Design and Enantioselective Synthesis of a Peptidomimetic of the Turn in the Helix-Turn-Helix DNA-Binding Protein Motif

Article abstract of DOI:10.1021/jo971077h

A peptidomimetic of the turn in the helix-turn-helix (HTH) motif of DNA-binding proteins was designed and synthesized. Conformational constraint was achieved by an unusual linking of two amino acids with a side chain carbon-carbon bond. A phenyl ring provides the potential for new hydrophobic contacts with the hydrophobic core of the HTH motif. In the mimic, the peptide backbone and the central residue were retained in native form within a 12-membered cyclic tripeptide. The target compound 1b was synthesized by two sequential Horner-Wittig couplings followed by enantioselective hydrogenation with Rh(MeDuPHOS) in eight steps and 35% overall yield. The stereochemical outcome of the key hydrogenation was determined by aromatic ring oxidation with RuO₂/NaIO₄ to give 2 equiv of Boc-Asp-OMe.

Full text of DOI:10.1021/jo971077h

J . Org. Chem. 1997, 62, 8387-8393
8387
Design a n d En a n tioselective Syn th esis of a P ep tid om im etic of th e
Tu r n in th e Helix-Tu r n -Helix DNA-Bin d in g P r otein Motif
J eremy M. Travins and Felicia A. Etzkorn*
University of Virginia, Department of Chemistry, McCormick Road, Charlottesville, Virginia 22901
Received J une 16, 1997X
A peptidomimetic of the turn in the helix-turn-helix (HTH) motif of DNA-binding proteins was
designed and synthesized. Conformational constraint was achieved by an unusual linking of two
amino acids with a side chain carbon-carbon bond. A phenyl ring provides the potential for new
hydrophobic contacts with the hydrophobic core of the HTH motif. In the mimic, the peptide
backbone and the central residue were retained in native form within a 12-membered cyclic
tripeptide. The target compound 1b was synthesized by two sequential Horner-Wittig couplings
followed by enantioselective hydrogenation with Rh(MeDuPHOS) in eight steps and 35% overall
yield. The stereochemical outcome of the key hydrogenation was determined by aromatic ring
oxidation with RuO2/NaIO4 to give 2 equiv of Boc-Asp-OMe.
The helix-turn-helix (HTH) structural motif is found
in DNA-binding proteins. From prokaryotic repressors
of transcription to eukaryotic homeodomain transcription
activators, the motif is well-suited to bind DNA. Of the
two helixes, one binds DNA directly in the major groove
and is stabilized by a second backing helix at right angles
to the first (Figure 1). Between these two helixes is a
hydrophobic core that serves to stabilize the tertiary
interactions between the two helixes. The helixes are
connected via a tight turn that resembles the seven-
membered hydrogen-bonded ring of the classical γ-turn.
The turn of the HTH motif is distinct from the more
common â-turn. The HTH-turn is tighter, containing
only three amino acid residues, while â-turns have four
residues.1 The HTH turn lacks the hydrogen bond of the
γ-turn, and consequently the backbone Φ and Ψ angles
differ from the classical γ-turn angles.
To study the molecular recognition process between
proteins and DNA, we wish to synthesize peptidomimet-
ics corresponding to the small DNA-binding HTH motif.
Since short peptides are usually not conformationally
stable in the absence of the constraints from the whole
protein, we designed a small peptidomimetic turn, 1, to
stabilize the HTH motif in small peptides (Figure 1). The
mimic was designed to stabilize the tertiary structure of
the HTH motif found in DNA-binding proteins such as
the bacteriophage 434 Cro protein2,3 or eukaryotic ho-
meodomain proteins, such as Oct-1, that are key devel-
opmental transcription factors.4,5
F igu r e 1. Illustration of the HTH motif bound to DNA.
Beneath is the HTH turn mimic 1a used in molecular modeling
design work. Sequence of the Cro HTH native peptide 2a and
peptide with HTH turn mimic 2b used in modeling.
Two principles were used to design the turn
peptidomimetic: (1) conformational constraint via side
chain covalent bonds and (2) introduction of additional
conformation, but the new bonds must not interfere with
the existing sterics or electronics of the motif. Mullen
and Bartlett designed a template for the outside of the
turn region of the HTH motif.7 Our template is designed
to be tucked into the hydrophobic core of the HTH motif,
inside the turn, so that the nativelike backbone is
retained. We chose to make a hydrocarbon-bridged turn
mimic because the carbon-carbon bond would be stable
and would suit the steric constraints in the core of the
tight turn. Although the mimic is rigidified relative to
the natural turn motif, the 12-membered ring is some-
what flexible. This flexibility is desirable when the
conformation of the natural motif cannot be matched
hydrophobic contacts in the core of the HTH motif.6
A
wide variety of covalent bonds, from disulfides and
amides to carbon-carbon bonds, could constrain the
* To whom correspondence should be addressed. Phone: (804) 924-
3135. FAX (804) 924-3567. email: fae8m@virginia.edu.
X Abstract published in Advance ACS Abstracts, October 1, 1997.
(1) Richardson, J . S. Adv. Protein Chem. 1981, 34, 167-339.
(2) Ptashne, M. A Genetic Switch, 2nd ed.; Cell Press & Blackwell
Scientific Publications: Cambridge, MA, 1992.
(3) Mondrago´n, A.; Harrison, S. C. J . Mol. Biol. 1991, 219, 321-
334.
(4) Klemm, J . D.; Rould, M. A.; Aurora, R.; Herr, W.; Pabo, C. O.
Cell 1994, 77, 21-32.
(5) Ruvkun, G.; Finney, M. Cell 1991, 64, 475-478.
(6) Schneider, J . P.; Kelly, J . W. Chem. Rev. (Washington, D.C.)
1995, 95, 2169-2187.
(7) Mullen, D. G.; Bartlett, P. A. Pept. 1990, Proc. Eur. Pept. Symp.,
21st 1991, 364-365.
S0022-3263(97)01077-3 CCC: $14.00 © 1997 American Chemical Society

8388 J . Org. Chem., Vol. 62, No. 24, 1997
Travins and Etzkorn
F igu r e 3. Turn mimic used in molecular modeling showing
vectors superimposed on Cro peptide, 1a : R1 ) CH3CO, R2 )
NHMe. Synthetic target 1b: R1 ) tBuOCO, R2 ) OH.
of which there are many examples,13-17 but few that are
carbon-carbon linked.18,19 Synthesis of unnatural amino
acids was greatly advanced by the development of ste-
reoselective catalysts for the hydrogenation of didehy-
droamino acids.20-22 A C2-symmetric chiral catalyst,
rhodium 2(S),5(S)-dimethyl-1,2-bisphospholanobenzene
(MeDUPHOS), cleanly reduces didehydroamino acids to
the corresponding L-amino acid in high yield and high
enantioselectivity.23-25 Synthesis of the precursor dide-
hydroamino acids was facilitated by the development of
an R-amino acid Horner-Wittig reagent.26 The Wittig
reaction and Horner-Wadsworth-Emmons modifica-
tions were reviewed in 1989.27 We have taken advantage
of the powerful enantioselective hydrogenation of dide-
hydroamino acids to synthesize turn mimic 1b. Simul-
taneous enantioselective hydrogenation of two aminoacry-
late centers in a single molecule is a key feature of our
synthesis.28
F igu r e 2. CPK model of the HTH turn mimic minimized in
the context of the X-ray structure of Cro peptide,3 2b. The
phenyl ring and three of the helix side chains involved in
hydrophobic interactions are labeled. Ala21 is buried.
exactly or is bounded by the uncertainty of the X-ray or
NMR structure that the mimic was modeled upon.
Hydrophobic contacts between side chains within the
two helixes and the turn mimic are expected to stabilize
the hydrophobic core of this peptide fragment taken from
the Cro protein. The phenyl ring thus satisfies the
second criterion of introducing hydrophobic contacts, as
shown in Figure 2. Hydrophobic contacts between pep-
tide side chains and â-turn mimics have been used
successfully to stabilize â-sheet peptides6,8-10 and a
â-sheet/R-helix tertiary motif.11 The central amino acid
residue of the HTH turn mimic can be readily varied in
casette fashion to accommodate the various hydrophilic
side chains found in different members of the HTH
structural family. The molecule was designed in an
iterative process of molecular modeling and synthetic
considerations.
Resu lts a n d Discu ssion
Design of th e Mim ic. A model of the HTH turn
mimic 1a (Figure 3), was constructed and the global
minimum conformation was explored using Macromodel
v 3.5. First, the X-ray crystal structure of the Cro protein
bound to OR1 DNA3 was edited to remove the DNA and
most of the protein, leaving a 23-residue HTH peptide,
Cro16-38 (Figure 1). Only helical residues were in-
cluded at the C- and N-termini. The C-terminus was
modified to a carboxamide and the N-terminus was
acetylated for all modeling. The turn mimic 1a was built
by removing all but the â-carbons from the side chains
(13) Ovchinnikov, Y. A.; Ivanov, V. T. Tetrahedron 1975, 31, 2177-
2209.
The synthetic considerations in the design of mimic 1
required construction of a unique R,ω-diaminodicarbox-
ylate with â-carbons from two R-amino acid backbones
attached meta- to a single phenyl side chain. Because
of this requirement, each of the four functional groups
had to be orthogonally protected. Two stereocenters
corresponding to L-amino acid configurations had to be
incorporated to retain the designed hydrophobic interac-
tions. Finally, a medium-sized ring had to be closed
either at a carbon-carbon bond or at an amide bond. We
chose the amide-bond closure for synthetic simplicity.
Stereoselective amino acid synthesis has been re-
viewed.12 The turn mimic is an example of the less
common side-chain bridged R,ω-diaminodicarboxylates,
(14) Manesis, N. J .; Goodman, M. J . Org. Chem. 1987, 52, 5331-
5341.
(15) O¨ sapay, G.; Taylor, J . W. J . Am. Chem. Soc. 1990, 112, 6046-
6051.
(16) Sugg, E. E.; Castrucci, A. M. d.; Hadley, M. E.; van Binst, G.;
Hruby, V. Biochemistry 1988, 27, 8181-8188.
(17) Creighton, T. E. Methods Enzymol. 1984, 107, 305-329.
(18) Miller, S. J .; Grubbs, R. H. J . Am. Chem. Soc. 1995, 117, 5855-
5856.
(19) Stachel, S. J .; Habeeb, R. L.; Van Vranken, D. L. J . Am. Chem.
Soc. 1996, 118, 1225-1226.
(20) Marko´, L.; Ungva´ri, F. J . Organomet. Chem. 1992, 432, 1-214.
(21) Kagan, H. B.; Sasaki, M. Optically Active Phosphines: Prepara-
tion, Uses, and Chiroptical Properties; J . Wiley & Sons: New York,
1990; Vol. 1.
(22) Williams, R. M. Synthesis of Optically Active R-Amino Acids;
Pergamon Press: Oxford, 1989; Vol. 7.
(23) Burk, M. J .; Feaster, J . E.; Nugent, W. A.; Harlow, R. L. J .
Am. Chem. Soc. 1993, 115, 10125-10138.
(24) Burk, M. J .; Lee, J . R.; Martinez, J . P. J . Am. Chem. Soc. 1994,
116, 10847-10848.
(8) Diaz, H.; Tsang, K. Y.; Choo, D.; Kelly, J . W. Tetrahedron 1993,
49, 3533-45.
(25) Burk, M. J .; Gross, M. F.; Martinez, J . P. J . Am. Chem. Soc.
1995, 117, 9375-9376.
(9) Tsang, K. Y.; Diaz, H.; Graciani, N.; Kelly, J . W. J . Am. Chem.
Soc. 1994, 116, 3988-4005.
(26) Schmidt, U.; Lieberknect, A.; Wild, J . Synth. Commun. 1984,
53-60.
(10) Schneider, J . P.; Kelly, J . W. J . Am. Chem. Soc. 1995, 117,
2533-2546.
(27) Maryanoff, B. E.; Reitz, A. B. Chem. Rev. (Washington, D.C.)
1989, 89, 863-927.
(11) Struthers, M. D.; Cheng, R. P.; Imperiali, B. Science 1996, 271,
342-348.
(28) Schmidt, U.; Meyer, R.; Leitenberger, V.; Griesser, H.; Lie-
berknecht, A. Synthesis 1992, 1025-1030.
(12) Duthaler, R. O. Tetrahedron 1994, 50, 1539-1650.

A Peptidomimetic of the Turn in the HTH Motif
J . Org. Chem., Vol. 62, No. 24, 1997 8389
Sch em e 1. Syn th esis of Or th ogon a lly P r otected
Am in o Acid Syn th on
the range of conformational flexibility designed into the
turn mimic (Figure 4).
In the lowest energy conformer of the turn mimic, no
hydrogen bond was formed between the CdO and NH
that normally would form in a γ-turn. Since this
hydrogen bond is also absent in the native turn, this
implies that the CdO and NH of the mimic will be
available for hydrogen bonding to the helixes as in the
native protein. In the X-ray structure, the extra-turn
hydrogen bond to the R2 helix is implied by a distance of
2.01 Å between the helix Ala21 CdO and the turn Val26
NH. The analogous hydrogen bond is also found in the
HTH peptidomimic 2b. In the second lowest energy
conformer of 1a the γ-turn hydrogen bond is present. The
existence of this structure in a nearby minimum indicates
that the HTH turn mimic may make a satisfactory γ-turn
mimic as well.
F igu r e 4. Stereodrawing of the superposition of the ΦAla24
and ΨVal26 backbone vectors of the turn residues of 1a , shown
with arrows, with the corresponding vectors of the two low-
energy conformers in 2b, rmsd 0.38 Å with the lowest energy
conformer (top) and 0.16 Å with the second lowest (bottom).
Five hydrophobic contacts between the turn mimic and
helix side chains were found in the minimized peptido-
mimetic 2b (Figure 2), specifically, (1) Ph1-proS-âH to
Leu20-proS-δCH3, 2.1 Å; (2) Ph2H to Ala21-RH, 1.4 Å;
(3) Ph2H to Ile31-δCH3, 2.2 Å; (4) Ph3-proR-âH to Ile31-
δCH3, 2.3 Å; and (5) Ph3-proR-âH to Ser30-âH, 1.8 Å.
Two hydrophobic contacts were found in the turn of the
native Cro HTH peptide 2a : (1) Val26-γH to Ile31-δCH3,
2.3 Å, and (2) Ala24-âCH3 to Leu20-proS-δCH3, 3.1 Å.
The design of the turn mimic thus incorporates potential
hydrophobic stabilization of the HTH motif using the aryl
side chain conformational constraint.
of Ala21 and Val23 and connecting a phenyl group to the
two â-carbons. To search for the global minimum, 1000
starting structures of 1a were generated by the Macro-
model Monte Carlo dihedral angle conformational search
option and minimized using the Amber force field with
water solvation. Of the 129 unique conformations found,
65 minimized with good convergence. The lowest energy
conformation, found 15 times with good convergence, was
6 kJ /mol lower than the next conformer, found 17 times.
The two lowest energy conformers were incorporated
into the backbone of the HTH Cro peptide 2a and
minimized using MacroModel Amber and water solva-
tion. Superposition of all R-carbons of the resulting HTH
peptidomimetics 2b with Cro HTH peptide 2a (lowest
energy root mean squared deviation (rmsd) of 0.80 Å;
second lowest, 0.71 Å) suggests minimal alteration of the
three-dimensional structure of the HTH motif (data not
shown). The ΦAla24 and ΨVal26 backbone vectors29-31
set the direction for the two helixes with respect to the
ring. Figure 4 shows the superposition of these vectors
in the turn of 1a with the corresponding vectors of the
two low-energy conformers in 2b, rmsd 0.38 Å with the
lowest energy conformer, and 0.16 Å with the second
lowest. The two conformers have different orientations
of the phenyl ring with respect to the backbone, reflecting
Syn th esis
r,ω-Dia m in od ia cr yla te. Despite the apparent sim-
plicity, the synthesis of the aryl turn mimic requires
orthogonal protection on four functional groups, two
amines and two carboxylic acids. The amino acid syn-
thon, 3, was first reported in a series of protected amino
phosphoryl acetate esters.26 Trimethylsilylethyl 2-Boc-
2-(dimethoxyphosphoryl)acetate (5) fits our criteria of
ease of synthesis and orthogonal protection. The use of
5, without its synthesis, has been reported.32 We now
report the synthesis and characterization of this key
starting material (Scheme 1).
Preparation of the aminoacrylate 6 was achieved in
94% yield by reaction of a 10-fold excess of the com-
mercially available benzene-1,3-dialdehyde with phos-
phonate 3 (Scheme 2). Reaction of phosphonate 5 with
monoaldehyde 6 produced orthogonally protected 7 in
(29) Bartlett, P. A.; Shea, G. T.; Telfer, S. J .; Waterman, S. In
Molecular Recognition: Chemical and Biological Problems; Roberts,
S. M., Ed.; Royal Society of Chemistry: London, 1989; pp 182-96.
(30) Etzkorn, F. A.; Guo, T.; Lipton, M.; Goldberg, S.; Bartlett, P.
A. J . Am. Chem. Soc. 1994, 116, 10412-10425.
(31) Bartlett, P. A.; Etzkorn, F. A.; Guo, T.; Lauri, G.; Liu, K.; Lipton,
M.; Morgan, B. P.; Shea, G. T.; Shrader, W. D.; Waterman, S. Proc.
Robert A. Welch Conf. Chem. Res. 1991, XXXV, 45-68.
(32) Schmidt, U.; Griesser, H.; Leitenberger, V.; Lieberknecht, A.;
Mangold, R.; Meyer, R.; Riedl, B. Synthesis 1992, 487-490.

8390 J . Org. Chem., Vol. 62, No. 24, 1997
Travins and Etzkorn
Sch em e 2. Syn th esis of Bis-Alk en e
Sch em e 4. P r oof of Ster eoch em istr y by Ch em ica l
Degr a d a tion to Boc-Asp -OMe
Sch em e 3. Asym m etr ic Hyd r ogen a tion of 7
(Scheme 4). Preparation of a symmetrical derivative
followed by oxidative cleavage of the aromatic ring
produced Boc-Asp-OMe of known configuration. Boc-
protected amines are stable to RuO4 oxidative condi-
tions.36 We avoided the presence of Cbz because it is not
stable to the reaction conditions.37 Thus we chose to
produce the Boc and methyl ester protected symmetric
derivative 11 for oxidative cleavage.
83% yield.33 Under conditions with tetramethylguani-
dine as base, we observed predominantly one diastere-
omer in the H NMR (see the Supporting Information),
1
presumably Z based on the literature precedence.28 The
one-step synthesis of a meta-substituted phenyl R,ω-
diaminodiacrylate has been reported, though it is not
orthogonally protected.26 Hydrogenation to the R,ω-
diamino dicarboxylate was not reported.
Removal of the trimethylsilylethyl (TMSE) ester with
fluoride gave acid 9, also the next step in the synthesis
(Scheme 4). Under rigorously dry conditions in the
desilylation, we observed epimerization. However, de-
liberate addition of water to the reaction resulted in clean
removal of the TMSE ester without epimerization (see
the Supporting Information). Esterification with (tri-
methylsilyl)diazomethane gave the bis-methyl ester 10.
The Cbz amino protecting group was switched to Boc by
hydrogenolysis in the presence of di-tert-butyl dicarbon-
ate (Boc2O) to afford 11. The optical rotation of this
intermediate, [R]D ) +46.3° ( 0.5°, demonstrated that
the major product was one of the C2 symmetric stereo-
isomers (R,R- or S,S-) and not the meso isomer. Oxida-
tion of 11 with RuO2 and an excess of NaIO4 for 12 h
produced Boc-Asp-OMe with [R]D ) -17.9° ( 0.5° (lit.
-17.8°38 and -19°39 ). We have thus synthesized the
unusual R,ω-diamino dicarboxylate 8 with the desired
S,S-stereochemistry.
En an tioselective Hydr ogen ation . Rh(MeDUPHOS)34
gives excellent enantioselectivity and high yields for both
Z and E olefins regardless of the amino acid protecting
groups.23 The double enantioselective hydrogenation of
7 produced R,ω-diamino dicarboxylate 8 in excellent yield
(Scheme 3). Hydrogenation using Wilkinson’s catalyst
gave a racemic mixture of diastereomers 8. The aryl
singlets were well-resolved in the 1H NMR spectra of
partially purified diastereomeric mixtures of 8. Integra-
tion35 of the aryl singlet of the hydrogenated product gave
a 98:2 diastereomeric ratio (see the Supporting Informa-
tion). Since the catalyst is known to produce L-amino
acid stereochemistry,23 this indicates that small amounts
of S,R- or R,S-isomer were formed in the hydrogenation,
with the major product being the S,S-diastereomer,
mimicking the natural L,L-stereochemistry of the turn.
Proof of the S,S-stereochemistry was obtained by
chemical degradation of the reduced intermediate 8
Cycliza tion . Coupling of the tosylate salt of benzyl
glycinate with acid 9 proceeded smoothly to give the
protected acyclic precursor 12 (Scheme 5). Simultaneous
hydrogenolysis of the Cbz and Bn protecting groups
followed by cyclization with diphenylphosphoryl azide
(DPPA) in dilute DMF solution afforded cyclic tripeptide
methyl ester 13 in high yield.40 In anticipation of future
incorporation of amino acids besides Gly into the mimic,
DPPA was chosen for cyclization because of the low
(33) Efficient synthesis of the cyclization precursor 12 (Scheme 5)
could conceivably have been achieved by reaction of a dipeptide
phosphonate with the monoaldehyde 6, followed by hydrogenation. The
Horner-Wittig reaction of dipeptide synthons having the phosphonate
in the C-terminal position such as methyl Boc-(S)-Ile-2-(dimethoxy-
phosphoryl)acetate with aryl aldehydes has been reported.25 However,
in our sterically congested target, the reaction of benzyl 2-(Cbz)amino-
2-(dimethoxyphosphoryl)acetylglycinate with 6 gave a very poor yield
in the second Horner-Wittig reaction. Inverting the order by reacting
the dipeptide phosphonate in the first step and 3 in the second step
did not improve the yield. We concluded that stepwise synthesis of
the tripeptide backbone was necessary and suitable for cassette
synthesis with amino acids other than Gly.
(34) The activity of the catalyst from Strem varied from batch to
batch and the turnover ratio did not approach that reported in the
literature. In our experience, rigorously oxygen-free conditions are
required to attain total conversion to product. In this initial work, Rh-
(MeDUPHOS) was used, but Rh(EtDUPHOS) gives slightly better
enantioselectivity.
(36) Tanaka, K.-i.; Yoshifuji, S.; Nitta, Y. Chem. Pharm. Bull. 1988,
36, 3125-3129.
(37) Schuda, P. F.; Cichowicz, M. B.; Heimann, M. R. Tetrahedron
Lett. 1983, 24, 3829-3830.
(38) Tohdo, K.; Hamada, Y.; Shiori, T. Synlett 1994, 105-106.
(39) Cantacuzene, D.; Guerreiro, C. Tetrahedron 1989, 45, 741-748.
(40) NMR samples at normal concentrations in CDCl3 were found
to form gels upon standing for a few hours. The target ester 13 was
soluble only in DMF, DMSO, or mixtures of THF/MeOH, CHCl3/DMSO.
Boc-protected intermediates 8 and 13 were found to decompose upon
standing in CDCl3 due to the presence of DCl.
(35) The recycle delay was increased to 10 s and no line-broadening
was used in processing to minimize integration errors due to the long
relaxation times for aromatic protons. The spectra were acquired at
500 MHz for enhanced resolution.

A Peptidomimetic of the Turn in the HTH Motif
J . Org. Chem., Vol. 62, No. 24, 1997 8391
and brine (10 mL) and dried over MgSO4. Concentration
followed by recrystallization (EtOAc/pet. ether) yielded 616 mg
(74%) of a white solid. mp 155-156 °C. 1H NMR (CDCl3): δ
9.21 (br s, 2H), 5.57 (d, 1H, J ) 9 Hz), 4.94 (dd, 1H, J ) 9.0,
22.6 Hz), 3.88 (d, 3H, J ) 11.1 Hz), 3.84 (d, 3H, 11.6 Hz), 1.45
(s, 9H). 13C NMR (CDCl3): δ 168.05, 155.45, 81.41, 55.17 (d,
J COP ) 7.6 Hz), 55.07 (d, J COP ) 7.6 Hz), 52.12 (d, J CP ) 150.4
Hz) 26.69. The acid (2.25 g, 7.94 mmol) and (trimethylsilyl)-
ethanol (1.42 mL, 9.93 mmol) were dissolved in CH2Cl2 (30
mL) and cooled to 0 °C. DCC (1.80 g, 8.74 mmol) was added
and the reaction was stirred for 8 h. The urea byproduct was
removed by filtration and the solution was concentrated. The
residue was taken up in EtOAc (60 mL), washed with citric
acid (2 × 10 mL), NaHCO3 (2 × 10 mL), water (10 mL), and
brine (20 mL), and dried over MgSO4. Concentration followed
by chromatography (6:4 pet. ether/EtOAc) gave 3.01 g (98%)
of the TMSE ester 5 as a colorless oil that slowly solidified at
-20 °C. 1H NMR (CDCl3): δ 5.31 (br d, 1H, J ) 8.4 Hz), 4.81
(dd, 1H, J ) 9.4, 22.1 Hz), 4.28 (m, 2H), 3.82 (d, 3H, J ) 3.1
Hz), 3.78 (d, 3H, J ) 3.1 Hz), 1.43 (s, 9H), 1.05 (m, 2H), 0.03
(s, 9H). 13C NMR (CDCl3): δ 166.98, 148.44, 89.49, 64.94,
53.88 (d, J COP ) 5.5 Hz), 51.96 (d, J CP ) 147.1 Hz), 28.18, 17.37,
-1.63.
Meth yl (Z)-2-[[(Ben zyloxy)car bon yl]am in o]-3-(3-for m yl-
p h en yl)-2-p r op en oa te (6). Phosphorylglycinate 3 (4.00 g,
12.1 mmol) was dissolved in 150 mL of CH2Cl2, and tetra-
methylguanidine (1.67 mL, 13.3 mmol) was added. The
reaction was stirred for 30 min, then isophthalaldehyde (16.20
g, 120.8 mmol) was added. The reaction was judged to be
complete after 10 min by monitoring the disappearance of 3
by TLC (Rf ) 0.3, 1:1 pet. ether/EtOAc, UV, veratraldehyde
stain). The solution was diluted with 50 mL of CH2Cl2, washed
with 10% citric acid (2 × 30 mL), NaHCO3 (30 mL), water (30
mL), and brine (80 mL), and then dried over MgSO4. Con-
centration followed by chromatography (90:10 pet. ether/EtOAc
eluted excess isophthalaldehyde; 85:15 pet. ether/EtOAc eluted
product) gave 3.87 g (94%) of 6 as a thick colorless oil which
solidified after several days at 4 °C. mp ) 77-83 °C. 1H NMR
(CDCl3): δ 9.89 (s, 1H), 7.94 (s, 1H), 7.80 (d, 1H, J ) 7.7 Hz),
7.72 (d, 1H, J ) 7.7 Hz), 7.47 (t, 1H, J ) 7.7 Hz), 7.37 (s, 1H),
7.32 (br s, 5H), 6.67 (s, 1H), 5.08 (s, 2H), 3.84 (s, 3H). 13C
NMR (CDCl3): δ 191.71, 165.34, 153.28, 136.47, 135.69,
135.01, 134.92, 134.56, 130.88, 129.73, 129.13, 128.50, 128.34,
128.27, 125.12, 67.60, 52.85. Anal. Calcd for C19H17NO5: C,
67.25 H, 5.05; N, 4.13. Found: C, 67.05; H, 5.18; N, 4.04.
Meth yl (Z)-2-[[(Ben zyloxy)ca r bon yl]a m in o]-3-[3-[(Z)-
2-[(t er t -b u t oxyca r b on yl)a m in o]-3-[2-(t r im e t h ylsilyl)-
eth oxy]-3-oxo-1-pr open yl]ph en yl]-2-pr open oate (7). Phos-
phorylglycinate 5 (2.80 g, 7.30 mmol) was dissolved in 10 mL
of CH2Cl2, tetramethylguanidine (1.25 mL, 9.96 mmol) was
added, and the mixture was stirred for 30 min at rt. Aldehyde
6 (2.40 g, 7.08 mmol), dissolved in 10 mL of CH2Cl2, was then
added and the reaction was stirred for 24 h. The solution was
diluted to 50 mL with CH2Cl2, washed with 10% citric acid (2
× 10 mL), NaHCO3 (10 mL), water (10 mL), and brine (10 mL),
and then dried over MgSO4. Concentration followed by
chromatography (9:1 pet. ether/EtOAc) gave 3.50 g (83%) of 7
as a colorless oil. 1H NMR (CDCl3): δ 7.63 (s, 1H), 7.47 (d, 2
H, J ) 6 Hz), 7.33-7.26 (m, 7H), 7.18 (s, 1H), 6.52 (br s, 1H),
6.27 (br s, 1H), 5.1 (s, 2H), 4.34 (m, 2H), 3.79 (s, 3H), 1.37 (s,
9H), 1.1 (m, 2H), 0.08 (s, 9 H). 13C NMR (CDCl3): δ 165.55,
165.42, 153.78, 152.51, 135.91, 134.81, 133.87, 131.02, 130.91,
130.29, 129.73, 128.64, 128.45, 128.18, 128.14, 126.87, 125.51,
124.84, 80.97, 67.48, 64.10, 52.58, 28.03, 17.42, -1.52. Anal.
Calcd for C31H40N2O8Si: C, 62.39; H, 6.76; N, 4.69. Found:
C, 62.14; H, 6.96; N, 4.60.
Sch em e 5. Tr ip ep tid e F or m a tion , Cycliza tion , a n d
Ester Hyd r olysis of Tu r n Mim ic
propensity of acyl azides to racemize.41 This cyclization
has also been performed with Ala in the central position
in 80% yield (J .M.T and F.A.E, unpublished results).
The methyl ester was hydrolyzed with 1 equiv of 1 N
NaOH in THF/MeOH to give 74% yield of the free acid
product 1b. This brief, high-yielding synthesis has
allowed the preparation of the quantities of 1b necessary
for peptide synthesis (825 mg). Peptidomimetic 2b has
been synthesized on solid phase using compound 1b
(J .M.T, Hart, S., and F.A.E., unpublished results).
Con clu sion
We have synthesized a mimic of either the HTH-turn
or γ-turn that is designed to stabilize the HTH motif by
both conformational constraint and hydrophobic interac-
tions. The mimic was synthesized in eight steps from
phosphonates 3 and 5, using two sequential Horner-
Wittig reactions followed by simultaneous enantioselec-
tive hydrogenation of both aminoacrylates to give the
nativelike stereochemistry. Macrocyclization of the trip-
eptide proceeded in 80% yield. Turn mimic 1b was
synthesized in eight steps and 35% overall yield.
Exp er im en ta l Section
All chemicals were reagent grade and commercially avail-
able. Rh(MeDUPHOS) was obtained from Strem Chemicals
Inc.34 Unless otherwise indicated, all reactions were carried
out under a N2 atmosphere, in flame-dried flasks. THF and
Et2O were distilled from dark blue solutions with K and
benzophenone. Benzene, toluene, CH2Cl2, CH3CN, Et3N, and
diisopropylethylamine (DIEA) were distilled from CaH2 under
a N2 atmosphere. DMF and MeOH were dried with 3 Å
molecular sieves. Chromatography was on 32-63 µm silica
gel with reagent grade solvents. TLC was performed using
aluminum-backed silica gel plates. Melting points are uncor-
rected. Proton (300 MHz) and carbon-13 (75 MHz) NMR
spectral data were recorded on a General Electric spectrometer
using deuterated solvents at approximately 22 °C.
2-(Tr im eth ylsilyl)eth yl 2-[(ter t-Bu toxycar bon yl)am in o]-
2-(d im eth oxyp h osp h or yl)a ceta te (5). Methyl phospho-
rylglycinate 4 (876 mg, 2.95 mmol) was dissolved in dioxane
(3 mL) and cooled to 0 °C. NaOH (1 N, 2.95 mL) was added
dropwise over 15 min and the reaction was stirred until the
starting material disappeared (∼1 h) (TLC: Rf ) 0.5 1:1 pet.
ether/EtOAc, ninhydrin stain). The solution was acidified to
∼pH 1 with 20% HCl and extracted with EtOAc (3 × 10 mL).
The combined EtOAc extracts were washed with water (5 mL)
Met h yl (S)-2-[[(Ben zyloxy)ca r b on yl]a m in o]-3-[3-[(S)-
2-[(t er t -b u t oxyca r b on yl)a m in o]-3-[2-(t r im e t h ylsilyl)-
eth oxy]-3-oxop r op yl]p h en yl]p r op a n oa te (8). The diacry-
late 7 (1.50 g, 2.51 mmol) was dissolved in 8 mL of MeOH
and degassed for 30 min with N2. Rh(MeDUPHOS) catalyst
was added (16 mg) and the reaction was subjected to 45-50
psi of H2 on
a Parr shaker for 12 h. The MeOH was
evaporated. The residue was dissolved in CHCl3 and passed
through a layer of silica (7:3 pet. ether/EtOAc) to give 1.45 g
(96%) of a colorless oil. Diastereomeric ratios were determined
(41) Shiori, T.; Ninomiya, K.; Yamada, S. J . Am. Chem. Soc. 1972,
94, 6203-5.
(42) Zoller, U.; Ben-Ishai, D. Tetrahedron 1975, 31, 863-866.

8392 J . Org. Chem., Vol. 62, No. 24, 1997
Travins and Etzkorn
1
to be 98:2 by H NMR analysis (see the Supporting Informa-
CH3CN (8 mL), CCl4 (8 mL), and water (14 mL). RuO2 (12
mg, 0.090 mmol) was added and the mixture was stirred
vigorously. NaIO4 (2.00 g, 9.35 mmol) was added in three
portions over the course of 4 h, each time causing a gas
emission and solution color change from black/gray to yellow.
Additional NaIO4 was added in three ∼300 mg portions until
all starting material had disappeared by TLC (Rf ) 0.4, 7:3
pet. ether/EtOAc, ninhydrin stain). The mixture was diluted
with 10 mL of CH3CN and 10 mL of CHCl3 and filtered. iPrOH
was added to reduce the RuO4, during which time the solution
turned from yellow to black, and the mixture was filtered and
concentrated. The residue was taken up in 10 mL of saturated
NaHCO3 and extracted with CH2Cl2 (2 × 4 mL). The water
layer was acidified to ∼pH 1 with 2 N HCl and extracted with
CH2Cl2 (5 × 4 mL). The organic extracts were washed with
water (5 mL) and brine (10 mL) and dried (MgSO4). Concen-
tration of the solution left 48 mg (36%) of a colorless oil that
slowly solidified at 4 °C. mp 81-87 °C, lit. mp 84-88 °C,38 89
tion). 1H NMR (CDCl3): δ 7.33 (m, 5H), 7.19 (t, 1H, J ) 7.5
Hz), 7.00 (d, 1H, J ) 11.4 Hz), 6.98 (d, 1H, J ) 11.4 Hz), 6.88
(s, 1H), 5.27 (br d, 1H, J ) 7.9 Hz), 5.10 (s, 2H), 4.98 (br d,
1H, J ) 7.91 Hz), 4.63 (m, 1H), 4.50 (m, 1H), 4.17 (m, 2H),
3.71 (s, 3H), 3.07-2.96 (m, 4H), 1.41 (s, 9H), 0.95 (m, 2H), 0.03
(s, 9H). 13C NMR (CDCl3): δ 172.36 (two peaks), 156.12,
155.58, 137.06, 136.80, 136.41, 130.99, 129.23, 129.01, 128.66,
128.61, 128.38, 127.45, 80.33, 67.46, 64.23, 55.32, 55.03, 52.82,
38.67 (two peaks), 28.79, 17.89, -1.03. Anal. Calcd for
C31H44N2O8Si: C, 61.98; H, 7.38; N, 4.66. Found: C, 61.88;
H, 7.38; N, 4.69.
3-[3-[(S)-2-[[(Ben zyloxy)ca r bon yl]a m in o]-3-m eth oxy-3-
oxop r op yl]p h en yl]-(S)-2-[(ter t-b u t oxyca r bon yl)a m in o]-
p r op a n oic Acid (9). TMSE ester 8 (1.00 g, 1.66 mmol) was
dissolved in 15 mL of DMF and cooled to 0 °C. Water (8 drops)
was added to the solution. Bu4N+F- (1.10 g, 3.49 mmol) was
added and the reaction was stirred for 12 h, during which time
gas was evolved. The reaction was concentrated and the
residue was dissolved in 40 mL of EtOAc. HCl (0.5 N, 20 mL)
was added to the organic layer and the layers were separated.
The aqueous layer was extracted with EtOAc (3 × 10 mL),
and the combined organic extracts were dried over MgSO4.
Concentration gave 802 mg (97%) of crude product which was
used without purification for all subsequent transformations.
1H NMR (CDCl3) δ 8.7-9.1 (br s, 1H), 6.8-7.4 (m, 9H), 6.06
(d, 0.5 H, J ) 8.4 Hz), 5.46 (d, 0.5 H, J ) 7.9 Hz), 5.0-5.3 (m,
3H), 4.5-4.7 (m, 2H), 3.70 (m, 3H), 2.8-3.2 (m, 4H), 1.4 (m,
9H). 13C NMR (CDCl3) δ 174.74, 174.33, 172.51, 170.72,
157.40, 156.41, 155.76, 155.45, 136.91, 136.51, 136.04, 135.37,
131.65, 130.73, 129.21, 129.07, 128.84, 128.75, 128.61, 128.55,
128.06, 80.53, 80.29, 68.60, 68.43, 67.75, 55.62, 55.23, 54.77,
52.95, 38.88, 38.58, 38.49, 28.88, 26.09. Compound 9 appears
to exist in two conformations in chloroform as indicated by
the doubling of some resonances in the 1H and 13C spectra.
(See the Supporting Information.)
°C.39 [R]20 ) -17.9° ( 0.5 (c ) 1.45 MeOH); lit. [R]22
)
D
D
-17.8° (c ) 1 MeOH),38 [R]23 ) -19.0° (c ) 1 MeOH).39 1H
D
NMR (CDCl3) δ 8.81 (br s, 1H), 5.54 (d, 1H, J ) 8 Hz), 4.58
(m, 1H), 3.75 (s, 3H), 3.04 (dd, 1H, 3.8, J ) 17.2 Hz), 2.85 (dd,
1H, 4.3, J ) 17.2 Hz), 1.44 (s, 9H). 13C NMR: (CDCl3) δ
175.85, 171.50, 155.50, 80.38, 52.75, 49.71, 36.62, 28.23.
Meth yl (S)-2-[[(Ben zyloxy)ca r bon yl]a m in o]-3-[3-[3-[[2-
(ben zyloxy)-2-oxoeth yl]a m in o]-(S)-2-[(ter t-bu toxyca r bo-
n yl)a m in o]-3-oxop r op yl]p h en yl]p r op a n oa te (12). Crude
acid 9 (515 mg, 1.03 mmol) was dissolved in DMF (12 mL)
along with HOBT (192 mg, 1.42 mmol), triethylamine (190 µL,
1.37 mmol), and glycine benzyl ester hydrotosylate (462 mg,
1.25 mmol). The reaction was cooled to 0 °C and DCC (249
mg, 1.20 mmol) was added. The reaction was allowed to warm
to rt and was stirred for 24 h. The solution was diluted with
50 mL of EtOAc and cooled to 0 °C to precipitate salts. The
solution was filtered through a bed of Celite and then extracted
with 10% citric acid (2 × 10 mL), saturated NaHCO3 (2 × 10
mL), water (5 × 5 mL), and brine (25 mL). Concentration of
the solution followed by column chromatography (65:35 pet.
ether/EtOAc) gave the 567 mg (85%) of the coupled product
12 as a colorless glass. 1H NMR (CDCl3): δ 7.26-7.32 (m,
10H), 7.16 (t, 1H, J ) 7.5 Hz), 7.06 (d, 1H, J ) 7.9 Hz), 7.04
(s, 1H), 6.95 (d, 1H, J ) 7.5 Hz), 6.46 (br s, 1H), 5.54 (d, 1H,
J ) 8.4 Hz), 5.21 (br s, 1H), 5.12 (s, 2H), 5.08 (br m, 1H), 5.01
(s, 2H), 4.63 (m, 1H), 3.90 (br d, 2H, J ) 4.9 Hz), 3.73 (s, 3H),
2.93-3.17 (m, 4H), 1.40 (s, 9H). 13C NMR (CDCl3): δ 172.39,
172.02, 169.95, 156.13, 155.76, 137.46, 136.81, 136.65, 135.64,
130.76, 129.16, 129.03, 128.92, 128.81, 128.61, 128.49, 128.32,
80.52, 67.52, 67.28, 56.16, 55.48, 52.81, 41.67, 39.00, 38.68,
28.70. Anal. Calcd for C35H41N3O9: C, 64.90; H, 6.38; N, 6.49.
Found: C, 65.01; H, 6.39; N, 6.35.
Met h yl (S)-2-[[(Ben zyloxy)ca r bon yl]a m in o]-3-[3-[(S)-
2-[(ter t-bu toxyca r bon yl)a m in o]-3-m eth oxy-3-oxop r op yl]-
p h en yl]p r op a n oa te (10). Crude acid 9 (119 mg, 0.238 mmol)
was dissolved in 4.0 mL of CHCl3:MeOH (3:1) and stirred.
(Trimethylsilyl)diazomethane (2 M in hexanes) was added
dropwise until evolution of N2 ceased and the solution re-
mained yellow. Concentration followed by chromatography
(3:1 pet. ether/EtOAc) yielded 101 mg (84%) of product as a
colorless glass. [R]25 ) +57.6° ( 0.8° (c ) 1.25 CHCl3). 1H
D
NMR (CDCl3) δ 7.33 (s, 5H), 7.19 (t, 1H, J ) 7.5 Hz), 6.98 (m,
2H), 6.86 (s, 1H), 5.29 (d, 1H, J ) 7.5 Hz), 5.09 (s, 2H), 5.00
(d, 1H, J ) 7.5 Hz), 4.64 (m, 1H), 4.54 (m, 1H), 3.71 (s, 3H),
3.67 (s, 3H), 3.04 (m, 4H), 1.41 (s, 9H); 13C NMR (CDCl3) δ
172.74, 172.38, 156.15, 155.57, 136.88, 136.77, 136.51, 130.87,
129.29, 129.01, 128.64, 128.45, 80.46, 67.48, 55.32, 54.90,
52.83, 52.70, 38.67, 28.78. Anal. Calcd for C27H34N2O8: C,
63.02; H, 6.66; N, 5.44. Found: C, 60.42; H, 6.49; N, 5.13.
Meth yl (S)-2-[(ter t-Bu toxyca r bon yl)a m in o]-3-[3-[(S)-2-
[(ter t-bu t oxyca r bon yl)a m in o]-3-m et h oxy-3-oxop r op yl]-
p h en yl]p r op a n oa te (11). The ester 10 (94 mg, 0.18 mmol)
was dissolved in MeOH (5 mL). Boc2O (81 mg, 0.37 mmol)
and 5% Pd/C (20 mg) were added, followed by H2 addition via
balloon while the reaction stirred for 10 h. In reactions where
the reprotection of the amine was slow, another equivalent of
Boc2O was added along with a catalytic amount of DMAP (5%).
When judged to be complete by TLC (disappearance of free
amine, Rf ) 0.1, 7:3 pet. ether/EtOAc, ninhydrin stain) the
reaction was filtered, the solvent was evaporated, and the
residue was chromatographed (excess Boc2O eluted with 9:1
pet. ether/EtOAc; product eluted with 7:3 pet. ether/EtOAc)
Meth yl (S)-9-[(ter t-Bu toxyca r bon yl)a m in o]-5,8-d ioxo-
(S)-4,7-d ia za bicyclo[9.3.1]p en ta d eca -1(15),11,13-tr ien e-3-
ca r boxyla te (13). The fully protected dipeptide 12 (1.61 g,
2.49 mmol) was dissolved in MeOH (40 mL) and 5% Pd/C (300
mg) was added. H2 was added via balloon and the reaction
was stirred vigorously for 12 h. TLC analysis showed the
disappearance of starting material (Rf ) 0.5, 50:50 pet. ether/
EtOAc, ninhydrin stain). The catalyst was removed by filtra-
tion and the solution was concentrated. Redissolving the
semisolid in MeOH followed by addition of EtOAc and evapo-
ration gave 1.00 g (95%) of a hygroscopic granular white solid.
Proton NMR showed the absence of benzylic groups. 1H NMR
(CDCl3) (peaks broad due to zwitterion): δ 7.9 (br s, 3H), 6.8-
7.2 (m, 5H), 5.4 (br s, 1H), 4.4 (br s, 1H), 4.0 (br s, 1H), 3.62
(s, 3H), 3.58 (br s, 2H), 2.7-3.2 (m, 4H), 1.23 (s, 9H). 13C NMR
(CDCl3): δ 174.21, 172.32, 155.94, 138.90, 137.26, 131.05,
128.77, 128.45, 127.90, 78.89, 56.33, 55.69, 52.52, 42.51, 25.38,
28.95. Anal. Calcd for C20H29N3O7 H2O: C, 54.41; H, 7.08; N,
9.52. Found: C, 54.52; H, 7.15; N, 8.94. The amino acid (1.00
g, 2.36 mmol) was dissolved in DMF (1.25 L) and cooled to 0
°C. DPPA (709 µL, 2.83 mmol) was added, followed by DIEA
(1.03 mL, 5.90 mmol). The reaction was stirred under N2 for
7 d at 4 °C. The solvent was removed in vacuo, dissolved in
minimal CHCl3, and chromatographed (120:2:1 CHCl3/MeOH/
AcOH), to give 800 mg (80%) of a white solid. mp 245 °C (dec).
1H NMR (CDCl3/DMSO-d6): δ 7.95 (t, 1H, J ) 6.4 Hz), 7.12
to give 88 mg (100%) of a colorless glass. [R]25 ) +46.3° (
D
1
0.5° (c ) 2.2 CHCl3). H NMR (CDCl3): δ 7.19 (t, 1H, J ) 7.3
Hz), 6.98 (d, 2H, J ) 7.9 Hz), 6.86 (s, 1H), 5.01 (d, 2H, J ) 7.3
Hz), 4.53 (m, 2H), 3.68 (s, 6H), 3.02 (m, 4H), 1.39 (s, 18H). 13
C
NMR: (CDCl3) δ 172.15, 154.97, 136.19, 130.37, 128.66,
127.83, 79.85, 54.36, 52.10, 38.11, 28.19. Anal. Calcd for
C24H36N2O8: C, 59.98; H, 7.55; N, 5.83. Found: C, 59.83; H,
7.52; N, 5.75.
r-Meth yl N-(ter t-Bu toxycar bon yl)aspar tate. The amino
ester 11 (131 mg, 0.273 mmol) was dissolved in a mixture of

A Peptidomimetic of the Turn in the HTH Motif
J . Org. Chem., Vol. 62, No. 24, 1997 8393
(d, 1H, J ) 8.8 Hz), 7.06 (t, 1H, J ) 7.5 Hz), 6.90 (d, 1H, J )
7.5 Hz), 6.82 (d, 1H, J ) 7.5 Hz), 6.60 (s, 1H), 5.98 (d, 1H, J
) 7.9 Hz), 4.52 (m, 1H), 4.11 (m, 1H), 3.72 (dd, 1H, J ) 6.8,
14.1 Hz), 3.61 (s, 3H), 3.40 (dd, 1H, J ) 6.2 , 14.1 Hz), 3.01
(dd, 1H, J ) 3.1, 13.8 Hz), 2.86 (d, 2H, J ) 6.7 Hz), 2.60 (dd,
1H, J ) 7.5 Hz, 12.7 Hz), 2.31 (dd, 1H, J ) 5.7, 13.2 Hz), 1.30
(s, 9H). 13C NMR (CDCl3): δ 172.07 (two peaks), 169.27,
155.45, 136.98, 136.40, 133.26, 129.31, 128.69, 127.83, 79.41,
56.57, 53.01, 52.49, 44.00, 39.06, 37.05, 29.11. Anal. Calcd for
C20H27N3O6: C, 59.25; H, 6.71; N, 10.36. Found: C, 59.15; H,
6.76; N, 10.29.
(S)-9-[(ter t-Bu t oxyca r b on yl)a m in o]-5,8-d ioxo-(S)-4,7-
dia za bicyclo[9.3.1]p en ta d eca -1(15),11,13-tr ien e-3-ca r box-
ylic Acid (1b). Cyclic peptide ester 13 (190 mg, 0.45 mmol)
was dissolved in warm THF/MeOH (40 mL) and cooled to rt.
NaOH (1 N, 0.5 mL) was slowly added with stirring. After 1
h, the completed reaction (TLC showed the disappearance of
13; 120:2:1 CHCl3/MeOH/AcOH, ninhydrin stain) was concen-
trated and the residue was dissolved in water. Dropwise
addition of 2 N HCl induced precipitation of the product which
(m, 1H), 4.05 (m, 1H), 3.85 (dd, 1H, J ) 8.8, 14.0 Hz), 3.27
(1H, obscured by H2O peak), 3.02-2.78 (m, 3H), 2.67 (dd, 1H,
8.2 Hz, J ) 12.9 Hz), 1.34 (s, 9H); 13C NMR (DMSO-d6; CD3-
OD): δ 173.27, 173.05, 169.97, 156.14, 137.34, 136.43, 133.51,
129.55, 128.80, 128.18, 80.24, 56.76, 52.70, 49.45, 44.16, 37.58,
28.73. Anal. Calcd for C19H25N3O6‚H2O: C, 55.74; H, 6.65; N,
10.26. Found: C, 55.16; H, 6.59; N, 9.90.
Ack n ow led gm en t. We thank Loretta Yang for
HPLC purification of diastereomers. Our thanks to
Professor Mark Lipton of Purdue University for helpful
suggestions for stereochemical proof. This work was
supported in part by ACS-PRF Grant #28634-G4 and
by NIH Grant #1 R29 GM52516-01.
Su p p or tin g In for m a tion Ava ila ble: Experimental pro-
cedures for compound 3 and its precursors, the 1H and 13C
NMR spectra of compounds 5,7,8,9,10,13, and 1b, and the
1
HPLC and H NMR of enriched diastereomeric mixtures of 8
(17 pages). This material is contained in libraries on micro-
fiche, immediately follows this article in the microfilm version
of the journal, and can be ordered from the ACS; see any
current masthead page for ordering information.
was collected on
a filter and dried under vacuum in a
desiccator to give a white solid (140 mg, 74%). mp >250 °C.
1H NMR (CDCl3/DMSO-d6): δ 8.30 (br s, 1H), 7.98 (t, 1H, J )
6.3 Hz), 7.23 (d, 1H, J ) 9.4 Hz), 7.14 (t, 1H, J ) 7.0 Hz), 6.96
(d, 1H, J ) 7.0 Hz), 6.66 (s, 1H), 6.37 (d, 1H, J ) 7.8 Hz), 4.39
J O971077H