Study of on-resin convergent synthesis of N-linked glycopeptides containing a large high mannose N-linked oligosaccharide

Article abstract of DOI:10.1021/ja9104073

Here we present a convergent on-resin glycosylamine coupling strategy for solid phase N-linked glycopeptide synthesis, and apply it to the synthesis of high mannose containing glycopeptides. In this strategy, the 2-phenylisopropyl protecting group is used as an orthogonal handle to create glycosylation sites on-resin after synthesis of nonglycosylated peptides. In addition to allowing selective deprotection of aspartic acid residues for creation of glycosylation sites, the 2-phenylisopropyl protecting group also efficiently suppresses aspartimide formation during peptide synthesis. The key step of on-resin glycosylamine coupling to an aspartic acid residue was first optimized for a small sugar, N-acetylglucosamine, and then applied to a much larger high mannose oligosaccharide, Man₈GlcNAc₂. Satisfying coupling yields were obtained for both small and large sugars. The use of on-resin glycosylamine coupling simplifies purification of N-linked glycopeptides, and also allows convenient recovery of unreacted valuable large oligosaccharides. This approach was applied to the solid phase synthesis of glycosylated forms of the 34 amino acid HIV-1 gp41 C34 glycopeptide, which is an HIV-1 entry inhibitor. The HIV-1 entry inhibition assay of synthesized glycopeptides showed the retention of bioactivity of high mannose Man₈GlcNAc₂-C34.

Full text of DOI:10.1021/ja9104073

Published on Web 02/16/2010
Study of On-Resin Convergent Synthesis of N-Linked
Glycopeptides Containing a Large High Mannose N-Linked
Oligosaccharide
Rui Chen and Thomas J. Tolbert*
Department of Chemistry, Indiana UniVersity, 800 East Kirkwood AVenue, Bloomington, Indiana 47405
Received December 9, 2009; E-mail: tolbert@indiana.edu
Abstract: Here we present a convergent on-resin glycosylamine coupling strategy for solid phase N-linked
glycopeptide synthesis, and apply it to the synthesis of high mannose containing glycopeptides. In this
strategy, the 2-phenylisopropyl protecting group is used as an orthogonal handle to create glycosylation
sites on-resin after synthesis of nonglycosylated peptides. In addition to allowing selective deprotection of
aspartic acid residues for creation of glycosylation sites, the 2-phenylisopropyl protecting group also efficiently
suppresses aspartimide formation during peptide synthesis. The key step of on-resin glycosylamine coupling
to an aspartic acid residue was first optimized for a small sugar, N-acetylglucosamine, and then applied to
a much larger high mannose oligosaccharide, Man8GlcNAc2. Satisfying coupling yields were obtained for
both small and large sugars. The use of on-resin glycosylamine coupling simplifies purification of N-linked
glycopeptides, and also allows convenient recovery of unreacted valuable large oligosaccharides. This
approach was applied to the solid phase synthesis of glycosylated forms of the 34 amino acid HIV-1 gp41
C34 glycopeptide, which is an HIV-1 entry inhibitor. The HIV-1 entry inhibition assay of synthesized
glycopeptides showed the retention of bioactivity of high mannose Man8GlcNAc2-C34.
Introduction
strategies that utilize Lansbury aspartylation to couple N-linked
glycosylamines6a to protected peptides in solution maximize the
Interest in understanding the specific effects of glycosylation
and in developing glycopeptide-based vaccines and therapeutics
stimulate the study of N-linked glycopeptide synthesis.1 How-
ever, it is still a challenge to synthesize glycopeptides containing
large N-linked oligosaccharides, not only because it is difficult
to obtain sufficient amounts of the N-linked oligosaccharides
themselves2 but also because reactions coupling bulky N-linked
oligosaccharides to peptides are often low yielding and prone
to aspartimide formation.3 Several strategies have been utilized
for N-linked glycopeptide synthesis with each strategy having
its own advantages and drawbacks.4 Solid phase synthesis of
N-linked glycopeptides using glycosylated asparagine (Asn)
building blocks takes advantage of the benefits of solid phase
peptide synthesis (SPPS), but is quite costly in terms of the
amounts of oligosaccharides necessary to first synthesize the
building blocks and then couple them to peptides.5 Convergent
yield of glycopeptides produced based on the amount of
N-linked oligosaccharide used.6b-d Enzymatic transglycosylation
can also be used to synthesize N-linked glycopeptides in
solution, although the preparation of the sugar oxazoline
substrates is required for appreciable transglycosylation yields.7
These in solution strategies allow N-linked oligosaccharides to
be introduced at a late stage, but at the expense of introducing
additional purification steps into the synthesis and eliminating
some of the advantages of SPPS.
To combine the advantages of SPPS and convergent synthesis,
researchers have utilized on-resin Lansbury aspartylation to
couple glycosylamines to peptides on the solid phase. This
strategy involves synthesizing a peptide by SPPS with orthogo-
nally protected aspartic acid residues that can be selectively
deprotected on-resin at the end of peptide synthesis. Subsequent
on-resin coupling of a glycosylamine to the Asp side chain and
resin cleavage results in the desired glycopeptide (Figure 1).
This on-resin Lansbury aspartylation strategy has only previ-
ously been applied to the synthesis of glycopeptides containing
(1) (a) Dove, A. Nat. Biotechnol. 2001, 19, 913–917. (b) Scanlan, C. N.;
Offer, J.; Zitzmann, N.; Dwek, R. A. Nature 2007, 446, 1038–1045.
(2) (a) Lee, J. C.; Greenberg, W. A.; Wong, C. H. Nat. Protoc. 2006, 1,
3143–3152. (b) Lis, H.; Sharon, N. J. Biol. Chem. 1978, 253, 3468–
3476. (c) Dudkin, V. Y.; Miller, J. S.; Danishefsky, S. J. J. Am. Chem.
Soc. 2004, 126, 736–738. (d) Demchenko, A. V. Lett. Org. Chem.
2005, 2, 580–589. (e) Chen, R.; Pawlicki, M. A.; Hamilton, B. S.;
Tolbert, T. J. AdV. Synth. & Catal. 2008, 350, 1689–1695. (f)
Seeberger, P. H. Chem. Soc. ReV. 2008, 37, 19–28.
(5) (a) Meinjohanns, E.; Meldal, M.; Paulsen, H.; Dwek, R. A.; Bock, K.
J. Chem. Soc.-Perkin Trans. 1 1998, 549–560. (b) Bejugam, M.;
Flitsch, S. L. Org. Lett. 2004, 6, 4001–4004. (c) Yamamoto, N.;
Takayanagi, Y.; Yoshino, A.; Sakakibara, T.; Kajihara, Y. Chem.sEur.
J. 2007, 13, 61325.
(3) (a) Nicolas, E.; Pedroso, E.; Giralt, E. Tetrahedron Lett. 1989, 30,
497–500. (b) Yang, Y.; Sweeney, W. V.; Schneider, K.; Thornqvist,
S.; Chait, B. T.; Tam, J. P. Tetrahedron Lett. 1994, 35, 9689–9692.
(b) Kan, C.; Trzupek, J. D.; Wu, B.; Wan, Q.; Chen, G.; Tan, Z.;
Yuan, Y.; Danishefsky, S. J. J. Am. Chem. Soc. 2009, 131, 5438–
5443. (c) Tan, Z.; Shang, S.; Halkina, T.; Yuan, Y.; Danishefsky, S. J.
J. Am. Chem. Soc. 2009, 131, 5424–5431.
(6) (a) Likhosherstov, L. M.; Novikova, O.; Derevetskaja, V. A.; Kochet-
kov, N. K. Carbohydr. Res. 1986, C1–5. (b) Cohen-Anisfeld, S. T.;
Lansbury, P. T. J. Am. Chem. Soc. 1993, 115, 10531–10537. (c)
Mandal, M.; Dudkin, V. Y.; Geng, X. D.; Danishefsky, S. Angew.
Chem., Int. Ed. 2004, 43, 2557–2561. (d) Kaneshiro, C. M.; Michael,
K. Angew. Chem., Int. Ed. 2006, 45, 1077–1081.
(4) Hojo, H.; Nakahara, Y. Biopolymers 2007, 88, 308–324.
(7) Wang, L. X. Carbohydr. Res. 2008, 343, 1509–1522.
9
10.1021/ja9104073  2010 American Chemical Society
J. AM. CHEM. SOC. 2010, 132, 3211–3216 3211

A R T I C L E S
Chen and Tolbert
and stained by vanillin (15 g vanillin in 250 mL ethanol and 2.5
1
mL concentrated sulfuric acid.). The 400 and 500 MHz H NMR
was carried out on a Varian Inova NMR Spectrometer. Masses of
peptides were determined either by a Bruker Autoflex III MALDI-
TOF or an API III ESI Mass spectrometer. HRESI-TOF-MS of
Man8GlcNAc2 was performed on a Waters/Micromass LCT Mass
spectrometer.
General Procedure for the Synthesis of Glycosylamines. A
total of 20-200 mg free sugar (N-acetyl-D-glucosamine or
Man8GlcNAc2) were dissolved in 25 mL saturated NH4HCO3 in
water and stirred for 5-6 days at room temperature. The reaction
solution was diluted with water and rotary evaporated to dryness.
This procedure was repeated several times to remove the residual
ammonium hydrogen carbonate. Then, the sample was dissolved
in water and lyophilized. The lyophilization was repeated until the
sample reached constant mass. 1H NMR was employed to monitor
the conversion of free sugars into their corresponding ꢀ-D-
glycosylamines.
General Procedure for Peptide Synthesis. Five equivalents of
Fmoc-amino acids, DEPBT, and 3 equiv. of DIEA were used for
manually loading of Rink Amide PEGA resin. Synthesis of fully
protected peptides was carried out following standard Fmoc/tBu
chemistry protocols on the Applied Biosystems 433A Peptide
Synthesizer. For the glycosylation site, Fmoc-Asp(O-2-PhiPr)-OH
or Fmoc-Asp(OAll)-OH was used. All of the other amino acids
side chains were protected by most commonly used protecting
groups, including tBu for Tyr, Ser, Thr, Glu, and Asp, Pbf
(2,2,4,6,7-pentamethyldihydrobenzofuran-5-sulfonyl) for Arg, trt
(trityl) for Gln, Asn and Cys, His, Boc for Trp. The Fmoc protecting
group was removed by 20% piperidine in DMF (10 min × 3). The
loading for the amino acid was determined by UV monitoring of
the Fmoc removal at 301 nm (ꢀ ) 7800 cm-1M-1).15 The capping
of unreacted amino groups was carried out using a large excess of
acetic anhydride with 3 equivalents of N-methylmorpholine in
DCM. The cleavage of resin was accomplished with neat TFA
containing 1% TIS scavenger for 3 h. The crude peptide was
precipitated and washed twice by diethyl ether anhydrous before
RP-HPLC purification.
Figure 1. N-linked glycopeptide synthesis using on-resin Lansbury
aspartylation. (a) synthetic strategy, (b) glycosylamines, and (c) glycopeptides.
monosaccharides and disaccharides, and can suffer from as-
partimide formation difficulties during peptide synthesis.8 Herein
we report studies of on-resin convergent synthesis of N-linked
glycopeptides containing a large high mannose oligosaccharide.
To optimize the coupling of the costly N-linked oligosaccharide
we utilized the aspartimide-prone peptide AKSANE, first
working out conditions on the monosaccharide N-acetylglu-
cosamine (glycosylamine (GlcNAcNH2, 1) and then applying
those conditions to a high mannose oligosaccharide glycosyl-
amine (Man8GlcNAc2NH2, 2).
Experimental Section
Materials and Equipment. All normally used amino acid
building blocks and coupling reagents 3-(Diethoxyphosphoryloxy)-
1,2,3-benzotriazin-4(3H)-one (DEPBT), O-(Benzotriazol-1-yl)-
N,N,N′,N′ tetramethyluronium Hexafluorophosphate (HBTU), and
1-hydroxybenzotriazole (HOBt) for solid phase peptide synthesis
were purchased from AAPPTEC. Methyl sulfoxide (DMSO,
99.9%), N-acetyl-D-glucosamine and N,N-Diisopropylethylamine
(DIEA) were purchased from Sigma-Aldrich. Rink Amide PEGA
resin (50-100 mesh, 0.20-0.50 mmol/g) and special Aspartic acid
building blocks, Fmoc-Asp(O-2-PhiPr)-OH and Fmoc-Asp-
(OAll)-OH were purchased from EMD Biosciences. The Pierce
Economy Mini-Spin Column (0.8 mL resin capacity) from Thermo
Scientific was used for on-resin coupling of glycosylamines. Solid
phase peptide synthesis was carried out following standard Fmoc/
tBu chemistry protocols on the Applied Biosystems 433A Peptide
Synthesizer. Preparative RP-HPLC systems (hyper prep, 120 C18
8u; length, 250 mm; I.D., 10 mm) were used to purify peptides.
Isolated yields refer to chromatographically pure compounds.
Theoretical yields for peptide synthesis were calculated based on
the first amino acid loadings of the resins, which were determined
by the UV monitoring of the Fmoc removal at 301 nm (ꢀ ) 7800
cm-1M-1).15 Purity of purified peptides were determined by
analytical HPLC at wavelength of 214 nm (0-80% B, 10 min,
Beckman SGB 0.46 × 5 cm Zorbax C8 column, A buffer: Water,
0.1% TFA; B buffer: 90% Acetonitrile 10% Water 0.1% TFA).
Lyophilization was carried on VirTis Sentry 2.0 lyophilizer. 60-200
mesh silica gel 62 (EMD) was employed for flash chromatography.
TLC analysis was on silica gel plates (EMD Chemical, Si 60 F254)
General Procedure for on-Resin Coupling of Glycosylamines.
In a typical coupling reaction, fully protected peptides with Asp(O-
2-PhiPr) at the N-glycosylation site of Asn were synthesized on
Rink Amide PEGA resin. The PhiPr protecting group was selec-
tively removed by the treatment of 94:1:5 DCM/TFA/TIS, 2 min
repeated 4 times. After washing with DCM and DMF thoroughly,
the resin was transferred into the reaction column and preswelled
with DMSO. Different and various equivalents of glycosylamines,
corresponding equivalents of coupling reagents and bases in DMSO
were added. The total reaction volume is typically less than one-
third of the column capacity. The reaction column was attached to
a rotator (30 rpm) and the coupling was conducted typically for
12 h. For analytical HPLC analyzed trials, the N-terminus Fmoc
was left on in order to absorb at wavelength of 254 nm. For C34
analog production, the N-terminus Fmoc was cleaved by 20%
piperidine in DMF (10 min × 3). Peptides then were cleaved from
resin by treatment with neat TFA containing 1% TIS scavenger
for 3 h. Crude peptides after resin cleavage were precipitated and
washed twice by diethyl ether anhydrous and then analyzed by
analytical HPLC or purified by RP-HPLC.
General Procedure for Recovery of Unreacted High
Mannose Glycan. Reaction solutions in DMSO from different
Man8GlcNAc2NH2 on-resin coupling synthesis trials were pooled
together and precipitated by diethyl ether anhydrous. The precipitate
was then redissolved in a loading buffer (10:3:3 ethyl acetate/
methanol/water) and purified by silica gel flash chromatography
(ethyl acetate/methanol/ water, 10:3:3 to 5:3:3). Fractions containing
glycan were detected positively by phenol-sulfuric acid test,16 then
pooled together and lyophilized to give Man8GlcNAc2, yield 78%
(based on the extra equivalents of Man8GlcNAc2 in the pooled
reactions).
(8) (a) Kates, S. A.; Delatorre, B. G.; Eritja, R.; Albericio, F. Tetrahedron
Lett. 1994, 35, 1033–1034. (b) Offer, J.; Quibell, M.; Johnson, T.
J. Chem. Soc., Perkin Trans. 1 1996, 175–182. (c) Packman, L. C.
Tetrahedron Lett. 1995, 36, 7523–7526.
9
3212 J. AM. CHEM. SOC. VOL. 132, NO. 9, 2010

On-Resin Synthesis of N-Linked Glycopeptides
A R T I C L E S
protecting group was slightly more susceptible to aspartimide
formation than standard t-butyl esters in their aspartimide prone
peptide syntheses,9 the PhiPr protecting group performed very
well in our trial peptide syntheses. The PhiPr protecting group
even suppressed aspartimide formation in the highly aspartimide
prone AKSADE peptide (Figure 2 and Figure S1-2 of the
Supporting Information), and allowed selective, on-resin depro-
tection in 1% trifluoroacetic acid (TFA). These properties make
PhiPr an excellent orthogonal handle for on-resin introduction
of N-linked glycosylation.
The other important element for glycopeptide synthesis,
glycosylamines 1 and 2, were prepared from the corresponding
free sugars, N-acetyl-D-glucosamine and the high mannose
N-linked oligosaccharide Man8GlcNAc2, by treatment with
ammonium hydrogen carbonate as described previously.6a
Repetitive rotary evaporation and lyophilization were utilized
to remove the residual ammonium hydrogen carbonate, which
if left in the sample could disrupt the glycosylation reaction.
1H NMR spectra (Figure S5-6 of the Supporting Information)
showed the conversion of free sugars into their corresponding
ꢀ-D-glycosylamines with close to 100% yields.
Next, the glycosylation reaction was optimized for glycosy-
lamines 1 and 2 using the aspartimide-prone sequence AK-
SANE. In order to quantify the glycosylation yields in these
experiments, N-terminal Fmoc protecting groups were left on
at the end of peptide synthesis to aid UV detection. After resin
cleavage from each of the synthesis trials, crude peptides were
precipitated and washed twice with anhydrous diethyl ether, and
then analyzed by analytical HPLC. Individual peaks detected
at 254 nm were identified by mass spectrometry and product
distribution (glycopeptide: uncoupled peptide: aspartimide) was
determined by the integrated analytical HPLC absorption signals.
Example analytical HPLC traces from different synthesis trials
are shown in Figure 3, and the results of synthesis trials analyzed
in this manner are summarized in Table 1.
Using the approach described above, PEG-polyacrylamide-
based (PEGA) and polystyrene-based (PS) rink amide resins
were tested as solid supports for glycopeptide synthesis (entries
1, 2, 9, and 10, Table 1). Excellent glycosylation yields were
achieved on both PS and PEGA resins with the monosaccharide
glycosylamine 1 (80% and 87%, respectively). In contrast, for
coupling with the high mannose oligosaccharide 2, no glyco-
sylation at all occurred on PS resin and a 43% glycosylation
yield was achieved on the more hydrophilic PEGA resin.5a,10
Because of this, PEGA resin was utilized as the solid support
for on-resin glycosylamine coupling in the rest of this study.
Due to the poor solubility of large oligosaccharides in most
organic solvents, we explored the use of DMSO/water mixtures
in addition to the neat DMSO that has been reported previously
for in solution glycosylamine coupling reactions.6 Unfortunately,
DMSO/water mixtures resulted in a significant increase in
aspartimide formation for monosaccharide 1 and no product
formation at all for oligosaccharide 2 (see entries 2, 4, 11, and
12 of Table 1), therefore neat DMSO was utilized in the
remainder of this study. The selection of coupling reagent is
also crucial to achieve high glycosylation yields. We found that
DEPBT, 3-(Diethoxyphospho-ryloxy)-1,2,3-benzotriazin-4(3H)-
one, was more efficient than HBTU (O-(Benzotriazol-1-yl)-
N,N,N′,N′-tetramethyluronium Hexafluoro-phosphate) and slightly
Figure 2. Aspartimide formation during Fmoc SPPS of peptides Fmoc-
AKSADE-NH2 and Fmoc-CDISR-NH2 using either Allyl or PhiPr orthogo-
nal protecting groups to protect Asp residues.
Analytical HPLC for on-Resin Glycosylation Yields. Analyti-
cal HPLC (0-80% B, 10 min, Beckman SGB 0.46 × 5 cm Zorbax
C8 column, A buffer: Water, 0.1% TFA; B buffer: 90% Acetonitrile
10% Water 0.1% TFA, detection at 254 nm). Individual peaks were
identified by mass spectrometry and product distribution (glyco-
peptide: uncoupled peptide: aspartimide) was determined by the
integrated analytical HPLC absorption signal using Peaksimple 2000
version 2.83 (SRI Instruments).
In Vitro HIV-1 Entry Inhibition Assay. TZM-bl and HL2/3
cells were obtained from the NIH AIDS Research and Reference
Reagent Program.13 A luciferase-based cell-cell fusion assay was
performed as previously described.13a The target cells, TZM-bl cells,
express CD4 and CCR5 and are cotransfected with a luciferase
reporter gene linked to the HIV-1 promoter which can be activated
by HIV-1 Tat from the effector HL2/3 cells. Cells were cultured
in DMEM medium containing 10% FBS, 100 units of Penicillin
and 0.1 mg/mL Streptomycin at 37 °C, 5% CO2. In the assay, serial
dilutions of C34 peptide analogs (20 µL per well) were added to a
96-well Microplate (Perkin-Elmer, Isoplate-96 TC) containing
TZM-bl cells (40 µL per well, 2.5 × 104 cells per well, seeded and
incubated in 37 °C, 5% CO2 overnight). Then, the HL2/3 cells (40
µL per well, 2.5 × 104 cells per well) were added and the plate
was incubated for 6 h at 37 °C, 5% CO2. After incubation, LucLite
luminescence substrate reagent (100 µL per well, Perkin-Elmer)
was added and the plate was shaken for 3 min at 600 rpm before
reading. Light output was measured on MicroBeta-1450 liquid
scintillation counter (Perkin-Elmer, Wellesley, MA). IC50 values
were calculated by Origin software (OriginLab).
Results and Discussion
In our initial studies of on-resin convergent synthesis of
N-linked glycopeptides, we attempted to use the allyl protecting
group for selective on-resin deprotection of Asp side chains,
but as others8 have reported, we observed significant aspartimide
formation during Fmoc SPPS (Figure 2 and Figure S1-4 of
the Supporting Information). The amount of aspartimide forma-
tion that occurs during Fmoc SPPS renders the allyl protecting
group unsuitable for on-resin N-linked glycopeptide synthesis
without amide backbone protection.8 Since amide backbone
protection can introduce additional complications into peptide
syntheses, we sought a protecting group that could be used for
on-resin N-linked glycopeptide synthesis without backbone
protection, but which would also be compatible with backbone
protection if desired. It has been previously noted by Offer et
al.8 that the sterically hindered t-butyl ester suppresses aspartim-
ide formation under conditions where the allyl protecting group
does not. Because of this, we explored the use of the sterically
hindered 2-phenylisopropyl (PhiPr)9 protecting group for this
application, which can be selectively deprotected unlike t-butyl
esters. Although a previous work reported that this more bulky
(9) (a) Yue, C. W.; Thierry, J.; Potier, P. Tetrahedron Lett. 1993, 34,
323–326. (b) Mergler, M.; Dick, F.; Sax, B.; Weiler, P.; Vorherr, T.
J. of Pept. Sci. 2003, 9, 36–46.
(10) Garcia-Martin, F.; Quintanar-Audelo, M.; Garcia-Ramos, Y.; Cruz,
L. J.; Gravel, C.; Furic, R.; Cote, S.; Tulla-Puche, J.; Albericio, F.
J. Comb. Chem. 2006, 8, 213–220.
9
J. AM. CHEM. SOC. VOL. 132, NO. 9, 2010 3213

A R T I C L E S
Chen and Tolbert
coupling reagent selective for amide bond formation with great
resistance to racemization,11 DEPBT was chosen as the on-resin
coupling reagent for use with the large glycosylamine 2. We
also investigated the effect of the presence of base on glyco-
sylation yield since it has been reported that omission of base
can reduce aspartimide formation.8 The omission of base during
coupling of glycosylamine 1 (Table 1 entry 3) did not have
much of an effect on glycosylamine coupling efficiency, but
removing base from the coupling of glycosylamine 2 (Table 1,
entries 11, 13, 14, and 15) resulted in significant reduction in
coupling efficiency making it desirable to include base in
reactions containing the large glycosylamine 2. Investigation
into the amount of glycosylamines utilized in on-resin coupling
reactions suggested that changing the amount of glycosylamine
1 (entries 2, 5, and 6) did not have significant effects on coupling
yields (87%, 84%, and 83% respectively). The amount of
glycosylamine 2, however, did appear to have a significant effect
on glycopeptide yield, with 1.5 and 3 equiv. of glycosylamine
2 resulting in coupling yields of 49% and 58%, respectively
(Table 1 entries 11 and 14). This result agrees with studies of
in-solution coupling of bulky, large sugar glycosylamines,3c
where the aspartimide formation side reaction has been found
to compete with glycosylation to reduce overall glycosylation
yields. Using the optimized reaction conditions developed here,
shown in Table 1 entry 14, an isolated yield12 of 50% was
achieved for the synthesis of the aspartimide prone, high
mannose N-linked glycopeptide 4.
To explore how these on-resin glycosylation conditions
worked with other small glycopeptide sequences, we applied
this on-resin coupling strategy to the synthesis of high mannose
containing glycopeptides 5 and 6 (Table 1, entries 16 and 17)
obtaining excellent glycosylation yields of 86% and 53%
respectively. Taken together, under similar reaction conditions
we have observed glycosylamine coupling yields ranging from
49% to 86% for different peptide sequences, indicating that the
coupling efficiency of large glycosylamines may be affected
by sequence.
An advantage of the on-resin approach taken here is the
possibility of easy recovery of valuable unreacted oligosaccha-
rides after on-resin coupling. This is possible because of the
hydrophilic nature of oligosaccharides, which allows them to
be easily distinguished from the base and coupling reagents used
Figure 3. Analytical HPLC trace examples for on-resin glycosylation
(detection at 254 nm); (a) glycosylamine 1 coupling to produce glycopeptide
3, crude sample from synthesis entry 5, Table 1; (b) glycosylamine 2
coupling to produce glycopeptide 4, crude sample from synthesis entry 10,
Table 1; and (c) glycosylamine 2 coupling to produce glycopeptide 5, crude
sample from synthesis entry 16, Table 1.
more efficient than HBTU/HOBt(1-hydroxybenzotriazole) in on-
resin glycosylation of glycosylamine 1 (see entries 2, 7, and 8
of Table 1). Because DEPBT has also been reported as a
Table 1. Optimization of on-Resin Glycosylation with Glycosylamines 1 and 2
entrya
(glycopeptides)
resin
solvent
sugar-NH2
(equiv)
coupling reagents
(equiv)
DIEA
(equiv)
product distribution %b
glycopeptide: uncoupled peptide: aspartimide
1 (3)
2 (3)
3 (3)
4 (3)
5 (3)
6 (3)
7 (3)
8 (3)
9 (4)
10 (4)
11 (4)
12 (4)
13 (4)
14 (4)
15 (4)
16 (5)
17 (6)
PS
DMSO
DMSO
DMSO
DMSO/water (4/1)
DMSO
DMSO
DMSO
DMSO
DMSO
DMSO
DMSO
DMSO/water (4/1)
DMSO
DMSO
DMSO
DMSO
DMSO
1 (5)
1 (5)
1 (5)
1 (5)
1 (3)
1 (1)
1 (5)
1 (5)
2 (1)
2 (1)
2 (1.5)
2 (1.5)
2 (1.5)
2 (3)
2 (3)
2 (1.5)
2 (1.5)
DEPBT (3)
DEPBT (3)
DEPBT (3)
DEPBT (3)
DEPBT (3)
DEPBT (3)
HBTU (3)
HBTU/HOBt (3/3)
DEPBT (3)
DEPBT (3)
DEPBT (3)
DEPBT (3)
DEPBT (3)
DEPBT (3)
DEPBT (3)
DEPBT (3)
DEPBT (3)
1.5
1.5
0
80: 7: 13
87: 9: 4
88: 9: 3
PEGA
PEGA
PEGA
PEGA
PEGA
PEGA
PEGA
PS
PEGA
PEGA
PEGA
PEGA
PEGA
PEGA
PEGA
PEGA
1.5
1.5
1.5
1.5
1.5
1.5
1.5
1.5
1.5
0
69: 5: 26
84: 9: 9 (75%c)
83: 3: 14
79: 4: 17
84: 6: 10
0: 39: 61
43: 8: 49 (36%c)
49: 3: 48
0: 28: 72
31: 4: 65
1.5
0
1.5
1.5
58: 7: 35 (50%c)
43: 3: 53
86: 5: 9 (76%c)
53: 11: 36 (43%c)
a Synthesis: scale 1-5 µmol; coupling reagent and glycosylamine equivalents based on loading values reported by the resin manufacturer; DIEA
equivalents based on amounts of glycosylamine added; reaction time ≈ 12 h. b Product distributions for crude peptides were determined by integrated
analytical HPLC absorbance signals at 254 nm. c Isolated yields based on Fmoc quantification for the first amino acid loaded.
9
3214 J. AM. CHEM. SOC. VOL. 132, NO. 9, 2010

On-Resin Synthesis of N-Linked Glycopeptides
A R T I C L E S
yield of 78% of the excess glycosylamine 2 utilized in on-resin
1
coupling reactions. The H NMR spectra (shown in Figure S7
of the Supporting Information) of the recovered high mannose
oligosaccharide was identical to the starting material
Man8GlcNAc2 (also shown in Figure S7 of the Supporting
Information), indicating the success of recovery, and also the
conversion of the glycosylamine back into the free sugar form
during the purification process. To demonstrate reuse of the
recovered material, we further converted it back into glycosy-
lamine 2 using ammonium hydrogen carbonate, and tested its
coupling reactivity with rink amide PEGA resin bound Fmoc-
Asp and Fmoc-Glu with free side chain carboxylic acids. As
shown in Figure 4 and Figure S8 of the Supporting Information,
the recovered glycosylamine 2 provided excellent coupling
yields, above 85% for both on-resin couplings. Using this
approach we have routinely been able to recover and reuse
valuable unreacted oligosaccharides from on-resin glycosylation
reactions. This allows the use of excess oligosaccharide gly-
cosylamines to drive on-resin glycosylation reactions without
complete loss of the unreacted, excess oligosaccharide.
To demonstrate that this approach works for large N-linked
glycopeptides as well as small glycopeptides, it was utilized to
synthesize different glycosylated forms of the 34 amino acid HIV-1
gp41 C34 glycopeptide which is active as a HIV entry inhibitor
(Figure 5). The on-resin glycosylamine coupling step was con-
ducted in DMSO using 3 equiv. of DEPBT, 1.5 equiv. of DIEA,
and 3 equiv. of glycosylamine 1 or 2 respectively for the synthesis
of GlcNAc-C34 8 and Man8GlcNAc2-C34 9. Excellent isolated
yields12 of 17% and 10% were obtained for GlcNAc-C34 8 and
Man8GlcNAc2-C34 9 respectively (Figure 5). To determine if the
C34 glycopeptides retained their anti-HIV activities they were tested
by a luciferase-based cell-cell fusion assay13 (Figure 6). The
estimated IC50 values for C34 7, GlcNAc-C34 8, and
Man8GlcNAc2-C34 9 were 3.7, 5.2, and 8.2 nM, respectively.
Although the inhibitory activity of C34 was reduced slightly by
glycosylation, these results are similar to what has been observed
in previous studies of glycosylated analogues of C34.14
Figure 4. Recovery and reuse of the unreacted, excess high mannose
oligosaccharide Man8GlcNAc2 after on-resin coupling. (a) Process scheme;
(b) MALDI-TOF-MS and analytical HPLC for crude sample from on-
resin coupling of recovered and reproduced glycosylamine 2 to rink amide
PEGA resin bound Fmoc-Asp, expected mass for Fmoc-Asp(Man8GlcNAc2)-
CONH2, [M + Na]+ ) 2078.71.
in on-resin coupling reactions. We have investigated the
recovery of unused high mannose oligosaccharide 2 from our
syntheses. After on-resin coupling of glycosylamine 2, the
coupling solution containing unused 2 was collected by filtration
from the reaction column, precipitated and washed twice with
anhydrous diethyl ether. The resulting precipitate was then
purified by silica gel flash chromatography, providing high
mannose oligosaccharide Man8GlcNAc2 with the approximate
Conclusions
In conclusion, we have successfully developed an efficient
convergent glycosylamine coupling approach for solid phase
Figure 5. MALDI-TOF-MS and analytical HPLC for Synthetic HIV-1 gp41 C34 Analogues: (a) C34, 7, expected mass:[M + H]+ ) 4248.58; (b) GlcNAc-
C34, 8, expected mass: [M + H]+ ) 4451.77; (c) Man8GlcNAc2-C34, 9, expected mass: [M + H]+ ) 5952.09.
9
J. AM. CHEM. SOC. VOL. 132, NO. 9, 2010 3215

A R T I C L E S
Chen and Tolbert
monosaccharide 1 and the large oligosaccharide 2, emphasizing
that care must be taken when adapting reaction conditions
worked out on monosaccharides to large oligosaccharides. This
strategy allows the rapid and efficient synthesis of biologically
active N-linked glycopeptides by SPPS such as the 34 amino
acid C34 HIV entry inhibitor glycopeptides shown here.
Acknowledgment. This work was supported by Indiana Uni-
versity. We thank Dr. Richard DiMarchi for his generous support
for SPPS. We also thank Mr. Junpeng Xiao for helping with the
HIV-1 entry inhibition assay. The following reagents were obtained
through the AIDS Research and Reference Reagent Program,
Division of AIDS, NIAID, NIH: HL2/3 from Dr. Barbara K. Felber
and Dr. George N. Pavlakis; TZM-bl from Dr. John C. Kappes,
Dr. Xiaoyun Wu, and Tranzyme Inc.
Supporting Information Available: Effects of Asp orthogonal
protecting groups on the quality of Fmoc SPPS,1H NMR for
glycosylamine conversions, recovery of unreacted Man8GlcNAc2,
characterizations of peptides and high mannose glycan
Man8GlcNAc2. This material is available free of charge via the
Internet at http://pubs.acs.org.
Figure 6. HIV-1 entry inhibition assay (luciferase-based cell-cell fusion
assay13) and estimated IC50 values for nonglycosylated and glycosylated
HIV-1 gp41 C34 analogs.
synthesis of N-linked glycopeptides that utilizes PhiPr protected
side chains to generate N-linked glycosylation sites on-resin.
The PhiPr protecting group is ideal for on-resin N-linked
glycopeptide synthesis, suppressing aspartimide formation dur-
ing peptide synthesis and allowing selective deprotection of Asp
side chains for glycosylamine coupling. On-resin glycosylamine
coupling allows glycosylation reactions to be driven by an excess
of glycosylamines, with the possibility of high-yield recovery
of unused valuable oligosaccharides. We have observed sig-
nificant differences between on-resin coupling behavior of the
JA9104073
(13) (a) Wexler-Cohen, Y.; Shai, Y. FASEB J. 2007, 21, 3677–3684. (b)
Ciminale, V.; Felber, B. K.; Campbell, M.; Pavlakis, G. N. AIDS Res.
Hum. RetroVir. 1990, 6, 1281–1287. (c) Takeuchi, Y.; McClure, M. O.;
Pizzato, M. J. Virol. 2008, 82, 12585–12588. (d) Wei, X.; Decker,
J. M.; Liu, H.; Zhang, Z.; Arani, R. B.; Kilby, J. M.; Saag, M. S.;
Wu, X.; Shaw, G. M.; Kappes, J. C. Antimicrob. Agents Chemother.
2002, 46, 1896–1905. (e) Derdeyn, C. A.; Decker, J. M.; Sfakianos,
J. N.; Wu, X.; O’Brien, W. A.; Ratner, L.; Kappes, J. C.; Shaw, G. M.;
Hunter, E. J. Virol. 2000, 74, 8358–8367.
(14) (a) Huang, W.; Groothuys, S.; Heredia, A.; Kuijpers, B. H.; Rutjes,
F. P.; van Delft, F. L.; Wang, L. X. Chembiochem 2009, 10, 1234–
1242. (b) Wang, L. X.; Song, H.; Liu, S.; Lu, H.; Jiang, S.; Ni, J.; Li,
H. Chembiochem 2005, 6, 1068–1074.
(11) (a) Yamamoto, N.; Takayanagi, A.; Sakakibara, T.; Dawson, P. E.;
Kajihara, Y. Tetrahedron Lett. 2006, 47, 1341–1346. (b) Li, H. T.;
Jiang, X. H.; Ye, Y. H.; Fan, C. X.; Romoff, T.; Goodman, M. Org.
Lett. 1999, 1, 91–93. (c) Ye, Y. H.; Li, H. T.; Jiang, X. H. Biopolymers
2005, 80, 172–178.
(15) Kay, C.; Lorthioir, O. E.; Parr, N. J.; Congreve, M.; McKeown, S. C.;
Scicinski, J. J.; Ley, S. V. Biotechnol. Bioeng. 2000, 71, 110.
(16) Dobois, M.; Gilles, K. A.; Hamilton, J. K.; Rebers, P. A.; Smith, F.
Anal. Chem. 1956, 23, 350.
(12) Based on Fmoc quantification for the first amino acid loaded.
9
3216 J. AM. CHEM. SOC. VOL. 132, NO. 9, 2010