and synthetic 6-O-methyl-d-glu-
cose-containing polysaccharides (sMGP).[4,5] During spectro-
scopic characterization of TE-FAs, we observed an interesting
phenomenon: C20-t,t,t,t-TE-FA (1c) exhibits an expected UV
Figure 1. UV absorption spectra of C20-t,t,t,t-TE-FA (1c; 2”10ꢁ6 m) at
(23ꢂ2)8C. A) UV spectra in a) aqueous solution (0.05m sodium
phosphate buffer, pH 7) and in b) methanol. B) UV spectra in aqueous
solution with increasing methanol content (0!25% v/v).
back to the normal absorption of tetraenes.[6,7] These results
support the view that the blue-shifted UV absorption of 1c in
aqueous solution is indeed associated with aggregation.
The examples known in the literature suggest that the
observed blue-shifted UV absorption may be attributed to an
H-aggregation of 1c in buffer solution. An H-aggregate
contains a card-pack orientation of 1c, and according to
exciton theory[8] the H-aggregate excitonic splitting results in
a high-energy allowed transition and a low-energy forbidden
transition (Figure 2). Thus, the UV absorption of 1c is blue-
absorption at 303 nm with fine structure in methanol but a
dramatically different UV absorption at 250 nm without fine
structure in aqueous solution (Figure 1A).
The observed blue shift in the UVabsorption appeared to
be associated with aggregation of 1c in buffer solution.
Aggregate formation is driven by hydrophobic interaction
and is dependent on the polarity of solvents. Figure 1B shows
overlaid UV spectra of 1c in buffer containing an increasing
amount of methanol. As expected, with the increase of
methanol content the unique UV absorption at 250 nm
changed to the normal absorption of tetraene with fine
structure. Similarly, the UV experiment demonstrated that
aggregate formation of 1c is concentration-dependent, with
the estimated threshold concentration of aggregation being
approximately 0.3mm.[6] Lastly, sMMP and sMGP were shown
to form a 1:1 complex with 1c in aqueous solution, resulting in
disaggregation and bringing the blue-shifted UV absorption
Figure 2. Explanation for the blue-shifted UV absorption of H-aggre-
gates according to exciton interaction theory.
shifted. Related to this work, we should specifically mention
that Whitten and co-workers have shown that FAs which
incorporate various chromophores form H-aggregates char-
acterized by a blue-shifted UV absorption.[9,10]
[*] Dr. Y. Wang, Dr. J. Ma, Dr. H.-S. Cheon, Prof. Dr. Y. Kishi
Department of Chemistry and Chemical Biology
Harvard University
With the unique observations on C20-t,t,t,t-TE-FA (1c), we
selected TE-FAs 1a–f and 2a–c systematically to study the
effects of chain length and tetraene geometry on aggregation
behaviors.[1] All the TE-FAs exhibited the expected UV
absorption with fine structure in methanol (1a–f: lmax
ꢀ 300 nm; 2a–c: lmax ꢀ 304 nm). The unique blue-shifted
UV absorption was observed for the C20 (1c), C22 (1d), C24
12 Oxford Street, Cambridge, MA 02138 (USA)
Fax: (+1)617-495-5150
E-mail: kishi@chemistry.harvard.edu
[**] Financial support from the National Institutes of Health (NS12108)
and Eisai Research Institute is gratefully acknowledged.
Supporting information for this article is available on the WWW
under http://www.angewandte.org or from the author.
Angew. Chem. Int. Ed. 2007, 46, 1333 –1336
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1333

Communications
(1e), and C26 (1 f) fatty acids in the t,t,t,t-TE-FA series, but
freshly prepared sample of 1c displayed a blue-shifted
absorption at 250 nm (curve a, Figure 3B) corresponding to
the K-aggregate of 1e. However, on allowing the sample to
stand at room temperature for several hours or overnight, the
UV spectrum changed to a broad absorption (curve b,
Figure 3B) corresponding to the T2-aggregate of 1e.[13]
Although no direct experimental support is available, we
assume that, like in the C24-TE-FA (1e) series, the T2-
aggregate was formed via the T1-aggregate but it did not
accumulate in the 1c TE-FA series.
only for the C24 fatty acid (2c) in the c,t,t,c-TE-FA series,[6]
thereby demonstrating that both chain length and tetraene
geometry affect aggregate formation.[11]
To gain insight into the aggregate structure, we studied the
UV spectral behavior of three C24 trienoic fatty acids (TriE-
FAs, 3a–c) with the chromophore located at a different
With these results, we then studied the properties of three
different types of aggregates formed from t,t,t,t-TE-FAs. First,
we conducted the doping experiments of K-, T1-, and T2-
aggregates with saturated (sat-) FAs. As discussed, the unique
blue-shifted UV absorption of K- and T1-aggregates can be
attributed to the intermolecular electronic interactions of the
tetraene chromophore within the head-to-head H-aggregates.
Then, on doping these aggregates with a sat-FA, one expects
that the blue-shifted UV absorption should be affected if the
sat-FA penetrates into these aggregates. On addition of sat-
FA, the H-aggregate of 1c was conserved, but its composition
changed from a homogeneous to a heterogeneous aggregate
in which the intermolecular electronic interaction between
the tetraene chromophores was prevented. Upon addition of
sat-C20-FA, the K-aggregate that formed from C20-t,t,t,t-TE-
FA (1c) exhibited the UV behavior fully consistent with this
expectation (Figure 4A). This property of K-aggregates is
referred to as “dynamic”. Interestingly, a clear trend was
observed that the efficiency of penetration is dependent on
the size-matching of sat-FA with C20-t,t,t,t-TE-FA (1c): C20-
sat-FA was far more effective than C16-sat-FA, and C12-sat-FA
had virtually no effect.[6]
position of the FA backbone.[1] Regardless of the location of
the chromophore, all the C24-TriE-FAs exhibited a character-
istic blue-shifted UV absorption, thereby demonstrating that
the aggregate in question is formed in a head-to-head manner
but not a head-to-tail manner.[12]
During the UV spectroscopic study, we noticed an
intriguing behavior of C24-t,t,t,t-TE-FA (1e) in aqueous
solution: the sample revealed a UV absorption maximum at
248 nm immediately after its preparation, but after allowing
the sample to stand overnight the maximum was shifted to
235 nm (Figure 3A, curves a and b). Apparently, the initially
Upon addition of C20-sat-FA, the broad UV absorption of
the T2-aggregate of 1c showed virtually no change (Fig-
ure 4B). This UV spectral behavior is indicative that the T2-
aggregate is not penetrated by C20-sat-FA, and this property is
referred to as “static”.
Figure 3. UV spectral behavior of TE-FAs in aqueous solution (0.05m
sodium phosphate buffer, pH 7.0) at (23ꢂ2)8C. A) C24-TE-FA (1e):
a) freshly prepared solution; b) after 3 h; c) after one week. B) C20-TE-
FA (1c): a) freshly prepared solution; b) after 3 h.
formed aggregate (designated as the K-aggregate) changed
into the thermodynamically more stable aggregate (desig-
nated as the T1-aggregate). The difference in the UV
absorption maximum indicated that the T1-aggregate displays
a higher degree of intermolecular electronic interaction
between the tetraene chromophores than the K-aggregate,
thereby suggesting that the T1-aggregate exhibits a tighter
packing than the K-aggregate. Interestingly, the third aggre-
gate state (designated as the T2-aggregate) was also detected;
when the buffer solution was allowed to stand at room
temperature for one week, the UV spectrum changed to the
spectrum shown as curve c in Figure 3A. The observed broad
absorptions indicate that this aggregate is not well ordered.[13]
Interestingly, the UV spectral behavior of C20-t,t,t,t-TE-
FA (1c) is different to that of C24-t,t,t,t-TE-FA (1e). The
Figure 4. Doping experiments of TE-FA aggregates with sat-FA in
aqueous solution (0.05m sodium phosphate buffer, pH 7.0) at
(23ꢂ2)8C. A) Doping the K-aggregate of C20-t,t,t,t-TE-FA (1c) with C20-
sat-FA. B) Doping the T2-aggregate of 1c with C20-sat-FA. C) Doping
the mixture of K- and T1-aggregates of C24-t,t,t,t-TE-FA (1e) with C24-
sat-FA. D) Doping the T2-aggregate of 1e with C24-sat-FA.
1334
www.angewandte.org
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2007, 46, 1333 –1336

Angewandte
Chemie
Figure 4C illustrates the UV spectral behavior of the
aggregates of C24-TE-FA (1e). Before addition of C24-sat-FA,
the UV spectrum showed that 1e existed as a mixture of the
K-aggregate (lmax = 248 nm) and the T1-aggregate (lmax
=
235 nm). Upon addition of C24-sat-FA, only the K-aggregate
exhibited the same behavior as observed for the K-aggregate
of C20-TE-FA (1c), whereas the T1-aggregate remained
virtually unchanged. Interestingly, as with the C20-TE-FA
series, a clear trend in the size-matching preference was again
noticed.[6] Lastly, the UV experiment summarized in Fig-
ure 4D showed that the T2-aggregate is static. Overall, these
doping experiments demonstrated that the K-aggregate is
dynamic, whereas the T1- and T2-aggregates are almost and
practically static, respectively.[14] This conclusion is further
supported by titration experiments of the K-, T1-, and T2-
aggregates with sMGP and sMMP.
sMMP and sMGP are the structural analogues of natural
MMP and MGP.[4] Note that, unlike natural MMP and MGP,
these synthetic analogues are available as chemically well-
defined and homogeneous materials. Also relevant to the
Figure 5. Titration of TE-FAs (2”10ꢁ6 m) with sMGP 20-mer in aque-
ous solution (0.05m sodium phosphate buffer, pH 7.0) at (23ꢂ2)8C.
A) Titration of the K-aggregate of C20-t,t,t,t-TE-FA (1c) with sMGP 20-
mer. B) Titration of the T2-aggregate of C20-t,t,t,t-TE-FA (1c) with sMGP
20-mer. C) Titration of the mixture of K- and T1-aggregates of 1e with
sMGP 20-mer. D) Titration of the T1-aggregate of 1e with sMGP 20-
mer. See Supporting Information for the titration of the T2-aggregate of
1e with sMGP 20-mer.
K-, T1-, and T2-aggregates are lamellar-type in a local sense
but different in the packing-mode of the FA backbone: the K-
aggregate is a melted type, the T1-aggregate is a crystalline
type, and the T2-aggregate is a disordered type with no
defined mode of intermolecular electronic interaction
between the tetraene chromophores (Figure 6). The experi-
present work, we should mention that, like natural MMP and
MGP, these synthetic analogues form a 1:1 complex with at
least monomeric TE-FAs. This unique property was used to
study the dynamic and static natures of K-, T1-, and T2-
aggregates.
Figure 5A summarizes the titration experiment of the K-
aggregate of C20-TE-FA (1c) with sMGP 20-mer (n = 20). On
addition of sMGP, the characteristic blue-shifted UV absorp-
tion changed to that normally found for tetraenes. Owing to
the complexation with sMGP, the K-aggregate is disaggre-
gated which gives rise to the normal absorption for tetraenes.
Figure 5B presents the titration result of the T2-aggregate of
C20-TE-FA (1c) with sMGP 20-mer. On addition of sMGP,
the characteristic broad UV absorption of 1c was almost
unaffected. These results support the dynamic and static
natures as concluded from the doping experiments with sat-
FAs. Likewise, the titration experiments of C24-TE-FA (1e)
with sMGP 20-mer (Figure 5C,D) gave the results consistent
with the previous conclusion; that is, the K-aggregate is
dynamic, whereas T1- and T2-aggregates are almost and
practically static, respectively.[14]
Figure 6. Schematic representation of the formation process and
characteristics of the aggregates.
ments on C24-TE-FA (1e) uncovered the aggregate-forming
processes; that is, an initial formation of the dynamic K-
aggregate, which is transformed first into the static T1-
aggregate and then into the static T2-aggregate.[15] The step
of dynamic K-aggregate formation is reversible, whereas the
following steps (dynamic K-aggregate!static T1-aggregate
and static T1-aggregate!static T2-aggregate) are (almost)
irreversible.[14]
We suggest that these steps are universally involved in the
aggregation of TE-FAs and likely sat-FAs as well. To test this
notion, we studied the UV spectral behavior of a series of TE-
FAs at a concentration of 2mm in aqueous solution (sodium
phosphate buffer, pH 7.0) at (23 ꢂ 2)8C (Figure 7). Under this
condition, C16-TE-FA (1a) exists as a monomer, C18-TE-FA
(1b) exists as a mixture of monomer and K-aggregate, and
Overall, three K-, T1-, and T2-aggregates were detected by
UV spectroscopy. The reported results demonstrate that
1) both K- and T1-aggregates of TE-FAs adopt head-to-head
H-aggregates, 2) the T1-aggregate is more tightly packed than
the K-aggregate, 3) the K-aggregate is dynamic, but the T1-
aggregate is almost static, and 4) the T2-aggregate, charac-
terized by its broad UV absorption, is thermodynamically
most preferred and static. At present, we have no data to
suggest a global structure or even an average molecular size
for an aggregate. However, we would speculate that all of the
Angew. Chem. Int. Ed. 2007, 46, 1333 –1336
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
1335

Communications
[3]For the isolation and characterization of MMP, see: a) G. R.
Gray, C. E. Ballou, J. Biol. Chem. 1971, 246, 6835; b) M. Ilton,
A. W. Jevans, H. B. White III, D. Vance, K. Bloch, Proc. Natl.
Acad. Sci. USA 1971, 68, 87; c) S. K. Maitra, C. E. Ballou, J. Biol.
Chem. 1977, 252, 2459, and references therein.
[4]The design and synthesis of sMMP and sMGP will be reported
elsewhere. a) M. C. Hsu, J. Lee, Y. Kishi, unpublished results;
b) M. Meppen, Y. Wang, H.-S. Cheon, Y. Kishi, unpublished
results.
[5]For the isolation and characterization of MGLP/MGP, see:
a) Y. C. Lee, C. E. Ballou, J. Biol. Chem. 1964, 239, PC3602;
b) L. S. Forsberg, A. Dell, D. J. Walton, C. E. Ballou, J. Biol.
Chem. 1982, 257, 3555, and references therein. On the basis of
the structure of the polysaccharide isolated from Mycobacterium
bovis BCG, Rivi›re suggested a revision on the structure
proposed by Ballou; see: G. Tuffal, R. Albigot, M. Rivi›re, G.
Puzo, Glycobiology 1998, 8, 675. Regarding the structural
heterogeneity of MGP, Ballou commented that at least two
forms of MGP containing 21 hexoses exist; see: K. Kamisango,
A. Dell, C. E. Ballou, J. Biol. Chem. 1987, 262, 4580; see also: G.
Tuffal, R. Albigot, B. Monsarrat, C. Ponthus, C. Picard, M.
Rivi›re, G. Puzo, J. Carbohydr. Chem. 1995, 14, 631.
Figure 7. Summary of the steps involved in the aggregation of TE-FAs
1a–f in aqueous solution (0.05m sodium phosphate buffer, pH 7.0) at
2”10ꢁ6 m at (23ꢂ2)8C. A solid vertical line represents an aggregate
that is detectable by UV spectroscopy, whereas a dashed vertical line
represents an aggregate that is not detectable by UV spectroscopy.
C20- (or longer) TE-FAs exist as aggregates. The behavior/
properties of C20-and C22-TE-FAs compared well. However,
in the C26-TE-FA series the K-aggregate was not detected by
UV spectroscopy, and the T1-aggregate formed was found to
be stable at room temperature at least for one week.[13]
Overall, on the balance of enthalpy and entropy, each TE-
FA exhibited a unique behavior.
Last, we note that the insights gained through this study
have led us to 1) uncover intriguing behavior of the chiral
aggregates formed from TE-FA derivatives[16] and 2) develop
a simple method to follow the association/dissociation events
between TE-FAs and sat-FAs with sMMP and sMGP.[7]
[6]See Supporting Information for details.
[7]UV spectroscopy studies showed that 1c forms a 1:1 complex
with sMMP 16-mer and sMGP 16-mer: H.-S. Cheon, Y. Wang, J.
Ma, Y. Kishi, ChemBioChem 2007, DOI: 10.1002/
cbic.200600499.
[8]a) R. M. Hochstrasser, M. Kasha, Photochem. Photobiol. 1964,
3, 317; b) M. Kasha, H. R. Rawls, M. A. El-Bayoumi, Pure Appl.
Chem. 1965, 11, 371; c) for a review on this subject, see: P. W.
Bohn, Annu. Rev. Phys. Chem. 1993, 44, 37.
[9]a) D. G. Whitten, Acc. Chem. Res. 1993, 26, 502; b) N. C. Maiti,
S. Mazumdar, N. Periasamy, J. Phys. Chem. B 1998, 102, 1528.
[10]For the aggregation of fatty acids containing an aromatic system,
see: a) trans-stilbene: X. Song, C. Geiger, M. Farahat, J.
Perlstein, D. G. Whitten, J. Am. Chem. Soc. 1997, 119, 12481;
b) trans-azobenzene: X. Song, J. Perlstein, D. G. Whitten, J. Am.
Chem. Soc. 1997, 119, 9144; c) diphenylpolyene: S. P. Spooner,
D. G. Whitten, J. Am. Chem. Soc. 1994, 116, 1240; d) phenyl/
biphenyl/terphenyl: H. C. Geiger, J. Perlstein, R. J. Lachicotte,
K. Wyrozebski, D. G. Whitten, Langmuir 1999, 15, 5606.
[11]In preliminary studies, we have noticed that 1) the position of the
chromophore in the backbone, 2) the cation of phosphate buffer
(i.e. Na+ vs Mg2+), and 3) the pH (pH 3.0, 7.0, and 10.0)
significantly affect the formation and properties of the aggre-
gates.
[12]The aggregate formed from a 1:1 mixture of 3a and 3c did not
give a characteristic blue-shifted UV peak, but rather gave a
broad UV absorption. See Supporting Information for details.
[13]On dilution with methanol, the sample with a broad UV
absorption gave the normal UVabsorption for tetraenes, thereby
verifying that the broad UV absorption does not result from
decomposition of the TE-FA.
Received: September 28, 2006
Revised: November 10, 2006
Published online: January 9, 2007
Keywords: aggregation · chromophores · doping · fatty acids ·
.
micelles
[1]See Supporting Information for the syntheses of TE-FAs and
TriE-FAs.
[2]C 18-c,t,t,c- and C18-t,t,t,t-TE-FAs I and II are known as a- and b-
parinaric acids, respectively. Both I and II were isolated from the
plant Parinarium glaberrimum [a) J. P. Riley, J. Chem. Soc. 1950,
12; b) L. A. Sklar, B. S. Hudson, R. D. Simoni, Proc. Natl. Acad.
Sci. USA 1975, 72, 1649]and were used as a fluorescent probe for
monitoring the complexation between methylated polysaccha-
rides and fatty acids: c) K. K. Yabusaki, C. E. Ballou, Proc. Natl.
Acad. Sci. USA 1978, 75, 691; d) T. Kiho, C. E. Ballou,
Biochemistry 1988, 27, 5824.
[14]On the basis of the doping experiment with sat-FAs, we suggest
that the process of K- to T1-aggregates is virtually irreversible.
From the titration experiment with sMMPs/sMGPs, however, we
noticed that the T1-aggregate probably disaggregates slowly.
Interestingly, in the C26-TE-FA series, both experiments dem-
onstrated that the T1-aggregate is irreversible. For details, see
Figure 10S in the Supporting Information.
[15]In the aggregation of C 20-TriE-FAs 4a and 4c, all of the K-, T1-,
and T2-states were detected by UV spectroscopy. See Figure 7S
in the Supporting Information.
[16]The aggregation behavior of chiral tetraenoic fatty acids will be
reported elsewhere: J. Ma, H.-S. Cheon, Y. Kishi, Org. Lett. 2007,
in press.
1336
www.angewandte.org
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2007, 46, 1333 –1336