Rigid-rod β-barrels as lipocalin models: Probing confined space by carotenoid encapsulation

Article abstract of DOI:10.1002/(SICI)1521-3765(20000515)6:10<1739::AID-CHEM1739>3.0.CO;2-Y

Herein, we describe the design, synthesis, structure, and function of synthetic, supramolecular β-barrel models. Assembly of octi(p-phenylene)s with complementary -Lys-Leu-Lys-NH₂ and Glu-Leu-Glu-NH₂ side chains yielded water-soluble

Full text of DOI:10.1002/(SICI)1521-3765(20000515)6:10<1739::AID-CHEM1739>3.0.CO;2-Y

FULL PAPER
Rigid-Rod b-Barrels as Lipocalin Models: Probing Confined Space by
Carotenoid Encapsulation
Bodo Baumeister[a] and Stefan Matile*[a,b]
Abstract: Herein, we describe the de-
sign, synthesis, structure, and function of
synthetic, supramolecular b-barrel mod-
els. Assembly of octi(p-phenylene)s with
complementary -Lys-Leu-Lys-NH2 and -
Glu-Leu-Glu-NH2 side chains yielded
water-soluble rigid-rod b-barrels of pre-
cise length and with flexible diameter. A
hydrophobic interior was evidenced by
guest encapsulation. Host ± guest com-
J-aggregates surrounded by about do-
decameric rigid-rod ªbicycle tiresº were
prepared from mixed micelles by dia-
lytic detergent removal. The significance
of these findings for future bioorganic
chemistry in confined, intratoroidal
space is discussed in comparison with
pertinent biological examples.
plexes with planarized, monomeric b-
carotene within tetrameric rigid-rod b-
barrels, and disc micellar astaxanthin
Keywords: bioorganic chemistry
cage compounds carotenoids
oligomers ´ peptidomimetics
´
´
´
Introduction
Toroidal structures are ubiquitous in nature.[1±9] With respect
to their remarkable functional diversity, the confined interior
of toroidal structures is of particular interest. For instance,
there is mounting evidence that the large protein assemblies
of molecular machines utilize intratoroidal space to mediate
macromolecular processes such as protein folding,[1] protein
degradation,[2] oligonucleotide binding,[3] and gene replica-
tion.[4] Among other examples, the mediation of molecular
transport across cell membranes by various membrane
proteins on the one hand[5] and binding of hydrophobic
natural products by lipocalins on the other[6±9] illustrate
biological functionality of both hydrophilic and hydrophobic
confined intratoroidal cores.
In nature, confined interiors are often created by b-
barrels.[1±9] In order to examine the uniqueness of intra-
toroidal chemistry beyond the limitations of peptide chem-
istry, we recently focused our attention on the development of
artificial b-barrels.[10] A general strategy for the design of
supramolecular rigid-rod b-barrels has been implied by
LaBrenz and Kellyꢀs pioneering study with dibenzofuran
peptide 1.[11] This artificial receptor binds the complementary
peptide 2 by forming antiparallel b-sheets with interdigitating
peptide strands in host ± guest complex 3 (Figure 1).[11] Com-
2
O
O
1
3
= -CH2CH2CO-Val-Lys-Leu-Lys-NHCH2CH2N(CH3)2
= HOOCCH2CH2CO-Glu-Leu-Glu-Leu-NH-benzyl
R
4: R = R' =
5: R = R' =
6: R = R' =
O
OH
OH
HO
R
O
O
R'
OH
OH
OH
H
N
O
7: R = R' =
8: R =
R'?
R'?
R
H
O
N
H
R'
R
O
O
R' = Me
H
N
H
N
H
O
O
=
O
H
N
O
N
H
H
72
O
(
= Na+)
Figure 1. Structures of LaBrenz and Kellyꢀs model 3, rigid-rod polyols
4 ± 6, and self-assembled rigid-rod b-barrel 72.[10±12]
[a] Prof. Dr. S. Matile, Dipl.-Chem. B. Baumeister
Department of Organic Chemistry, University of Geneva
1211 Geneva 4 (Switzerland)
parable antiparallel strand interdigitations were observed in
our laboratory during the study of rigid-rod ion-channel
models 4 ± 8.[10, 12] Rigid-rod polyol 4[12f,g] (but not the more
flexible 5 and 6)[12e] was initially suspected to form transient
toroidal supramolecules in highly polarized black lipid
membranes.[12d] To obtain permanently stable rigid-rod b-
Fax: ( ‡41)22-328-7396
E-mail: stefan.matile@chiorg.unige.ch
[b] Prof. Dr. S. Matile
Department of Chemistry, Georgetown University
Washington, DC 20057 (USA)
Chem. Eur. J. 2000, 6, No. 10
ꢁ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2000
0947-6539/00/0610-1739 $ 17.50+.50/0
1739

S. Matile, B. Baumeister
FULL PAPER
Table 1. Characteristics of carotenoid encapsulation.
Entry
Starting materials
Products
GF (SEC)[g]
cartenoid[a]
l [nm][b]
13[c]
14[d]
yield [%][e]
25[j]
l [nm][f]
x[h]
FQ [%][i]
48
1
11
436
0.5
0.5
480
‡ (‡)
4:1
[k]
2
3
11
11
436
436
0
0
0
0
±
±
O
4
5
18
18
446
446
0.5
0
0.5
0
±
18
464
OH
6
9
450
0.25
0.25
±
HO
7
8
9
9
450
450
0
0
0.5
0
40
36
388 (450)[l]
388 (450)[l]
O
OH
9
10
445
0.1
0.1
10
576
‡ (À)
1:1
29
HO
O
10
11
10
10
445
445
0
0
0.2
0
6
17
560 (524)
560 (524)
À
À
[a] 9: zeaxanthin; 10: astaxanthin; 11: b-carotene; 18: b-apo-8'-carotenal. [b] Absorption maxima of carotenoid in mixed cholate micelles. [c] [13]/
[carotenoid]. [d] [14]/[carotenoid]. [e] Determined by carotenoid absorption after dialysis and filtration. [f] Absorption maxima of isolated carotenoid in
buffer, pH 6.4; significant maxima of fine structure are given in parenthesis. [g] GF: Gel filtration (SEC: Size exclusion chromatography). [h] x ˆ stoichi-
ometry oligo(p-phenylene)/carotenoid (determined by spectroscopy). [i] FQ: Quenching of oligo(p-phenylene) fluorescence. [j] Quantitative with respect to
oligo(p-phenylene) concentration. [k] Yields <5% were not considered. [l] Relative intensity of the two maxima was concentration dependent as in ref. [18].
barrels, we subsequently replaced the lateral glycol side chains
in 4 by the lateral diamides in 7.[10] The resulting dimer 72 was
stable in polar and nonpolar solvents.[10] The lack of similar
self-organization of octi(p-phenylene) 8 with four instead of
eight lateral diamide chains demonstrated the importance of
multiple intermolecular interactions along the preorganizing
rigid-rod scaffold for molecular architecture.[10]
Self-assembled rigid-rod b-barrel 72 contains a lipophilic
surface and hydrophilic intratoroidal space for multiple cation
binding, that is, toroidal amphiphilicity suitable for trans-
membrane ion transport.[5, 10] In the present study, we focus on
both inversion of the above toroidal amphiphilicity and
expansion of rigid-rod b-barrels, that is, on rigid-rod lipocalin
models.
Lipocalins are relatively small proteins associated with
pheromone activity, invertebrate coloration, olfaction, and
gustation and have an eight-stranded antiparallel b-barrel in
common.[6±9] Lipocalins are attracting increasing attention as
universal ligand-binding proteins, a feature considered as
unique for immunoglobulins until recently.[6] Among lipoca-
lins, carotenoproteins were of particular interest for the
present study.[7, 8] Photoprotective xanthophyll-cycle enzymes
that catalyze epoxidation and de-epoxidation of zeaxanthin 9
and violaxanthin, respectively, were identified as lipocalins
with intratoroidal carotenoid binding-sites up to 40 Š deep
(see Table 1 for carotenoid structures).[7] The classical car-
otenoproteins from the lobster carapace that account for its
characteristic blue color, crustacyanins, are composed of
Â
Á
Abstract in French: Nous decrivons la conception, la synthese,
Á
Â
la structure et la fonction de modeles supramoleculaires et
lipocalin heterodimers with two internal astaxanthins 10.[8]
A
Â
synthetiques pour des barriques b. Lꢀassemblage des octa(p-
comparable case was reported for b-lactoglobulin, a lipocalin
abundant in milk that binds and stabilizes one vitamin A per
b-barrel.[9]
Á
Â
Â
phenylene)s complementaires avec les chaînes laterales -Lys-
Â
Leu-Lys-NH2 et -Glu-Leu-Glu-NH2 a donne des barriques b
Â
Â
Â
faites de »baguettes moleculaires« qui sont caracterisees par
The dependence of the spectroscopic properties of carot-
enoids on their local environment identified these chromo-
phores as superb, hydrophobically matching probes for the
interior of rigid-rod b-barrels.[13±18] Because of its significance
in photosynthetic systems, the spectroscopic properties of b-
carotene 11 in particular have been studied in detail.[13±17] The
S0 ! S2 transition of b-carotene shows a hypsochromic shift
with increasing solvent polarity.[13] Practically no shifts but loss
of vibronic fine structure plus strong exciton coupled circular
dichroism (CD) were reported for b-carotene aggregates in
chiral environments such as lipid bilayers[14c] or reconstituted
low-density lipoprotein capsules.[14b] An increase in vibronic
Â
Á
Â
une longueur precise, un diametre flexible, et une solubilite
Â
 Â
Â
dans lꢀeau. Lꢀinterieur hydrophobe a ete prouve par encapsu-
Â
lation de substrats. Des complexes substrat-recepteur avec un b-
Á
Â
Â
carotene planarise dans une barrique b tetramerique, et des
Â
Â
disques micellaires composes dꢀagregats J de lꢀastaxanthine
entoures par des »pneus de velo« faits parÐprobablement
douzeлbaguettes« ont ete prepares a partir de micelles mixtes
en enlevant les detergents par dialyse. Lꢀimportance de ces
decouvertes pour une future chimie bioorganique dans lꢀespace
confine intratoroidal est discutee en comparaison avec des
exemples pertinents de la biologie.
Â
Â
   Á
Â
Â
Â
Â
Â
1740
ꢁ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2000
0947-6539/00/0610-1740 $ 17.50+.50/0
Chem. Eur. J. 2000, 6, No. 10

Lipocalin Models
1739±1749
fine structure together with a bathochromic shift has been
found at low temperature,[15] in single crystals,[16] and upon
ring ± chain co-planarization.[9, 17] The most striking spectro-
scopic features of carotenoids other than b-carotene include a
strong blue-shift of zeaxanthin and astaxanthin upon aggre-
gation.[18] On the other hand, a bathochromic effect has been
found for lipocalin-bound astaxanthin[8] that may be compa-
rable to the opsin shift underlying the chemistry of vision.[19]
Thus, the spectroscopic properties of carotenoids encapsu-
lated by hydrophobically matching rigid-rod b-barrels are
likely to reveal the nature of their interior.
72.[10] The length of an octi(p-phenylene) scaffold is practi-
cally identical to that of an eight-stranded b-sheet (Fig-
ure 2).
2. Spontaneous self-assembly has been identified as a major
limitation of rigid-rod b-barrel 72 for several reasons, which
include purification of the target molecule, access to
thermodynamic data, and guest encapsulation. Control
over the assembly of rigid-rod b-barrels 12n by using, in
analogy to LaBrenz and Kellyꢀs model 3,[11] two rigid-rod
molecules with complementary lateral peptides was thus
desirable.
In the following, we describe the design, synthesis, and
structure of supramolecular rigid-rod b-barrels 124 and 126
(Figure 2). We also report encapsulation of b-carotene and
astaxanthin by these artificial lipocalins, and structural studies
of the resulting host ± guest complexes.
3. Charged lateral side chains are advantageous to ascertain
solubility in water, to prevent spontaneous intra- and
intermolecular side chain interactions in the absence of the
complementary rod, and for constructive electrostatic
interactions upon peptide-strand interdigitation.
4. Hydrophobic amino acid residues alternating with charged
ones are needed to create the hydrophobic interior upon b-
barrel formation in water.
The polycationic Lys-Leu-Lys-rod 13 and the complemen-
tary polyanionic Glu-Leu-Glu-rod 14 fulfill these prerequi-
sites (Scheme 1).
R
O
O
OH
15: R =
R
c)
O
O
R
R
R
O
O
a)
b)
Glu(OtBu)-Leu-Glu(OtBu)-NH2
17: R =
16: R =
14: R =
13: R =
d)
R
R
R
Lys(Boc)-Leu-Lys(Boc)-NH2
O
O
Glu-Leu-Glu-NH2
Lys-Leu-Lys-NH2
O
O
Scheme 1. a) H-Lys(Boc)-Leu-Lys(Boc)-NH2, PyBOP, DIPEA, 2 h, 57%;
b) TFA/CH2Cl2, 5 min, quantitative; c) H-Glu(OtBu)-Leu-Glu(OtBu)-
NH2, PyBOP, DIPEA, 2 h, 57%; d) TFA/CH2Cl2, 45 min, quantitative.
Molecular models of rigid-rod supramolecules 12n com-
posed of 13 and 14 revealed b-barrels with lipocalin-type
toroidal amphiphilicity, precisely defined length, and a
flexible diameter (Figure 2A and B). This architecture is
opposite to that of thoroughly investigated nanotubes with
fixed diameter and flexible length formed by stacked cyclic
peptides, carbohydrates, and phenylacetylenes.[20] Flexibility
in internal diameter, however, seems essential for the
encapsulation of hydrophobic guests of different volume.
Three clearly distinct situations were identified by molec-
ular modeling of 12n. Dimeric supramolecules with an
architecture analogous to that of self-assembled ionophore
72 (Figure 1) could not be assembled because of the steric
demand of the intratoroidal Leu-residues (not shown). The
same was true for the electrostatically disfavored intramolec-
ular b-sheet formation in monomeric 13 and 14 (not shown).
Figure 2. Molecular architecture of rigid-rod b-barrels. Schematic side
view of tetramer 124 and top view of molecular models of 124 (A) and
hexamer 126 (B); only the two top strands are shown for clarity. Molecular
modeling was done with Cerius2 (Molecular Simulations) and Insight II/
Discover (BIOSYM Technologies).
Results and Discussion
Design and molecular modeling of rigid-rod b-barrels: Rigid-
rod b-barrels 12n were designed based on the following
considerations:
1. Rigid-rod scaffolds were selected to preorganize the
interdigitation of peptide strands, as in rigid-rod b-barrel
Chem. Eur. J. 2000, 6, No. 10
ꢁ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2000
0947-6539/00/0610-1741 $ 17.50+.50/0
1741

S. Matile, B. Baumeister
FULL PAPER
In tetramer 124, the Leu-residues pointing towards the
center of the b-barrel create a central hydrophobic channel of
about 8 Š diameter with four adjacent peripheral pockets
(Figure 2A). The central tube seems well preorganized to
accommodate one rod-shaped, hydrophobic guest of up to
34 Š length; this guest can be firmly clamped by the four
internal Leu side-chain iso-butyl arrays through multiple
hydrophobic interactions.
In hexamer 126, the ªinnermolecularº space is significantly
enlarged and less structured (Figure 2B). In sharp contrast to
tetramer 124, hexamer 126 can encompass disc micelles rather
than encapsulate monomeric guests. Molecular models of
higher oligomers (octamer 128, decamer 1210, etc.) gave
expanded ªbicycle tiresº that resembled hexamer 126 with re-
spect to general appearance and potential function (not shown).
sistent with the calculated molecular weight of hexamer 126,
while the major product (81%) corresponded to tetramer 124.
In the absence of the complementary polyanion 14, polycation
13 was monomeric (Figure 3c). Self-assembly of polyanion 14
into tetrameric pinwheels 144, with central p,p-stacked arene
arrays surrounded by amphiphilic tripeptides, and its signifi-
cance for nanoarchitecture has been described elsewhere
(Figure 3b).[12a]
The rigid-rod b-barrels 124 (126) in Figure 3a formed
spontaneously upon addition of concentrated 13 and 14 in
methanol (e.g., 20 mL) to saline buffer at pH 6.4 (e.g., 2 mL).
The results were independent of sequence of addition and
octi(p-phenylene) concentration. Rigid-rod b-barrels 124 (126)
were soluble at mircomolar but insoluble at millimolar
concentrations.[22] About identical results were obtained by
solubilization of equimolar amounts of 13 and 14 in cholate
micelles followed by dialytic removal of cholate (Figure 3a).
However, dialytic detergent removal from mixed n-octyl b-d-
glucopyranoside/octi(p-phenylene) micelles yielded higher
oligomers 12n (Figure 3d, n ꢀ 26). Dialytic detergent removal
applied to mixed cholate/octi(p-phenylene)/b-carotene mi-
celles gave tetramers with encapsulated b-carotene and traces
of the corresponding hexamers and oligomers (Figure 3e, see
below for discussion).
Oligo(p-phenylene) synthesis: The complementary oligo(p-
phenylene) peptides 13 and 14 were prepared by coupling of
octaacid 15 with the tripeptides
H-Lys(Boc)-Leu-Lys(Boc)-NH2
and
(OtBu)-NH2,
H-Glu(OtBu)-Leu-Glu-
respectively
(Scheme 1). Tripeptides H-Lys-
(Boc)-Leu-Lys(Boc)-NH2 and
H-Glu(OtBu)-Leu-Glu(OtBu)-
NH2 were prepared by standard
solution-phase peptide synthe-
sis with Boc-protection for Lys-
residues, OtBu-protection for
Glu-residues, and orthogonal
Z-protection for the N-terminal
amino function. Coupling of the
final tripeptides H-Lys(Boc)-
Leu-Lys(Boc)-NH2 and H-Glu-
(OtBu)-Leu-Glu(OtBu)-NH2
with octaacid 15 by using Py-
BOP/DIPEA-methodology[21]
gave rigid-rod peptides 16 and
17 in 67% and 56% yield.
Peptide deprotection with
TFA to give polyanion 14 and
polycation 13 was quantitative.
The synthesis of polyanion 14
has been previously communi-
cated.[12a]
Structural studies of rigid-rod b-barrels: The position of the
absorption maximum of the oligo(p-phenylene) 1L transitions
at 316 nm (e ꢀ 28.6 mmÀ1 cmÀ1)[12f] did not significantly change
during the assembly of rigid-rod b-barrels. The circular
dichroism (CD) spectra of polycationic 13 (Figure 4, dotted
Figure 3. SEC of rigid-rod su-
pramolecules in saline buffer,
pH 6.4, prepared by a) addition
of concentrated 13 and 14 in
MeOH or dialysis of cholate
micelles with 13 and 14, b) ad-
dition of concentrated 14 in
MeOH, c) addition of concen-
trated 13 in MeOH, d) dialysis
of n-octyl b-d-glucopyranoside
micelles with 13 and 14, e) dial-
ysis of cholate micelles with 13,
14, and b-carotene 11. Accord-
ing to protein standards, peaks
at ꢁ24 min correspond to tet-
ramers [124: MWcalcd 16644;
144: MWcalcd 16664; 19 (see
Figure 6 for structure): MWcalcd
17181], peaks at ꢁ22 min to
hexamers (126: MWcalcd 24966),
and peaks at ꢁ15 min to
Figure 4. CD spectra of rigid-rod lipocalin 124/126 (ꢁ4:1, solid line), rigid-
rod pinwheel 144 (dashed line), and polycation 13 (dotted line) at pH 6.4.
Programmed assembly of rigid-
rod b-barrels: Size-exclusion
chromatography (SEC) of an
equimolar mixture of polyca-
tion 13 and complementary
polyanion 14 in saline buffer
(pH 6.4) revealed the presence
of two suprastructures (Fig-
ure 3a). Direct comparison with
the retention times of protein
standards implied that the ap-
parent molecular weight of the
minor product (19%) was con-
curve) and tetrameric pinwheel 144 (Figure 4, dashed curve)
were not distinct. These CD spectra differed clearly from that
of rigid-rod b-barrels 124/126 (ꢀ4:1) in saline buffer at
pH 6.4 (Figure 4, solid curve). That for 124/126 contains six
distinct CD Cotton effects (CEs). Judged from of their
comparable red and blue shift from the absorption at 316 nm,
the first two CEs at 333 nm and 295 nm are likely to originate
from exciton coupling.[10, 12b,c] Consistent with intramolecular
coupling, the amplitudes seen for 124 (126) (A ˆ À6.1) are
similar in magnitude (but opposite in sign) compared with
that observed for dimer 72 (Figure 1, A ˆ ‡9.9).[10]
oligomers
(12n:
MWfound
ꢁ110000; n ꢀ 26). Low molec-
ular weight peaks around the
cut-off of the column are not
shown for clarity.
1742
ꢁ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2000
0947-6539/00/0610-1742 $ 17.50+.50/0
Chem. Eur. J. 2000, 6, No. 10

Lipocalin Models
1739±1749
The CD spectrum of 124 (126) further exhibits four high-
energy CEs with increasing magnitude and alternating sign.
The CEs at 231 nm (De ˆ À9.6) and 216 nm (De ˆ ‡5.4) are
red-shifted relative to the expected values for b-sheets.[23±25] If
the negative CE at 231 nm originates from a b-sheet
conformation, then it indicates flattened and/or twisted b-
sheets in 124 (126).[24] However, since additional contributions
from oligo(p-phenylene) 1B transitions were also expected in
this region,[12b,c] further support for the b-barrel architecture of
124 (126) from the dependence of its CD spectrum on
stoichiometry, ionic strength, pH, temperature, and concen-
tration was essential (Figure 5).
corroborated the presence of positive cooperativity (nH ˆ 2.0,
not shown).[26]
The dependence of the CD spectrum of 124 (126) on ionic
strength and pH evidenced the importance of constructive
ionic interactions, that is, the interaction of Glu and Lys
residues (Figure 5b and c). The only conceivable rationaliza-
tion of these findings was the formation of rigid-rod b-barrels
by multiple tripeptide interdigitation. Relatively minor
changes of the CD spectrum of 124 (126) at pH < 3.5 and
>10.5 and at salt concentrations of up to 1m, as well as
stability toward heat (up to 658C, not shown) and dilution (nm
concentrations, not shown), evidenced the robust supramo-
lecular architecture of rigid-rod b-barrels 124 (126).
Interestingly, oligomer 12n (n ꢀ 26, Figure 3d) exhibited a
blue-shift for the 1L absorptions (278 nm) and the corre-
sponding first CD Cotton effect (288 nm, not shown). These
hypsochromic effects were indicative for intermolecular p ± p
interactions between oligophenylenic arene arrays.[12a±c] The
underlying, evidently more complex, suprastructure was not
further investigated.
In summary, the spectroscopic properties of rigid-rod b-
barrels 124 (126), but not those of oligomer 12n, were in good
agreement with the molecular models (Figure 2) with respect
to molecular weight (Figure 3a), stoichiometry (Figure 5a),
and extratoroidal interaction of complementary Lys and Glu
residues (Figure 5b and c). However, these results contained
no information on the presence and properties of a hydro-
phobic interior. The capacity of rigid-rod b-barrels to
encapsulate carotenoids of different length and hydrophobic-
ity was therefore explored.
Carotenoid encapsulation by rigid-rod b-barrels: Molecular
models of rigid-rod lipocalin 124 revealed an intratoroidal,
hydrophobic space that is roughly complementary to the rod-
shape of a single carotenoid, while hexamer 126 and higher
oligomers appreared as ªbicycle tiresº that might possibly
surround disc micellar carotenoids (Figure 2). These two
possibilities for guest encapsulation were studied with the
hydrophobic b-carotene 11, the truncated b-apo-carotenal 18,
and the bolaamphiphilic,[27] biologically relevant zeaxanthin 9
and astaxanthin 10 (Table 1).
For encapsulation, carotenoids were solubilized in cholate
micelles at pH 6.4. Mixed carotenoid/cholate micelles in
hexane have blue-shifted absorption maxima between 436
and 450 nm (Table 1). Dialytic cholate removal resulted in
complete precipitation for b-carotene (Table 1, entry 2), the
formation of well-known H-aggregates for zeaxanthin (Ta-
ble 1, entry 8),[18] and presumably unprecedented J-aggregates
for astaxanthin (Table 1, entry 11). The presence of polyanion
14 during dialytic cholate removal did not change these
outcomes significantly (Table 1, entries 3, 7, and 10). The
presence of both polyanion 14 and the complementary
polycation 13 inhibited aggregation of b-apo-carotenal (Ta-
ble 1, entry 4) and zeaxanthin (Table 1, entry 6), inhibited
precipitation of b-carotene (Table 1, entry 1), and had appa-
rently little influence on the J-aggregation of astaxanthin
(Table 1, entry 11). These last two effects were studied in
more detail.
Figure 5. Dependence of the first CD Cotton effect of 124/126 (ꢁ4:1) on
a) monomer stoichiometry, b) ionic strength, and c) pH. Spectra for a) were
obtained in NanH3ÀnPO4 (10mm), NaCl (100mm), pH 6.4, for b) in
NanH3ÀnPO4 (10mm), pH 6.4, and for c) in NanH3ÀnPO4 (10mm), NaCl
(100mm). x ˆ [13]/([13] ‡ [14]). All results were independent of the total
oligo(p-phenylene) concentration (4 ± 35mm).
The mixing curves for rigid-rod lipocalin 124 (126) demon-
strated 1:1 stoichiometry (Figure 5a). The non-linearity of
these mixing curves was in good agreement with cooperative
formation of 124 (126); Hill plots of the above data fully
Chem. Eur. J. 2000, 6, No. 10
ꢁ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2000
0947-6539/00/0610-1743 $ 17.50+.50/0
1743

S. Matile, B. Baumeister
FULL PAPER
Rigid-rod carotenolipocalin 19: Spectroscopic studies of b-
carotene solubilized in detergent-free water (Table 1, entry 1)
were consistent with encapsulation of one polyene by
tetramer 124 to give rigid-rod carotenolipocalin 19 (Figure 6).
A
CH3
H
H
CH3
CH3
s-trans planarized
B
CH3
H
H3C
34 Å
H
CH3
s-cis planarized
C
H
H
19
twisted
Figure 6. Rigid-rod carotenolipocalin 19. Possible conformations with
respect to the torsion angle between the polyene chain and the b-ionylidene
ring of b-carotene 11 are shown on the left: s-trans (A) and s-cis planarized
(B) conformers with maximal and twisted conformer (C) with minimal
conjugation and nonbonded repulsion between b-ionylidene methyls and
polyene hydrogens.
b-Carotene encapsulation was distinguishable from simple
binding by the absence of detectable interactions with
preformed tetramer 124 and, most importantly, the failure to
extract b-carotene from capsule 19 into hexane.[28, 9b] It was,
however, possible to deconstruct supramolecule 19 in THF.
The relative chromophore absorption and fluorescence in-
tensities in THF and water were consistent with an oligo-
(p-phenylene) to b-carotene ratio of ꢀ4:1.
Figure 7. a) CD, b) UV/Vis, and c) fluorescence emission spectra of rigid-
rod carotenolipocalin 19 in detergent-free water at pH 6.4. b) The
normalized absorption spectrum of capsule 19 (solid line) in comparison
to those of b-carotene 11 in cholate micelles (pH 6.4, dashed line) and in
hexane (dotted line). The red shift of the maxima of 19 relative to that of 11
in hexane was determined using 2nd derivative spectra (not shown) and is
given in parenthesis. c) Fluorescence emission spectrum of capsule 19 (solid
line) in comparison to that of rigid-rod b-barrel 124 (126) (dotted line) at
pH 6.4 (excitation: 316 nm). Both samples had identical emission intensity
in THF.
Rigid-rod carotenolipocalin 19 passed through Sephadex
columns as a single band together with the oligo(p-phenylene)
peptides. Size exclusion chromatography (SEC) revealed
tetramers 124 with traces of hexamers 126 and oligomers 12n
(Figure 3e). Suppression of hexameric b-barrels 126 suggested
that b-carotene may act as hydrophobic template during the
assembly of capsule 19. The confirmed need of guest excess
for successful assembly of 19 supported this role of b-
carotene. An insufficient template effect could further explain
why comparable host ± guest complexes were not formed with
the hydrophobically mismatched b-apo-carotenal 18 (Table 1,
entry 4).
The interaction of oligo(p-phenylene) peptides and polyene
in host ± guest complex 19 was evidenced by CD spectroscopy
and fluorescence quenching. On the one hand, the presence of
monomeric b-carotene in a chiral environment was implied by
induced CD (ICD) in the polyene region (Figure 7a). The
slight offset of CD and absorption maxima (Figure 7b) of
encapsulated b-carotene implied the additional presence of
more complexand not yet understood effects. Oligo-
(p-phenylene)s with neighboring, energy-accepting b-carotene,
on the other hand, were indicated by quenching of their
fluorescence emission (Figure 7c). Note that 48% quenching
is remarkable considering the short lifetime of the excited
state of oligo(p-phenylene)s.[12f]
The absorption spectrum of lipocalin model 19 contained
information on the nature of the intratoroidal space of rigid-
rod b-barrel 124 as expected. Namely, the b-carotene S0 ! S2
transition in capsule 19 (Figure 7b, solid spectrum) was up to
40 nm red-shifted relative to that of b-carotene in hexane
(Figure 7b, dotted spectrum), and its fine structure was well
resolved. The symmetry-forbidden low-energy S0 ! S1 tran-
sition was not observed.[14±16]
These spectroscopic features indicate that b-carotene is a) in
hydrophobic environment (no blue shift)[13] and b) monomeric
(no broadening).[14] To explain the observed bathochromic
effect, consideration of solvent polarity is clearly not suffi-
1744
ꢁ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2000
0947-6539/00/0610-1744 $ 17.50+.50/0
Chem. Eur. J. 2000, 6, No. 10

Lipocalin Models
1739±1749
cient; it further indicates extension of b-carotene conjugation
by ring ± chain co-planarization (Figure 6A and B). This
implies that the polyene chromophore with twisted confor-
mation in solution is flattened by rigid-rod b-barrel 124 to fit
the internal hydrophobic channel in capsule 19. Very similar
changes in the absorption spectrum of retinol upon binding to
lipocalin b-lactoglobulin have been explained by ring ± chain
co-planarization.[9] Moreover, the observed red shift and
increase in fine structure for encapsulated b-carotene are in
excellent agreement with theoretical models for carotenoid
planarization.[17]
In summary, we have shown that monomeric b-carotene can
be encapsulated by rigid-rod lipocalin 124 to quantitatively
yield host ± guest complex 19 in detergent-free water. The
usefulness of b-carotene as probe for the hydrophobic interior
of rigid-rod b-barrel 124 was evidenced by a remarkable
bathochromic effect (presumably) due to planarization of the
polyene within the complementary ªinnermolecularº[28f]
space.
Figure 9. Absorption (bottom) and CD (top) spectra of astaxanthin
aggregates at pH 6.4 prepared in the presence (solid lines) and absence
(dotted lines) of octi(p-phenylene)s 13 and 14.
Mixed rigid-rod disc micelles 20: The influence of comple-
mentary oligo(p-phenylene) peptides 13 and 14 on formation
of astaxanthin J-aggregates by dialytic detergent removal
seemed initially negligible. A remarkable red shift of ꢀ90 nm
with respect to astaxanthin absorption in hexane was ob-
served with and without rods (Table 1, entries 9 ± 11). Indica-
tions for the formation of mixed disc micelles 20 (Figure 8)
of twist of these astaxanthin aggregates (A ˆ ‡84). Control
experiments corroborated that addition of rigid-rod b-barrels
124 (126) to preformed astaxanthin aggregates (A ˆ À110) and
dialysis with polyanionic rod 14 only do not yield astaxanthin
aggregates with positive amplitude. These findings were
consistent with the formation of mixed-disc micelles 20 during
dialytic detergent removal.
O
OH
HO
O
The quenching efficiency of astaxanthin in disc micelles 20
is 50% of that of b-carotene in rigid-rod carotenolipocalin 19
(Table 1, entries 9 and 1). Gel filtration of supramolecule 20
yielded a single product that contained both oligo(p-phenyl-
ene)s and astaxanthin, while astaxanthin J-aggregates without
both rods were retained on Sephadexcolumns due to dilution
below critical aggregation concentration (Table 1, entries 9 ±
11). SEC with disc micelles 20 gave not reproducible results.
However, an occasionally observed broad peak with MWꢀ
84000 and the spectroscopically estimated 1:1 stoichiometry
suggests that disc micelles 20 may consist of dodecameric
rigid-rod b-barrels surrounding twelve J-aggregated astaxan-
thin molecules (Figure 8). Similar ªbicycle tireº structures
have been observed before with amphiphilic proteins (e.g.,
apolipoproteins), peptides (e.g., melittin), and detergents
(e.g., lysophosphatidylcholine).[29]
OH
OH
O
O
O
OH
The capacity of rigid-rod b-barrels to stabilize small
J-aggregates is of high interest with respect to the significance
of J-aggregates in biological pigmentation in general[30] and
that of marine invertebrates in particular.[8] As mentioned in
the introduction, the blue color of the lobster Humarus
gammarus may originate from binding of two astaxanthins
(lmax ꢀ 472 nm) to b-crustacyanidin dimers (lmax ꢀ 585 nm,
A ꢀÀ 30) followed by aggregation into a-crustacyanidin
(lmax ꢀ 630 nm, A ꢀÀ 95).[8] Here we have shown that astax-
anthin J-aggregates with similar spectroscopic properties
(lmax ꢀ 560 nm, A ꢀÀ 110) can be prepared by dialytic deter-
gent removal, and that their structure, stability and stereo-
20
Figure 8. Tentative structure of astaxanthin/oligo(p-phenylene) disc mi-
celles 20.
were found by CD spectroscopy (Figure 9). Strong bisignate
CD Cotton effects (A ˆ À110) centered around the red-
shifted astaxanthin absorption evidenced exciton coupling
that was consistent with carotenoid aggregation. The presence
of complementary oligo(p-phenylene) peptides during dia-
lytic detergent removal caused inversion of the absolute sense
Chem. Eur. J. 2000, 6, No. 10
ꢁ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2000
0947-6539/00/0610-1745 $ 17.50+.50/0
1745

S. Matile, B. Baumeister
FULL PAPER
performed over SephadexG-50 (1 Â 27 cm). Size-exclusion chromatogra-
phy (SEC) was performed on Superdexꢂ 75 HR 10/30 prepacked column
from Pharmacia Biotech (MW 70000 ± 3000, 1 mL buffer per min) coupled
with a Jasco PU-980 pump and a Jasco UV-970 UV-Vis detector. MWs were
determined by using the protein standards albumin (67.0 kDa), ovalbumin
(43.0 kDa), chymotrypsinogen A (25.0 kDa), ribonuclease A (13.7 kDa),
and aprotinin (6.5 kDa) from Pharmacia Biotech. Fluorescence spectra
were recorded on FluoroMax-2 (Jobin Yvon-Spex). Both emission and
excitation spectra were not corrected. CD-spectra were recorded on
JASCO-710 and JASCO-715 spectropolarimeters and reported as Demax
[mÀ1cmÀ1] at lmax [nm]. UV-Vis spectra were recorded on a Hewlett ±
Packard 8452A diode array spectrophotometer or on a Varian Cary1 Bio
spectrophotometer and are reported as lmax [nm] (emax [mmÀ1cmÀ1]).
chemistry can be controlled by surrounding rigid-rod b-
barrels (lmax ꢀ 576 nm, A ꢀ‡ 84). These findings may thus
contribute to the understanding of the molecular mechanisms
of invertebrate coloration.
Conclusion
With the present study, we demonstrate the capacity of
complementary oligo(p-phenylene) peptides to form water-
soluble, expanded rigid-rod b-barrels with defined length,
flexible diameter, and lipocalin-type toroidal amphiphilicity.
Spectroscopic evidence was obtained for the designed role of
multiple extratoroidal electrostatic and intratoroidal hydro-
phobic interactions. Encapsulation of hydrophobic guests by
rigid-rod b-barrels was shown to depend on hydrophobic
matching and guest polarity. Both tetrameric rigid-rod
carotenolipocalins with one encapsulated hydrophobic guest
and more fragile rigid-rod disc micelles with J-aggregated
bolaamphiphiles were observed.
These results suggest that rigid-rod b-barrels may serve as
general models for biological b-barrels; this implies an
extremely diverse potential applicability of these unique
toroidal supramolecules for biomimetic pigmentation,[6±8, 30]
membrane-protein solubilization,[1, 5] gene transfection,[3, 5]
and catalysis in the broadest sense.[2, 4, 31] Studies along these
lines are ongoing.
Z-Leu-Lys(Boc)-NH2 (general procedure A): 1-(3-Dimethylaminopropyl)-
3-ethyl-carbodiimide ´ HCl (288 mg, 1.5 mmol), 1-hydroxybenzotriazole
(172 mg, 1.27 mmol), Z-Leu-OH (282 mg, 1.06 mmol) and triethylamine
(0.441 mL, 6 mmol) were added to a solution of H-Lys(Boc)-NH2 ´ HCl
(300 mg, 1.06 mmol) in CH2Cl2 (5.31 mL) at 08C. After stirring for 6 h in
the dark at RT, the reaction mixture was diluted with CH2Cl2, extracted
with saturated aqueous NaHCO3, washed with brine, extracted with 1m
aqueous KHSO4, washed with brine, dried over anhydrous Na2SO4, and
concentrated in vacuo. Purification of the crude product by column
chromatography (CH2Cl2/MeOH 10:1, Rf ˆ 0.4) yielded pure Z-Leu-
Lys(Boc)-NH2 (489 mg, 93%) as a white powder. [a]2D0 ˆ À15.8 (c ˆ 1.00
in MeOH); m.p. 165.1 ± 167.08C; IR (KBr): nÄ ˆ 3392 (s), 3374 (s), 3215 (m),
3070 (w), 3039 (w), 2956 (m), 2935 (m), 2867 (w), 1685 (s), 1643 (s), 1534 (s),
1452 (m), 1420 (w), 1390 (w), 1363 (m), 1275 (s), 1254 (s), 1171 (s), 1124 (w),
1052 (w), 912 (w), 865 (w), 782 (w), 735 (w), 694 (m), 647 cmÀ1 (w); 1H NMR
(300 MHz, [D]6DMSO, 258C, TMS): d ˆ 7.80 ± 7.73 (m, 1H; Lys-NH,
exchange with D2O), 7.52 ± 7.43 (m, 1H; Leu-NH, exchange with D2O),
7.34 ± 7.29 (m, 6H; ArH and Lys-CONH2, partial exchange with D2O),
7.04 ± 6.97 (m, 1H; Lys-CONH2, exchange with D2O), 6.78 ± 6.71 (m, 1H;
Boc-NH, exchange with D2O), 5.04 ± 5.00 (m, 2H; ArCH2), 4.20 ± 4.12 (m,
1H; Lys-Ha), 4.06 ± 3.96 (m, 1H; Leu-Ha), 2.96 ± 2.78 (m, 2H; Lys-He),
1.70 ± 1.20 (several m, 9H; Leu-Hb,g, Lys-Hb,g,d), 1.35 (s, 9H; Boc-CH3),
0.90 ± 0.80 (m, 6H; Leu-CH3).
Experimental Section
H-Leu-Lys(Boc)-NH2 (general procedure B): A catalytic amount of Pd/C
was added to a solution of Z-Leu-Lys(Boc)-NH2 (498 mg, 1.01 mmol) in
MeOH (5 mL). The suspension was degassed at least four times and set
under an H2 atmosphere. After stirring for 2 h, the Pd/C was filtered off and
the crude product was concentrated in vacuo. Column chromatography
(CH2Cl2/MeOH/iPr-NH2 90:9:1, Rf ˆ 0.4) yielded pure H-Leu-Lys(Boc)-
NH2 as a white powder (284 mg, 0.79 mmol, 80%). [a]2D0 ˆ À0.4 (c ˆ 1.00 in
MeOH); m.p. 109.3 ± 110.88C; IR (KBr): nÄ ˆ 3352 (s), 3325 (s), 3200 (w),
2933 (m), 2868 (w), 1688 (s), 1669 (s), 1615 (s), 1536 (s), 1465 (w), 1441 (w),
1389 (w), 1367 (m), 1283 (m), 1250 (m), 1180 (m), 1123 (w), 1006 (w), 659
(w), 588 cmÀ1 (w); 1H NMR (300 MHz, [D]6DMSO, 258C, TMS): d ˆ 8.03 ±
7.87 (m, 1H; Lys-NH, exchange with D2O), 7.50 ± 7.30 (m, 1H; Lys-CONH2,
exchange with D2O), 7.10 ± 6.98 (m, 1H; Lys-CONH2, exchange with D2O),
6.76 ± 6.72 (m, 1H; Boc-NH, exchange with D2O), 4.18 ± 4.10 (m, 1H; Lys-
Ha), 3.22 ± 3.15 (m, 1H; Leu-Ha), 2.90 ± 2.81 (m, 2H; Lys-He), 1.80 ± 1.10
General: Reagents for synthesis were purchased from Aldrich. PyBOP was
from Calbiochem-Novabiochem. Amino acid derivatives were obtained
from Calbiochem-Novabiochem and Bachem. Zeaxanthin was from
Indofine Chemical Company, astaxanthin from Alexis Biochemicals, b-
apo-8'-carotenal from Fluka, and b-carotene from Aldrich. SephadexG-50,
sodium cholate and other salts were of the best grade available from Sigma
Chemicals. Solvents were distilled and, if necessary, dried before use. All
reactions were performed under nitrogen atmosphere. Column chroma-
tography was carried out over silica gel (Selecto Scientific, 32 ± 6 mm).
Analytical thin-layer chromatography (TLC) was performed on AL SIL G/
UV (Whatman). Preparative TLC (PTLC) was performed on silica gel 60F-
254 (Merck) or silica gel GF-2 (Aldrich). The values of optical rotation
([a]2D0† were determined at room temperature on a Perkin ± Elmer 241
Polarimeter and are given in degrees. Melting points (mp) were determined
on a heating table from Reichert (Austria). IR-spectra were recorded in
KBr pellets on a Perkin ± Elmer FT-IR Spectrometer Paragon 500 and are
(several m, 9H; Leu-Hb,g, Lys-Hb,g,d), 1.35 (s, 9 Boc-CH3), 0.87 (d,
3J(H,H) ˆ 6.6 Hz, 3H; Leu-CH3), 0.83 (d, 3J(H,H) ˆ 6.6 Hz, 3H; Leu-CH3).
1
reported in cmÀ1 (intensity: w ˆ weak, m ˆ medium, s ˆ strong). H NMR
spectra were recorded on a Varian Mercury 300 spectrometer, a Bruker
AMX400 spectrometer, and a Varian UNITY INOVA 500 MHz spec-
trometer. Chemical shifts are reported in ppm relative to TMS (d ˆ 0). Spin
multiplicities are reported as singlet (s), doublet (d), triplet (t), quartet (q)
or multiplet (m), and coupling constants (J) are given in Hz. 1H NMR
resonances were assigned with the aid of additional information from
pertinent 2D NMR spectra (H, H-COSY, NOESY, ROESY, and TOCSY).
The presence of solvent in analytical samples was corroborated by 1H NMR
spectroscopy. FAB-HRMS was performed on a Fenningan 3200 twin-
quadrupole mass spectrometer at the University of Maryland, College
Park. ESI-MS was performed on a PE-SciexAPI100 electrospray instru-
ment by the Lombardi Cancer Centerꢀs Macromolecular Analysis Shared
Resource. MALDI-TOF-MS was performed on a ProflexBruker mass
spectrometer at the University of Maryland, College Park, using a DHB
(2,5 dihydroxy benzoic acid) matrix. Dialysis was performed with a Mini
Lipoprepꢂ (Sialomed) or dialysis cells from Fisher Scientific with dialysis
membranes from Diachema (MW-cutoff ˆ 5000). Gel filtration (GF) was
Z-Lys(Boc)-Leu-Lys(Boc)-NH2: Coupling of H-Leu-Lys(Boc)-NH2
(284 mg, 0.79 mmol) and Z-Lys(Boc)-OH (452mg, 1.19 mmol) by following
procedure A and purification of the crude product by column chromatog-
raphy (CH2Cl2/MeOH 10:1, Rf ˆ 0.3) yielded pure Z-Lys(Boc)-Leu-Lys-
(Boc)-NH2 (464 mg, 82%) as a white powder. [a]2D0 ˆ À23.0 (c ˆ 1.00 in
MeOH); m.p. 177.5 ± 178.98C; IR (KBr): nÄ ˆ 3314 (s), 2934 (m), 2869 (w),
2373 (w), 1686 (s), 1635 (s), 1534 (s), 1458 (w), 1391 (w), 1367 (m), 1269 (m),
1252 (m), 1174 (m), 1057 (w), 868 (w), 778 (w), 754 (w), 699 (w), 641 cmÀ1
(w); 1H NMR (300 MHz, [D]6DMSO, 258C, TMS): d ˆ 7.98 ± 7.92 (m, 1H;
Lys-NH, exchange with D2O), 7.73 ± 7.66 (m, 1H; Leu-NH, exchange with
D2O), 7.43 ± 7.25 (several m, 6H; ArH and Lys-NH, partial exchange with
D2O), 7.25 ± 7.20 (m, 1H; Lys-CONH2, exchange with D2O), 7.03 ± 6.97 (m,
1H; Lys-CONH2, exchange with D2O), 6.80 ± 6.70 (m, 2H; Boc-NH), 5.00
(s, 2H; ArCH2), 4.32 ± 4.20 (m, 1H; Lys-Ha), 4.20 ± 4.05 (m, 1H; Leu-Ha),
4.02 ± 3.88 (m, 1H; Lys-Ha), 2.95 ± 2.78 (m, 4H; Lys-He), 1.70 ± 1.10 (several
3
m, 15H; Leu-Hb,g, Lys-Hb,g,d), 1.34 (s, 18H; Boc-CH3), 0.86 (d, J(H,H) ˆ
3
6.6 Hz, 3H; Leu-CH3), 0.81 (d, J(H,H) ˆ 6.6 Hz, 3H; Leu-CH3).
1746
ꢁ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2000
0947-6539/00/0610-1746 $ 17.50+.50/0
Chem. Eur. J. 2000, 6, No. 10

Lipocalin Models
1739±1749
H-Lys(Boc)-Leu-Lys(Boc)-NH2: Deprotection of Z-Lys(Boc)-Leu-Lys-
(Boc)-NH2 (464 mg, 0.64 mmol) by following procedure B and purification
by column chromatography (CH2Cl2/MeOH/iPr-NH2 90:9:1, Rf ˆ 0.4)
yielded pure H-Lys(Boc)-Leu-Lys(Boc)-NH2 as a white powder (328 mg,
87%). [a]2D0 ˆ À17.0 (c ˆ 1.00 in MeOH); m.p. 76.0 ± 77.28C; IR (KBr): nÄ ˆ
3342 (s), 3061 (w), 2937 (m), 2864 (w), 2356 (w), 1685 (s), 1529 (s), 1457 (w),
1394 (w), 1368 (m), 1275 (w), 1249 (m), 1171 (m), 1046 (w), 1010 (w), 865
(w), 782 (w), 652 cmÀ1 (w); 1H NMR (300 MHz, [D]6DMSO, 258C, TMS):
d ˆ 8.03 ± 7.94 (m, 1H; Lys-NH, exchange with D2O), 7.85 ± 7.78 (m, 1H;
Leu-NH, exchange with D2O), 7.27 ± 7.21 (m, 1H; Lys-CONH2, exchange
with D2O), 7.00 ± 6.93 (s, 1H; Lys-CONH2, exchange with D2O), 6.78 ± 6.68
(m, 2H; Boc-NH, exchange with D2O), 4.37 ± 4.22 (m, 1H; Lys-Ha), 4.18 ±
4.05 (m, 1H; Leu-Ha), 3.20 ± 3.08 (m, 1H; Lys-Ha), 2.92 ± 2.78 (m, 4H; Leu-
He), 1.70 ± 1.10 (several m, 15H; Leu-Hb,g, Lys-Hb,g,d), 1.36 (s, 18H; Boc-
CH3), 0.86 (d, 3J(H,H) ˆ 6.6 Hz, 3H; Leu-CH3), 0.81 (d, 3J(H,H) ˆ 6.6 Hz,
3H; Leu-CH3); HRMS (FAB): m/z calcd for C28H54N6O7: 587.41321; found
587.41586.
(w), 666 cmÀ1 (w); 1H NMR (300 MHz, [D]6DMSO, 258C, TMS): d ˆ 8.04 ±
7.89 (m, 2H; Glu-NH, Leu-NH, exchange with D2O), 7.28 ± 7.20 (s, 1H;
Glu-CONH2, exchange with D2O), 7.08 ± 7.00 (s, 1H; Glu-CONH2,
exchange with D2O), 4.34 ± 4.22 (m, 1H; Glu-Ha), 4.20 ± 4.09 (m, 1H;
Leu-Ha), 3.19 ± 3.08 (m, 1H; Glu-Ha), 2.29 ± 2.12 (m, 4H; Glu-Hg), 2.00 ±
1.30 (m, 9H; Leu-Hb,g, Glu-Hb), 1.37 (s, 18H; tBu), 0.87 (d, 3J(H,H) ˆ
6.6 Hz, 3H; Leu-CH3), 0.84 (d, 3J(H,H) ˆ 6.6 Hz, 3H; Leu-CH3); HRMS
(FAB): m/z calcd for C24H44N4O7: 501.32883; found 501.32838.
83,72,63,52,43,32,23,13-Octa(hydroxycarbonylmethoxy)-p-octiphenyl
(general procedure C): TFA (1 mL) was added to solution of
(15)
a
83,72,63,52,43,32,23,13-octa(tert-butoxycarbonylmethoxy)-p-octiphenyl[10, 12a,g]
(6.8 mg, 4.2 mmol) in CH2Cl2 (1 mL). After stirring for 45 min at RT, the
reaction mixture was concentrated in vacuo to give pure 15 (5.0 mg, 100%)
as a white solid. 1H NMR (300 MHz, CDCl3/CD3OD 1:1, 258C, TMS): d ˆ
7.48 ± 7.41 (m, 6H; ArH), 7.37 ± 7.37 ± 7.12 (m, 16H; ArH), 6.91 (dd,
3J(H,H) ˆ 5.8 Hz, 4J(H,H) ˆ 2.2 Hz, 2H; ArH), 4.69 (s, 16H; ArOCH2);
MS (MALDI-TOF): m/z calcd for C64H50O24: 1203.08; found 1203.66.
Z-Leu-Glu(OtBu)-NH2: Coupling of H-Glu(OtBu)-NH2 ´ HCl (240 mg,
1.25 mmol) and Z-Leu-OH (331 mg, 1.25 mmol) by following procedure A
and purification of the crude product by column chromatography (CH2Cl2/
MeOH 20:1, Rf ˆ 0.3) yielded pure Z-Leu-Glu(OtBu)-NH2 (474 mg, 84%)
as a white powder. [a]2D0 ˆ À22.0 (c ˆ 1.00 in MeOH); m.p. 140.6 ± 141.28C;
IR (KBr): nÄ ˆ 3388 (s), 3316 (s), 3209 (m), 3068 (w), 3036 (w), 2959 (m),
2933 (m), 2874 (w), 1730 (s), 1674 (s), 1644 (s), 1531 (s), 1468 (w), 1455 (w),
1420 (w), 1393 (w), 1368 (m), 1287 (m), 1261 (m), 1236 (m), 1159 (m), 1123
(w), 1051 (w), 993 (w), 958 (w), 913 (w), 849 (w), 780 (w), 744 (w), 696 (m),
645 cmÀ1 (w); 1H NMR (300 MHz, [D]6DMSO, 258C, TMS): d ˆ 7.85 (d,
3J(H,H) ˆ 8.0 Hz, 1H; Glu-NH, exchange with D2O), 7.48 (d, 3J(H,H) ˆ
8.2 Hz, 1H; Leu-NH, exchange with D2O), 7.38 ± 7.24 (m, 6H; , ArH and
Glu-CONH2, partial exchange with D2O), 7.10 ± 7.04 (m, 1H; Glu-CONH2,
exchange with D2O), 5.01 (s, 2H; ArCH2), 4.24 ± 4.13 (m, 1H; Glu-Ha),
4.07 ± 3.96 (m, 1H; Leu-Ha), 2.23 ± 2.12 (m, 2H; Glu-Hg), 1.96 ± 1.33 (m,
5H; Glu-Hb and Leu-Hb), 1.37 (s, 9H; tBu), 0.90 ± 0.79 (m, 6H; Leu-CH3).
83,72,63,52,43,32,23,13-Octa(NH2-Glu(OtBu)-Leu-Glu(OtBu)-carbonylmeth-
oxy)-p-octiphenyl (17): Coupling of 15 (4.5 mg, 3.74 mmol) in DMSO
(1 mL, instead of CH2Cl2) and H-Glu(OtBu)-Leu-Glu(OtBu)-NH2 (45 mg,
89 mmol) by following procedure A, and purification of the crude product
by PTLC (CH2Cl2/MeOH/toluene 10:1:1, Rf ˆ 0.1, and CH2Cl2/MeOH 10:1,
Rf ˆ 0.5) yielded pure 17 (12.9 mg, 67%) as a white solid. 1H NMR
(500 MHz, CDCl3/CD3OD 1:1, 258C, TMS): d ˆ 7.58 ± 7.24 (m, 24H; ArH),
6.98 (d, 3J(H,H) ˆ 7.7 Hz, 2H; ArH), 4.80 ± 4.20 (several m, 40H; ArOCH2,
Leu-Ha, Glu-Ha), 2.40 ± 2.20 (m, 32H; Glu-Hg), 2.16 ± 1.50 (several m,
56H; Leu-Hb,g, Glu-Hb), 1.41 ± 1.29 (m, 144H; tBu), 0.95 ± 0.80 (m, 48H;
Leu-CH3); MS (ESI, MeOH, 1% formic acid): m/z (%): 1289 (97)
[M‡Na]4‡, 1711 (100) [M‡Na]3‡, 2555 (20) [M‡Na]2‡
.
83,72,63,52,43,32,23,13-Octa(NH2-Glu-Leu-Glu-carbonylmethoxy)-p-octi-
phenyl (14): Deprotection of 17 (5.5 mg, 1.08 mmol) by following procedure
C gave pure 14 (4.5 mg, 100%) as a white solid. UV/Vis (H2O): 314 nm
(28.6); 1H NMR (300 MHz, CDCl3/CD3OD 1:1, 258C, TMS): d ˆ 7.50 ± 7.28
(m, 24H; ArH), 6.95 (d, 3J(H,H) ˆ 8.0, 2H; ArH), 4.82 ± 4.20 (several m,
40H; ArOCH2, Leu-Ha, Glu-Ha), 2.44 ± 2.26 (m, 32H; Glu-Hg), 2.24 ± 1.46
(several m, 56H; Leu-Hb,g, Glu-Hb), 0.94 ± 0.78 (m, 48H; Leu-CH3);
1H NMR (500 MHz, D2O/CD3OD 5:1, pH ˆ 6.4, 258C, TMS): d ˆ 7.6 ± 6.8
(very broad m, 26H; ArH), 4.50 ± 4.00 (several m, 40H; ArOCH2, Leu-H,
Glu-H ), 2.55 ± 2.28 (several m, 32H; Glu-Hg), 2.20 ± 1.40 (several m, 56H;
Leu-Hb,g, Glu-Hb), 0.90 ± 0.00 (very broad m, 48H; Leu-CH3); fluorescence
(H2O): 314 (excitation), 385 nm (emission); CD (H2O): see Figure 4.
H-Leu-Glu(OtBu)-NH2: Deprotection of Z-Leu-Glu(OtBu)-NH2 (474 mg,
1.05 mmol) by following procedure B and purification by column chroma-
tography (CH2Cl2/MeOH/iPr-NH2 90:9:1, Rf ˆ 0.3) yielded pure H-Leu-
Glu(OtBu)-NH2 (274 mg, 83%) as a colorless gum. [a]2D0 ˆ À2.2 (c ˆ 1.00 in
MeOH); IR (KBr): nÄ ˆ 3345 (s), 3202 (m), 2953 (s), 2868 (m), 1726 (s), 1660
(s), 1517 (m), 1468 (w), 1454 (w), 1419 (w), 1392 (w), 1369 (m), 1321 (w),
1290 (w), 1255 (m), 1157 (s), 1034 (w), 957 (w), 920 (w), 846 (w), 753 (w),
650 cmÀ1 (w); 1H NMR (300 MHz, [D]6DMSO, 258C, TMS): d ˆ 8.04 ± 7.91
(m, 2H; Glu-NH, Leu-NH, exchange with D2O), 7.42 ± 7.33 (m, 1H; Glu-
CONH2 exchange with D2O), 7.14 ± 7.04 (m, 1H; Glu-CONH2 exchange
with D2O), 4.25 ± 4.14 (m, 1H; Glu-Ha), 3.20 ± 3.12 (m, 1H; Leu-Ha), 2.22 ±
2.09 (m, 2H; Glu-Hg), 1.98 ± 1.60 (m, 5H; Glu-Hb, Leu-Hb,g), 1.37 (s, 9H;
83,72,63,52,43,32,23,13-Octa(NH2-Lys(Boc)-Leu-Lys(Boc)-carbonylme-
thoxy)-p-octiphenyl (16): Coupling of 15 (6.3 mg, 5.2 mmol) in DMSO
(1 mL, instead of CH2Cl2) and H-Lys(Boc)-Leu-Lys(Boc)-NH2 (64 mg,
11 mmol) by following procedure A, and purification of the crude product
by PTLC (CH2Cl2/MeOH 10:1, Rf ˆ 0.1 first run, Rf ˆ 0.5 second run)
yielded pure 16 (16.8 mg, 56%) as a white solid. 1H NMR (500 MHz,
CDCl3/CD3OD 1:1, 258C, TMS): d ˆ 7.55 ± 7.24 (m, 24H; ArH), 6.96 (d,
3J(H,H) ˆ 7.4 Hz, 2H; ArH), 4.80 ± 4.20 (several m, 40H; ArOCH2, Leu-
3
3
tBu), 0.86 (d, J(H,H) ˆ 6.6 Hz, 3H; Leu-CH3), 0.83 (d, J(H,H) ˆ 6.6 Hz,
3H; Leu-CH3).
Z-Glu(OtBu)-Leu-Glu(OtBu)-NH2: Coupling of H-Leu-Glu(OtBu)-NH2
(274 mg, 0.87 mmol) and Z-Glu(OtBu)-OH (440 mg, 1.31 mmol) by
following procedure A and purification of the crude product by column
chromatography (CH2Cl2/MeOH 20:1, Rf ˆ 0.25) yielded pure Z-Glu-
Ha, Lys-Ha), 3.20 ± 2.70 (m, 32H; Lys-He), 1.90 ± 1.00 (m, 120H; Leu-Hb,g
,
Lys-Hb,g,d), 1.48 ± 1.21 (several s, 144H; Boc-CH3), 0.94 ± 0.75 (m, 48H; Leu-
(OtBu)-Leu-Glu(OtBu)-NH2 (486 mg, 88%) as a white powder. [a]D20
ˆ
CH3). MS (ESI, MeOH, 1% formic acid): m/z (%): 1461.6 (94) [M‡Na]4‡
,
À29.0 (c ˆ 1.00 in MeOH); m.p. 172.8 ± 174.28C; IR (KBr): nÄ ˆ 3307 (s),
3067 (w), 2978 (m), 2934 (m), 1731 (s), 1670 (s), 1637 (s), 1534 (s), 1454 (m),
1393 (m), 1368 (s), 1256 (s), 1156 (s), 1054 (w), 953 (w), 850 (w), 753 (w),
698 cmÀ1 (m); 1H NMR (300 MHz, [D]6DMSO, 258C, TMS): d ˆ 7.96 (d,
3J(H,H) ˆ 8.0 Hz, 1H; Glu-NH, exchange with D2O), 7.85 (d, 3J(H,H) ˆ
1942 (100) [M‡Na]3‡, 2896 (16) [M‡Na]2‡
.
83,72,63,52,43,32,23,13-Octa(NH2-Lys-Leu-Lys-carbonylmethoxy)-p-octi-
phenyl (13): Deprotection of 16 (2.6 mg, 0.35 mmol) by following procedure
C gave pure 13 (1.4 mg, 100%) as a white solid. UV/Vis (H2O): 316 nm
(28.6); 1H NMR (300 MHz, CDCl3/CD3OD 1:1, 258C, TMS): d ˆ 7.55 ± 7.22
(m, 24H; ArH), 7.00 ± 6.90 (m, 2H; ArH), 4.80 ± 4.20 (several m, 40H;
ArOCH2, Leu-Ha, Lys-Ha), 2.98 ± 2.62 (m, 32H; Lys-He), 1.95 ± 1.00 (m,
120H; Leu-Hb,g, Lys-Hb,g,d), 0.95 ± 0.75 (m, 48H; Leu-CH3); 1H NMR
(500 MHz, D2O/CD3OD 5:1, pH ˆ 6.4, 258C, TMS): d ˆ 7.60 ± 7-26 (m,
24H; ArH), 7.05 (d, 3J(H,H) ˆ 8.5 Hz, 2H; ArH), 4.85 ± 4.22 (several m,
40H; ArOCH2, Leu-Ha, Lys-Ha), 3.02 ± 2.70 (m, 32H; Lys-He), 1.90 ± 1.10
(m, 120H; Leu-Hb,g, Lys-Hb,g,d), 1.00 ± 0.68 (m, 48H; Leu-CH3); fluores-
cence (H2O, pH 6.4): 316 (excitation), 385 nm (emission); CD (H2O,
pH 6.4): see Figure 4.
3
8.0 Hz, 1H; Glu-NH, exchange with D2O), 7.46 (d, J(H,H) ˆ 8.2 Hz, 1H;
Leu-NH, exchange with D2O), 7.34 (m, 5H; ArH), 7.24 ± 7.18 (m, 1H; Lys-
CONH2, exchange with D2O), 7.08 ± 7.00 (m, 1H; Lys-CONH2, exchange
with D2O), 5.06 ± 4.95 (m, 2H; ArCH2), 4.34 ± 4.22 (m, 1H; Glu-Ha), 4.23 ±
4.09 (m, 1H; Leu-Ha), 4.05 ± 4.09 (m, 1H; Glu-Ha), 2.29 ± 2.12 (m, 4H; Lys-
Hg), 1.96 ± 1.20 (m, 7H; Leu-Hb,g, Glu-Hb), 1.37 (s, 18H; tBu), 0.87 (d,
3J(H,H) ˆ 6.6 Hz, 3H; Leu-CH3), 0.81 (d, 3J(H,H) ˆ 6.6 Hz, 3H; Leu-CH3).
H-Glu(OtBu)-Leu-Glu(OtBu)-NH2: Deprotection of Z-Glu(OtBu)-Leu-
Glu(OtBu)-NH2 (486 mg, 0.77 mmol) by following procedure
B and
purification by column chromatography (CH2Cl2/MeOH/iPr-NH2 90:9:1,
Rf ˆ 0.4) yielded pure H-Glu(OtBu)-Leu-Glu(OtBu)-NH2 (358 mg, 93%)
as a white powder. [a]2D0 ˆ -38.5 (c ˆ 1.00 in MeOH); m.p. 85.5 ± 86.28C; IR
(KBr): nÄ ˆ 3306 (s), 3070 (w), 2977 (m), 2934 (m), 2870 (w), 2354 (w), 1729
(s), 1650 (s), 1542 (m), 1453 (w), 1368 (s), 1257 (m), 1156 (s), 956 (w), 849
Rigid-rod b-barrels 124 (126) (general procedure D): Solutions of 13 (0 ±
32 mmol) in MeOH (20 mL) and/or solutions of 14 (0 ± 32 mmol) in MeOH
(20 mL) were added to 2.5 mL of buffer (10mm NanH3ÀnPO4, pH 6.4).
Products were characterized by gel filtration (GF). SEC: (H2O, pH 6.4): see
Chem. Eur. J. 2000, 6, No. 10
ꢁ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2000
0947-6539/00/0610-1747 $ 17.50+.50/0
1747

S. Matile, B. Baumeister
FULL PAPER
Figure 3a; UV/Vis (H2O, pH 6.4): 316 nm (28.6); CD (H2O): see Figures 4
and 5; fluorescence (H2O, pH 6.4): see Figure 7c. Dependence of b-barrel
formation on rod stoichiometry was assessed by addition of different mole
fractions of 13 and 14 and subsequent measurement of the CD spectra
(Figure 5a). Dependence of b-barrel formation on ionic strength (0 ± 1m
NaCl) and pH (pH 2 ± 12) was assessed with correspondingly changed
buffers (Figures1 5a and c). b-Barrel stability was assessed by heating the
water-jacketed CD-cell and by sample dilution with continuous measure-
ment of the CD spectra (not shown).
versity, and the University of Geneva for support of this work. Further
support by NSF (CHE-9601976) through the Georgetown Molecular
Modeling Center is gratefully acknowledged. Mass spectrometry was
performed by Dr. N. Leffers at Georgetown University Lombardi Cancer
Centerꢀs Macromolecular Analysis Shared Resource, which is supported in
part by U.S. PHS Grant P30-CA-51008, and by Dr. K. Lu and C. Ladd at
University of Maryland, College Park.
Rigid-rod b-barrels 124 (126) (general procedure E): Identical concentra-
tions of 13 and 14 were dissolved in MeOH/chloroform, and sodium cholate
(64.5 mg, 0.15 mmol) and increasing amounts of MeOH/chloroform were
added until a clear solution was obtained. The organic solvents were slowly
evaporated at RT, and the resulting film was dried for at least 2 h in vacuo.
Then NanH3ÀnPO4 (10mm, 1.0 mL, pH 6.4) was added. Hydration of the
resulting suspension was eventually accelerated by sonication or heating.
The mixed micellar mixture was filtered through cotton. This micellar
mixture (1 mL) was then dialyzed 12 times against buffer (1 mL) for >2 h
at RT in the dark by using dialysis cells (Fisher) with membranes from
[1] a) K. K. Kim, R. Kim, S.-H. Kim, Nature 1998, 394, 595; b) A. L.
Horwich, H. R. Saibil, Nat. Struct. Biol. 1998, 5, 333; c) L. Ditzel, J.
Löwe, D. Stock, K.-O. Stetter, H. Huber, R. Huber, S. Steinbacher,
Cell 1998, 93, 125; d) B. Bukau, A. L. Horwich, Cell 1998, 92, 351.
[2] a) W. Baumeister, J. Walz, F. Zühl, E. Seemüller, Cell 1998, 92, 367;
b) J. Löwe, D. Stock, B. Jap, P. Zwickl, W. Baumeister, R. Huber,
Science 1995, 268, 534.
[3] a) L. Joshua-Tor, H. E. Xu, S. A. Johnston, D. C. Rees, Science 1995,
269, 945; b) T. S. R. Krishna, X.-P. Kong, S. Gary, P. M. Burgers, J.
Kuriyan, Cell 1994, 79, 1233; c) C. D. Lima, J. C. Wang, A. Mondra-
Diachema (MW-cutoff ˆ 5000) mounted on
a Thermomixer (Fisher)
Â
gon, Nature 1994, 367, 138; d) X.-P. Kong, R. Onrust, M. OꢀDonnell, J.
shaking at about 1300 rpm. The resulting mixture was filtered twice
through cotton and characterized by gel filtration (GF). SEC: (H2O,
pH 6.4): see Figure 3a; UV/Vis (H2O, pH 6.4): 316 nm (28.6); CD (H2O):
see Figure 4; fluorescence (H2O, pH 6.4): see Figure 7b.
Kuriyan, Cell 1992, 69, 425; e) D. B. Nikolov, S.-H. Hu, J. Lin, A.
Gasch, A. Hoffmann, M. Horikoshi, N.-H. Chua, R. G. Roeder, S. K.
Burley, Nature 1992, 360, 40.
[4] T. A. Baker, S. P. Bell, Cell 1998, 92, 295.
[5] a) L.-Q. Gu, O. Braha, S. Conlan, S. Cheley, H. Bayley, Nature 1999,
Rigid-rod b-barrels 12n (general procedure F): Identical concentrations of
13 and 14 were dissolved in MeOH/chloroform, and n-octyl b-d-glucopyr-
anoside (36.5 mg, 125 mmol) and increasing amounts of MeOH/chloroform
were added until a clear solution was obtained. The organic solvents were
slowly evaporated at RT, the resulting film was dried for at least 2 h in
vacuo. Then NanH3ÀnPO4 (10mm, 1.0 mL, pH 6.4) was added. Hydration of
the resulting suspension was eventually accelerated by sonication or
heating. The mixed micellar mixture was filtered through cotton. This
micellar mixture (1 mL) was then dialyzed once against buffer (1 L) for
>2 h at RT in the dark by using Mini Lipoprepꢂ (Sialomed) with the
dialysis chamber rotating at 20 rpm. The resulting mixture was filtered
twice through cotton and characterized by gel filtration (GF). SEC: (H2O,
pH 6.4): see Figure 3d; UV/Vis (H2O, pH 6.4): 2781 nm (24.0). CD (H2O,
pH 6.4): 288 nm (À2.3).
Rigid-rod lipocalin 19: By following procedure E, b-carotene 11, 13, and 14
(molar ratio ˆ 1:0.5:0.5) were assembled to give 19 in 25% yield with
respect to 11 and 100% yield with respect to 13 and 14 (Table 1, entry 1).
Carotenolipocalin 19 was characterized by gel filtration (GF). SEC: (H2O,
pH 6.4): see Figure 3e; UV/Vis (H2O, pH 6.4): see Figure 7b; CD (H2O,
pH 6.4): see Figure 7a; fluorescence (H2O, pH 6.4): see Figure 7c. Control
experiments: a) excess b-carotene (1:0.1:0.1) gave about identical results;
b) excess rods (1:2.5:2.5): <5% yield; c) without 13 (1:0:1): <5% yield;
d) without 13 and 14 (1:0:0): <5% yield.
Â
Â
398, 686; b) C. Lesieur, B. Vecsey-Semjen, L. Abrami, M. Fivaz, F. G.
van der Goot, Mol. Membrane Biol. 1997, 14, 45; c) L. Song, M. R.
Hobaugh, C. Shustack, S. Cheley, H. Bayley, J. E. Gouaux, Science
1996, 274, 1859; d) T. Schirmer, T. A. Keller, Y.-F. Wang, J. P.
Rosenbusch, Science 1995, 267, 512; e) S. R. Durell, H. R. Guy, N.
Arispe, E. Rojas, H. B. Pollard, Biophys. J. 1994, 67, 2137; f) M. S.
Weiss, A. Kreusch, E. Schiltz, U. Nestel, W. Welte, J. Weckesser, G. E.
Schulz, FEBS Lett. 1991, 280, 379.
[6] a) G. Beste, F. S. Schmidt, T. Stibora, A. Skerra, Proc. Natl. Acad. Sci.
USA 1999, 96, 1898; b) D. R. Flower, Biochem. J. 1996, 318, 1;
Compare also: c) M. Ormö, A. B. Cubitt, K. Kallio, L. A. Gross, R. Y.
Tsien, S. J. Remington, Science 1996, 273, 1392.
[7] a) R. C. Bugos, A. D. Hieber, H. Y. Yamamoto, J. Biol. Chem. 1998,
273, 15321; b) K. K. Niyogi, O. Björkman, A. R. Grossman, Proc.
Natl. Acad. Sci. USA 1997, 94, 14162; c) H. A. Frank, R. J. Cogdell,
Photochem. Photobiol. 1996, 63, 257; d) Y. Koyama, M. Kuki, P. O.
Anderson, T. Gillbro, Photochem. Photobiol. 1996, 63, 243.
[8] a) G. Britton, R. J. Weesie, D. Askin, J. D. Warburton, L. Gallardo-
Guerrero, F. J. Jansen, H. J. M. de Groot, J. Lugtenburg, J.-P. Cornard,
J. C. Merlin, Pure Appl. Chem. 1997, 69, 2075; b) P. F. Zagalsky, Pure
Appl. Chem. 1994, 66, 973; c) J. N. Keen, I. Caceres, E. E. Eliopoulos,
P. F. Zagalsky, J. B. C. Findlay, Eur. J. Biochem. 1991, 202, 31;
d) R. J. H. Clark, N. R. DꢀUrso, P. F. Zagalsky, J. Am. Chem. Soc.
1980, 102, 6693.
[9] a) M. Z. Papiz, L. Sawyer, E. E. Eliopoulos, A. C. T. North, J. B. C.
Findlay, R. Sivaprasadarao, T. A. Jones, M. E. Newcomer, P. J. Kraulis,
Nature 1986, 324, 383; b) R. Hemley, B. E. Kohler, P. Siviski,
Biophys. J. 1979, 28, 447; c) S. Futterman, J. Heller, J. Biol. Chem.
1972, 247, 5168.
[10] N. Sakai, N. Majumdar, S. Matile, J. Am. Chem. Soc. 1999, 121, 4294.
[11] a) S. R. LaBrenz, J. W. Kelly, J. Am. Chem. Soc. 1995, 117, 1655; b) J. P.
Schneider, J. W. Kelly, Chem. Rev. 1995, 95, 2169; For ªnon-rigid-rodº
scaffolds, see: c) J. S. Nowick, Acc. Chem. Res. 1999, 32, 287.
[12] a) B. Baumeister, A. C. de Dios, S. Matile, Tetrahedron Lett. 1999, 40,
4623; b) B. Ghebremariam, S. Matile, Enantiomer 1999, 4, 121; B.
Ghebremariam, S. Matile, Enantiomer 1999, 4, 131 (A. M. Rouhi,
Chem. Eng. News 1999, 77(22), 32); c) M. M. Tedesco, B. Ghebremar-
iam, N. Sakai, S. Matile, Angew. Chem. 1999, 111, 523; Angew. Chem.
Int. Ed. 1999, 38, 540; d) N. Sakai, C. Ni, S. M. Bezrukov, S. Matile,
Bioorg. Med. Chem. Lett. 1998, 8, 2743; e) C. Ni, S. Matile, Chem.
Commun. 1998, 755; f) L. A. Weiss, N. Sakai, B. Ghebremariam, C. Ni,
S. Matile, J. Am. Chem. Soc. 1997, 119, 12142; g) N. Sakai, K. C.
Brennan, L. A. Weiss, S. Matile, J. Am. Chem. Soc. 1997, 119, 8726;
excellent reviews on synthesis and properties of rigid-rod molecules
have appeared very recently: h) P. F. H. Schwab, M. D. Levin, J. Michl,
Chem. Rev. 1999, 99, 1863; i) A. J. Berresheim, M. Müller, K. Müllen,
Rigid-rod disc micelle 20: By following procedure E, astaxanthin 10, 13,
and 14 (molar ratio ˆ 1:0.1:0.1) were assembled to give 20 in 10% yield
with respect to 10 and about 50% yield with respect to 13 and 14 (Table 1,
entry 9). Disc micelles 20 were characterized by gel filtration (GF). SEC:
(H2O, pH 6.4): MW ꢀ84000 (not 100% reproducible); UV/Vis (H2O,
pH 6.4): see Figure 9; CD (H2O, pH 6.4): 602 (‡46), 442 nm (À38);
fluorescence (H2O, pH 6.4): 316 (excitation), 385 nm (emission, 29%
quenching). Control experiments: a) excess rods (1:1:1) gave about
identical results; b) without rods (1:0:0): 17% yield; GF (H2O, pH 6.4):
not eluted; SEC: (H2O, pH 6.4): not eluted; UV/Vis (H2O, pH 6.4): see
Figure 9; CD (H2O, pH 6.4): 572 (À46), 424 nm (‡64); fluorescence (H2O,
pH 6.4): 316 (excitation), 385 nm (emission); c) without 13 (1:0:0.2): 6%
yield; data as in b; d) without 14 (1:0.2:0): <5% yield. Control experiments
with carotenoids 9 and 18 were performed by following procedure E (see
Table 1, entries 4 ± 8).
Acknowledgement
We are grateful to Dr. N. Majumdar for synthesizing 15, and to Dr. A. C.
de Dios for measurement of 500 MHz NMR spectra. We thank NIH
(GM56147), Research Corporation (Research Innovation Award), Suntory
Institute for Bioorganic Research (SUNBOR Grant), Georgetown Uni-
1748
ꢁ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2000
0947-6539/00/0610-1748 $ 17.50+.50/0
Chem. Eur. J. 2000, 6, No. 10

Lipocalin Models
1739±1749
Chem. Rev. 1999, 99, 1747; j) R. E. Martin, F. Diederich, Angew.
Chem. 1999, 111, 1440; Angew. Chem. Int. Ed. 1999, 38, 1351; k) N.
Sakai, S. Matile, Chem. Eur. J. 2000, 6, 1731.
Supramolecules 124 (126) were also not detectable with confidence by
ESI-MS.
[23] a) D. V. Waterhous, W. C. Johnson, Jr., Biochemistry 1994, 33, 2121;
b) L. Zhong, W. C. Johnson, Jr., Proc. Natl. Acad. Sci. USA 1992, 89,
4462; c) D. S. Yang, C. C. Wu, H. M. Martinez, Methods Enzymol.
1986, 130, 208.
[24] a) L. V. Najbar, D. J. Craik, J. D. Wade, D. Salvatore, M. McLeish,
Biochemistry 1997, 36, 11525; b) M. C. Manning, M. Illangasekare,
R. W. Woody, Biophys. Chem. 1988, 31, 77.
[25] a) J. Reed., T. A. Reed, Anal. Biochem. 1997, 254, 36; b) D. W. Choo,
J. P. Schneider, N. R. Graciani, J. W. Kelly, Macromolecules 1996, 29,
355; c) A. Brack, G. Spach, J. Am. Chem. Soc. 1981, 103, 6319; d) P. K.
Sarkar, P. Doty, Proc. Natl. Acad. Sci. USA 1966, 55, 901.
[26] K. A. Connors, BindingConstants , Wiley, New York, 1987, p. 61.
[27] J.-H. Fuhrhop, J. Köning, Membranes and Molecular Assemblies: The
Cytokinetic Approach, The Royal Society of Chemistry, Cambridge
(UK), 1994.
[28] For comparison of toroidal with the widely studied spherical capsules,
see, for example: a) J. Rebek, Jr., Acc. Chem. Res. 1999, 32, 278;
b) L. R. MacGillivray, J. L. Atwood, Angew. Chem. 1999, 111, 1080;
Angew. Chem. Int. Ed. Engl. 1999, 38, 1018; c) A. Jasat, J. C. Sherman,
Chem. Rev. 1999, 99, 931; d) J. de Mendoza, Chem. Eur. J. 1998, 4,
1373; e) T. Douglas, M. Young, Nature 1998, 393, 152; f) R. Warmuth,
Angew. Chem. 1997, 109, 1406; Angew. Chem. Int. Ed. Engl. 1997, 36,
1347.
[13] H. Torii, M. Tasumi, J. Chem. Phys. 1993, 98, 3697, and references cited
therein.
[14] a) J. L. Haley, A. N. Fitch, C. Lambert, T. G. Truscott, J. N. Chacon, D.
Stirling, W. Schalch, J. Chem. Soc. Chem. Commun. 1992, 1175;
b) G. C. Chen, M. Krieger, J. P. Kane, C.-S. C. Wu, M. S. Brown, J. L.
Goldstein, Biochemistry 1980, 19, 4706; c) H. Y. Yamamoto, A. D.
Bangham, Biochim. Biophys. Acta 1978, 507, 119.
[15] a) P. O. Anderson, S. M. Bachilo, R. L. Chen, T. Gillbro, J. Phys.
Chem. 1995, 99, 16199; b) S. L. Bondarev, S. M. Bachilo, S. S.
Dvornikov, S. A. Tikhomirov, J. Photochem. Photobiol. A 1989, 46,
315.
[16] K. Gaier, A. Angerhofer, A. C. Wolf, Chem. Phys. Lett. 1991, 187, 103.
[17] R. Hemley, B. E. Kohler, Biophys. J. 1977, 20, 377.
[18] A. Milon, G. Wolff, G. Ourisson, Y. Nakatani, Helv. Chim. Acta 1986,
69, 12, and references cited therein.
[19] H. Houjou, Y. Inoue, M. Sakurai, J. Am. Chem. Soc. 1998, 120, 4459,
and references cited therein.
[20] a) D. Ranganathan, C. Lakshimi, I. L. Karle, J. Am. Chem. Soc. 1999,
121, 6105; b) J. D. Hartgerink, T. D. Clark, M. R. Ghadiri, Chem.
Eur. J. 1998, 4, 1367; c) T. D. Clark, L. K. Buehler, M. R. Ghadiri, J.
Am. Chem. Soc. 1998, 120, 651; d) D. Seebach, J. L. Matthews, A.
Meden, T. Wessels, C. Baerlocher, L. B. McCusker, Helv. Chim. Acta
1997, 80, 173; e) K. Benner, P. Klüfers, J. Schuhmacher, Angew. Chem.
1997, 109, 783; Angew. Chem. Int. Ed. Engl. 1997, 36, 743; f) G.
Gattuso, S. Menzer, S. A. Nepogodiev, J. F. Stoddart, D. J. Williams,
Angew. Chem. 1997, 109, 1615; Angew. Chem. Int. Ed. Engl. 1997, 36,
1451; g) D. Venkataraman, S. Lee, J. Zhang, J. S. Moore, Nature 1994,
371, 591; h) X. C. Sun, G. P. Lorenzi, Helv. Chim. Acta 1994, 77, 1520;
i) A. Harada, J. Li, M. Kamachi, Nature 1993, 364, 516; j) I. L. Karle,
B. K. Handa, C. H. Hassall, Acta Crystallogr. Sect. B 1975, 31, 555.
[21] J. Coste, D. Le-Nguyen, B. Castro, Tetrahedron Lett. 1990, 31, 205.
[22] Structural studies of 124 (126) were thus restricted to techniques more
sensitive than 500 MHz 1H NMR, 13C NMR and IR spectroscopy.
[29] C. E. Dempsey, Biochim. Biophys. Acta 1990, 1031, 143, and
references cited therein.
[30] T. Goto, T. Kondo, Angew. Chem. 1991, 103, 17; Angew. Chem. Int. Ed.
Engl. 1991, 30, 17.
[31] a) S. J. Wieczorek, K. A. Kalivoda, J. G. Clifton, D. Ringe, A. Petsko,
J. A. Gerlt, J. Am. Chem. Soc. 1999, 121, 4540; b) M. Henning, A.
DꢀArcy, I. C. Hampele, M. P. G. Page, Nature Struct. Biol. 1998, 5, 357;
c) P. C. Babbitt, J. A. Gerlt, J. Biol. Chem. 1997, 272, 30591.
Received: August 2, 1999 [F1955]
Chem. Eur. J. 2000, 6, No. 10
ꢁ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2000
0947-6539/00/0610-1749 $ 17.50+.50/0
1749