Synthesis of a fluorine-substituted puromycin derivative for Bronsted studies of ribosomal-catalyzed peptide bond formation

Article abstract of DOI:10.1021/jo802611t

The mechanism by which the ribosome catalyzes peptide bond formation remains controversial. Here we describe the synthesis of a nucleoside that can be used in Bronsted experiments to assess the transition state of ribosome catalyzed peptide bond formation. This substrate is the nucleoside 3 -amino-3 -deoxy-3 -[(3 R)-3-fluoro-L-phenyl-alanyl]-N⁶,N ⁶-dimethyladenosine, which was prepared from (1R,2R)-2-amino-1- phenylpropane-1,3-diol. This substrate is active in peptide bond formation on the ribosome and is a useful probe for Bronsted analysis experiments on the ribosome.

Full text of DOI:10.1021/jo802611t

To explain this acceleration, several catalytic strategies have
been proposed such as general acid-base catalysis,1,4 substrate-
assisted catalysis,5 and catalysis derived solely from substrate
alignment.6 To distinguish these possibilities, additional chemi-
cal probes that facilitate mechanistic analysis are needed.
Synthesis of a Fluorine-Substituted Puromycin
Derivative for Brønsted Studies of
Ribosomal-Catalyzed Peptide Bond Formation
Kensuke Okuda,*,† Takashi Hirota,‡ David A. Kingery,§ and
Hideko Nagasawa†
Gifu Pharmaceutical UniVersity, Gifu 502-8585, Japan,
Faculty of Pharmaceutical Sciences, Okayama UniVersity,
Okayama 700-8530, Japan, and Department of Molecular
Biophysics and Biochemistry, Yale UniVersity, New HaVen,
Connecticut 06520-8114
okuda@gifu-pu.ac.jp
ReceiVed NoVember 25, 2008
FIGURE 1. The puromycin derivatives used as A-site substrates.
To determine the chemical and catalytic mechanism of the
ribosome, a detailed understanding of the transition state of
peptidyl transfer is necessary. Brønsted analysis is a direct
experimental technique that reports the charge state of the
transition state. The Brønsted coefficient is a linear free energy
relationship between the log of the reaction rate and the pKa of
a particular reactive group and reflects the change in charge on
that group between the ground state and the transition state.7
This information can then be used to determine the change in
bonding that occurs in the transition state, which helps to define
the reaction mechanism. We previously reported a limited
Brønsted analysis of the PT reaction on 50S ribosomes using
dinucleotide C-puromycin (C-Pmn) derivatives (Figure 1, parts
A and B) as A-site substrates and the P-site substrate CC-3′-
(biotinyl-ε-aminocaproyl-L-phenylalanyl) A.8 Using these small
molecules as substrates, this model reaction enabled us to
examine the rate difference due to the perturbed nucleophilic
pKa values of the C-Pmn derivatives. The rates between both
substrates did not vary significantly, leading to an initial estimate
of the Brønsted coefficient for the nucleophile (ꢀnuc) value to
be close to zero. These results suggest that the transition state
of the ribosomal PT reaction was fairly different from that of
the simple ester aminolysis reaction in solution where ꢀnuc values
are 0.8-0.9. However, two substrates alone are insufficient to
The mechanism by which the ribosome catalyzes peptide
bond formation remains controversial. Here we describe the
synthesis of a nucleoside that can be used in Brønsted
experiments to assess the transition state of ribosome
catalyzed peptide bond formation. This substrate is the
nucleoside 3′-amino-3′-deoxy-3′-[(3″R)-3-fluoro-L-phenyl-
alanyl]-N6,N6-dimethyladenosine, which was prepared from
(1R,2R)-2-amino-1-phenylpropane-1,3-diol. This substrate is
active in peptide bond formation on the ribosome and is a
useful probe for Brønsted analysis experiments on the
ribosome.
The ribosome is the ribonucleoprotein complex responsible
for protein synthesis in all living systems. The structure of the
50S ribosomal subunit, which was determined by X-ray
crystallography,1,2 revealed that RNA surrounds the peptidyl
transferase (PT) center. The ribosome catalyzes peptide bond
formation between two tRNA substrates, the aminoacyl tRNA
(A-site tRNA) and the peptidyl tRNA (P-site tRNA). These two
substrates bind to the ribosome in their respective active sites,
and then a nucleophilic substitution reaction of the R-amino
group of the A-site tRNA on to the ester-carbonyl linkage of
the P-site tRNA occurs.
define the ꢀnuc
.
Measuring ꢀnuc for ribosomal peptide bond formation requires
more A-site substrates with perturbed pKa values in the middle
of the pKa range of the available Pmn derivatives (<5.0 (B) to
7.0 (A)). The first candidate was a Pmn derivative with a
pentafluorophenyl moiety substituted for the phenyl moiety
(Figure 1C). This pentafluorophenyl derivative was prepared
The ribosome enhances the rate of peptide bond formation
by more than 106-fold compared to nonenzymatic reactions.3
(4) Muth, G. W.; Ortoleva-Donnelly, L.; Strobel, S. A. Science 2000, 289,
947–50.
(5) (a) Dorner, S.; Panuschka, C.; Schmid, W.; Barta, A. Nucleic Acids Res.
2003, 31, 6536–6542. (b) Weinger, J. S.; Parnell, K. M.; Dorner, S.; Green, R.;
Strobel, S. A. Nat. Struct. Mol. Biol. 2004, 11, 1101–1106. (c) Trobro, S.; Aqvist,
J. Proc. Natl. Acad. Sci. U.S.A. 2005, 102, 12395–400.
(6) Sievers, A.; Beringer, M.; Rodnina, M. V.; Wolfenden, R. Proc. Natl.
Acad. Sci. U.S.A. 2004, 101, 7897–7901.
† Gifu Pharmaceutical University.
‡ Okayama University.
§ Yale University.
(1) Nissen, P.; Hansen, J.; Ban, N.; Moore, P. B.; Steitz, T. A. Science 2000,
289, 920–30.
(2) Ban, N.; Nissen, P.; Hansen, J.; Moore, P. B.; Steitz, T. A. Science 2000,
289, 905–20.
(7) Jencks, W. P. Catalysis in Chemistry and Enzymology; McGraw-Hill:
New York, 1969.
(3) Katunin, V. I.; Muth, G. W.; Strobel, S. A.; Wintermeyer, W.; Rodnina,
M. V. Mol. Cell 2002, 10, 339–346.
(8) Okuda, K.; Seila, A. C.; Strobel, S. A. Biochemistry 2005, 44, 6675–
6684.
10.1021/jo802611t CCC: $40.75
Published on Web 02/20/2009
2009 American Chemical Society
J. Org. Chem. 2009, 74, 2609–2612 2609

SCHEME 1. Retrosynthesis for the Production of
Nonracemic 3F-L-Phe (1)
SCHEME 2. Preparation of Pmn(3-F-Phe) (12)
from the commercially available (S)-2-Fmoc-amino-3-(perfluo-
rophenyl)propanoic acid by an analogous method described
previously.8 The R-amino group pKa of this substrate is
perturbed to 5.9 and it bound to the ribosomal PT center, but it
was found to be inactive as a substrate (data not shown). The
reason for the inactivity of the pentafluorophenyl-Pmn derivative
is unknown. An alternative target is a monofluorine-substituted
L-phenylalanine analogue at the 3-position (Figure 1D). This is
expected to reduce the pKa of the amine, but by an amount less
than that of the disubstituted analogue. Here we describe the
synthesis of the mononucleoside and demonstrate it is an active
ribosomal substrate for peptide bond formation by the 70S
initiation complex assay, a standard ribosomal peptide bond
assay.3
The first synthetic goal was the synthesis of chiral 3-substi-
tuted L-phenylalanine (3F-L-Phe). An efficient asymmetric
synthesis of chiral 3F-D-Phe was already reported by Davis.9
Their stereodirecting group at the 2-position (R-position) was
(R)-(-)-p-toluenesulfinamide. Therefore, using (S)-(+)-p-tolu-
enesulfinamide should lead to a chiral 3F-L-Phe, but the reagents
are rather expensive. Furthermore, we felt that the easily
racemizable R-fluoro aldehyde preparation by Dess-Martin
oxidation was technically too demanding. There are two other
reported synthetic strategies for producing chiral 3F-L-Phe. One
used the enzymatic hydrolysis of the racemic precursor ester,
but was inefficient for asymmetric synthesis.10,11 The other
strategy used the Scho¨llkopf’s bis lactim ether strategy and
resulted in low overall yields and mixtures of products.12 The
infeasibility of these precedents required that we explore another
synthetic route to produce chiral 3F-L-Phe.
Our retrosynthetic strategy for the production of nonracemic
3F-L-Phe (1) is outlined in Scheme 1. Deprotection of the
hydroxyl group, oxidation of the alcohol to the carboxylic acid,
and deprotection of the amino group of the N,O-protected 3F-
L-phenylalaninol 2 will afford the amino acid 1. Compound 2
is obtained by diethylaminosulfur trifluoride (DAST) deoxo-
fluorination of 3, usually with stereoinversion.13 3 is derived
by N,O3-protection from the commercially available, nonracemic
(1R,2R)-2-amino-1-phenylpropane-1,3-diol (4).
equiv) solution in CH2Cl2 at -78 °C and then the solution was
warmed to rt for 18-19 h under N2 atmosphere. TBS, DPS,
and Tr groups were used to protect the hydroxyl group and the
amino group was protected with Boc, trifluoroacetyl, formyl,
and dibenzyl groups. All of these combinations resulted in
diastereomeric mixtures14-16 determined by H and 19F NMR.
1
Although some of the diastereomers were separable by careful
silica gel column chromatography purification, these procedures
were inefficient. Perhaps SN1 mechanism competed with the
SN2 mechanism16 due to benzyl position.
Finally, we found that tert-butyl (1R,2R)-3-(allyloxycarbo-
nyloxy)-1-hydroxy-1-phenylpropan-2-ylcarbamate (Scheme 2,
5) could be converted to a single diastereomer (37%).17 To
improve the relatively low yield, we added i-Pr2EtN (3 equiv)
to boost reactivity of DAST to give the product with slightly
better yield (54%) and some recovered starting material (18%).
Increasing the equivalents of DAST and i-Pr2EtN (each 3.5
equiv) gave the best result in 58% yield with some recovered
starting material (6%).
We screened the DAST deoxofluorination reaction of 3. A
solution of 3 in CH2Cl2 was added dropwise to the DAST (3
At first, it was assumed that this reaction proceeded via SN2
reaction, so the product should be erythro-(1S,2R). However,
the product was later confirmed to be the threo isomer judged
by the analytical data of 3F-L-Phe (9),9,10 so it must be threo-
(1R,2R) that has retention of its stereocenter (described later).
We speculate the mechanism of this stereochemistry is either
(1) participation of the carbonate protecting group via a 6-exo-
tet intermediate or (2) a SN1-like pathway. The first scenario is
plausible because only the carbonate protective group (Alloc)
(9) Davis, F. A.; Vaidyanathan, S.; Titus, D. D. J. Org. Chem. 1999, 64,
6931–6934.
(10) Tsushima, T.; Kawada, K.; Nishikawa, J.; Sato, T.; Tori, K.; Tsuji, T.;
Misaki, S. J. Org. Chem. 1984, 49, 1163–1169.
(11) There is some literature describing the synthesis of racemic 3-F-Phe:
(a) Shi, G. Q.; Zhao, Z. Y.; Zhang, X. B. J. Org. Chem. 1995, 60, 6608–6611.
(b) Kaneko, C.; Chiba, J.; Toyota, A.; Sato, M. Chem. Pharm. Bull. 1995, 43,
760–765. (c) O’Donnell, M. J.; Barney, C. L.; McCarthy, J. R. Tetrahedron
Lett. 1985, 26, 3067–3070, and references cited in this paper.
(12) Groth, U.; Scho¨llkopf, U. Synthesis 1983, 673–675.
(13) (a) Singh, R. P.; Shreeve, J. n. M. Synthesis 2002, 2561–2578. (b)
Demange, L.; Cluzeau, J.; Me´nez, A.; Dugave, C. Tetrahedron Lett. 2001, 42,
651–653. (c) Manthati, V. L.; Murthy, A. S. K.; Caijo, F.; Drouin, D.; Lesot, P.;
Gre´e, D.; Gre´e, R. Tetrahedron: Asymmetry 2006, 17, 2306–2310. (d) Shi, J.;
Du, J.; Ma, T.; Pankiewicz, K. W.; Patterson, S. E.; Tharnish, P. M.; McBrayer,
T. R.; Stuyver, L. J.; Otto, M. J.; Chu, C. K.; Schinazi, R. F.; Watanabe, K. A.
Bioorg. Med. Chem. 2005, 13, 1641–1652. (e) Yi-Lin, L.; Luthman, K.; Hacksell,
U. Tetrahedron Lett. 1992, 33, 4487–4490. (f) Sutherland, A.; Vederas, J. C.
Chem. Commun. 1999, 1739–1740.
(14) Pansare, S. V.; Vederas, J. C. J. Org. Chem. 1987, 52, 4804–4810.
(15) Hudlicky, M.; Merola, J. S. Tetrahedron Lett. 1990, 31, 7403–7406.
(16) Iacazio, G.; Re´glier, M. Tetrahedron: Asymmetry 2005, 16, 3633–3639.
(17) TLC analysis of the crude reaction mixture did not show any considerable
spot indicating possible diastereomer. Rf of spots on silica gel TLC (ethyl acetate/
n-hexane, 25:75): 0.38 (6), 0.15 (5), 0.02 (a byproduct), 0 (byproducts).
2610 J. Org. Chem. Vol. 74, No. 6, 2009

gave the threo diastereomer. Incorporating other protective
groups (TBS, DPS, and Tr) did not lead to similar results. If
the reaction does proceed by the first scenario a 6-membered
cyclic carbonate byproduct should be observed upon loss of
the allyl group. However, this byproduct was not detected during
the reaction. Alternatively, the second scenario is supported by
some of the R-fluoro phosphonate literature. While simple
secondary R-hydroxyphosphonates usually show stereoinversion
by DAST deoxofluorination reaction,18 this reaction proceeds
via SN1 pathways in R-hydroxy-R-phenylphosphonates.19 Like
the Shibuya study, shielding of one face of the carbocation
intermediate gives some diastereomeric excess. If our reaction
went via a SN1-like pathway, one might expect the erythro
diastereomer as a byproduct in the reaction mixture. We did
not detect such a spot during TLC analysis of the crude reaction
mixtures.17
After deprotection of the allyloxycarbonyl group by Pd
catalysis (94%), the oxidation reaction of Boc-amino alcohol
(7) to Boc-amino acid (8) was screened. Jones oxidation (3
equiv)20 gave a poor result (27%). Under Parikh-Doering
oxidation (5 equiv) condition, only starting material was
recovered.21 The RuCl3 (5.5 mol %)-NaIO4 (4 equiv) oxidation
system gave moderate results (79%).22 The TEMPO (1 mol
%)-n-Bu4NBr (0.5 mol %)–NaOCl (3.5 equiv) oxidation system
gave the best result (95%).23 Furthermore, the TEMPO-n-
Bu4NBr–NaOCl oxidation system only needed simple recrys-
tallization to be purified, in contrast to RuCl3-NaIO4 oxidation,
which required purification by silica gel column chromatography.
The Boc group of 8 was removed by HCl (99%) to give the
previously characterized compound, 3F-L-Phe (9). The most
distinguishable feature was the 14.0 Hz JH2-F for 3F-L-Phe (9),
which was in good accordance with the 14.3 Hz reported for
threo-(2S,3S)-3F-D-Phe, as opposed to the 27.4 Hz reported for
erythro-(2S,3R)-3F-D-Phe.9
reported by Robins when they prepared a related compound.26
Boc protection (8) was substituted with Fmoc protection (10)
(89% for 2 steps). Bromotris(pyrrolidino)phosphonium hexaflu-
orophosphate was used to couple 10 to puromycin aminonucleo-
side to give 11 (78%). Finally, deprotection of the Fmoc group
by tris(2-aminoethyl)amine gave the target compound (12) in a
quantitative yield.
The pKa value of the R-amino group of 12 was measured by
the titration method. As anticipated, the pKa of 12 is 5.6, which
fills in the gap between the 7.0 pKa of the Phe derivative and
<5.0 pKa of the 3′′,3′′-difluoroPhe derivative.
Finally, this pKa perturbed probe (12) was an active substrate
for the ribosomal peptidyl transferase reaction as measured by
the 70S initiation complex assay (see the Supporting Information
for details).3 KM of 12 was similar to KM of puromycin (2.9
mM vs. 2.5 mM), and reaction rates between 12 and puromycin
were no more than 3-fold different (37.5 s-1 vs. 12 s-1). These
kinetic data were included in the determination of the Brønsted
coefficient for the nucleophile of ribosomal peptidyl trans-
ferase.27
In conclusion, an efficient synthesis of a monofluorinated
puromycin analogue was accomplished and the product was
active as a ribosomal substrate in peptide bond formation. This
analogue provided essential data in the middle of the pKa range
used for the Brønsted analysis that was unattainable by any other
puromycin derivative. In the process of producing the monof-
luorinated puromycin analogue, a diastereoselective synthesis
of 3-fluoroPhe was accomplished in 6 steps.
A protected threo-1-substituted 2-amino-1,3-propanediol as
well as Ph group should be available via addition of the
appropriate organometallics to a Garner aldehyde with chelation
control.28 Therefore, this deoxofluorination methodology of
protected threo 3-hydroxy-L-phenylalaninol with stereoretention
would pave a general procedure for syntheses of threo selective,
chiral 3-fluorinated amino acids.
To confirm its enantio purity, we converted 8 to (1R)-R-
methylbenzylamide.24 Standard DCC coupling failed, but the
acyl fluoride method produced (2R,3R)-2-[(tert-butoxycarbo-
nyl)amino]-3-fluoro-3-phenyl-N-[(1R)-1-phenylethyl]propana-
mide 8′ in good yield (70%).25 Analogously, we prepared
(2S,3S)-8 (ent-8) starting from (1S,2S)-4 (ent-4) to give di-
asetero-8′. We confirmed that the newly introduced chiral centers
Experimental Section
tert-Butyl (1R,2R)-3-(Allyloxycarbonyloxy)-1-hydroxy-1-phe-
nylpropan-2-ylcarbamate (5). To a cooled solution (0 °C) of
(1R,2R)-2-amino-1-phenylpropane-1,3-diol (4, 2.51 g, 15.0 mmol)
in methanol (15 mL) was added Boc2O (3.77 mL, 15.9 mmol)
dropwise over 1 min, then the mixture was stirred at rt for 13 h.
The resulting yellow solution was evaporated in vacuo to give an
orange viscous oil. Silica gel chromatography (ethyl acetate/n-
hexane, 1:1) gave N-Boc-4 as a colorless oil. To the cooled solution
(0 °C) of N-Boc-4 and pyridine (4.3 mL, 53.3 mmol) in CH2Cl2
(200 mL) was added allyl chloroformate (3.97 mL, 37.5 mmol)
dropwise over 1 min, then the mixture was stirred at rt for 20 min.
Water was added and then extracted. The aqueous phase was further
extracted with CH2Cl2. The combined organic phase was washed
with saturated brine, re-extracted with CH2Cl2, dried over Na2SO4,
and then evaporated in vacuo. The resulting yellow oily residue
was repeatedly coevaporated with toluene, followed by silica gel
column chromatography (ethyl acetate/n-hexane, 12:88 to 100:0)
to give 5 as a white waxy solid (3.74 g, 71%) with recovered
N-Boc-4 (19%).
1
provided different sets of signals for each diastereomer in H
NMR which showed their ee were >99% (see the Supporting
Information for details).
Next this amino acid was coupled to puromycin amino-
nucleoside. 8 was coupled by the acyl fluoride method, but the
Boc group deprotection gave disappointing results likely due
to the acid cleavage conditions. This is similar to the problems
(18) Berkowitz, D. B.; Bose, M.; Pfannenstiel, T. J.; Doukov, T. J. Org.
Chem. 2000, 65, 4498–4508.
(19) Yokomatsu, T.; Yamagishi, T.; Matsumoto, K.; Shibuya, S. Tetrahedron
1996, 52, 11725–11738.
(20) Yonezawa, Y.; Shimizu, K.; Yoon, K.-S.; Shin, C.-G. Synthesis 2000,
634–636.
(21) Okamoto, N.; Hara, O.; Makino, K.; Hamada, Y. J. Org. Chem. 2002,
67, 9210–9215.
(22) Hallinan, E. A.; Kramer, S. W.; Houdek, S. C.; Moore, W. M.; Jerome,
G. M.; Spangler, D. P.; Stevens, A. M.; Shieh, H. S.; Manning, P. T.; Pitzele,
B. S. Org. Biomol. Chem. 2003, 1, 3527–3534.
(23) Osada, S.; Ishimaru, T.; Kawasaki, H.; Kodama, H. Heterocycles 2006,
67, 421–431.
(26) Robins, M. J.; Miles, R. W.; Samano, M. C.; Kaspar, R. L. J. Org.
Chem. 2001, 66, 8204–8210.
(27) Kingery, D. A.; Pfund, E.; Voorhees, R. M.; Okuda, K.; Wohlgemuth,
I.; Kitchen, D. E.; Rodnina, M. V.; Strobel, S. A. Chem. Biol. 2008, 15, 493–
500.
(24) Pelagatti, P.; Carcelli, M.; Calbiani, F.; Cassi, C.; Elviri, L.; Pelizzi,
C.; Rizzotti, U.; Rogolino, D. Organometallics 2005, 24, 5836–5844.
(25) Tong, Y.; Fobian, Y. M.; Wu, M.; Boyd, N. D.; Moeller, K. D. J. Org.
Chem. 2000, 65, 2484–93.
(28) (a) Van Overmeire, I.; Boldin, S. A.; Dumont, F.; Van Calenbergh, S.;
Slegers, G.; De Keukeleire, D.; Futerman, A. H.; Herdewijn, P. J. Med. Chem.
1999, 42, 2697–2705. (b) Nishida, A.; Sorimachi, H.; Iwaida, M.; Matsumizu,
M.; Kawate, T.; Nakagawa, M. Synlett 1998, 389–390.
J. Org. Chem. Vol. 74, No. 6, 2009 2611

tert-Butyl (1R,2R)-3-(Allyloxycarbonyloxy)-1-fluoro-1-phe-
nylpropan-2-ylcarbamate (6). A solution of 5 (1.23 g, 3.50 mmol)
in anhydrous CH2Cl2 (35 mL) and i-Pr2EtN (2.13 mL, 12.2 mmol)
was added dropwise to a cooled solution (-78 °C) of DAST (1.62
mL, 12.3 mmol) in anhydrous CH2Cl2 (35 mL) under N2 atmosphere
over 1 h. The reaction mixture was then warmed to rt over 18 h.
Water was added and then extracted. The aqueous phase was further
extracted with CH2Cl2. The combined organic phase was washed
with saturated brine, dried over Na2SO4, and then evaporated in
vacuo to give a dark reddish brown oil. This oily residue was
purified by silica gel column chromatography (ethyl acetate/n-
hexane, 1:19 to 1:4) to give 6 as a white solid (711.3 mg, 58%)
with 5 also recovered (5%).
tert-Butyl (1R,2R)-1-Fluoro-3-hydroxy-1-phenylpropan-2-yl-
carbamate (7). To a solution of 6 (713.7 mg, 2.02 mmol) in THF
(30 mL) was added Pd(Ph3P)4 (94.5 mg, 0.081 mmol) and
morpholine (0.27 mL, 3.09 mmol) successively, then the mixture
was stirred at rt for 1 h under N2 atmosphere. Saturated brine was
added, and then extracted with ethyl acetate. The combined organic
phase was dried over Na2SO4, and then evaporated in vacuo to give
a light orange semisolid. This residue was purified by silica gel
column chromatography (ethyl acetate/n-hexane, 1:4) to give 7 as
a white solid (512.6 mg, 94%).
(2R,3R)-2-(tert-Butoxycarbonylamino)-3-fluoro-3-phenylpro-
panoic Acid ((3R)-N-Boc-3-fluoro-L-Phe) (8). 7 (512.6 mg, 1.90
mmol) was dissolved in ethyl acetate (4 mL), then water (2 mL),
TEMPO (2.9 mg), (n-Bu)4NBr (2.9 mg), and NaHCO3 (0.60 g)
were added successively. NaOCl aq (10%, 5 mL) was added
dropwise to this mixture (ice bath) with vigorous stirring, and then
the solution was stirred for 30 min at 0 °C. Ethyl acetate was added,
and then extracted with saturated NaHCO3 aq. To the combined
aqueous phase was added NaHSO3 (0.80 g) slowly, followed by
dropwise addition of 10% HCl aq to pH ca. 3. This cloudy mixture
was extracted by ethyl acetate, washed with saturated brine, dried
over Na2SO4, then evaporated in vacuo. The white solid residue
was recrystallized from cyclohexane to give 8 as a colorless feather
crystal (513.9 mg, 95%).
(3R)-3-Fluoro-L-Phe (9). To a solution of 8 (425.1 mg. 1.50
mmol) in 1,4-dioxane (36 mL) was added 12% HCl aq (12 mL)
dropwise, then the mixture was stirred at rt for 3.5 h. After
evaporation in vacuo at 30 °C, excess water was removed through
repeated coevaporation with 1,4-dioxane. The resulting white solid
was added to 2-propanol (24 mL) and then propylene oxide (0.6
mL) was added dropwise with stirring at rt for 3.5 h to give a white
precipitate. After evaporation in vacuo, the resulting white powder
was filtrated and washed with 2-propanol repeatedly to give 9 as a
white powder (271.7 mg, 99%).
(3R)-N-Fmoc-3-fluoro-L-Phe (10). To a solution of Fmoc-OSu
(608.1 mg, 1.77 mmol) in 1,2-dimethoxyethane (6 mL) was added
9 (220.1 mg, 1.20 mmol), then the mixture was stirred at rt for
12 h. The resulting white precipitate was filtrated and washed
successively with 1,4-dioxane. The filtrate was slightly acidified
by dropwise addition of 2 N HCl aq to pH 2-3, followed by ethyl
acetate extraction. The collected organic layer was washed with
saturated brine, re-extracted by ethyl acetate, dried over Na2SO4,
and then evaporated in vauo. The resulting oily residue was purified
by silica gel column chromatography (methanol/CH2Cl2, 0:100 to
1:99) to give a white solid residue, which was washed successively
by cyclohexane to give pure 10 as a white powder (437.2 mg, 90%).
3′-Amino-3′-deoxy-3′-[(3′′R)-N-Fmoc-3″-fluoro-L-phenylala-
nyl]-N6,N6-dimethyladenosine (11). Puromycin aminonucleoside
(30.6 mg, 0.104 mmol) was dried by repeated coevaporation with
pyridine, then solvated in anhydrous CH2Cl2 (1.0 mL). To this
mixture was added 10 (44.5 mg, 0.110 mmol), PyBroP (51.2 mg,
0.107 mmol), i-Pr2EtN (0.06 mL, 0.344 mmol), and DMF (1.0 mL)
successively at 0 °C. The solution was stirred at rt for 1.5 h. After
evaporation in vacuo, the oily residue was purified by silica gel
column chromatography (methanol/CH2Cl2, 0:100 to 2:98) to give
a white solid powder, which was washed successively by water
followed by methanol to give pure 11 as a white powder (55.2 mg,
78%).
3′-Amino-3′-deoxy-3′-[(3′′R)-3′′-fluoro-L-phenylalanyl]-N6,N6-
dimethyladenosine (12). 11 (195.0 mg, 0.286 mmol) was dissolved
in pyridine (15 mL) and tris(2-aminoethyl)amine (0.46 mL, 3.07
mmol) was added, then the solution was stirred at rt for 4 h. The
resulting cloudy reaction mixture was evaporated in vacuo and
coevaporated with toluene. The oily residue was purified by silica
gel column chromatography (methanol/CH2Cl2, 4:96 to 6:94) to
give a white solid, which was recrystallized from CHCl3-n-hexane
to give pure 12 as a white powder (131.9 mg, 100%).
Acknowledgment. We would like to thank Prof. Scott A.
Strobel (Yale University) for comments on the manuscript.
Funding for this work was provided partly by a Ryobi Teien
Memory Foundation grant to K.O. We are grateful to the SC-
NMR Laboratory of Okayama University for NMR experiments.
Supporting Information Available: Experimental proce-
1
dures for 8′, characterization, H and 13C NMR spectra of all
compounds, and details for the 70S ribosome initiation complex
assay. This material is available free of charge via the Internet
at http://pubs.acs.org.
JO802611T
2612 J. Org. Chem. Vol. 74, No. 6, 2009