Synthesis of N-Fmoc-O-(N′-boc-N′-methyl)-aminohomoserine, an amino acid for the facile preparation of neoglycopeptides

Article abstract of DOI:10.1021/jo026641p

The synthesis of N-Fmoc-O-(N′-Boc-N′-methyl)- aminohomoserine in 35% overall yield from L-homoserine is described. This amino acid can be efficiently incorporated into peptides using Fmoc-chemistry-based solid-phase peptide synthesis, and the resulting peptides can be chemo-selectively glycosylated at the aminooxy side chains to generate neoglycopeptides. The synthesis of this derivative greatly expands the availability of a previously developed neoglycopeptide synthesis strategy.

Full text of DOI:10.1021/jo026641p

SCHEME 1. Gen er a l Neoglycop ep tid e Syn th esis
Syn th esis of N-F m oc-O-
(N ′-Boc-N ′-m eth yl)-a m in oh om oser in e, a n
Am in o Acid for th e F a cile P r ep a r a tion of
Neoglycop ep tid es†
Michael R. Carrasco,* Ryan T. Brown,
Iana M. Serafimova, and Oscar Silva
Department of Chemistry, Santa Clara University,
Santa Clara, California 95053-0270
mcarrasco@scu.edu
Received October 31, 2002
Abstr a ct: The synthesis of N-Fmoc-O-(N ′-Boc-N ′-methyl)-
aminohomoserine in 35% overall yield from L-homoserine
is described. This amino acid can be efficiently incorporated
into peptides using Fmoc-chemistry-based solid-phase pep-
tide synthesis, and the resulting peptides can be chemo-
selectively glycosylated at the aminooxy side chains to
generate neoglycopeptides. The synthesis of this derivative
greatly expands the availability of a previously developed
neoglycopeptide synthesis strategy.
acid 1b in protected form suitable for Fmoc-chemistry
based SPPS to make our approach available to a broader
community. Two routes were considered for making 1b
based on our synthesis of 1a , which proceeded in a few,
mild, and high-yielding steps. Initial protection of ho-
moserine with the Fmoc group was considered and
pursued as the most expedient route to 1b; however,
several issues were envisioned as potential drawbacks
to this approach. For one, the reported yield for N-Fmoc-
homoserine was low (49%).5 Additionally, N-Fmoc amino
acid derivatives often exhibit poor solubility. Most im-
portantly, the Fmoc carbamate might prove unstable to
the basic conditions of the hydroxylamine oxyanion
substitution reaction used for side chain elaboration. As
a result, initial protection of homoserine with the allyl-
oxycarbonyl (Alloc) group and its conversion to an Fmoc
group in the last step of the synthesis was pursued as
an alternative route to avoid such difficulties.
The chemoselective reaction of aminooxy-derivatized
peptides with reducing sugars has proven an attractive
method for the site-specific glycosylation of peptides.1-3
This general strategy for neoglycopeptide synthesis al-
lows the coupling of a given peptide with a variety of
sugars and therefore facilitates the formation of combi-
natorial neoglycopeptide arrays (Scheme 1). Recently, we2
and others3 have developed aminooxy amino acid deriva-
tives designed to create neoglycopeptides that are much
closer to their natural counterparts. The key features of
these amino acids are that they contain short side chains
and incorporate N-alkylation of the aminooxy moiety to
ensure cyclic sugar conformations. Our previously re-
ported derivative, 1a , has been efficiently incorporated
into peptides using Boc-chemistry based solid-phase
peptide synthesis (SPPS) procedures, and the resulting
peptides were efficiently glycosylated.
Because an estimated 60% of all peptide research
laboratories employ 9-fluorenylmethoxycarbonyl (Fmoc)
based SPPS,4 we have now synthesized aminooxy amino
† In memory of Professor Henry Rapoport.
(1) (a) Cervigni, S. E.; Dumy, P.; Mutter, M. Angew. Chem., Int. Ed.
Engl. 1996, 35, 1230-1232. (b) Zhao, Y.; Kent, S. B. H.; Chait, B. T.
Proc. Natl. Acad. Sci. U.S.A. 1997, 94, 1629-1633. (c) Peri, F.; Dumy,
P.; Mutter, M. Tetrahedron 1998, 54, 12269-12278. (d) Lang, I.; Donze´,
N.; Garrouste, P.; Dumy, P.; Mutter, M. J . Peptide Sci. 1998, 4, 72-
80. (e) Spetzler, J . C.; Hoeg-J ensen, T. J . Peptide Sci. 1999, 5, 582-
592. (f) Peluso, S.; Imperiali, B. Tetrahedron Lett. 2001, 42, 2085-
2087.
(2) Carrasco, M. R.; Nguyen, M. J .; Burnell, D. R.; MacLaren, M.
D.; Hengel, S. M. Tetrahedron Lett. 2002, 43, 5727-5729.
(3) (a) Peri, F.;Venturini, S.; Nicotra, F. Presented at the 27th
European Peptide Symposium, Sorrento, Italy, August 31-September
7, 2002: J . Peptide Sci. 2002, 8, S91. (b) Filira, F.; Biondi, L.; Gobbo,
M.; Giannini, E.; Goldin, D.; Rocchi, R. Presented at the 27th European
Peptide Symposium, Sorrento, Italy, August 31-September 7, 2002:
J . Peptide Sci. 2002, 8, S121.
In the direct approach, L-homoserine was converted to
its N-Fmoc, allyl ester derivative 2a by treatment with
Fmoc-hydroxy succinimide (Fmoc-OSu) and NaHCO3
followed by allyl bromide (Scheme 2). Because 2a could
not be conveniently purified from the O-allylhydroxysuc-
cinimide byproduct, the mixture was reacted directly with
MsCl in CH2Cl2 followed by LiBr in acetone to produce
bromide 3a in 37% overall yield from L-homoserine. The
next step was the critical nucleophilic displacement of
the bromide with the anion of tert-butyl N-methyl-N-
hydroxycarbamate.6 In the event, our fears of the insta-
bility of the Fmoc carbamate were realized. The reaction
proved unsuccessful, and from the complex product
(4) Sheppard, R. C. Presented at the 27th European Peptide
Symposium, Sorrento, Italy, August 31-September 7, 2002: J . Peptide
Sci. 2002, 8, S65.
10.1021/jo026641p CCC: $25.00 © 2003 American Chemical Society
Published on Web 11/21/2002
J . Org. Chem. 2003, 68, 195-197
195

SCHEME 2
might be complexed as the pyrrolidinium salt of the
carboxylic acid, our solution was to treat the reaction
mixture with 100 mol % of NaHCO3 to ensure the free-
base form of all the amines, remove the solvents, and heat
the residue under vacuum at 50 °C overnight. This
procedure effectively removed the remaining pyrrolidine
and the majority of the N-allylpyrrolidine. The remaining
residue was treated directly with Fmoc-OSu to provide
78% of 1b. The overall yield of 1b from L-homoserine was
35%.
We then verified the efficacy of 1b by using it to
synthesize the peptide H2N-FKAZSK-NH2, where Z )
O-(N-methyl)-aminohomoserine, by standard Fmoc-chem-
istry-based SPPS.10 We chose this model sequence be-
cause it had been synthesized by Boc-chemistry-based
SPPS previously and because its glycosylated derivative
is easily separable from the parent peptide.2 After the
peptide was deprotected and cleaved from the resin,
HPLC analysis of the crude peptide product showed a
single major peak (>90%), which corresponded to the
expected mass by electrospray ionization MS. A purified
sample of the peptide was treated with D-glucose (0.1 M
NaOAc, pH 4.0, 45 °C) to verify the availability of the
N-methyl-aminooxy side chain. Glycosylation proceeded
in a manner consistent with that reported for the peptide
synthesized by Boc-chemistry-based SPPS,2 and the
peptide underwent 72% conversion to its glycosylated
derivative in 4 h without the formation of any side
products. Importantly, the retention times for the parent
and glycosylated peptides made by the Fmoc-chemistry-
based SPPS were identical with those made previously
by Boc-chemistry-based SPPS.
We have demonstrated a practical synthesis of N-
Fmoc-O-(N ′-Boc-N ′-methyl)-aminohomoserine from L-
homoserine. This amino acid derivative is efficiently
incorporated into peptides using Fmoc-chemistry-based
SPPS, and the resulting peptides can be glycosylated
chemoselectively at the N-methyl-aminooxy side chains.
Because a growing majority of peptide syntheses are
performed using Fmoc-based chemistry, this new deriva-
tive greatly expands the ability of others to take advan-
tage of our method for the synthesis of neoglycopeptides.
mixture a substantial amount of dibenzofulvene from
elimination of the Fmoc group was recovered.
By contrast, the indirect, Alloc approach proved suc-
cessful (Scheme 2). Treatment of L-homoserine with allyl
chloroformate and NaHCO3 followed by allyl bromide
yielded the N-Alloc, allyl ester 2b. Despite experimenting
with a variety of reaction conditions, we were unable to
produce 2b without also forming 5-8% of N-Alloc-
homoserine lactone. This byproduct proved inseparable
from 2b in all the chromatography conditions attempted.
Because all other byproducts could be removed by simple
extractive workup procedures, we avoided chromatogra-
phy at this stage and proceeded with the impure 2b. The
2b/lactone mixture was reacted with MsCl in CH2Cl2
followed by LiBr in acetone to produce pure bromide 3b
in 56% overall yield from L-homoserine. At this point the
lactone, which was unreactive in the bromination condi-
tions, was easily separated by chromatography from 3b
and characterized by NMR. Unlike with 3a , nucleophilic
displacement of the bromine of 3b with the anion
(generated by NaH) of tert-butyl N-methyl-N-hydroxy-
carbamate proceeded cleanly to afford 4 in 81% yield. At
this point, what remained was to convert the Alloc to
Fmoc and deprotect the carboxylic acid. We removed the
Alloc and the allyl ester simultaneously with catalytic
Pd(PPh3)4. After trying a variety of nonbasic allyl scav-
engers, including dimedone7 and formic acid,8 we found
pyrrolidine9 to be the only additive that effected the
deprotection cleanly and rapidly. Cognizant that residual
pyrrolidine would stymie our subsequent Fmoc protec-
tion, we developed a method to ensure its removal
without using chromatography or aqueous extractions,
both of which would be extremely difficult with the
deprotected intermediate. Because residual pyrrolidine
Exp er im en ta l Section
Chromatographic separations were performed using silica gel
(230-400 mesh). Organic solutions were dried with Na2SO4, and
solvents were removed using standard rotary evaporation under
reduced pressure. Products were dried under high vacuum.
Commercial reagents were used in all cases without further
purification. Spectral characterizations of 1b were performed at
elevated temperatures because of the presence of distinct
rotational isomers under most conditions.
SPPS was performed using a Rink amide resin using standard
Fmoc-chemistry-based procedures10 with the following varia-
tions. The amino acids were activated with 2-(1H-benzotriazole-
1-yl)-1,1,3,3-tetramethyluronium hexafluorophosphate (HBTU)
and NEt(i-Pr)2 in DMF. All commercial amino acids were used
in a 10-fold excess to the resin loading and allowed to react for
10 min. 1b was used in a 2.5-fold excess and allowed to react
for 25 min. Fmoc deprotections were carried out using 20%
piperidine in DMF. Resin washings between steps were per-
(5) St. Hilaire, P. M.; Cipolla, L.; Franco, A.; Tedebark, U.; Tilly, D.
A.; Meldal, M. J . Chem. Soc., Perkin Trans. 1 1999, 3559-3564.
(6) (a) House, H. O.; Richey, F. A., J r. J . Org. Chem. 1969, 34, 1430-
1439. (b) Freeman, J . P.; Lillwitz, L. D. J . Org. Chem. 1970, 35, 3107-
3110. (c) Baldwin, J . E.; Adlington, R. M.; Birch, D. J . Tetrahedron
Lett. 1985, 26, 5931-5934.
(7) Kunz, H.; Unverzagt, C. Angew. Chem., Int. Ed. Engl. 1984, 23,
436-437.
formed with
a continuous flow of DMF for 1 min. Final
(8) Minami, I.; Ohashi, Y.; Shimizu, I.; Tsuji, J . Tetrahedron Lett.
1985, 26, 2449-2452.
(10) Fmoc solid-phase peptide synthesis: a practical approach; Chan,
W. C., White, P. D., Eds.; Oxford University Press: New York, 2000.
(9) Deziel, R. Tetrahedron Lett. 1987, 28, 4371-4372.
196 J . Org. Chem., Vol. 68, No. 1, 2003

deprotection and resin cleavage was accomplished by treatment
with 95:2.5:2.5 TFA:H2O:triisopropylsilane.
O-[N-ter t-Bu toxyca r bon yl-N-m eth yl]-a m in o-N-(9-flu or -
en ylm eth oxyca r bon yl)-h om oser in e (1b). A solution of 4 (464
mg, 1.25 mmol), PPh3 (16 mg, 0.063 mmol), Pd(PPh3)4 (3.6 mg,
0.031 mmol), and pyrrolidine (313 µL, 3.75 mmol) in CH2Cl2 was
stirred for 40 min. The solvent was evaporated, and the residue
was redissolved in CH3CN (10 mL). A solution of NaHCO3 (105
mg, 1.25 mmol) in H2O (5 mL) was added and the mixture was
stirred for 30 min. The solvents were evaporated, and the
resulting mixture was dried overnight under vacuum at 50 °C.
The resulting solids were dissolved in DMF (30 mL) and H2O
(30 mL), treated with NaHCO3 (210 mg, 2.5 mmol) and Fmoc-
OSu (464 mg, 1.37 mmol), and stirred for 24 h. The solvents
were removed, and the residue was dissolved in EtOAc (150 mL)
and washed with 0.1 M KHSO4 (4 × 50 mL), H2O (4 × 50 mL),
and brine (100 mL). After drying and removal of the solvent,
the residue was chromatographed (acetone:CH2Cl2:AcOH, 5:95:
0.5 to 10:90:0.5) and then purified by size exclusion chromatog-
raphy (LH-20, CH2Cl2) to yield 1b (456 mg, 0.969 mmol, 78%)
as a glassy solid: 1H NMR (CD3CN, 55 °C, 400 MHz) δ 7.80 (d,
J ) 8.5 Hz, 2H), 7.65 (d, J ) 7.3 Hz, 2H), 7.39 (t, J ) 7.5 Hz,
2H), 7.30 (dt, J ) 1.2, 7.5 Hz, 2H), 6.44 (br s, 1H), 4.34 (m, 3H),
4.21 (t, J ) 6.8 Hz, 1H), 3.88 (m, 2H), 3.03 (s, 3H), 2.07 (m, 1H),
1.94 (m, 1H), 1.44 (s, 9H); 13C NMR (CD3OD, 55 °C, 100 MHz)
δ 175.5, 158.6, 145.4, 145.3, 142.6, 128.8, 128.2, 126.3, 121.0,
82.9, 71.7, 68.1, 53.0, 48.5, 36.9, 31.3, 28.7. Anal. Calcd for
C25H30N2O7: C, 63.82; H, 6.43; N, 5.95. Found: C, 63.46; H, 6.65;
N, 6.17.
Allyl 2-(N-Allyloxyca r bon yl)-a m in o-4-br om obu ta n oa te
(3b). L-homoserine (1.01 g, 8.49 mmol) and Na2CO3 (899 mg,
8.49 mmol) were dissolved in H2O (10 mL) and CH3CN (5 mL)
and treated with allyl chloroformate (897 µL, 8.49 mmol). The
mixture was stirred for 21 h, and the solvents were removed.
The resulting residue was dissolved in DMF (20 mL), treated
with allyl bromide (807 µL, 9.34 mmol) and NaHCO3 (713 mg,
8.49 mmol), and stirred for 65 h. The solvent was removed, and
the residue was dissolved in EtOAc. This organic layer was
washed with H2O (50 mL), sat. NaHCO3 (3 × 50 mL), H2O (50
mL), 0.1 M KHSO4 (2 × 50 mL), and brine (50 mL) and then
dried. Removal of the solvent left a residue that corresponded
to >90% pure 2b by 1H NMR.
The residue was dissolved in CH2Cl2 (30 mL), treated with
MsCl (514 µL, 6.64 mmol) and NEt3 (998 µL, 7.20 mmol), and
stirred for 45 min. LiBr (4.81 g, 55.4 mmol) and acetone (30 mL)
were added, and the mixture was stirred for an additional 17 h.
The solvents were removed, and the residue was redissolved in
EtOAc (125 mL) and washed with H2O (2 × 75 mL), sat.
NaHCO3 (75 mL), and brine (75 mL). After drying and removal
of the solvent, the residue was chromatographed (EtOAc:
hexanes, 20:80) to yield 3b (1.47 g, 4.79 mmol, 56%) as a clear
oil: 1H NMR (CDCl3, 400 MHz) δ 5.91 (m, 2H), 5.51 (d, J ) 7.8
Hz, 1H), 5.38-5.21 (m, 4H), 4.66 (dd, J ) 1.1, 5.9 Hz, 2H), 4.59
(d, J ) 5.5 Hz, 2H), 4.52 (m, 1H), 3.45 (t, J ) 7.1 Hz, 2H), 2.45
(m, 1H), 2.26 (m, 1H); 13C NMR (CDCl3, 100 MHz) δ 171.4, 156.0,
132.6, 131.4, 119.3, 118.1, 66.5, 66.1, 53.0, 35.6, 28.3. Anal. Calcd
for C11H16BrNO4: C, 43.15; H, 5.27; N, 4.58. Found: C, 43.06;
H, 5.43; N, 4.52.
Allyl 2-(N-Allyloxyca r bon yl)-a m in o-4-(O-[N-ter t-bu toxy-
ca r b on yl-N-m et h yl]-a m in o)-h yd r oxyb u t a n oa t e (4). tert-
Butyl N-methyl-N-hydroxycarbamate (904 mg, 6.15 mmol) was
dissolved in anhydrous DMF (5 mL), treated with NaH (236 mg,
60% dispersion in mineral oil, 5.90 mmol), and stirred for 1 h
under N2. The mixture was cooled to 0 °C, treated with a solution
of 3b (1.51 g, 4.92 mmol) in DMF (10 mL), and stirred at 0 °C
for an additional 3 h. The solvents were removed, and the residue
was dissolved in EtOAc (150 mL) and poured into a separatory
funnel. The organic layer was washed with 0.1 M NaOH (5 ×
50 mL), H2O (50 mL), 0.1 M KHSO4 (2 × 50 mL), and brine (50
mL) and then dried. After removal of the solvent, the residue
was chromatographed (EtOAc:hexanes, 30:70) to yield 4 (1.48
g, 3.98 mmol, 81%) as a clear oil: 1H NMR (CDCl3, 400 MHz) δ
6.29 (d, J ) 7.6 Hz, 1H), 5.90 (m, 2H), 5.32 (m, 2H), 5.21 (dd, J
) 10.4, 17.3 Hz, 2H), 4.64 (d, J ) 5.7 Hz, 2H), 4.57 (d, J ) 5.3
Hz, 2H), 4.51 (m, 1H), 3.95 (m, 2H), 3.08 (s, 3H), 2.18 (m, 1H),
2.09 (m, 1H), 1.49 (s, 9H); 13C NMR (CDCl3, 100 MHz) δ 171.6,
156.8, 156.0, 132.7, 131.6, 118.5, 117.3, 81.5, 70.2, 65.7, 65.5,
51.8, 36.4, 30.0, 28.1. Anal. Calcd for C17H28N2O7: C, 54.83; H,
7.58; N, 7.52. Found: C, 54.50; H, 7.74; N, 7.26.
Ack n ow led gm en t is made to the donors of the
Petroleum Research Fund, administered by the Ameri-
can Chemical Society, for partial support of this re-
search. We also gratefully acknowledge financial sup-
port from the Camille and Henry Dreyfus Faculty Start-
up Grant Program for Undergraduate Institutions and
the National Science Foundation (NSF-REU grant
CHE01-39527 and NSF-CAREER grant CHE01-34818).
This research was also supported by an award to Santa
Clara University under the Undergraduate Biological
Sciences Education Program of the Howard Hughes
Medical Institute.
Su p p or t in g In for m a t ion Ava ila b le: 1H NMR and 13C
NMR spectra for 2b, 3b, 4, and 1b; experimental procedure
and data for tert-butyl N-methyl-N-hydroxycarbamate. This
material is available free of charge via the Internet at
http://pubs.acs.org.
J O026641P
J . Org. Chem, Vol. 68, No. 1, 2003 197