Conformational effects of the non-natural α-methylserine on small peptides and glycopeptides

Article abstract of DOI:10.1021/jo901988w

(Chemical Equation Presented) The synthesis and the conformational analysis in aqueous solution of a peptide and a glycopeptide containing the sequence threonine-alanine-alanine (Thr-Ala-Ala) are reported. Furthermore, the threonine residue has been repla

Full text of DOI:10.1021/jo901988w

pubs.acs.org/joc
Conformational Effects of the Non-natural r-Methylserine on Small
Peptides and Glycopeptides
ꢀ
Alberto Fernandez-Tejada, Francisco Corzana,* Jesus H. Busto, Alberto Avenoza, and
ꢀ
ꢀ
Jesus M. Peregrina*
´
ꢀ
´
Departamento de Quımica, Universidad de La Rioja, Centro de Investigacion en Sıntesis Quımica, Grupo de
´
´ ´
~
Sıntesis Quımica de La Rioja, UA-CSIC, 26006 Logrono, Spain
francisco.corzana@unirioja.es; jesusmanuel.peregrina@unirioja.es
Received September 17, 2009
The synthesis and the conformational analysis in aqueous solution of a peptide and a glycopeptide
containing the sequence threonine-alanine-alanine (Thr-Ala-Ala) are reported. Furthermore, the
threonine residue has been replaced by the quaternary amino acid R-methylserine (MeSer) and their
corresponding non-natural peptide and glycopeptide are also studied. The conformational analysis
in aqueous solution combines NOEs and coupling constants data with Molecular Dynamics (MD)
simulations with time-averaged restraints. The study reveals that the β-O-glycosylation produces a
remarkable and completely different effect on the backbone of the peptide derived from Thr and
MeSer. In the former, the β-O-glycosylation is responsible for the experimentally observed shift from
extended conformations (peptide) to folded ones (glycopeptide). In contrast, the β-O-glycosylation
of the MeSer-containing peptide, which clearly shows two main conformations in aqueous solution
[extended ones (70%) and β-turn (30%)], causes a high degree of flexibility for the backbone.
Introduction
the simplest non-natural β-O-glycosylated amino acids as
well as a small peptide and glycopeptide containing a non-
natural residue.6 The studies showed that these non-natural
molecules can stabilize conformations present in naturally
occurring molecules or exhibit some atypical conformations.
On the other hand, the recent improvements in the syn-
thetic methodology make it possible to gain access to novel
synthetic glycopeptides that can be useful to study the
structural influence of protein glycosylation. One of the
main influences that glycosylation exerts is the stabilization
of protein conformations, which is of pivotal importance for
the biological activity of glycopeptides. On this basis, we
have selected a glycosylated tripeptide (Thr*-Xaa-Xaa) and
its analogue containing the noncoded R-amino acid MeSer
(MeSer*-Xaa-Xaa) to unravel the role of carbohydrate
The synthesis of nonproteinogenic R-amino acids has
received enormous attention because it represents a great
advance in the design of modified peptides with altered
biological activity. Particular attention has been paid to R,
R-disubstituted R-amino acids;1 for example, R-methylserine
(R-MeSer) can be regarded as a potential R-helix-stabilizing
building block.2 By contrast, the field of modified glycopep-
tides remains relatively unexplored, and most of the modi-
fications studied are centered around the carbohydrate
moiety.3 On this basis, and considering that various biolo-
gically active O-glycoproteins, such as epidermal growth
factor (EGF) domains of different serum proteins4 and
Notch receptor,5 incorporate the β-O-glucosylserine (βGlc-
Ser) substructure, we have recently synthesized and studied
DOI: 10.1021/jo901988w
r
Published on Web 11/19/2009
J. Org. Chem. 2009, 74, 9305–9313 9305
2009 American Chemical Society

ꢀ
Fernandez-Tejada et al.
JOCArticle
attachment in the stabilization of peptide conformations. In
previous works,6 commented on above, we have studied this
feature in the simplest model glycopeptides: amino acid
diamides Ac-Thr-NHMe and Ac-MeSer-NHMe glyco-
sylated with β-O-Glc. In the present paper, we want to
demonstrate if the characteristics observed in these short
models are kept when the peptide is bigger. To this aim, we
have selected natural amino acids to add to the β-hydroxy-
R-amino acid. Taking into account that some amino acids
(Pro, Phe, etc.) can exert a structural role in the conforma-
tions of the backbone, we have decided to use an amino acid
that does not cause these conformational effects. Moreover,
the selected amino acid must be present around glycosylation
sites in proteins.7 In this context, alanine (Ala) appears as a
good candidate for this purpose, therefore it was selected
as the amino acid to increase the size of the peptidic chain
(Xaa=Ala).
Herein, wehave synthesized andstudied the peptideand the
corresponding β-O-glycopeptide derived from the sequence
MeSer-Ala-Ala, where MeSer is the non-natural amino acid
R-methylserine (Figure 1). In addition, the analogues mole-
cules derived from the natural sequence Thr-Ala-Ala are also
studied for a comparative purpose. Interestingly, this small
peptide is the repeating tandem sequence found in antifreeze
glycoproteins, which allow polar fish to survive in seas where
the temperature is subzero.8 In addition, this substructure
FIGURE 1. Compounds studied in this work.
appears to be important for the activity of a family of
antimicrobial peptides called Alyteserin.9 In fact, the deriva-
tives that incorporate the Thr-Ala-Ala sequence display more
activity against gram-positive bacteria than their analogue
without this sequence.
Results and Discussion
Synthesis. The synthesis of peptide 1 was carried out as
described in the Supporting Information. The corresponding
glycopeptide 2 was obtained following the modified condi-
tions of the Koenigs-Knorr glycosylation.10 The treatment
of peptide 1 with 2,3,4,6-tetra-O-benzoyl-R-D-glucopyrano-
syl bromide in the presence of AgTfO led to derivative 5 with
a moderate yield. The hydroxyl groups of this compound
were deprotected with MeONa to give glycopeptide 2
(Scheme 1). It is important to notice that these basic condi-
tions did not produce racemization in the final compound 2,
since no other product was observed in the NMR spectra.
Concerning the non-natural peptide and glycopeptide, the
synthesis of peptide 3 was started from dipeptide 6, which
was previously synthesized by our group.6b This com-
pound was first deprotected with trifluoroacetic acid
(TFA) to give the amine 7. MeSer derivative 8, also obtained
in our group,11 was then treated with 7, and N,N,N0,N0-
tetramethyl-O-(benzotriazol-1-yl)uranium tetrafluoroborate
(TBTU) as a coupling agent in the presence of diisopropy-
lethylamine (DIEA) as a base to give the N-Boc protected
peptide 9, which was purified by column chromatography on
silica gel. Deprotection of the terminal amino group and
further acetylation with acetic anhydride (Ac2O) and pyri-
dine (Py) followed by subsequent hydrolysis of the acetyl
group with MeONa gave the desired peptide 3 (Scheme 2).
Glycopeptide 4 was synthesized from non-natural amino
acid 10, previously obtained in our group, whose carboxylic
acid group was protected as allyl ester (a selectively remo-
vable carboxy-protecting function), and the hydroxyl group
as silyl ether (TBDMS) to give compound 11. Further
acetylation of the amino group (Ac2O/Py) and subsequent
deprotection of the hydroxyl group led to derivative 12 in a
moderate yield. The key step in the synthesis of glycopeptide
(1) (a) Cativiela, C.; Dıaz-de-Villegas, M. D. Tetrahedron: Asymmetry
1998, 9, 3517–3599. (b) Cativiela, C.; Dıaz-de-Villegas, M. D. Tetrahedron:
Asymmetry 2000, 11, 645–732. (c) Cativiela, C.; Dıaz-de-Villegas, M. D.
€
Tetrahedron: Asymmetry 2007, 18, 569–623. (d) Vogt, H.; Brase, S. Org.
Biomol. Chem. 2007, 5, 406–430. (e) Aydillo, C.; Jimenez-Oses, G.; Busto, J.
ꢀ
ꢀ
H.; Peregrina, J. M.; Zurbano, M. M.; Avenoza, A. Chem.;Eur. J. 2007, 13,
ꢀ
4840–4848. (f) Jimenez-Oses, G.; Aydillo, C.; Busto, J. H.; Zurbano, M. M.;
Peregrina, J. M.; Avenoza, A. J. Org. Chem. 2007, 72, 5399–5402. (g)
Avenoza, A.; Busto, J. H.; Canal, N.; Peregrina, J. M. J. Org. Chem. 2005,
ꢀ
70, 330–333. (h) Avenoza, A.; Busto, J. H.; Canal, N.; Peregrina, J. M.; Perez-
ꢀ
Fernandez, M. Org. Lett. 2005, 7, 3597–3600. (i) Avenoza, A.; Busto, J. H.;
Cativiela, C.; Corzana, F.; Peregrina, J. M.; Sucunza, D.; Zurbano, M. M.
Preparation and synthetic applications of (S)- and (R)-N-Boc-N,O-isopropyli-
dene-R-methylserinals. In Targets in Heterocyclic Systems. Chemistry and
Properties; Attanasi, O. A., Spinelli, D., Eds.; Italian Society of Chemistry:
Roma, Italy, 2002; Vol. 6, pp 231-244. (j) Williams, R. M.; Sinclair, P. J.;
DeMong, D. E.; Chen, D.; Zhai, D. Org. Synth. 2003, 80, 18–30. (k) Williams, R.
M.; Sinclair, P. J.; DeMong, D. E.; Chen, D.; Zhai, D. Org. Synth. 2003, 80, 31–
37. (l) Williams, R. M. Asymmetric Synthesis of R-Amino Acids. In Advances in
Asymmetric Synthesis; Hassner, A., Ed.; JAI Press: New York, 1995; Vol. 1, pp
45-94. (m) Williams, R. M.; Im, M.-N. J. Am. Chem. Soc. 1991, 113, 9276–
9286. (n) Williams, R. M. Aldrichim. Acta 1992, 25, 11–25.
(2) (a) Toniolo, C.; Benedetti, E. Trends Biochem. Sci. 1991, 16, 350–353.
(b) Toniolo, C.; Crisma, M.; Formaggio, F.; Peggion, C. Biopolymers
(PepideScience) 2001, 60, 396–419. (c) Sissi, C.; Rossi, P.; Felluga, F.;
Formaggio, F.; Palumbo, M.; Tecilla, P.; Toniolo, C.; Scrimin, P. J. Am.
€
Chem. Soc. 2001, 123, 3169–3170. (d) Gessman, R.; Bruckner, H.; Petratos,
K. J. Pept. Sci. 2003, 9, 753–762. (e) Crisma, M.; Moretto, A.; Rainaldi, M.;
Formaggio, F.; Broxterman, Q. B.; Kaptein, B.; Toniolo, C. J. Pept. Sci.
2003, 9, 620–637.
(3) (a) Wilstermann, M.; Kononov, L. O.; Nilsson, U.; Ray, A. K.;
ꢀ
Magnusson, G. J. Am. Chem. Soc. 1995, 117, 4742–4754. (b) Alibes, R.;
Bundle, D. R. J. Org. Chem. 1998, 63, 6288–6301. (c) Geyer, A.; Muller, M.;
€
Schmidt, R. R. J. Am. Chem. Soc. 1999, 121, 6312–6313.
(4) (a) Shao, L.; Luo, Y.; Moloney, D. J.; Haltiwanger, R. S. Glycobiology
2002, 12, 763–770. (b) Harris, R. J.; Spellman, M. W. Glycobiology 1993, 3,
219–224.
(5) Bruckner, K.; Perez, L.; Clausen, H.; Cohen, S. Nature 2000, 406, 411–
415.
ꢀ
ꢀ
(6) (a) Fernandez-Tejada, A.; Corzana, F.; Busto, J. H.; Jimenez-Oses,
G.; Peregrina, J. M.; Avenoza, A. Chem.;Eur. J. 2008, 14, 7042–7058. (b)
Fernandez-Tejada, A.; Corzana, F.; Busto, J. H.; Avenoza, A.; Peregrina, J.
M. Org. Biomol. Chem. 2009, 7, 2885–2885.
(9) Conlon, J. M.; Demandt, A.; Nielsen, P. F.; Leprince, J.; Vaudry, H.;
Woodhams, D. C. Peptides 2009, 30, 1069–1073.
(10) Hanessian, S.; Banoub, J. Carbohydr. Res. 1977, 53, C13–C16.
(11) Avenoza, A.; Cativiela, C.; Corzana, F.; Peregrina, J. M.; Sucunza,
D.; Zurbano, M. M. Tetrahedron: Asymmetry 2001, 12, 949–957.
(7) Christlet, T. H. T.; Veluraja, K. Biophys. J. 2001, 80, 952–960.
(8) (a) Yeh, Y.; Feeney, R. E. Chem. Rev. 1996, 96, 601–618. (b)
Tachibana, Y.; Fletcher, G. L.; Fujitani, N.; Tsuda, S.; Monde, K.;
Nishimura, S.-I. Angew. Chem., Int. Ed. 2004, 43, 856–862.
9306 J. Org. Chem. Vol. 74, No. 24, 2009

ꢀ
Fernandez-Tejada et al.
JOCArticle
SCHEME 1. Synthesis of Glycopeptide Diamide 2
SCHEME 2. Synthesis of Peptide Diamide 3
The torsional angles and the atom labels used herein for
the compounds are shown in Figure 2.
4 was the glycosylation of MeSer derivative 12 following the
modified conditions of the Koenigs-Knorr glycosylation10
with the same glycosyl donor as commented on above, to give
building block 13 in 50% yield. Selective removal of the allyl
group in compound 13 was carried out with Pd0-catalyzed
allyl transfer to morpholine. Under these conditions, the
deprotection proceeded without affecting other labile
groups.12 Further coupling of the acid group with the amine
group of dipeptide 7 led to intermediate 14. Finally, the
hydrolysis of the benzoyl groups with sodium methoxide in
methanol gave the desired glycopeptide 4 derived from MeSer
(Scheme 3). In this case, racemization was not observed.
Conformational Analysis of the Desired Peptides and
Glycopeptides. The conformational analysis in aqueous solu-
tion of all the compounds was carried out combining NMR
datawithmolecular dynamicssimulations with time-averaged
restraints (MD-tar), following a protocol13 identical with that
recently applied by our group on different glycopeptides.
In the first step, full assignment of the protons of peptides
1 and 3, as well as the corresponding glycopeptides 2 and
4, was carried out with standard COSY and HSQC experi-
ments (see the Supporting Information). Then, selective
1D-NOESY experiments in D2O (25 ꢀC, pH 4.8) and
2D-NOESY experiments in H2O/D2O (9/1) (25 ꢀC, pH 4.8)
were carried out for all the compounds.
FIGURE 2. Molecular structure of glycopeptide 4, showing the
atom labels and the most relevant torsional angles. The letter “s”
makes reference to the sugar moiety. The blue and green numbers
refer to the numeration of the NH and methyl groups, respectively.
The same atom labels were used for the rest of the molecules.
Conformational Analysis of Thr-Containing Derivatives.
Figure 3a shows the amide region of the 2D NOESY spectra
of glycopeptide 2. The presence of consecutive NH-NH
cross peaks suggests the existence of an important popula-
tion of helix-like conformation. On the contrary, despite the
overlapping of signals in the spectrum of the parent peptide,
it seems clear that there is an absence of these NOEs,
indicating that the backbone adopts almost exclusively ex-
tended conformations. This fact, along with the slight upfield
shift of the NH protons of the glycosylated residues (CSD),14
corroborates that β-O-glycosylation of peptides derived
from natural amino acids promotes folded conformations
for the peptide backbone within the glycopeptide. This
finding confirms our previous work,15 where β-O-glycosyla-
tion was found to be responsible for the experimental shift
from extended conformations (model peptides derived from
Ser and Thr) toward the folded ones (corresponding model
glycopeptides).
(12) (a) Kunz, H.; Waldmann, H. Angew. Chem., Int. Ed. 1984, 23, 71–72.
(b) Kunz, H.; Unverzagt, C. Angew. Chem., Int. Ed. 1984, 23, 436–437.
(14) Wishart, D. S.; Bigam, C. G.; Holm, A.; Hodges, R. S.; Sykes, B. D.
J. Biol. NMR 1995, 5, 67–81.
ꢀ
ꢀ
ꢀ
ꢀ
(15) Corzana, F.; Busto, J. H.; Engelsen, S. B.; Jimenez-Barbero, J.;
Asensio, J. L.; Peregrina, J. M.; Avenoza, A. Chem.;Eur. J. 2006, 12, 7864–
7871.
(13) Corzana, F.; Busto, J. H.; Jimenez-Oses, G.; Asensio, J. L.; Jimenez-
Barbero, J.; Peregrina, J. M.; Avenoza, A. J. Am. Chem. Soc. 2006, 128,
14640–14648.
J. Org. Chem. Vol. 74, No. 24, 2009 9307

ꢀ
Fernandez-Tejada et al.
JOCArticle
FIGURE 3. Amide region of the 800 ms 2D NOESY spectrum (400 MHz) in H2O/D2O (9:1) at 25 ꢀC of glycopeptides 2 (a) and 4 (b).
SCHEME 3. Synthesis of Glycopeptide Diamide 4
The next step was to experimentally determine some
proton-proton distances relevant to the conformational
analysis. They were obtained from the corresponding NOE
build-up curves16 (see the Supporting Information). Addi-
tionally, distances involving NH protons were semiquan-
titatively determined by integrating the volume of the
corresponding cross-peaks in the 2D NOESY spectrum.
3
On the other hand, JNH,HR coupling constants were mea-
sured from the splitting of the resonance signals in the 1D
spectrum. All these experimentally determined distances, as
well as the measured 3JNH,HR, were used as restraints in MD-
tar simulations17 with the aim of obtaining a distribution of
(16) Haselhorst, T.; Weimar, T.; Peters, T. J. Am. Chem. Soc. 2001, 123,
10705–10714.
(17) (a) Pearlman, D. A. J. Biomol. NMR 1994, 4, 1–16. (b) Torda, A. E.;
Scheek, R. M.; van Gunsteren, W. F. J. Mol. Biol. 1990, 214, 223–235.
9308 J. Org. Chem. Vol. 74, No. 24, 2009

ꢀ
Fernandez-Tejada et al.
JOCArticle
FIGURE 4. φp/ψp (p = peptide) distributions obtained from the MD-tar for compounds 1 (a) and 2 (b).
TABLE 1. Comparison of the Experimental and MD-tar Simulations
found for the Cbz-Thr-Ala-Ala-CO2Me derivative in
DMSO.19
Derived Distances and 3J Coupling Constants for Compound 2
exptla
MD-tar (ε = 80)
In contrast, for glycopeptide 2, the three amino acids show
a significant population corresponding to helix-like confor-
mations (close to 36% for Thr and around 20% for the
Ala residues). This result is in accordance with the NMR
data and also with our previous results obtained for small
peptides.
dNH1,NH2
dNH2,NH3
dNH3,NH4
dNH4,HR3
dNH3,HR3
dNH3,HR2
dNH2,HR2
dNH1,HR1
dH1s,Hβ
2.9
2.9
2.9
2.2
2.5
2.2
2.5
>2.7
2.8
4.3
6.9
5.9
5.5
3.0
3.0
2.9
2.3
2.6
2.1
2.4
2.8
2.8
3.8
6.5
5.4
5.7
With regard to the glycosidic linkage (Figure 5a), the
conformational preferences of the glycopeptide 2 are very
similar to those observed for the Thr model glycopeptide
previously studied by our group.15 The φs dihedral angle was
found to be rather rigid, exhibiting values close to -60ꢀ, in
agreement with the exoanomeric effect.20 The most signifi-
cant feature is that ψs takes values around 120-140ꢀ, result-
ing in an eclipsed conformation for the Hβ-Cβ and O1-C1
bonds. This conformer avoids the steric repulsion between
the methyl group at Cβ and the anomeric H1 proton, which
would be present when this ψs angle is close to 180ꢀ. At this
point, it has to be mentioned that the eclipsed conformation
has been previously reported by our group for different
model glycopeptides derived from Thr6a,21 but this is the
first time that we confirm this finding on a larger glyco-
peptide.
3JHR1,Hβ1
3JNH1,HR1
3JNH2,HR2
3JNH3,HR3
a
˚
Distances are given in A and coupling constants in Hz.
conformers in aqueous solution able to reproduce quantita-
tively the experimental NMR data. All these data, and those
theoretically obtained from the simulations, are shown in
Table 1 for glycopeptide 2.
The φi/ψi (i=1, 2, or 3) distribution obtained from the
MD-tar simulations for peptide 1 and glycopeptide 2 is
shown in Figure 4. As can be seen, peptide 1 presents φi/ψi
typical values of extended conformations for each residue,
in accordance with the NOEs commented on above. On
the other hand, this result is in agreement with those ob-
served by other groups for small peptides, for which polypro-
line II (PPII) conformation seems to be the predominant
one.18 Additionally, this finding is similar to that previously
As far as the lateral chain is concerned, in glycopeptide 2,
the χ1 torsional angle exhibits mainly the g(þ) conformation
with only 10% of the anti conformer (Figure 5c). This result
(20) Thatcher, G. R. J. The Anomeric Effect and Associated Stereoelec-
tronic Effects; American Chemical Society: Washington, DC, 1993.
ꢀ
(21) (a) Corzana, F.; Busto, J. H.; Jimenez-Oses, G.; Garcıa de Luis, M.;
Asensio, J. L.; Jimenez-Barbero, J.; Peregrina, J. M.; Avenoza, A. J. Am.
ꢀ
ꢀ
Chem. Soc. 2007, 129, 9458–9467. (b) Fernandez-Tejada, A.; Corzana, F.;
(18) Mu, Y.; Kosov, D. S.; Stock, G. J. Phys. Chem. B 2003, 107, 5064–5073.
^ ^
(19) Mimura, Y.; Inoue, Y.; Maeji, N. J.; Chujo, R. Int. J. Pept. Protein
Res. 1989, 34, 363–368.
ꢀ
Busto, J. H.; Jimenez-Oses, G.; Jimenez-Barbero, J.; Avenoza, A.; Peregrina,
J. M. Chem.;Eur. J. 2009, 15, 7297–7301.
ꢀ
J. Org. Chem. Vol. 74, No. 24, 2009 9309

ꢀ
Fernandez-Tejada et al.
JOCArticle
FIGURE 6. Calculated ensembles obtained from the MD-tar
simulations for peptide 1 (a), glycopeptide 2 (b), peptide 3 (c), and
glycopeptide 4 (d), showing the RMSDs for heavy atoms super-
imposition.
extended conformations (β-sheet and PPII) were the most
populated ones for both Ala residues.
FIGURE 5. (a) φs/ψs (s = sugar) distribution obtained from the
MD-tar simulations for compound 2. (b) φs/ψs distribution ob-
tained from the MD-tar simulations for compound 4. (c) χ1dis-
tribution obtained from the MD-tar simulations for compounds 2
and 4.
As a consequence, and according to the MD simulations,
MeSer-containing peptide 3 is rather flexible in aqueous
solution, showing the coexistence of extended conformations
(70%) and folded ones (30%) (see the Supporting
Information). However, it is important to note the excep-
tionally high population of this folded conformation, repre-
sented by a type I β-turn, for a small peptide (Figure 8).
With regard to the lateral chain (see the Supporting
Information), peptide 3 exhibits a flexible behavior for the
χ1 angle, and the g(þ) conformation is the most populated
one (χ1 angle close to 60ꢀ).
Concerning glycopeptide 4, the medium-strong
MeR1-NH1 NOE (typical of folded conformations), along
with the also observed MeR1-NH2 NOE (characteristic of
extended ones), suggests the coexistence of both conforma-
tions in aqueous solution for the MeSer residue (see the
Supporting Information). On the other hand, the presence of
the NH2-NH3 and NH3-NH4 NOEs suggests the exis-
tence of a certain population of folded conformations for the
two alanine residues, as already occurred for its parent
peptide (Figure 3b). Moreover, the observation of the
NH1-Hβ and ΝH2-Hβ NOE interactions seems to indicate
some flexibility for the conformational preferences of the χ1
torsional angle for glycopeptide 4.
As shown in Figure 7b, the MeSer residue adopts mainly
helix-like conformations (60%) with preference for the RD
helix, similar to that observed for the nonglycosylated
compound. Regarding the two residues of alanine, they
exhibited dispersed values of φi/ψi. While for Ala2 the
helix-like conformations were the most populated ones
(about 53% of the total trajectory time), extended conforma-
tions (β-sheet and PPII) were predominant for Ala3. There-
fore, the conformational analysis of the backbone indicates
that, in general, both folded conformations and extended
ones coexist for the three residues, as suggested by the NOE
experiments. As a consequence, the β-turn observed for
peptide 3 is present only 18% of the total trajectory time
for compound 4. Moreover, an additional β-turn was also
is in agreement with the small value experimentally observed
3
for JHR,Hβ (less than 4 Hz). Finally, Figure 6 shows the
superimposition of 20 conformers randomly taken from the
MD-tar simulations that quantitatively reproduce the ex-
perimental NMR data. As can be seen, peptide 1 (Figure 6a)
˚
presents a rather rigid structure with a rmsd value of 1.32 A
and adopts clearly an extended conformation. On the con-
trary, in the case of the corresponding glycopeptide, com-
pound 2 (Figure 6b), a larger degree of flexibility was
˚
observed with a higher rmsd (close to 1.61 A), showing
mainly a folded conformation.
Conformational Analysis of MeSer-Containing Deriva-
tives. The protocol described for peptide 1 and glycopeptide
2 was also used to quantitatively determine the con-
formers present in aqueous solution for their non-natural
analogues derived from MeSer. Nevertheless, it is worth
noting that not only was the synthesis of these non-
natural target compounds difficult, but also the structural
analysis was complicated. The absence of the HR atom
in the quaternary amino acid (MeSer) as well as the over-
lapping of the signals in the NMR spectrum corres-
ponding to the HR atoms in both alanine residues implicates
the lack of some NOE interaction signals and coupling
constants, which are important in eliciting the conforma-
tions.
Regarding peptide 3, the medium NH2-NH3 and
NH3-NH4 NOEs suggest the presence of a significant
population of the helix-like conformations (Supporting
Information). In this sense, the φp/ψp distribution obtained
from the MD-tar simulations for this compound is shown in
Figure 7a. The φ1/ψ1 dihedral angle values exhibited by the
MeSer residue were typical for helix-like conformations,
such as the right- and left-handed R-helix (RD, 65%; RL,
27%), in agreement with the NOE experiments. However,
9310 J. Org. Chem. Vol. 74, No. 24, 2009

ꢀ
Fernandez-Tejada et al.
JOCArticle
FIGURE 7. (a) φp/ψp distributions (backbone) obtained from the MD-tar simulations for peptide 3. (b) φp/ψp distributions (backbone)
obtained from the MD-tar simulations for glycopeptide 4.
found (13% of the total time), in which the intramolecular
hydrogen bond is formed between i þ 1 and i þ 4 residues
sequence Thr-Ala-Ala as well as of their non-natural ana-
logues from R-MeSer. Some important conclusions can be
drawn from this study. Comparing both peptides 1 and 3, it
is apparent that the incorporation of a methyl group at the
CR position of the amino acid induces folded conforma-
tions for the peptide backbone. This result confirms our
recent work on small model peptides. However, in spite of
the structural rigidity conferred by the quaternary amino
acid (R-MeSer), neither peptide 3 nor glycopeptide 4 are
conformationally rigid. Indeed, both molecules are rather
flexible in aqueous solution. Concerning the glycosylated
compounds, the study reveals that the β-O-glycosylation
produces a remarkable and completely different effect on
the peptide backbone of the diamide tripeptide derived
from Thr and MeSer (compare panels b and d of Figure 6).
In the former, the β-O-glycosylation is responsible for the
experimentally observed shift from extended conforma-
tions (peptide) to folded ones (glycopeptide). This feature
supports the tendency observed in our previous study on
model peptides derived from Ser and Thr. However, the
β-O-glycosylation of the non-natural peptide with MeSer
does not seem to promote that shift toward helix-like
conformations. This result is in good agreement with our
recent study on non-natural model glycopeptides, indicat-
ing that the methyl group at the R-position, and not the
β-O-glycosylation, is responsible for the backbone con
formations in these non-natural compounds. In fact,
for the diamide MeSer-Ala-Ala tripeptide, which clearly
shows two main conformations in aqueous solution
[extended ones (70%) and β-turn (30%)], the β-O-glycosy-
lation causes a large degree of flexibility for the peptide
backbone.
(C2dO2 H4-N4, see the Supporting Information). These
3 3 3
results are in agreement with the observed MeC-NH3 and
MeR1-NH4 NOEs, respectively, for each turn.
Concerning the glycosidic linkage (Figure 5b), the values
of the φs dihedral angle are mainly determined by the
exoanomeric effect. On the other hand, the ψs angle showed
well-defined values around 180ꢀ (typical of alternate con-
formations). It is important to note that this is a completely
different behavior to that observed for the Thr-containing
glycopeptide 2. This result demonstrates that the distinct
substitution of the methyl group at the CR or Cβ position
(MeSer and Thr, respectively) drastically affects the geometry
of the glycosidic linkage also in larger glycopeptides.
Finally, as shown in Figure 5c, it is possible to see that the
lateral chain of the non-natural glycopeptide is rather flex-
ible, showing a significant population of each staggered
conformation for the χ1 torsional angle. This fact is in
accordance with the NH1-Hβ and NH2-Hβ NOE interac-
tions commented on above.
Therefore, from the exhaustive conformational study of
the MeSer-containing molecules, it can be concluded that in
this case, the β-O-glycosylation is responsible for the con-
formational flexibility within the glycopeptide, which exhi-
bits the highest root-mean-square deviation (rmsd) value
˚
(2.13 A), regarding not only the peptide backbone but also
the lateral chain (Figure 6d).
Conclusions
Herein we report the synthesis and the conformational
analysis of a peptide and glycopeptide derived from the
J. Org. Chem. Vol. 74, No. 24, 2009 9311

ꢀ
Fernandez-Tejada et al.
JOCArticle
C19H34N4O10: C, 47.69; H, 7.16; N, 11.71. Found: C, 47.79;
H, 7.20; N, 11.65.
Synthesis of Compound 9. TFA (2 mL) was added to a solution
of 6 (66 mg, 0.24 mmol) in CH2Cl2 (2 mL) at 0 ꢀC. The reaction
was maintained at 0 ꢀC for 30 min then at 25 ꢀC for 3 h, and then
concentrated to give derivative 7 (0.23 mmol). This compound
was then dissolved at 25 ꢀC in DMF (3 mL) and treated with R-
methylserine derivative 8 (43 mg, 0.20 mmol), DIEA (0.2 mL,
0.90 mmol), and TBTU (83 mg, 0.26 mmol). The reaction
mixture was stirred at 25 ꢀC for 8 h, the solvent was then
removed, and the residue was purified by silica gel column
chromatography, eluting with dichloromethane/MeOH (15:1)
to give compound 9 (105 mg, 70%) as a colorless oil. This
compound was used without further characterization in the next
step. Anal. Calcd for C16H30N4O6: C, 51.32; H, 8.08; N, 14.96.
Found: C, 51.40; H, 8.05; N, 14.92.
Synthesis of Compound 3. TFA (3 mL) was added to a solution
of 9 (100 mg, 0.27 mmol) in CH2Cl2 (3 mL) at 0 ꢀC. The reaction
was stirred at 0 ꢀC for 30 min then at 25 ꢀC for 2.5 h, and then
concentrated. The residue was dissolved in pyridine/acetic
anhydride (2:1, 6 mL) and the mixture was stirred at 25 ꢀC for
3 h to give the corresponding O,N-diacetylated compound,
which was used in the next step without purification. A solution
of this compound in MeOH (5 mL) was treated with sodium
methoxide/MeOH (0.5M) to pH 9. After being stirred at 25 ꢀC
for 2 h, the mixture was neutralized with Dowex 50-X8, filtered,
and concentrated. Finally, the residue was purified with a C18
reverse-phase sep-pak cartridge to give 3 (69 mg), as a colorless
oil, in 81% yield. [R]25D -15.0 (c 0.24, CH3OH). 1H NMR (400
MHz, D2O): δ 1.36-1.47 (m, 6H), 1.45 (s, 3H), 2.05 (s, 3H), 2.75
(s, 3H), 3.74-3.87 (m, 2H), 4.22-4.34 (m, 2H). 1H NMR (400
MHz, H2O/D2O): δ 1.31-1.37 (m, 6H), 1.40 (s, 3H), 2.00 (s,
3H), 2.71 (d, J=4.5 Hz, 3H), 3.69-3.84 (m, 2H), 4.15-4.31 (m,
2H), 7.66 (br s, 1H), 7.98 (d, J=6.0 Hz, 1H), 8.21 (d, J=5.6 Hz,
1H), 8.24 (s, 1H). 13C NMR (100 MHz, D2O): δ 15.9, 16.4, 19.3,
22.2, 25.8, 49.8, 50.1, 60.3, 64.5, 174.3, 175.1, 175.3, 175.4. Anal.
Calcd for C13H24N4O5: C, 49.36; H, 7.65; N, 17.71. Found: C,
49.44; H, 7.63; N, 17.74.
FIGURE 8. Scheme of the type I β-turn obtained from the MD-tar
simulations in aqueous solution for peptide 3. The β-turns are
formed when the polypeptide chain adopts a folded structure
involving four amino acid residues, with the carbonyl oxygen of
the first residue (i) forming a hydrogen bond with the amino group
hydrogen of the fourth (i þ 3). Type I β-turn shows characteristic
values for the dihedral angles φ/ψ of -60ꢀ, -30ꢀ (φiþ1/ψiþ1) and
-90ꢀ, 0ꢀ (φiþ2/ψiþ2).
Synthesis of Compound 11. A mixture of allylic alcohol (1.7
mL, 25.0 mmol), (S)-R-methylserine 10 (430 mg, 2.76 mmol), p-
TsOH (635 mg, 3.34 mmol), and toluene (20 mL) was refluxed
for 24 h by using a Dean-Stark trap and then concentrated. The
crude salt was suspended in CH2Cl2 (15 mL) and imidazol (946
mg, 13.9 mmol) and TBDMSCl (922 mg, 6.12 mmol) were
added. The resulting suspension was stirred at 25 ꢀC for 24 h.
The solvent was then removed and the mixture was partitioned
between NaOH (1 N, 10 mL) and ethyl acetate (10 mL). The
organic layer was washed with NaOH (1 N, 3 ꢀ 10 mL) and
brine (10 mL), dried, filtered, and evaporated. Finally, the
residue was purified by silica gel column chromatography
Experimental Section
Synthesis of Compound 5. Silver triflate (60 mg, 0.25 mmol)
was added to a suspension of 1 (35 mg, 0.15 mmol) and
˚
powdered molecular sieves (4 A, 20 mg) in CH2Cl2 (3 mL),
under an inert atmosphere. The mixture was stirred at -30 ꢀC
and then 2,3,4,6-tetra-O-benzoyl-R-D-glucopyranosyl bromide
(570 mg, 0.86 mmol) in CH2Cl2 (3 mL) was added. The mixture
was stirred at this temperature for 1 h and then was warmed to
25 ꢀC and stirred for an additional 14 h. The crude was filtered,
concentrated, and purified by silica gel column chromatogra-
phy, eluting with CH2Cl2/MeOH (95:5), to give 5 as a colorless
oil (31 mg, 31%), which was used without further characteriza-
tion in the next step. Anal. Calcd for C47H50N4O14: C, 63.08; H,
5.63; N, 6.26. Found: C, 63.14; H, 5.60; N, 6.30.
Synthesis of Compound 2. A solution of 5 (30 mg, 0.03 mmol)
in MeOH (5 mL) was treated with MeONa/MeOH (0.5M) to
pH 9. After being stirred for 3 h at 25 ꢀC, the mixture was
neutralized with Dowex 50-X8, filtered, and concentrated.
Purification of the residue with a C18 reverse-phase sep-pak
(ethyl acetate/hexane, 8:2) to give compound 11 (280 mg,
1
37%) as a yellow oil. [R]25 -5.9 (c 1.10, CH3OH). H NMR
D
(300 MHz, CDCl3): δ 0.00-0.11 (m, 6H), 0.87 (s, 9H), 1.25 (s,
3H), 1.90 (s, 1H), 3.48 (d, J=9.1 Hz, 1H) 3.90 (d, J = 9.1 Hz,
1H), 4.52-4.69 (m, 2H), 5.14-5.51 (m, 2H), 5.80-6.03 (m, 1H).
13C NMR (75 MHz, CDCl3): δ -5.8,-5.7,18.0, 22.3, 25.6, 59.5,
65.5, 69.9, 118.0, 131.9, 176.2. Anal. Calcd for C13H27NO3Si: C,
57.10; H, 9.95; N, 5.12. Found: C, 57.22; H, 9.90; N, 5.10.
Synthesis of Compound 12. A solution of 11 (260 mg, 0.95
mmol) in pyridine/acetic anhydride (2:1, 6 mL) was stirred at 25
ꢀC for 5 h and then concentrated. The residue was dissolved in
THF (7 mL) and a solution of TBAF 1 M in THF (1.25 mL, 1.25
mmol) was added. The reaction was stirred for 8 h, then the
solvent was removed and the mixture partitioned between
CHCl3/iPrOH (4:1, 15 mL) and saturated NH4Cl solution
(5 mL). The organic layer was dried over Na2SO4, filtered,
and concentrated to give a residue that was purified by silica
gel column chromatography, eluting with ethyl acetate/MeOH
1
cartridge gave 14 mg of 2, as a colorless oil in 88% yield. H
NMR (400 MHz, D2O): δ 1.27 (d, J=6.3 Hz, 3H), 1.35-1.43
(m,6H), 2.10 (s, 3H), 2.74 (s, 3H), 3.22-3.30 (m, 1H), 3.35-3.42
(m, 1H), 3.43-3.53 (m, 2H), 3.73 (dd, J1=12.3 Hz, J2=5.9 Hz,
1H), 3.92 (dd, J1 = 12.2 Hz, J2 = 1.8 Hz, 1H), 4.24 (q, J=7.2
Hz, 1H), 4.31-4.39 (m, 2H), 4.41 (d, J=4.6 Hz, 1H), 4.55 (d,
J = 7.9 Hz, 1H). 13C NMR (100 MHz, D2O): δ 15.6, 16.4,
16.5, 21.7, 25.8, 49.7, 49.8, 58.3, 60.8, 69.7, 73.0, 73.6, 75.6,
75.9, 99.7, 171.6, 174.4, 174.8, 175.2. Anal. Calcd for
9312 J. Org. Chem. Vol. 74, No. 24, 2009

ꢀ
Fernandez-Tejada et al.
JOCArticle
(95:5) to give 12 (170 mg, 88%), as a colorless oil. [R]25D -10.7
(c 1.10, CH3OH). 1H NMR (300 MHz, CDCl3): δ 1.47 (s, 3H),
1.97 (s, 3H), 3.59 (br s, 1H), 3.76 (d, J=11.3 Hz, 1H), 4.01 (d, J=
11.5 Hz, 1H), 4.52-4.72 (m, 2H), 5.11-5.41 (m, 2H), 5.73-5.97
(m, 1H), 6.36 (br s, 1H). 13C NMR (75 MHz, CDCl3): δ 20.3,
23.8, 62.3, 66.5, 66.6, 118.9, 131.4, 170.8, 172.9. Anal. Calcd for
C9H15NO4: C, 53.72; H, 7.51; N, 6.96. Found: C, 53.61; H, 7.54;
N, 7.01.
Synthesis of Compound 13. Silver triflate (149 mg, 0.58 mmol)
was added to a suspension of 12 (70 mg, 0.35 mmol) and
˚
powdered molecular sieves (4 A, 20 mg) in CH2Cl2 (4 mL),
under an inert atmosphere. The mixture was stirred at -30 ꢀC
and then 2,3,4,6-tetra-O-benzoyl-R-D-glucopyranosyl bromide
(323 mg, 0.49 mmol) in CH2Cl2 (4 mL) was added. The mixture
was stirred at this temperature for 1 h and then was warmed to
25 ꢀC and stirred for an additional 14 h. The crude was filtered,
concentrated, and purified by silica gel column chromatogra-
cartridge gave 14 mg of 4, as a colorless oil in 88% yield. [R]25
D
1
-26.0 (c 1.20, H2O). H NMR (400 MHz, D2O): δ 1.26-1.33
(m, 6H), 1.40 (s, 3H), 1.95 (s, 3H), 2.65 (s, 3H), 3.18-3.32 (m,
2H), 3.33-3.44 (m, 2H), 3.63 (dd, J1=12.3 Hz, J2=6.1 Hz, 1H),
3.73-3.89 (m, 2H), 4.04 (d, J=10.3 Hz, 1H), 4.12-4.25 (m, 2H),
4.39 (d, J=7.9 Hz, 1H). 13C NMR (100 MHz, D2O): δ 15.9, 16.4,
19.4, 22.3, 25.8, 49.8, 50.2, 59.4, 60.8, 69.7, 71.8, 73.1, 75.6, 76.0,
102.4, 174.3, 174.8, 175.1, 175.3. Anal. Calcd for C19H34N4O10:
C, 47.69; H, 7.16; N, 11.71. Found: C, 47.78; H, 7.13; N, 11.78.
2D NMR Experiments. NMR experiments were recorded on a
400 MHz spectrometer at 293 K. Magnitude-mode ge-2D
COSY spectra were recorded with gradients and with the
cosygpqf pulse program with 90 degree pulse width. Phase-
sensitive ge-2D HSQC spectra were recorded with z-filter and
selection before t1 removing the decoupling during acquisition
by use of the invigpndph pulse program with CNST2 (JHC)=
145. 2D NOESY experiments were made with phase-sensitive
ge-2D NOESY for CDCl3 spectra and phase-sensitive ge-2D
NOESY with WATERGATE for H2O/D2O (9:1) spectra. Se-
lective ge-1D NOESY experiments were carried out with the 1D-
DPFGE NOE pulse sequence. NOEs intensities were normal-
ized with respect to the diagonal peak at zero mixing time.
phy, eluting with ethyl acetate/hexane (1:1), to give 13 as a
colorless oil (135 mg, 50%). [R]25 þ3.7 (c 1.25, CH3OH). H
1
D
NMR (300 MHz, CDCl3): δ 1.38 (s, 3H), 1.56 (s, 3H), 4.01-4.08
(m, 1H), 4.11 (d, J=9.9 Hz, 1H), 4.18 (d, J=9.9 Hz, 1H), 4.39
(dd, J1=12.2 Hz, J2=4.8 Hz, 1H), 4.48-4.61 (m, 3H), 4.74 (d,
J=7.9 Hz, 1H), 5.12 (d, J=11.0 Hz, 1H), 5.17-5.28 (m, 1H),
5.40 (dd, J1=9.7 Hz, J2=8.0 Hz, 1H), 5.61 (“t”, J=9.7 Hz, 1H),
5.72-5.87 (m, 2H), 6.16 (br s, 1H), 7.16-7.23 (m, 2H),
7.24-7.30 (m, 2H), 7.31-7.38 (m, 5H), 7.39-7.53 (m, 3H),
7.73-7.79 (m, 2H), 7.80-7.86 (m, 2H), 7.87-7.92 (m, 2H),
7.93-8.01 (m, 2H). 13C NMR (75 MHz, CDCl3): δ 19.6, 23.4,
60.3, 63.0, 66.5, 69.5, 71.5, 71.8, 72.3, 72.6, 101.5, 118.5, 128.3,
128.4, 128.5, 128.7, 128.8, 129.3, 129.6, 129.8, 129.8, 129.9,
131.5, 133.2, 133.3, 133.4, 133.5, 165.0, 165.2, 165.8, 166.2,
169.7, 171.9. Anal. Calcd for C43H41NO13: C, 66.23; H, 5.30;
N, 1.80. Found: C, 66.16; H, 5.35; N, 1.77.
Experimental NOEs were fitted to a double exponential func-
tion, f(t)=p0(e-p t)(1 - e t) with p0, p1, and p2 being adjustable
-p2
1
parameters.16 The initial slope was determined from the first
derivative at time t = 0, f0(0)=p0p2. From the initial slopes,
interproton distances were obtained by employing the isolated
spin pair approximation.
MD Simulations: MD-tar Simulations. MD-tar simulations
were performed with AMBER22 6.0 and AMBER94 force
field,23 which was implemented with GLYCAM 0424 para-
meters to accurately simulate the conformational behavior of
the carbohydrate moiety. NOE-derived distances were included
as time-averaged distance constraints, and scalar coupling con-
Synthesis of Compound 14. A solution of derivative 13 (130
mg, 0.17 mmol) in THF (10 mL) was stirred under an argon
atmosphere at 25 ꢀC, then Pd(PPh3)4 (2 mg, 1.7 ꢀ 10-3 mmol)
and morpholine (52 μL, 0.60 mmol) were subsequently added.
After the mixture was stirred for 2 h, the solvent was evaporated
and the residue was taken up in 10 mL of ethyl acetate. The
organic layer was washed with 1 N HCl (5 mL), dried, and
concentrated to give the desired acid, which was directly used for
the next step. A solution of the acid in DMF (5 mL) was treated
with DIEA (0.13 mL, 0.80 mmol), amine 7 (63 mg, 0.22 mmol),
and TBTU (71 mg, 0.22 mmol). The reaction mixture was stirred
at 25 ꢀC for 10 h and then evaporated to give a residue that was
purified by silica gel column chromatography (CH2Cl2/MeOH,
stants J as time-averaged coupling constraints. A Ær-6 -1/6
æ
average was used for the distances and a linear average was
used for the coupling constants. Final trajectories were run with
an exponential decay constant of 8000 ps and a simulation
length of 80 ns with a dielectric constant ε=80.
ꢀ
Acknowledgment. We thank the Ministerio de Educacion
y Ciencia and FEDER (project CTQ2009-13814/BQU,
Ramon y Cajal contract of F.C., and grant of A.F.-T.) and
the Universidad de La Rioja (project EGI09/60). We thank
CESGA for computer support.
ꢀ
15:1) to give 14 as a colorless oil (60 mg, 41%). [R]25 þ1.0 (c
Supporting Information Available: 1H and 13C NMR spec-
tra as well as COSY and HSQC correlations for compounds
1-4, 1H and 13C NMR spectra for compounds 11-14, 2D
NOESY correlations for compounds 1-4, scheme of the second
β-turn obtained from the MD-tar simulations in aqueous solu-
tion for compound 4. This material is available free of charge via
the Internet at http://pubs.acs.org.
D
1.10, CH3OH). 1H NMR (300 MHz, CDCl3): δ 1.30-1.55 (m,
9H), 1.72 (s, 3H), 2.74 (d, J=4.8 Hz, 3H), 3.74 (d, J=9.9 Hz,
1H), 4.00 (d, J=9.9 Hz, 1H), 4.15-4.31 (m, 2H), 4.39-4.56 (m,
2H), 4.72 (dd, J1=12.0 Hz, J2=2.8 Hz, 1H), 4.94 (d, J=7.8 Hz,
1H), 5.42-5.55 (m, 1H), 5.72 (“t”, J=9.8 Hz,1H), 6.00 (“t”, J=
9.9 Hz, 1H), 6.60 (s, 1H), 6.72 (d, J=4.5 Hz, 1H), 6.85-6.98 (m,
1H), 7.22-7.62 (m, 13H), 7.78-8.07 (m, 8H). 13C NMR (75
MHz, CDCl3): δ 17.1, 17.3, 19.2, 23.1, 26.3, 29.7, 49.4, 51.4,
59.8, 62.6, 69.3, 72.1, 72.2, 72.4, 72.5, 101.0, 128.4, 128.5, 128.5,
128.6, 128.7, 128.8, 129.4, 129.7, 129.8, 133.4, 133.5, 133.6,
134.0, 165.2, 165.6, 165.7, 166.2, 171.6, 172.2, 172.7, 173.4.
Anal. Calcd for C47H50N4O14: C, 63.08; H, 5.63; N, 6.26.
Found: C, 63.18; H, 5.57; N, 6.22.
(22) (a) Pearlman, D. A.; Case, D. A.; Caldwell, J. W.; Ross, W. R.;
Cheatham, T. E. III; DeBolt, S.; Ferguson, D.; Seibel, G.; Kollman, P. A.
Comput. Phys. Commun. 1995, 91, 1–41. (b) Case, D. A.; Pearlman, D. A.;
Caldwell, J. W.; Cheatham, T. E., III; Ross, W. S.; Simmerling, C. L.; Darden, T.
A.; Merz, K. M.; Stanton, R. V.; Cheng, A. L.; Vincent, J. J.; Crowley, M.; Tsui, V.;
Radmer, R. J.; Duan, Y.; Pitera, J.; Massova, I.; Seibel, G. L.; Singh, U. C.; Weiner,
P. K.; Kollman, P. A. AMBER 6; University of California: San Francisco, CA,
1999.
(23) Cornell, W. D.; Cieplak, P.; Bayly, C. I.; Gould, I. R.; Merz, K. M.;
Ferguson, D. M.; Spellmeyer, D. C.; Fox, T.; Caldwell, J. W.; Kollman, P. A.
J. Am. Chem. Soc. 1995, 117, 5179–5197.
(24) Woods, R. J.; Dwek, R. A.; Edge, C. J.; Fraser-Reid, B. J. Phys.
Chem. 1995, 99, 3832–3846.
Synthesis of Compound 4. A solution of 14 (30 mg, 0.03 mmol)
in MeOH (5 mL) was treated with MeONa/MeOH (0.5M) to pH
9. After being stirred for 3 h at 25 ꢀC, the mixture was
neutralized with Dowex 50-X8, filtered, and concentrated.
Purification of the residue with an C18 reverse-phase sep-pak
J. Org. Chem. Vol. 74, No. 24, 2009 9313