Induced folding by chiral nonplanar aromatics

Article abstract of DOI:10.1021/jo9013047

(Chemical Equation Presented) We report a structural motif based on a C₃-symmetric bowl-shaped core, on which three substituted amino acids on the periphery adopt either a folded or a spread-out conformation. This class of chiral folded structu

Full text of DOI:10.1021/jo9013047

pubs.acs.org/joc
Induced Folding by Chiral Nonplanar Aromatics
Sri Kamesh Narasimhan, Deborah J. Kerwood, Lei Wu, Jun Li, Rosina Lombardi,
Teresa B. Freedman,* and Yan-Yeung Luk*
Department of Chemistry, Syracuse University, Syracuse, New York 13244
tbfreedm@syr.edu; yluk@syr.edu
Received June 17, 2009
We report a structural motif based on a C3-symmetric bowl-shaped core, on which three
substituted amino acids on the periphery adopt either a folded or a spread-out conformation.
This class of chiral folded structures is achieved by controlling the reactivity of the stereogenic
protons on the nonplanar aromatic rings of trioxatricornan to afford predominantly C3-
symmetric isomers. Bromination of trioxatricornan afforded a C1-symmetric and a C3-sym-
metric trisubstituted isomer, with the former being the major product as a statistical conse-
quence during the reaction cascade. To obtain the C3 symmetric isomer as the major product,
C-H activation by means of ortho-lithiation with the bulky tert-butyl lithium and tetramethyl-
ethylenediamine was followed by a nucleophilic substitution that successfully reversed the
statistically controlled regioselectivity. Further derivatization of the trioxatricornan with amino
acids or menthol afforded diastereomers that were resolved by preparative chromatography.
The absolute configurations of the diastereomers were determined by vibrational circular
dichroism (VCD) in combination with density functional theory (DFT) and electronic circular
dichroism (ECD). The folding structure of cysteine-derivatized trioxatricornan diastereomers
was determined by two-dimensional NMR spectroscopy and molecular dynamics calculation,
which revealed that one diastereomer has the amino acids folded toward the cavity of
trioxatricornan and the other has a “spread-out” structure.
Introduction
like structures.1-4 In particular, a recent report described
the folding of a peptide containing ten amino acid resi-
dues.5 In this work, we report the design and synthesis of
a dissymmetric nonplanar molecular core that induces
folding of its three substituents on the periphery. We
Folding of natural macromolecules is often a conse-
quence of linear polymers that can form helices or sheet-
*To whom correspondence should be addressed. Phone: (315) 443-7440
(Y.-Y.L.) and (315) 443-1134 (T.B.F.). Fax: (315) 443-4070 (T.B.F.).
(1) Chen, F.; Zhu, N.-Y.; Yang, D. J. Am. Chem. Soc. 2004, 126, 15980–
15981.
(4) Hill, D. J.; Mio, M. J.; Prince, R. B.; Hughes, T. S.; Moore, J. S. Chem.
Rev. 2001, 101, 3893–4011.
(2) Gellman, S. H. Acc. Chem. Res. 1998, 31, 173–180.
(3) Hecht, S.; Huc, I.; Eds.; Foldamers: Structure, Properties and Applica-
tions; Wiley-VCH : Weinheim, Germany, 2007.
(5) Honda, S.; Akiba, T.; Kato, Y. S.; Sawada, Y.; Sekijima, M.;
Ishimura, M.; Ooishi, A.; Watanabe, H.; Odahara, T.; Harata, K. J. Am.
Chem. Soc. 2008, 130, 15327–15331.
DOI: 10.1021/jo9013047
r
Published on Web 08/12/2009
J. Org. Chem. 2009, 74, 7023–7033 7023
2009 American Chemical Society

Narasimhan et al.
JOCArticle
FIGURE 1. Protons are transformed from achiral to prochiral by
making a planar (A) aromatic moiety nonplanar (B). Such a
nonplanar aromatic moiety can be introduced by a bowl-shaped
trioxatricornan (C). Substituting the prochiral protons with a
substitutent renders the molecule chiral.
FIGURE 2. Potential intramolecular interactions for trioxatricor-
nan core substituted with three amino acid residues at the
C3-symmetric positions.
describe the synthesis that reverses the statistically con-
trolled regiochemistry resulting in diastereomers with a
C3-symmetric core. The resolution, absolute configura-
tion determination, and folding characterization for
the diastereomers having a C3-symmetric core are also
elucidated.
A wide range of non-natural molecules have been
demonstrated to fold into preferred structures among a
large ensemble of possible conformations. Examples in-
clude β-peptide2,4,6 and m-phenylene ethynylene oligo-
mers.4,7-9 Other folded structures include synthetic
oligomers,1-4,10-12 synthetic R-peptide sequences,13-19
artificial proteins,20-28 nucleic acids,29-32 and helical
polymers.33-38 Most of these folding molecules involve
stereogenic centers on a linear polymer or oligomer. The
use of highly symmetric, but chiral moieties to build
molecules that fold into preferred conformations has
not been extensively studied.39 Many C3 symmetric, but
achiral molecules have been explored for their utility
toward recognition,40,41 molecular folding,42 and cataly-
sis.43,44 To make use of the C3 symmetry for designing
functions, Anslyn and co-workers have built C3v sym-
metric receptors for facial selectivity40 and for selective
binding of phosphate ions.41
We are interested in discovering new folding motifs and self-
assembly from non-natural structures,45 and exploring the
biocompatibility and new functions46 from these unique struc-
tures, as well as molecules with exotic symmetries.47-49 Here,
we explore using a chiral C3-symmetric core50-55 and intramo-
lecular interactions to induce a folded structure. We reason that
while protons on a planar aromatic ring are, in general, not
prochiral (Figure 1A), protons on a nonplanar aromatic ring
can be prochiral (Figure 1B). To study the effect of the chiral
microenvironment of a nonplanar aromatic on molecular
conformation and folding, we consider a bowl-shaped core
(6) Petersson, E. J.; Schepartz, A. J. Am. Chem. Soc. 2008, 130, 821–823.
(7) Elliott, E. L.; Ray, C. R.; Kraft, S.; Atkins, J. R.; Moore, J. S. J. Org.
Chem. 2006, 71, 5282–5290.
(8) Hartley, C. S.; Elliott, E. L.; Moore, J. S. J. Am. Chem. Soc. 2007, 129,
4512–4513.
(9) Kelley, R. F.; Rybtchinski, B.; Stone, M. T.; Moore, J. S.; Wasielewski,
M. R. J. Am. Chem. Soc. 2007, 129, 4114–4115.
(10) Li, X.; Yang, D. Chem. Commun. 2006, 3367–3379.
(11) Yang, D.; Zhang, D.-W.; Hao, Y.; Wu, Y.-D.; Luo, S.-W.; Zhu,
N.-Y. Angew. Chem., Int. Ed. 2004, 43, 6719–6722.
(12) Horne, W. S.; Gellman, S. H. Acc. Chem. Res. 2008, 41, 1399–1408.
(13) Balaram, P. Curr. Opin. Struct. Biol. 1992, 2, 845–851.
(14) Betz, S. F.; Bryson, J. W.; DeGrado, W. F. Curr. Opin. Struct. Biol.
1995, 5, 457–463.
(15) DeGrado, W. F.; Summa, C. M.; Pavone, V.; Nastri, F.; Lombardi,
A. Annu. Rev. Biochem. 1999, 68, 779–819.
(16) Karle, I. L.; Balaram, P. Biochemistry 1990, 29, 6747–6756.
(17) Venkatraman, J.; Shankaramma, S. C.; Balaram, P. Chem. Rev.
2001, 101, 3131–3152.
(18) Singh, Y.; Dolphin, G. T.; Razkin, J.; Dumy, P. ChemBioChem 2006,
7, 1298–1314.
(19) DeGrado, W. F. Adv. Protein Chem. 1988, 39, 51–124.
(20) Ansari, A. Z.; Mapp, A. K.; Nguyen, D. H.; Dervan, P. B.; Ptashne,
M. Chem. Biol. 2001, 8, 583–592.
(21) Kirby, A. J. Angew. Chem., Int. Ed. 1996, 35, 707–724.
(22) Lokey, R. S.; Iverson, B. L. Nature 1995, 375, 303–305.
(23) Mutter, M. Angew. Chem. 1985, 97, 639–654.
(37) Nolte, R. J. M. Chem. Soc. Rev. 1994, 23, 11–19.
(38) Yashima, E.; Maeda, K. Foldamers 2007, 331–366.
(39) Rump, E. T.; Rijkers, D. T. S.; Hilbers, H. W.; de Groot, P. G.;
Liskamp, R. M. J. Chem.;Eur. J. 2002, 8, 4613–4621.
(40) Hennrich, G.; Lynch, V. M.; Anslyn, E. V. Chem.;Eur. J. 2002, 8,
2274–2278.
(41) Tobey, S. L.; Jones, B. D.; Anslyn, E. V. J. Am. Chem. Soc. 2003, 125,
4026–4027.
(42) Kwak, J.; De Capua, A.; Locardi, E.; Goodman, M. J. Am. Chem.
Soc. 2002, 124, 14085–14091.
(24) Nelson, J. C.; Saven, J. G.; Moore, J. S.; Wolynes, P. G. Science 1997,
277, 1793–1796.
(25) Patch, J. A.; Barron, A. E. Curr. Opin. Chem. Biol. 2002, 6, 872–877.
(26) Qiu, J. X.; Petersson, E. J.; Matthews, E. E.; Schepartz, A. J. Am.
Chem. Soc. 2006, 128, 11338–11339.
(43) Kemp, D. S.; Petrakis, K. S. J. Org. Chem. 1981, 46, 5140–5143.
(44) Das, S.; Incarvito, C. D.; Crabtree, R. H.; Brudvig, G. W. Science
2006, 312, 1941–1943.
(45) Han, Y.; Cheng, K.; Simon, K. A.; Lan, Y.; Sejwal, P.; Luk, Y.-Y. J.
Am. Chem. Soc. 2006, 128, 13913–13920.
(27) Razeghifard, R.; Wallace, B. B.; Pace, R. J.; Wydrzynski, T. Curr.
Protein Pept. Sci. 2007, 8, 3–18.
(46) Sejwal, P.; Han, Y.; Shah, A.; Luk, Y.-Y. Org. Lett. 2007, 9, 4897–
4900.
(28) Wu, C. W.; Seurynck, S. L.; Lee, K. Y. C.; Barron, A. E. Chem. Biol.
2003, 10, 1057–1063.
(47) Narasimhan, S. K.; Lu, X.; Luk, Y.-Y. Chirality 2008, 20, 878–884.
(48) Stang, P. J.; Olenyuk, B. Acc. Chem. Res. 1997, 30, 502–518.
(49) Seidel, S. R.; Stang, P. J. Acc. Chem. Res. 2002, 35, 972–983.
(50) Canary, J. W.; Allen, C. S.; Castagnetto, J. M.; Wang, Y. J. Am.
Chem. Soc. 1995, 117, 8484–8485.
(51) Gibson, S. E.; Castaldi, M. P. Angew. Chem., Int. Ed. 2006, 45, 4718–
4720.
(52) Moberg, C. Angew. Chem., Int. Ed. 1998, 37, 248–268.
(53) Bolm, C.; Meyer, N.; Raabe, G.; Weyhermuller, T.; Bothe, E. Chem.
Commun. 2000, 2435–2436.
(29) Nielsen, P. E. Mol. Biotechnol. 2004, 26, 233–248.
(30) Rothemund, P. W. K. Nature 2006, 440, 297–302.
(31) Chworos, A.; Jaeger, L. Foldamers 2007, 291–329.
(32) LaBean, T. H.; Yan, H.; Kopatsch, J.; Liu, F.; Winfree, E.; Reif, J.
H.; Seeman, N. C. J. Am. Chem. Soc. 2000, 122, 1848–1860.
(33) Amabilino, D. B.; Serrano, J.-L.; Sierra, T.; Veciana, J. J. Polym.
Sci., Part A: Polym. Chem. 2006, 44, 3161–3174.
(34) Green, M. M.; Peterson, N. C.; Sato, T.; Teramoto, A.; Cook, R.;
Lifson, S. Science 1995, 268, 1860–1866.
(35) Maeda, K.; Yashima, E. Top. Curr. Chem. 2006, 265, 47–88.
(36) Nakano, T.; Okamoto, Y. Chem. Rev. 2001, 101, 4013–4038.
(54) Bolm, C.; Davis, W. M.; Halterman, R. L.; Sharpless, K. B. Angew.
Chem. 1988, 100, 882–883.
(55) Bolm, C.; Sharpless, K. B. Tetrahedron Lett. 1988, 29, 5101–5104.
7024 J. Org. Chem. Vol. 74, No. 18, 2009

Narasimhan et al.
JOCArticle
SCHEME 1
structure of trioxatricornan (Figure 1C).56-62 Introducing
three identical substituents at the dissymmetric positions of
its periphery, this bowl-shaped molecule becomes enantiomeric
(if the substitutent is achiral) or diastereomeric (if the substitu-
tent is chiral), and possesses three identical chiral microenvi-
ronments at the C3-symmetric substitution positions.
Here, we explored substituting the C3-symmetric positions
of trioxatricornan with amino acid residues (Figure 2). This
molecular structure provides a rich collection of intramole-
cular interactions including partial conjugation between the
aromatic rings and the carbonyl group from the peptide
bond, the potential intramolecular hydrogen bonds between
the amide protons and the bridge oxygen, and possibly the
steric effect between the side chains of the amino acid
residues and the trioxatricornan moiety. Because of the close
connectivity between these intramolecular interactions and
the dissymmetric core, the molecular interactions have the
potential to guide the overall molecule to adopt a well-
defined minimum-energy structure.
TABLE 1. Temperature Effect on Bromination of Trioxatricornan
entry
R
solvent temp, °C C1:C3 isomer (yield, %)
1
-CH2CH2OH CHCl3
-CH3 HOAc
0
100:0 (60%, 0%)b
100:10 (51%, 5%)
100:50 (61%, 34%)b
100:0 (61%, 0%)b
100:33 (42%, 14%)
2a
3
rt
rt
60
80
-CH2CH2OH CHCl3
-CH2CH2OH CHCl3
4
5a
-CH3
HOAc
aConditions reported previously.59 bratio for molecules 3 and 4.
4,8,12-trioxadibenzopyrene cation (Martin’s salt)60 to 12c-meth-
yl-4,8,12-trioxatricornan, which was further functionalized to
make a siderophore with C3v symmetry. To make an amino acid-
derivatized dissymmetric molecule with C3 symmetry, we first
coupled Martin’s salt with different alkenyl Grignard reagents to
afford bowl-shaped molecules (Scheme 1) followed by convert-
ing the alkenyl groups 1 to primary alcohols 2. Compound 2a
was chosen for subsequent studies as it is furnished in higher
yields when compared with 2b and 2c. Bromination of 2a in
chloroform resulted in 61% C1 tribromosubstituted 3 and 34%
C3 tribromosubstituted 4 trioxatricornan.
Statistical Control of Regioselectivity. The substitution
reactions of trioxatricornan appear to be ortho-directed by
the bridge oxygen, and lead to trisubstituted products that
are predominantly C1 symmetric rather than C3 symmetric.
This preference in regiochemistry is likely controlled by the
statistics of the substitution as reaction progresses from
monosubstitution to disubstitution and finally trisubstitu-
tion (see the Supporting Information). Without considering
the small energy difference, the statistics prescribe a product
ratio of 75:25 for C1-symmetric and C3-symmetric isomers.
Our own efforts in bromination and past examples by Siegel
and coworkers57 which included other electrophilic aromatic
reactions such as nitration with HNO3, bromination with
Br2/HOAc, or silylation with n-BuLi/TMSCl of trioxatricornan,
have resulted in ratios of the C1-symmetric to C3-symmetric
isomers all around 75:25 (C1:C3).
Results and Discussion
Bromination of Trioxatricornan. The trioxatricornan core
has been synthesized previously.59,63 Lofthagen et al.64 converted
(56) Laursen, B. W.; Krebs, F. C. Angew. Chem., Int. Ed. 2000, 39, 3432–
3434.
(57) Lofthagen, M.; Chadha, R.; Siegel, J. S. J. Am. Chem. Soc. 1991, 113,
8785–8790.
(58) Lofthagen, M.; Siegel, J. S. J. Org. Chem. 1995, 60, 2885–2890.
(59) Lofthagen, M.; VernonClark, R.; Baldridge, K. K.; Siegel, J. S. J.
Org. Chem. 1992, 57, 61–69.
(60) Martin, J. C.; Smith, R. G. J. Am. Chem. Soc. 1964, 86, 2252–2256.
(61) Mobian, P.; Nicolas, C.; Francotte, E.; Burgi, T.; Lacour, J. J. Am.
Chem. Soc. 2008, 130, 6507–6514.
(62) Laursen, B. W.; Soerensen, T. J. J. Org. Chem. 2009, 74, 3183–3185.
(63) Faldt, A.; Krebs, F. C.; Thorup, N. J. Chem. Soc., Perkin Trans. 2
1997, 2219–2227.
(64) Lofthagen, M.; Siegel, J. S.; Hackett, M. Tetrahedron 1995, 51, 6195–
6208.
J. Org. Chem. Vol. 74, No. 18, 2009 7025

Narasimhan et al.
JOCArticle
FIGURE 3. (A) The proposed model for preventing second lithiation on the neighboring proton (Ha) of trioxatricornan. The first ortho-
directed lithiation provides either steric or electronic influence by the presumed chelation or organolithium-based aggregates, which prevents
the second lithiation at the nearest neighboring site (Ha). The arrows indicate positions left possible for the second lithiation. (B) Substitution
statistics controlled by stereoelectronic influence of TMEDA/t-BuLi.
SCHEME 2
Table 1 shows the selected conditions and results for
bromination of trioxatricornan. For all conditions, the
C1-symmetric isomer is the predominant product over the
C3-symmetric isomer. The exact ratio of the two depends on
the experimental conditions. The most interesting observa-
tion for this reaction is its dependence on temperature.
At both a low temperature (0 °C) and a high temperature
(60-80 °C), the proportion of the C3-symmetric isomer
(0-14%) was significantly lower than the statistically pre-
dicted 25% (C1:C3=75:25), whereas at an ambient tempera-
ture the proportion of C3-symmetric isomer can increase to
34%, which is higher than the statistically predicted 25%
(see the Supporting Information). Calculations based on
density functional theory (DFT)65,66 suggest that the
energy difference between the C1-symmetric isomer and
C3-symmetric isomer is marginal, with the C3 isomer being
0.2 kcal/mol lower than the C1 isomer. However, based on
the temperature effect on the substitution reaction in chloro-
form, we believe that the C1-symmetric isomer is likely both
the kinetic and thermodynamic product. At low tempera-
ture, we believe kinetic control dominates for this reaction,
affording mostly the C1-symmetric product; at intermediate
temperature the C3-symmetric product becomes more acces-
sible while the reaction is still largely irreversible, and thus
the increase in yield of C3-symmetric product; at a suffi-
ciently high temperature, the reaction becomes reversible
and thermodynamic control dominates and affords predo-
minately the C1-symmetric product again.
Reversing the Regioselectivity. In this work, we are inter-
ested in how the repeated microenvironment at C3-sym-
metric positions can couple with the stereogenic centers of
the amino acid residues to induce folding in the molecule. To
break the statistical influence and reverse the regioselectivity
to obtain the desired C3-symmetric isomer, we seek to
introduce either steric and/or electronic influence on the tri-
oxatricornan ring to prevent the disubstitution on the closest
neighboring positions, which leads to an undesired C1-sym-
metric isomer (Figure 3A). With use of MOM-protected
trioxatricornan 7 (Scheme 2), ortho-lithiation67-69 by tert-
butyllithium and TMEDA followed by the addition of
1,2-diiodoethane validated this approach, which affords the
(67) Gschwend, H. W.; Rodriguez, H. R. Org. React. 1979, 26, 1–360.
(68) Shimano, M.; Meyers, A. I. J. Am. Chem. Soc. 1994, 116, 10815–
10816.
(65) Hohenberg, P.; Kohn, W. Phys. Rev. B 1964, 136, 864–871.
(66) Kohn, W.; Sham, L. J. Phys. Rev. A 1965, 140, 1133–1138.
(69) Snieckus, V. Chem. Rev. 1990, 90, 879–933.
7026 J. Org. Chem. Vol. 74, No. 18, 2009

Narasimhan et al.
JOCArticle
SCHEME 3
desired C3-symmetric isomer of triiodo-substituted trioxa-
tricornan, 9, as the major product (isolated yield: 55%), and
the diiodosubstituted isomers 8 containing a trace amount
of chiral disubstitution product (total yield: 32%). No
C1-symmetric isomer was observed. Because the two closest
protons on the two aromatic rings in trioxatricornan are
tert-butyllithium is used. While using normal butyllithium
afforded both C1-symmetric (∼8%) and C3-symmetric
(∼5%) trisubstituted product, using the bulkier tert-butyl-
lithium gave only C3-symmetric product (∼8-10%) (see
the Supporting Information). Using other electrophiles,
including carbon dioxide, diethylcarbonate, or diphenyl-
carbonate, afforded only disubstituted products.
˚
about 4.7 A apart, the approach of a second tert-butyllithium
reagent to the aromatic hydrogen (Ha) closest to the first
lithiation is prevented by either the steric hindrance of the
bulky chelate (or the putative formation of organolithium
aggregates)70-73 or the lack of further chelation from the
bridged oxygen for the second lithium cation.
Synthesis and Resolvability of C3-Dissymmetric Molecules.
Using this controlled regiochemistry, we derivatized the
trioxatricornan with esters and amino acids at C3-symmetric
positions, which resulted in a new class of C3-symmetric
diastereomers (Scheme 3). The trisubstituted trioxatricornan
6 (or 9) was lithiated by t-BuLi and treated with ethylchloro-
formate or (-)-(1R)-menthylchloroformate to afford race-
mic 12up and 12down. Basic hydrolysis of 10 followed by
amide bond formation with amino acid residues cysteine or
histidine resulted in diastereomers of 13up and 13down, as
well as 14up and 14down. The terms “up” or “down” denote
the low or high polarity, respectively, as measured on the
TLC, their correlation with the absolute configurations
as shown are determined by using vibrational circular
dichroism and density functional calculation, see below.
Resolution for Amino Acid Substituted Trioxatricornan Is
Easier than That for (-)-Menthol Substituted of Trioxa-
tricornan. L-Histidine and L-cysteine (1-(triphenylmethyl)-
L-histidine methyl ester and triphenylmethyl-L-cysteine ethyl
ester) were substituted on trioxatricornan at C3-symmetric
positions. For histidine and cysteine derivatized C3-trioxa-
tricornan, the top and bottom diastereomers 13up/13down
and 14up/14down are easily resolvable by flash column
chromatography. For trioxatricornan derivatized with
menthyl ester 12up/12down, the diastereomers are more diffi-
cult to separate than the amide diastereomers. The purifica-
tion of 12up and 12down was accomplished by 14 repetitive
runs on a preparative thin layer chromatography (PTLC)
plate that employed 5% EtOAc in pentane.
Figure 3B describes a proposed reaction cascade of
multiple lithiation that accounts for the product distribu-
tion. Two of the three possible positions for second lithia-
tion afford the chiral dilithiated trioxatricornan, and the
other one affords an achiral dilithiated trioxatricornan.
Whereas the chiral dilithiated product proceeds with the
third lithiation at the C3-symmetric position, the achiral
dilithiation product cannot proceed with a third lithiation
because neither of the ortho positions on the third ring are
accessible. This lithiation pattern leads to a ratio of 2:1
for the C3-symmetric product and diiodo-substituted
trioxatricornan, which is consistent with the observed
yields. Direct carbonylation of trioxatricornan without
any substituted halides also supported the proposed
stereochemical control. Using normal butyllithium and
tert-butyllithium followed by ethyl chloroformate re-
sulted in overall low yield for the products, but with
a ratio that showed reversal of the regiochemistry when
(70) Beak, P.; Musick, T. J.; Liu, C.; Cooper, T.; Gallagher, D. J. J. Org.
Chem. 1993, 58, 7330–7335.
(71) Gossage, R. A.; Jastrzebski, J. T. B. H.; van Koten, G. Angew.
Chem., Int. Ed. 2005, 44, 1448–1454.
(72) Pratt, L. M. Mini-Rev. Org. Chem. 2004, 1, 209–217.
(73) Gruver, J. M.; Liou, L. R.; McNeil, A. J.; Ramirez, A.; Collum, D. B.
J. Org. Chem. 2008, 73, 7743–7747.
J. Org. Chem. Vol. 74, No. 18, 2009 7027

Narasimhan et al.
JOCArticle
FIGURE 4. Assignment of handedness of the C3 core (upper frame) and calculated optimized geometries (lower frames, Gaussian 03: DFT
B3LYP/6-31G(d) level) for diastereomer models “R” and “S” for comparison with samples 14up and 14down. Cores are viewed here concave-
up, with the ethyl group below the core. The arrows and the numbering show the priority for assigning “R” and “S” absolute configurations.
SCHEME 4
Increasing the Distance between the Stereogenic Center and
the Trioxatricornan Core Reduces Resolvability. Substituting
a rigid phenyl linker between the amino acid and the trioxatri-
cornan moiety afforded diastereomers 17 (Scheme 4). Multiple
elutions with various solvent systems on the same TLC plates did
not afford any observable separation for these diastereomers.
Furthermore, whereas the chemical shifts in proton NMR are
different between each of the three pairs of diastereomers
described above, the chemical shift of aromatic protons in 17
showed a clean set of four doublets: two doublets arise from the
two chemically equivalent sets of protons on the aromatic rings
of trioxatricornan and the remaining two doublets arise from the
identical phenyl linkers. These results suggest that the stereogenic
centers of the amino acids are not coupled with the dissymmetry
of the central fused aromatic rings.
Vibrational Circular Dichroism (VCD) of the Diastreomers
of 13 and 14. To determine the absolute configurations of 14up,
14down, 13up, and 13down, vibrational circular dichroism
(VCD)74 spectra was measured for samples in CDCl3 solu-
tion and compared to the calculations (DFT level, B3LYP
functional, 6-31G(d) basis set) for the model structures for
diastereomers of 14 shown in Figure 4. For the purpose of
assigning the absolute configuration, we use the following
guidelines to describe the handedness of C3-dissymmetric
molecules. Our analysis involves (1) viewing along the C3 axis
through the concave face and (2) deducing the priority of atoms
on the periphery of the ring from high to low using the
(74) Freedman, T. B.; Cao, X.; Dukor, R. K.; Nafie, L. A. Chirality 2003,
15, 743–758.
7028 J. Org. Chem. Vol. 74, No. 18, 2009

Narasimhan et al.
JOCArticle
FIGURE 6. ECD spectra for “S” and “R” trioxatricornan cores
derivatized with (-)-menthol (12up and 12down) collected at room
temperature in a 1.0 cm path length quartz cell. ECD signals were
measured at 1 nm intervals with averaging times of 5s/nm. The
spectral bandwidth was 1.5 nm. Sample concentrations were 0.074
(12up) and 0.068 μM (12down) in THF.
In other regions with large contributions from aromatic ring
deformations, such as “c”, “d”, “e”, and “f”, the A and E
components are not resolved, and the VCD reflects the net
intensity bias of the couplet, yielding near mirror-image
VCD signals for the diastereomers. For regions “g”, “h”,
and “i”, the large contribution of methine deformations at
the L-amino acid chiral centers is the dominant determinant
of the VCD signal, and mode a is primarily the ester CdO
stretch; for these normal modes, the propeller sense changes
only the magnitude, but not the sign of the VCD. The
experimental VCD spectra for 13 show VCD patterns similar
to those of 14, but considerably weaker in magnitude.
FIGURE 5. IR (lower frame) and VCD (upper frame) spectra
observed for 13up/13down and 14up/14down (left axes, 3-5 mg
sample/150 μL CDCl3) compared to calculation for diastereomer
models “R” and “S” (right axes, DFT B3LYP functional/6-31G(d)
basis set). Solid lines are spectra for 13down, 14down, and “R”;
dotted lines are spectra for 13up, 14up, and “S”. Uppermost trace is
VCD noise for 14up.
guidelines established by Cahn-Ingold and Prelog.75 If the
direction of precedence along the periphery is clockwise, then
the handedness is designated as “R”; if the direction is anti-
clockwise, then the handedness is designated as “S”.
The experimental IR and VCD spectra of 14up, 14down,
13up, and 13down in CDCl3 solution are compared to the
calculation in Figure 5 for the 1800-1100 cm-1 region. The
IR spectra of each pair of samples overlay closely, indicative
of similar high purity for each pair, and the VCD spectra
show bands that are representative of diastereomers in the
given spectral range, indicative of similar optical purity for
each pair. The VCD spectra were obtained with a very good
signal-to-noise ratio (ΔA ≈ 10-6).
The calculated VCD spectra are for model diastereomers
for 14, each in a single conformation. The large size of the
molecules under investigation precludes VCD calculation of
the entire structure and identification of the large number of
solution conformers possible for the model structures. The
conformations of models “R” and “S” include stabilization
by intramolecular hydrogen bonding between the peptide
NH and diaryl ether oxygen of the C3 core (a 180° rotation of
the peptide produces a conformation much higher in energy)
and weak intramolecular hydrogen bonding between NH
and ester CdO, one of the many possible conformations for
the amino acid centers. For both 14up/14down and 13up/
13down, the NH stretch was measured at 3420 cm-1, con-
sistent with a moderate to weak nonlinear hydrogen bond.
Despite the limitations of the models, the calculated VCD
spectra reproduce many of the observed VCD intensity pat-
tern differences for the diastereomers (regions “c”, “d”, “e”,
“f”, “h”), and the overall VCD patterns and smaller differ-
ences in the region with large amino acid methine deforma-
tion contributions (below 1400 cm-1). For region b, the
calculated A and E components for the peptide CdO stretch
are not resolved as well as in the experimental spectrum;
however, the sense of the couplet, (-,þ) for “R”, (þ,-) for
“S”, is reproduced in the calculated intensities for the A and
E components, and this sense confirms the orientation of the
peptide and consequent NH--O intramolecular hydrogen
bonding for the experimental samples. For region “a”, the
VCD calculation is opposite in sign to that observed. The
VCD of this ester CdO stretch is sensitive to orientation of
the ester group. The observed positive VCD for region “a”
The calculated IR spectra agree well in pattern with the
observed, providing evidence that the normal vibrational
modes are well-produced at this DFT level of calculation.
Some of the differences between observed and calculated IR
spectra arise from modes of the peripheral alkyl groups
truncated to generate the model. The experimental VCD
spectra of 13 show regions of opposite sign for the diaster-
eomers (regions “b”, “c”, “d”, “e”, and “f” in Figure 5) that
reflect the opposite propeller configurations of the diastere-
omers, and regions with similar VCD pattern (regions “a”,
“g”, “h”, and “i” in Figure 5) that have large contributions
from vibrations of the L-amino acid substituents. The
coupled vibrations under C3 symmetry give rise to A and E
symmetry species components that produce a VCD couplet
of opposite sense for the two propeller configurations, as
demonstrated by couplet “b” for the peptide CdO stretches.
(75) Cahn, R. S.; Ingold, C.; Prelog, V. Angew. Chem., Int. Ed. Engl. 1966,
5, 385–415.
J. Org. Chem. Vol. 74, No. 18, 2009 7029

Narasimhan et al.
JOCArticle
FIGURE 8. Effect of chiral substituents on the ECD signal; the
trioxatricornan core and the amino acid residues are ECD active but
the menthol groups are not.
electronic transitions in this region, but the remainder of the
trioxatricornan group with three carbonyl groups are en-
antiomeric and are active for ECD (Figure 8).
The ECD of 13up/13down and 14up/14down show a com-
mon pattern, but do not exhibit mirror-image characteristics
(Figure 7). The ECD spectra of 13up and 14up show positive
Cotton effects with bands centered at 238, 260, 296 nm and 235,
266, 296 nm, respectively, and negative Cotton effects with
bands at 248, 280 nm and 253, 279 nm, respectively. Similarly,
13down and 14down show positive Cotton effects at bands
centered at 247, 279 nm and 243, 279 nm, respectively, and
negative Cotton effects with bands centered at 230, 262, 295 nm
and 230, 260, 295 nm, respectively. Overall, the electronic CD
transitions occur at similar wavelengths, but with different
intensities, which is consistent with the diastereomeric nature
between 13up/13down and 14up/14down where the peptide
conjugation with the trioxatricornan and proximity to the
chiral amino acid directly contribute in this spectral region.
We note that the bands for the high wavelengths from 279 to
295 nm did exhibit similar intensity with opposite Cotton effect
for each pair of diastereomers.
FIGURE 7. ECD spectra of for “S” and “R” trioxatricornan core
derivatized with L-histidine (13up and 13down) or with L-cysteine
(14up and 14down) collected at room temperature in a 1.0 cm path
length quartz cell. ECD signals were measured at 1 nm intervals with
averaging times of 5s/nm. The spectral bandwidth was 1.5 nm.
Sample concentrations were 138 μM for 13up, 123 μM for 13down,
10 μM for 14up, and 10 μM for 14down in THF.
suggests a prevalence of other orientations for the amino acid
ester substituents, such as those deduced for a conforma-
tionally averaged structure by NMR (structure derived by
molecular dynamics). The weaker VCD signal (relative to the
IR) for diastereomers 13 compared to 14 suggests lower
optical purity and/or much greater conformational flexibil-
ity for 13. However, the observed VCD pattern differences
between 13up and 13down also correspond to the calculated
VCD differences between models “R” and “S”.
The comparison between observed VCD spectral patterns
for 13 and 14 and the calculated VCD spectra for models “R”
and “S” provides assignment of 13up and 14up to “S”
and assignment of 13down and 14down to “R” for the
absolute configurations of the core of the C3-symmetric
trioxatricornan.
Compared with absolute configuration determined by the
VCD, these results suggest that there is a common relation
between the absolute configuration and the electronic circular
dichroism for these molecules. For example, both 13up and
14up have “S” configuration and a positive ECD band at
296 nm, and both 13down and 14down have “R” configuration
and a negative ECD band at 295 nm.
Determination of Absolute Configuration of the Diastereo-
mers 12up and 12down by ECD Inferences. Examining the
bands in the ECD suggests that the bands in 13up [(þ)
Cotton at 238, 260, and 296 nm and (-) Cotton at 248 and
280 nm] and 14up [(þ) Cotton at 235, 266, and 296 nm and
(-) Cotton at 253 and 279 nm] are very similar to bands
observed in 12up [(þ) Cotton at 238, 262, and 300 nm and (-)
Cotton at 252 and 281 nm]. Likewise, the bands in 13down
[(þ) Cotton at 247 and 279 nm and (-) Cotton at 230, 262,
and 295 nm] and 14down [(þ) Cotton at 243 and 279 nm and
(-) Cotton at 230, 260, and 295 nm] are very similar to bands
observed in 12down [(þ) Cotton at 252 and 279 nm and (-)
Cotton at 238, 261, and 300 nm]. Because the observed
correlation between the absolute configuration and the
pattern of ECD for 13up/13down and 14up/14down, and
because similar patterns in ECD are seen for 12up and
12down (especially at long wavelength, ∼296-300 nm), we
Electronic Circular Dichroism (ECD) of the Diastreomers
of 12, 13, and 14. The diastereomers of 12up and 12down
exhibited electronic circular dichroism with remarkable mir-
ror-image characteristics. For 12up, three bands in the ECD
spectra centered at 238, 262, and 300 nm show positive
Cotton effects; for 12down, three bands were also seen
centered at 238, 261, and 300 nm with the same intensity
magnitude, but exhibited negative Cotton effects. Further-
more, diastereomer 12up exhibited two bands centered at 252
and 281 nm with negative Cotton effects, and diatereomer
12down also exhibited two bands centered at 252 and 279 nm
but with positive Cotton effects (Figure 6). This remarkable
mirror-image characteristic for the ECD for 12up and
12down indicates that even though 12up and 12down are
diastereomers, the electronic transitions of the saturated
menthol group are too high in energy to contribute to the
7030 J. Org. Chem. Vol. 74, No. 18, 2009

Narasimhan et al.
JOCArticle
TABLE 2. Correlation between Observed Polarity, Cotton Effects from ECD, and Absolute Configuration of Diastereomers of 12, 13, and 14
14up 14down 13up 13down 12up 12down
apolarity (Rf)
Cotton effect (wavelengths) positive (296 nm) negative (295 nm) positive (296 nm) negative (295 nm) positive (300 nm) negative (300 nm)
absolute configuration
eS eR eS eR fS fR
b0.48
b0.32
c0.27
c0.19
d0.43
d0.38
aRf values from thin layer chromatography (TLC). bEluents for TLC: 50% EtOAc in Hexanes. cEluents: 4% MeOH in CH2Cl2. dEluents: 5% EtOAc
in pentane. eDetermined by vibrational circular dichroism and DFT calculation. fDetermined by indirect inference from electronic circular dichroism.
and CDCl3. The diastereomers exhibited identical patterns
of NOE correlations in both solvents. Figure 9 shows the
NOESY spectrum for 14up in DMSO-d6 with a mixing time
of 1.2 s; the NOESY spectrum for 14down is included in the
Supporting Information. No long-range NOEs between
aromatic protons (a and b) and amino acid residue protons
(d and e) were observed for both of the diastereomers (see the
Supporting Information).
Without ring constraint, a secondary amide bond, in
general, does not exist in the “E” configuration.76-78 The
“Z” configuration can be favored by roughly 2 kcal/mol
relative to the “E” configuration.77,79 Because of the partial
conjugation of the carbonyl group with the aromatic ring,
the “Z” amide bonds in diastereomers 14up and 14down can
adopt a conformation of either “A” or “B” form (Figure 9).
Both of the conformations “A” and “B” prevent the protons
on the amino acid residues (d and e) from being in close
proximity to aromatic protons (a and b) for an NOE
correlation. These NOEs were not observed. We note that,
if the diastereomers adopt “E” amide bonds for either rota-
mers concerning the aromatic-carbonyl C-C bond, the
amino acid protons and the aromatic protons on the triox-
atricornan ring will be in close enough promixity for an NOE
correlation.
FIGURE 9. NOESY spectrum of 14up in DMSO-d6 shows the
For conformation “A” with “Z” amide configuration, the
relayed NOE from protons a to b to c to d to e, which supports
amide proton (c) is closer to the aromatic proton (a) than to
conformation “A”. The NOESY spectrum for 14down exhibited
proton (b); whereas for conformation “B”, the amide proton
identical NOE correlations (see the Supporting Information).
(c) is closer to aromatic proton (b) than to proton (a). Thus,
the NOE correlation characterizing the distances between
these two sets of protons (amide proton c to aromatic
protons a or b) should distinguish between conformations
“A” and “B”. Examining the NOESY shown in Figure 9
indicated that the NOE signal between amide proton (c)
and aromatic proton (a) is about 1.5 time stronger than that
between proton (c) and aromatic proton (b). This bias in
NOE intensity was observed in the NOESY spectra for both
14up and 14down in DMSO-d6 and in CDCl3 (see the
Supporting Information). Thus, we conclude that conforma-
tion “A” rather than “B” is preferred for both of the
diastereomers. Because of the close promixity, conformation
“A” is also consistent with the existence of intramolecular
hydrogen bonding between the N-H proton and the bridged
oxygen on the trioxatricornan core.
infer that the menthyl ester diastereomer 12up has the “S”
absolute configuration for the trioxatricornan core, the same
as that for 13up and 14up. The diastereomer 12down has the
“R” absolute configuration, the same as that for 13down and
14down. Together with the TLC results, this inference also
results in the “S” diastereomers being less polar than the “R”
for the conformations on the TLC plate.
We note that the polarity, based on TLC, reveals that, for
the diastereomers studied in this work, the “R” configura-
tion in the core is consistently more polar than the “S”
configuration. Table 2 summarizes the observed polarity,
Cotton effect of ECD at long wavelength, and the absolute
configurations of the diastereomers.
NMR Analysis of the Structures of the Diastereomers.
Chemical shifts of protons in each pair of the resolved
diastereomers were different. For example, protons (b) ex-
hibited a chemical shift at 8.13 ppm for 14up, but 7.93 ppm for
14down in CDCl3. Even for diastereomers of menthol deriva-
tized trioxatricornan that were difficult to resolve by column
chromotography, the aromatic protons exhibited a difference
in chemical shifts of 0.03 ppm in CDCl3 (see the Supporting
Information). These results suggest that the protons are in
different microenvironments on each diastereomer.
To assess the local conformation and the spatial location
of substituted amino acid residues relative to the chiral
aromatic core of 14up and 14down, the relay NOEs between
protons (a/b to c), (c to d/e), and the torsional angles that
define the bond H-N-C*-H were examined. The NOEs
between protons (a/b to c) and (c to d/e) are similar in their
(76) Gardner, R. R.; McKay, S. L.; Gellman, S. H. Org. Lett. 2000, 2,
2335–2338.
(77) Stewart, W. E.; Siddall, T. H. III Chem. Rev. 1970, 70, 517–51.
(78) Weiss, M. S.; Jabs, A.; Hilgenfeld, R. Nat. Struct. Biol. 1998, 5, 676.
(79) Radzicka, A.; Pedersen, L.; Wolfenden, R. Biochemistry 1988, 27,
4538–4541.
The Nuclear Overhauser Enhancement Spectroscopy
(NOESY) experiments for amino acid derivatized diastere-
omers 14up and 14down were conducted in both DMSO-d6
J. Org. Chem. Vol. 74, No. 18, 2009 7031

Narasimhan et al.
JOCArticle
FIGURE 10. NMR derived structures for diastereomers in vacuum: (A) 14up is folded toward the interior of the cavity and (B) 14down has an
open folded structure.
intensities (see the Supporting Information). The averaged
coupling constants (3JH-N-C-H) for the N-C* bonds in the
amino acid residues of the two diastereomers were close; 7.43
Hz for 14up and 7.28 Hz for 14down, implying a similar range
of torsional angles for both of the diastereomers. Together,
these results suggest that the local conformations of the
amino acid residues in both diastereomers are comparable.
Because the substituted amino acids are of a pure L-
configuration, substitutions at C3-symmetric (chiral) posi-
tions are diastereotopic. Furthermore, because the local
conformations of the amino acids are similar, the absolute
“S” or “R” configuration of the core structure will govern
how the amino acid residues are oriented relative to the
trioxatricornan core. Specifically, the “S” enantiomer with
conformation “A” will orient the cysteine residue toward the
concave cavity of the trioxatricornan core, whereas the “R”
enantiomer will orient the cysteine outside the cavity of the
trioxatricornan core. We note that, although these NOE
signals, together with difference in polarity of the molecules,
suggest different folded structures for the two diastereomers,
the amino acid residues may afford more dynamic, albeit
restricted, local conformations due to the thermal fluctua-
tion and freedom in the rotation of N-C* bonds.
We applied molecular dynamics (MD) using experimental
parameters generated by NOESY experiments for the two
diastereomers, 14up and 14 down. These calculations were
performed in a vacuum (using DISCOVER module in
Insight Program)80 and with solvents including CHCl3 and
DMSO (using AMBER 10).81 The two diastereomers show
different global structures. In the vacuum, the energy-
minimized structures for both diastereomers have “Z” rota-
mers for the amide bonds, and identical core structures. In
diastereomer 14up, the side chains of the amino acids fold
toward the center of the chiral core; in 14down, the side
chains of the amino acids spread outward from the ring
periphery (Figure 10). We note that the amino acid residues
exhibit a more dynamic range although their orientations
remain differently relative to the trioxatricornan core. MD
simulations for both diastereomers in solvents are in agree-
ment with the minimized structures obtained in the vacuum
(see the Supporting Information). Furthermore, for all the
calculated structures, the amide proton (c) and the bridged
oxygen atoms are in close proximity to each other, within a
distance consistent with the existence of intramolecular
hydrogen bonds. The calculated N-H---O bond distances
for 14up are 2.0, 2.27, and 1.98 A, and for 14down they are
˚
2.27, 2.54, and 2.07 A. The existence of intramolecular
hydrogen bonds is further supported by the observation that
the downfield signals arising from amide N-H protons were
not perturbed significantly (Δδ(NH) < 0.25 ppm) when the
solvent was changed from CDCl3 to DMSO-d6.82,83
˚
Conclusions
We have synthesized a new class of chiral bowl-shaped
C3-symmetric trioxatricornan molecules. Without deliberate
control, the regioselectivity for the tribromination of trioxa-
tricornan is dominated by the statistical outcomes to afford
predominately C1-symmetric tribromo trioxatricornan. We
reversed this statistically dominating regioselectivity by
using bulky lithiation reagents to afford predominately
C3-symmetric triiodosubstituted trioxatricornan in good
yields without the C1-symmetric triiodo product. Derivatiz-
ing the C3-symmetric halogenated trioxatricornans with
L-amino acids or (-)-menthol afforded diastereomers that
can be resolved by column chromatography. The absolute
configurations for the resolved diastereomers have been
determined by a combination VCD and DFT calculation,
or ECD inference. Consistent among the three pairs of the
diastereomers, the “R” configuration for the trioxatricornan
core is more polar than the “S” configuration.
The similar NOE signals suggest that the two diastereo-
mers adopt highly similar enantiomeric core structures, in
which the amide hydrogen is close to the aryl hydrogen on
the trioxatricornan that is adjacent to the bridged oxygen.
(80) Accelrys Software Inc., San Diego, CA.
(81) Case, D. A.; Darden, T. A.; Cheatham, T. E.; Simmerling, C. L.;
Wang, J.; Duke, R. E.; Luo, R.; Crowley, M.; Walker, R. C.; Zhang, W.;
ꢀ
Merz, K. M.; Wang, B.; Hayik, S.; Roitberg, A.; Seabra, G.; Kolossvary, I.;
Wong, K. F.; Paesani, F.; Vanicek, J.; Wu, X.; Brozell, S. R.; Steinbrecher,
T.; Gohlke, H.; Yang, L.; Tan, C.; Mongan, J.; Hornak, V.; Cui, G.;
Mathews, D. H.; Seetin, M. G.; Sagui, C.; Babin, V.; Kollman, P. A.;
AMBER, 10th ed.; University of California: San Francisco, CA, 2008.
(82) Le Grel, P.; Salauen, A.; Mocquet, C.; Le Grel, B.; Roisnel, T.; Potel,
M. J. Org. Chem. 2008, 73, 1306–1310.
(83) Kopple, K. D.; Ohnishi, M.; Go, A. Biochemistry 1969, 8, 4087–4095.
7032 J. Org. Chem. Vol. 74, No. 18, 2009

Narasimhan et al.
JOCArticle
Because of the NOE relay between the amide hydrogen and
the protons in the amino acid residue, this core structure
further defines the conformation ensemble of the substituted
amino acids and the global folding of the structure. As a
result, this work identified a folding moiety that consists of
stereogenic centers connected by semirigid bonds to non-
planar aromatics. We believe that these chiral bowls have the
potential for use in building novel supramolecular assembly
and mimicking functional biological molecules.
12down: 1H NMR (300 MHz, CDCl3) δ 7.92 (d, J=8.7 Hz,
3H), 7.05 (d, J=8.7 Hz, 3H), 4.99 (m, 3H), 4.35 (s, 2H), 3.38 (t,
J=6.8 Hz, 2H), 3.18 (s, 3H), 2.14-1.07 (m, 27H), 0.98-0.80 (m,
27H); 13C NMR (300 MHz, CDCl3) δ 164.5, 155.2, 152.1, 132.2,
116.0, 113.0, 112.4, 97.1, 77.6, 63.8, 55.7, 47.8, 44.5, 41.4, 34.7,
31.9, 27.5, 26.7, 23.8, 22.5, 21.3, 16.7; HRMS calcd for
C56H72O11 [M]Naþ 943.4967, found 943.4946.
1,5,9-Tri(S-trt-L-cysteine ethyl ester)-12c-(2-(methoxymetho-
xy)ethyl)-4,8,12-trioxatricornan Diastereomers (14up and 14do-
wn). To a solution of the crude acid (0.021 g, 0.041 mmol) and S-
trt-L-cysteine ethyl ester (0.049 g, 0.125 mmol) in dichloro-
Experimental Section
methane (2.0 mL) at 0 °C was added EDC HCl (1-ethyl-3-(3-
3
dimethylaminopropyl)carbodiimide) (0.077 g, 0.401 mmol)
followed by HOBt (N-hydroxybenzotriazole monohydrate)
(0.065 g, 0.481 mmol). The reaction mixture was stirred over-
night and then quenched with water (5 mL) and extracted with
EtOAc (2ꢀ50 mL). The organic layers were collected, washed
with brine (25 mL), dried over Na2SO4, and concentrated in
vacuo. Flash column chromatography (SiO2, 40-60% EtOAc
in hexanes) yielded 0.032 g of 14up (49%) and 0.032 g of 14down
(49%) as off-white solids; Rf 0.48 (50% EtOAc in hexanes) for
14up and Rf 0.32 (50% EtOAc in hexanes) for 14down. Sub-
sequent purification and isolation was accomplished by pre-
parative thin-layer chromatography (PTLC) that employed
20% pentane in ether and afforded 14up and 14down diastereo-
mers as a white powder.
1,5,9-Triiodo-12c-(2-(methoxymethoxy)ethyl)-4,8,12-trioxa-
tricornan (9). To a suspension of 7 (0.110 g, 0.294 mmol) in 5 mL
of ether at -78 °C was added TMEDA (0.70 mL, 4.70 mmol).
Then, to the resulting mixture, t-BuLi (1.7 M, 2.76 mL) was
added dropwise to give a pale yellow turbid solution that was
slowly warmed to 0 °C and stirred for 12 h. At this juncture, 1,2-
diiodoethane (2.55 g, 9.05 mmol) in 5 mL of ether was intro-
duced via cannula into the reaction flask and the mixture was
stirred for another 12 h. The reaction mixture was quenched
with 20 mL of water and extracted with EtOAc (2ꢀ100 mL). The
organic layers were collected, washed with brine (25 mL), dried
over Na2SO4, and concentrated in vacuo. Flash column chro-
matography (SiO2, 10% EtOAc in hexanes) yielded 0.119 g of
compound 9 (55%) as a white solid. Rf 0.4 (15% EtOAc in
14up: 1H NMR (300 MHz, CDCl3) δ 8.58 (d, J=7.1 Hz, 3H),
8.13 (d, J=8.8 Hz, 3H), 7.44-7.20 (m, 48H), 4.96 (m, J=7.1 Hz,
3H), 4.30 (s, 2H), 4.34-4.24 (m, 6H), 3.48 (t, J=6.5 Hz, 2H),
3.08 (s, 3H), 2.91-2.80 (m, 6H), 2.25 (t, J=6.5 Hz, 2H), 1.31
(t, J=7.1 Hz, 9H); 13C NMR (300 MHz, CDCl3) δ 170.9, 163.0,
154.0, 150.6, 144.7, 132.8, 129.9, 128.4, 127.3, 117.5,
113.3, 112.5, 97.2, 67.1, 63.9, 62.4, 55.8, 52.3, 34.7, 28.1, 14.6;
HRMS calcd for C98H87N3O14S3 [M]Naþ 1648.5242, found
1648.5304.
1
hexanes); H NMR (300 MHz, CDCl3) δ 7.73 (d, J=8.7 Hz,
3H), 6.93 (d, J=8.7 Hz, 3H), 4.39 (s, 2H), 3.40 (t, J=6.6 Hz, 2H),
3.24 (s, 3H), 2.11 (t, J = 6.6 Hz, 2H); 13C NMR (300 MHz,
CDCl3) δ 153.2, 151.8, 138.2, 114.4, 114.0, 97.1, 63.7, 55.8, 44.5,
30.1, 29.1; HRMS calcd for C23H15I3O5 [M]Naþ 774.7945,
found 774.7960.
1,5,9-Tri[(-)-(1R)-menthyl ester]-12c-(2-(methoxymethoxy)-
ethyl)-4,8,12-trioxatricornan Diastereomers (12up and 12down).
To a solution of 6 (0.100 g, 0.109 mmol) in dry THF (1.5 mL)
was added n-butyllithium (1.6 M, 0.34 mL, 0.544 mmol) drop-
wise under argon atmosphere at -78 °C with stirring. After 3 h
(-)-(1R)-menthylchloroformate (0.62 mL, 2.89 mmol) was
added quickly at 0 °C. The solution was slowly warmed to rt
and stirred for a further 8 h. The reaction mixture was quenched
with saturated NH4Cl solution (20 mL) and extracted with
EtOAc (2ꢀ50 mL). The organic layers were collected, washed
with brine (25 mL), dried over Na2SO4, and concentrated in
vacuo. Flash column chromatography (SiO2, 5-10% EtOAc in
hexanes) yielded 0.066 g of a mixture of two diastereomers
(44%) as a white greasy solid. The two diastereomers give a
single spot (Rf 0.43; 10% EtOAc in hexanes). Further resolution
was accomplished by 5 TLC runs (4% EtOAc in hexanes)
yielding 12up (Rf 0.43) and 12down (Rf 0.38). Subsequent
purification and isolation was accomplished by Preparative
Thin Layer Chromatography (PTLC) that employed 5%
EtOAc in pentane and after 14 repeated runs afforded 12up
and 12down as a white powder.
14down: 1H NMR (300 MHz, CDCl3) δ 8.45 (d, J=7.4 Hz,
3H), 7.93 (d, J=8.8 Hz, 3H), 7.37-7.13 (m, 48H), 5.00 (m, J=
7.4 Hz, 3H), 4.33 (m, J=7.1 Hz, 8H), 3.38 (m, J=6.3 Hz, 2H),
3.17 (s, 3H), 2.94-2.74 (m, J=4.7, 12.1 Hz, 6H), 2.16 (t, J=
6.3 Hz, 2H), 1.35 (t, J = 7.1 Hz, 9H); 13C NMR (300 MHz,
CDCl3) δ 170.9, 162.9, 153.9, 150.5, 144.7, 132.7, 129.8, 128.5,
127.4, 117.5, 113.3, 112.5, 97.1, 67.2, 63.8, 62.5, 55.8, 52.0, 34.7,
28.0, 14.7; HRMS calcd for C98H87N3O14S3 [M]Naþ 1648.5242,
found 1648.5206.
Acknowledgment. We thank the Chemistry Department
at SU, the Syracuse Center for Excellence CARTI award
supported by the U.S. Environmental Protection Agency
[Grant X-83232501-0], and the NSF-CMMI (Grant No.
0727491) for partial financial support. We thank Professor
Michael B. Sponsler (SU) for fruitful discussions, Professor
S. Loh (SUNY Upstate Medical University) for the use of
the CD instrument, and Professor Laurence A. Nafie (SU)
for use of the VCD spectrometer.
12up: 1H NMR (300 MHz, CDCl3) δ 7.95 (d, J=8.7 Hz, 3H),
7.03 (d, J=8.7 Hz, 3H), 4.99 (m, 3H), 4.36 (s, 2H), 3.36 (t, J=6.8
Hz, 2H), 3.19 (s, 3H), 2.17-1.09 (m, 27H), 1.0-0.80 (m, 27H);
13C NMR (300 MHz, CDCl3) δ 164.6, 155.3, 152.1, 132.4, 115.9,
113.0, 112.4, 97.2, 77.6, 63.9, 55.7, 47.8, 44.6, 41.5, 34.7, 31.9,
27.6, 26.5, 23.7, 22.8, 21.4, 16.6; HRMS calcd for C56H72O11
[M]Naþ 943.4967, found 943.4952.
Supporting Information Available: Experimental proce-
dures and spectroscopic and analytical data for all new com-
pounds. This material is available free of charge via the Internet
at http://pubs.acs.org.
J. Org. Chem. Vol. 74, No. 18, 2009 7033