Exploring neoglycoprotein assembly through native chemical ligation using neoglycopeptide thioesters prepared via N→S acyl transfer

Article abstract of DOI:10.1039/b920535g

Sugars and simplified oligosaccharide "mimics" can be joined with protein fragments at pre-defined sites using reliable chemical reactions such as thiol alkylation and Cu(i) catalysed azide/acetylene ligation (click chemistry). These fragments have the po

Full text of DOI:10.1039/b920535g

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/obc | Organic & Biomolecular Chemistry
Exploring neoglycoprotein assembly through native chemical ligation using
neoglycopeptide thioesters prepared via N→S acyl transfer†‡
Jonathan P. Richardson, Chung-Hei Chan, Javier Blanc, Mona Saadi and Derek Macmillan*
Received 1st October 2009, Accepted 17th December 2009
First published as an Advance Article on the web 22nd January 2010
DOI: 10.1039/b920535g
Sugars and simplified oligosaccharide “mimics” can be joined with protein fragments at pre-defined
sites using reliable chemical reactions such as thiol alkylation and Cu(I) catalysed azide/acetylene
ligation (click chemistry). These fragments have the potential to be assembled into neoglycoprotein
therapeutics using native chemical ligation.
modified proteins as research tools and diagnostics in addition
Introduction
to potential therapies.8 While assembly of a glycoprotein can seem
Native chemical ligation (NCL) is a powerful method for protein
complex and laborious in terms of synthesis, strategic replacement
synthesis and semi-synthesis that is particularly suited to the
of a few synthetically challenging motifs with alternatives formed
preparation of proteins bearing posttranslational modifications.1
through less capricious chemical reactions may greatly simplify
In the case of glycosylated proteins NCL is presently the only
the production of analogues such that they can be assembled
method to have demonstrated an ability to deliver proteins
outwith a handful of specialised laboratories. The advantages of
bearing multiple forms of defined glycosylation at pre-designated
combining chemoselective strategies for glycoconjugate synthesis
positions employing entirely native linkages.2 While advances in
and assembly of biomolecular multimers have recently been
demonstrated4d,9 and in this manuscript we describe our own
progress towards glycoprotein analogues through combination
of NCL, cysteine sulfhydryl modification with a-haloacetamide
sugars10 and Cu(I) catalysed azide/acetylene ligation (“click”
chemistry).11 In this case chemoselective couplings are used not
only in glycoconjugation but also in the assembly of a simplified
oligosaccharide mimic.
carbohydrate chemistry have allowed elaborate oligosaccharide
syntheses,3 solid-phase oligosaccharide and glycopeptide synthesis
still lacks the simplicity of oligopeptide and oligonucleotide
assembly. The importance of oligosaccharides in living systems
demands that complex homogeneous replica structures need be
prepared to unravel biological functions though there are other
potential applications such as improvement of the pharmacoki-
netic profile of therapeutic proteins, analysis of saccharide-lectin
interactions and glycoconjugate targeting where the presentation
of similar structures may suffice.4 Furthermore, it is desirable
that simpler chemical building blocks are available such that a
variety of structures can be rapidly prepared by a non-expert
and in a small number of synthetic steps.5 Neoglycoproteins
formed through global or site-specific modification of recombi-
nant (usually bacterially derived) proteins have been extensively
investigated though concern as to the possible immunogenic
nature of these substances has driven the development of new
methods that exert more control on the degree of glycosylation
and introduce native sugar-protein linkages.6 Of the synthetic and
semi-synthetic methods available, glycoprotein assembly through
NCL2 and perhaps remodelling of simply glycosylated precursors,7
which themselves may be more generally accessible through
NCL, have emerged as being particularly effective. Significant
improvements in NCL methodology combined with the typically
mild reaction conditions employed and tolerance to densely
functionalised starting materials has also rendered NCL an
attractive method for assembling unnatural, and unnaturally
Neoglycopeptide thioester synthesis via N→S acyl
shift
Previously we prepared an acetylene-modified biologically active
variant of the 166 amino acid hormone erythropoietin (EPO)
through NCL between a bacterially derived N-cysteinyl protein
fragment and a synthetic acetylene-containing peptide thioester
(Scheme 1a).12 The thioester component was accessible, albeit in
relatively low (5%) yield using the sulfonamide safety-catch linker
for solid-phase peptide synthesis.13
The main difficulties we encountered were associated with
efficient loading of the first amino acid and release of the assembled
peptide chain from the solid support. Having recently developed
a novel method for the production of peptide thioesters for
NCL employing an N→S acyl shift14 we were keen to apply
an optimised procedure15 in the context of glycopeptides. We
believed that this approach might be more economical with
regards to precious carbohydrate appendages since only a slight
excess of sugar a-haloacetamide would be required to modify
free thiol groups within a preassembled precursor peptide in
solution (Scheme 2) whereas larger excesses of carbohydrates or
glycoaminoacids are usually committed in solid phase peptide
synthesis based on resin loading, before the peptide is recovered
from the solid support.16 Interestingly, during the course of our
experiments we happened upon an unexpected improvement in
Department of Chemistry, University College London, 20 Gordon Street,
London, WC1H 0AJ, UK. E-mail: d.macmillan@ucl.ac.uk
† This paper is part of an Organic & Biomolecular Chemistry web theme
issue on protein labelling and chemical proteomics. Guest editors: Ed Tate
and Matthew Bogyo
‡ Electronic supplementary information (ESI) available: UV-VIS and mass
spectra. See DOI: 10.1039/b920535g
This journal is
The Royal Society of Chemistry 2010
Org. Biomol. Chem., 2010, 8, 1351–1360 | 1351
©

View Article Online
Scheme 1 Semi-synthetic approaches to modified EPO analogues employing NCL.12
Scheme 2 Proposed synthesis of a neoglycopeptide thioester. A preassembled peptide is alkylated at free sulfhydryl groups with sugar a-iodoacetamides
(2 equivs per thiol). Next the remaining thiols, involved in disulfide bond formation and required for thioester formation via an N→S acyl shift, are
exposed and the C-terminal His-Cys motif is converted to a C-terminal histidyl thioester.15
our peptide thioesterification reaction in a model system. In
our preceding studies all peptides were prepared on Rink amide
resin and possessed C-terminal carboxamides (i.e. contained a C-
terminal Xaa-Cys-CONH2 motif) or were internal with respect
to the peptide sequence. We had not investigated the reaction in
the context of a C-terminal cysteine with a free carboxyl group
(i.e. containing Xaa-Cys-CO2H) though considered this subtle
difference a more accurate model for a recombinant sample that
might ultimately be engineered to possess a C-terminal His-Cys
motif. We were therefore very much surprised to find that when
the terminal cysteine residue possessed a free carboxyl group
(e.g. 1a, Fig. 1) the rate of conversion to the corresponding
thioesters approximately doubled relative to the carboxamide-
terminated counterparts (1b, Fig. 1). This raised the possibility
that the “+1” amide linkage dramatically influences a peptide’s
susceptibility to thioester formation in addition to the nature
of the Xaa-Cys motif. Conceivably, thioesterification could be
catalysed through intramolecular protonation of the scissile amide
N or carbonyl O atoms by the C-terminal carboxyl group. If a
similar explanation could be formulated for the role of acidic
protons on His (1) and Cys (3) then this might help account
for enhanced thioesterification at these sites. As the reaction
was already conducted at low pH (pH 2) we considered this
explanation unlikely though incorporated non-acidic variants of
His (2a/b) and Cys (4a/b) to probe the observation in more
detail. Importantly 2a/b and 4a/b facilitated thioesterification
suggesting no essential role for these residues as general acids.
Next we investigated whether the rate acceleration attributed to
the free carboxyl allowed motifs (such as Ala-Cys) that would not
be expected to form thioesters to any appreciable extent at 60 ◦C,14
to become more favourable. Indeed the Ala-Cys terminated
peptides 6a/b did undergo thioesterification though more readily
as terminal carboxylic acid 6a than as carboxamide 6b. Clearly
the terminal nature of the motif also serves to increase conversion
to thioester, perhaps through greater conformational flexibility
which, in part, may account for the ease with which Gly-Cys
motifs undergo thioesterification. The Ala-thioesters formed from
6a/b appeared to undergo significant epimerisation during the
course of the reaction. This was confirmed through synthesis and
thioesterification of the corresponding D-Ala containing peptides
(which also epimerised)17 while 1a/b, 2a/b and 4a/b showed no
signs of epimerisation. Furthermore when the optimal His-Cys
motif was transposed to the N-terminus of the peptide resulting
in sequence H2N-HCMEEYLKS-CONH2 (7, Fig. 1)18 the rate of
thioesterification was slowed a little further than when installed
as a C-terminal amide suggesting important stereochemical and
electronic requirements for the reaction. The resulting N-cysteinyl
peptide, itself a potentially useful product if unmasked in the
context of a recombinant protein, was isolated and its identity
confirmed by LC-MS and NCL reaction with the thioester formed
from 6a.
From this short study we concluded that the C-terminal car-
boxyl group can enhance the thioesterification reaction at Xaa-Cys
motifs such that it can relax the previously observed requirement
for Gly, His or Cys◦at the thioesterification site when the reaction
is performed at 60 C.14 The effect is likely attributable to both a
1352 | Org. Biomol. Chem., 2010, 8, 1351–1360
This journal is
The Royal Society of Chemistry 2010
©

View Article Online
Fig. 1 Model peptides prepared to test the effect of the terminal functional group on thioesterification (each peptide was prepared on 0.02 mmol
scale and isolated in 60–65% yield). Formation of thioester is represented with solid lines for C-terminal carboxylic acids and with broken lines for
carboxamides. Reagents and conditions: a) approx. 1 mM peptide, guanidine.HCl, 10% v/v AcOH, 10% w/v MESNa, 0.5% w/v TCEP.HCl, 60 ◦C, 96 h.
Note: there is generally good agreement between % conversion determined by HPLC peak integration at 230 nm and relative MS peak intensities of the
crude reaction mixture (shown above for conversion of 2a and 2b to thioester after 48 h).
catalytic role for the carboxyl group and increased conformational
flexibility associated with the terminal residue. Even Ala-Cys
motifs were transformed (albeit with least efficiency) to thioesters
at 60 ◦C when installed as C-terminal carboxylic acids, though
epimerisation (approx. 10–15%) was observed after 48 h. Acting
as a general acid catalyst, the high effective concentration of
the C-terminal carboxyl may accelerate the breakdown of a
carbinolamine intermediate as proposed for the essential histidine
residue in the protein splicing mechanism orchestrated by inteins,19
and in protein thioesterification catalysed by sortase.20 In contrast
to these systems our thioesterification reaction is tolerant to
denaturants such as 6 M guanidine hydrochloride.
Ultimately, the significant improvement attributed to the termi-
nal carboxyl group (Fig. 1) provided the impetus to prepare our
neoglycopeptide thioester precursors on hydroxyl functionalised
resins (e.g. Novasyn TGA) rather than the Rink amide resins
we had employed thus far.14,15 In anticipation of performing
the NCL reaction with a bacterially derived protein fragment
comprising residues 33–166 (Scheme 1b) we assembled EPO
residues 1–33 containing a C-terminal HC motif for acid-mediated
thioester formation (Fig. 2a).21 Preparation of residues 1–33 allows
incorporation of two glycosylation sites (Asn24 and Asn30) found
in the enhanced erythropoiesis stimulating agent AranespTM and
permits NCL to be performed using a native cysteine residue.22
For optimal compatibility with our thioesterification the synthetic
peptide sequence carried two mutations. The first, G28A was
introduced to prevent undesirable thioesterification occurring
across the Gly28-Cys29 junction, our results from Fig. 1 suggesting
that the difference in reactivity between thioesterification at a
His-Cys (terminal) motif and an Ala-Cys (internal) motif should
confer acceptable selectivity. The second mutation, D8E was
introduced to prevent slow hydrolysis at aspartic acid which can
occur at elevated temperature under the thioesterification reaction
conditions upon prolonged heating, which is completely abolished
by substitution with glutamic acid.23 Furthermore, at positions to
be modified with carbohydrates (residues 24 and 30) we introduced
orthogonally (trityl) protected cysteine residues to the native
cysteines (Cys 7 and Cys 29) and the C-terminal Cys, which
were SAcm protected. The desired peptide sequence was therefore
assembled in automated fashion on an Applied Biosystems 433A
synthesiser and cleaved from the solid support under standard
conditions. The crude product was purified by semi-preparative
reverse-phase HPLC and isolated in 18% yield (Fig. 2b). Peptide
8 was dissolved in 50 mM NH4HCO3; pH 8.0 (2.0 mL) and
incubated with the a-iodoacetamide of N-acetylglucosamine
(approximately 2 equivalents per thiol) at room temperature for
This journal is
The Royal Society of Chemistry 2010
Org. Biomol. Chem., 2010, 8, 1351–1360 | 1353
©

View Article Online
Fig. 2 a) Target EPO peptide sequence for synthesis compared with the wild-type sequence (mutations are highlighted). b) Reactions involved in the
transformation from peptide 8 to glycopeptide mimetic thioester 11 with accompanying ESI-MS data for isolated products. Reagents and conditions. a)
50 mM NH4HCO3, glycosyl iodoacetamide (2 eq per thiol), room temp. 2 h b) 10% v/v aqueous AcOH, AgOAc (10 equiv. per SAcm group), DTT, 30 ◦C,
16 h. c) Guanidine.HCl, 10% v/v AcOH, 10% w/v MESNa, TCEP. HCl, 60 ◦C, 48 h.
2 h. N-Acetylglucosamine was employed in the first instance to
ensure compatibility with the mimetic sugar-protein linkage and
thioesterification. Although the reaction appeared quantitative
by LC-MS 9 was isolated by HPLC in only 50% yield then
subjected to AgOAc/DTT in 10% aqueous acetic acid to cleave
acetamidomethyl protecting groups on the remaining cysteine
residues. Again the deprotection appeared efficient by HPLC
though the product was isolated after HPLC in 40% yield.
Thioester precursor 10 was re-dissolved at a concentration of
1.25 mg mL-1 in 10% w/v MESNa, 10% v/v AcOH, 0.5% w/v
TCEP in 6 M guanidine hydrochloride and heated to 60 ◦C for
48 h to produce the C-terminal thioester. The reaction afforded
two major species in approximately 1 : 1 ratio: the product thioester
11 and unreacted HC-terminated peptide 10. In agreement with
previous results there was no evidence of protein cleavage between
the internal Ile6-Cys7 and Ala28-Cys29 motifs. 10 and 11 could
not be separated by semi-preparative reverse-phase HPLC though
10 would not participate in NCL and could likely be recovered
after ligation.15 To confirm thioester formation, the partially
purified lyophilised product (5 mg) was added, in excess, to the
bacterially-derived C-terminal fragment (EPO 33-166, 1 mg)24 in
6 M guanidine hydrochloride containing 300 mM Na phosphate
buffer; pH 7.0, 20 mM 4-mercaptophenylacetic acid (MPAA),25
and 10 mM TCEP (0.25 mL) for 24 h as previously described.12
Encouragingly, the desired product was immediately observed by
LC-MS (Fig. 3b inset, calculated mass = 19016 Da, observed
mass = 19018 Da) suggesting that neoglycopeptide thioester
formation via the N→S acyl transfer method and subsequent
glycoprotein assembly appear feasible. In fact, an early aim had
been to compare the efficiency of this thioester synthesis with
that of an identical peptide released from the safety-catch linker
following solid-phase modification of selectively liberated thiols
on solid-support. However, as a consequence of several failed
attempts to efficiently unmask both thiols (at cys 24 and cys 30)
within the resin-bound 33mer, this route was ultimately abandoned
and the comparison no longer possible.
Glycomimetic synthesis
Having demonstrated the union of N-acetyl glucosamine with
a peptide thioester precursor, we envisaged that, in the simplest
case, synthesis of a tripeptide “scaffold” displaying saccharides
(exemplified by 12, Scheme 3) may serve as a mimic for branched
oligosaccharides of the type typically observed extending from
the pentasaccharide core structure of N-linked glycoproteins.4a
Following synthesis, routine enzymatic sialylation of the
1354 | Org. Biomol. Chem., 2010, 8, 1351–1360
This journal is
The Royal Society of Chemistry 2010
©

View Article Online
Fig. 3 a) An erythropoietin derived peptide bearing two oligosaccharide mimics. The positions of attachment of 12 to the acetylene-linked peptide are
highlighted. b) Schematic representation of green fluorescent protein (GFP) C-terminal thioester (modified from PDB entry 1w7s). c) NCL reaction
between GFP thioester and 18 after 48 h. Lane 1, molecular weight markers; lane 2, GFP (approx. mass = 28 kDa); lane 3, ligation reaction after 48 h.
d) Lectin binding assay: a) A solution of GFP (left) and crude GFP linked-18 (right); b) after addition of Ricinus communis RCA120 conjugated agarose
beads; c) after washing with lectin binding buffer.
terminal galactose residues is all that would be required to
complete the biantennary sialyl N-acetyllactosamine motif.26
While simplifying synthesis considerably, notably by removal of
the core b-mannoside, we retained the branched nature of the
N-acetyllactosamine antennae in a format we hoped would still
be sufficiently “biological” to permit clearance via the hepatic
asialoglycoprotein receptor providing a potential advantage over
polyethylene glycol (PEG)-linked proteins as they are not effi-
ciently degraded and accumulate in kidney vacuoles.27 12 was
assembled on Rink amide resin, initially employing commercially
available amino acids and standard peptide coupling methodology
using HBTU/HOBt as coupling reagents in the presence of
N,N- diisopropylethyl amine (DIPEA). 13 was then exposed
to dithiothreitol (DTT) to remove the S-tert-butylthio groups,
followed by N-acetyllactosamine bromoacetamide (14) to afford
15. The N-allyloxycarbonyl (alloc) group was removed with
Pd(PPh3)4/1,3-dimethylbarbituric acid and the glycosyl azide
16 was coupled to the free e-amino group of lysine using
standard peptide coupling methods.28 It was hoped that the
azide functionality would next facilitate the direct site-specific
modification of acetylenic peptides or proteins, circumventing the
often troublesome Lansbury aspartylation of glycosyl amines.29
Following cleavage from the solid support and acetyl ester
deprotection, 12 was obtained in 47% overall yield (based on initial
resin loading). To demonstrate that the “biantennary” saccharide
mimic 12 could be recognised as an oligosaccharide we wished
to evaluate it in a more biological context and so developed a
lectin binding experiment employing the commercially available
lectin Ricinus communis RCA120 immobilised on agarose beads.
This lectin was chosen since it binds to the monosaccharide
D-galactose and has been used to isolate glycoproteins from
natural sources. To this end we appended two saccharide mimics
(containing four terminal Gal residues) to a short peptide sequence
derived from erythropoietin residues 22–32 adorned with an N-
terminal cysteine residue (18, Fig. 3a). The oligopeptide sequence
of 18 was assembled on Novasyn-TGT resin. At positions where
oligosaccharide mimic 12 was to be installed SS-tBu protected
cysteine residues were introduced and, following reduction on
solid-phase, the two free thiol groups were capped with 2-
bromoacetyl propargylamide (19).28 After cleavage from the solid-
support we introduced two oligosaccharide mimics employing two
equivalents of 12 per acetylene. Briefly 12 was incubated with
the acetylene-linked peptide in phosphate buffered saline (0.1 M,
pH 8.0) containing the Sharpless tris-benzyltriazolylmethylamine
Cu(I) ligand,11 CuSO4, and sodium ascorbate for 24 h at room
temperature, then 18 was purified from the reaction mixture using
semi-preparative reverse-phase HPLC in 69% yield. 18 could also
be prepared by first reacting 12 with 19 and uniting this thiol-
selective oligosaccharide mimic with the same peptide sequence
displaying free cysteine residues (as described in Fig. 2b employing
N-acetyl glucosamine) in solution with very similar efficiency. 18
was then ligated to Green Fluorescent Protein (GFP) thioester
(Fig. 3b) which was easily obtained by expression of the GFP
gene from the commercially available DNA vector pTYB-1.30 Use
of GFP in this way provided a more glycoprotein-like context for
the oligosaccharide mimic and allowed protein immobilisation
to be easily observed. The native chemical ligation reaction
was conducted in 200 mM sodium phosphate buffer; pH 8.0,
containing 100 mM NaCl, 2% w/v 2-mercaptoethanesulfonic
acid (MESNa), 5 mM tris-carboxyethyl phosphine (TCEP) and
excess 18. SDS-PAGE analysis of the ligation reaction after
48 h (Fig. 3b) showed that the reaction had proceeded to over
50% completion with the ligated product (molecular weight of
approximately 32 kDa) clearly visible. The ligation reaction was
then concentrated and ultimately obtained in 200 ml of Dulbeccos
PBS solution; pH 7.4. This solution was tested for interaction
with the galactose-binding lectin. When compared with GFP
alone, which demonstrated no interaction with the immobilised
This journal is
The Royal Society of Chemistry 2010
Org. Biomol. Chem., 2010, 8, 1351–1360 | 1355
©

View Article Online
Scheme 3 Solid-phase synthesis of a branched oligosaccharide mimics. Reagents/conditions: a) 10% w/v DTT, DMF, 2.5% v/v DIPEA, 48 h. b) 14,
(2 equiv. per thiol), DMF, 2.5% v/v Et3N, 16 h, c) 1,3-dimethylbarbituric acid, Pd(PPh3)4, DMF–CHCl3, 2 h, quant. d) 16 (2.5 equiv.), PyBOP/HOBt,
DIPEA, DMF, 16 h. e) 95% v/v TFA, 2.5% v/v ethanedithiol, 2.5% v/v H2O, 3 h. f) 5% v/v hydrazine hydrate, 10 : 1 H2O–MeOH, 24 h. Overall yield =
47%.
1356 | Org. Biomol. Chem., 2010, 8, 1351–1360
This journal is
The Royal Society of Chemistry 2010
©

View Article Online
lectin, the 18-linked GFP was clearly retained by the beads after
extensive washing with Dulbeccos PBS solution suggesting an
interaction between the lectin and the terminal galactose residues
of 18 (Fig. 3d). Furthermore the fluorescence could be washed
from the beads by incubation with a solution of D-galactose
(100 mM) in Dulbeccos PBS solution, confirming the specificity
of the interaction.
mercially available Nova-Syn TGA resin (Merck Biosciences)
was loaded with Fmoc-Cys(Acm)-OH by double coupling with
HBTU/HOBt in the presence of DIPEA and then extended using
HBTU/HOBt (Fastmoc protocol) in automated fashion using
an Applied Biosystems 433A peptide synthesiser. The coupling
time was 0.5 h. Target sequence: APPR(Pmc)LIC(Acm)E(OtBu)
S(tBu)R(Pmc)VLE(OtBu)R(Pmc)Y(tBu)LLE(OtBu)AK(Boc)E-
¯
¯
¯ ¯ ¯ ¯ ¯ ¯ ¯ ¯ ¯ ¯ ¯
¯ ¯ ¯ ¯ ¯ ¯
(OtBu)AE(OtBu)C(Trt)IT(tBu)T(tBu)AC(Acm)C(Trt)E(OtBu)-
¯
¯
¯ ¯ ¯ ¯ ¯ ¯ ¯
H(Trt)C(Acm)-resin. Underlined residues were double coupled,
Conclusion
and residues shown in bold are substitutions from the wild-type
sequence. The dry resin was then transferred to a glass vial and
exposed to a solution comprised of 95% TFA, 2.5% EDT, 2.5%
H2O (6.0 mL), with stirring for 5 h. The crude product was then
collected by centrifugation at 3000 rpm following precipitation
from cold diethyl ether (50.0 mL). The fully deprotected and
precipitated product was redissolved in 25% aqueous MeCN and
purified by semi-preparative HPLC. The major peak (retention
time = 32.7 min, 35 mg, 18%) was analysed by ESI-MS and
was found to correspond to the desired product. ESI-MS (ES+)
observed mass = 3964.8 Da calculated mass = 3964.6 Da.
Alkylation with sugar iodoacetamides (9). The purified peptide
(30.0 mg 7.57 mmol) was dissolved in 50 mM NH4HCO3;
pH 8.0 (3.0 mL) and N-acetylglucosamine iodoacetamide (15 mg,
39 mmol) was added. The reaction mixture was incubated at room
temperature with shaking for 2 h, neutralised with AcOH then
loaded directly onto a semi-preparative HPLC column to afford
the product (retention time = 30.6 min, 17 mg, 3.78 mmol) as a
white fluffy solid following lyophilisation. ESI-MS (ES+) observed
mass = 4484.8 Da, calculated mass = 4485.1 Da.
Acm deprotection (10). The purified glycopeptide (11 mg,
2.77 mmol) was redissolved in 10% v/v aqueous acetic acid (2.0
ml) and AgOAc (13 mg, 83.2 mmol) was added. After shaking
at room temperature for 30 min DTT was added to a final
concentration of 10% w/v and shaking was continued at room
temperature overnight. The thick yellow precipitate was removed
by centrifugation and the supernatant was loaded directly onto
a semi-preparative HPLC column to afford the product (5 mg,
1.1 mmol) as a white fluffy solid following lyophilisation. ESI-MS
(ES+) observed mass = 4271.8 Da calculated mass = 4271.8 Da.
Thioester formation (11). Peptide 10 (5 mg, 1.1 mmol) was
dissolved in 6 M guanidine.HCl (3.2 mL) and acetic acid 400 mL
followed by MESNa (400 mg) and TCEP.HCl (5 mg). The
reaction mixture was then incubated at 60 ◦C with shaking at
700 rpm in an Eppendorf thermomixer for 48 h. After 48 h the
conversion was estimated as approximately 50% by LC-MS and
the reaction mixture was loaded directly onto a semi-preparative
HPLC column and thioester 11 was isolated along with unreacted
10 (5 mg) as a white fluffy solid following lyophilisation.
Significant progress has been made in a NCL-based route to
neoglycoproteins using thioesters derived from an N→S acyl shift-
triggered reaction. This allowed an EPO neoglycoform, decorated
with simple sugar candidates for enzymatic remodelling, to be
assembled. The next step is to replace simple sugars with oligosac-
charide mimetics such as 12 which was shown to be recognised
and bound by a lectin. This simplified saccharide-display scaffold
was prepared in good yield with minimal purification and can be
fused with protein fragments at pre-defined sites using reliable
thiol targeting or azide/acetylene chemistry. Following native
chemical ligation 12 might ultimately form part of useful protein
therapeutics.
Experimental details
General experimental details
1H NMR spectra were recorded at 500, 350 and 300 MHz, 13C
NMR spectra were recorded at 125, 63 and 75 MHz respectively on
a Bruker instrument. Electrospray mass spectrometry was carried
out on a Waters Acquity UPLC-SQD MS system with an applied
voltage of 60 V. Analytical HPLC was performed on a Dionex
Ulimate 3000 HPLC system equipped with a C18 reverse phase
column (Phenomenex Spherclone ODS (4.6 mm ¥ 250 mm))
employing a gradient of 5–95% acetonitrile containing 0.1%
TFA over 45 min (flow rate of 1.0 mL min-1). Semi-preparative
HPLC was performed using a Phenomenex Jupiter proteo column
(10 mm ¥ 250 mm) and a gradient of 5–60% acetonitrile containing
0.1% TFA over 45 min (flow rate of 4.0 mL min-1). All reagents
and solvents (excluding HPLC solvents) were standard laboratory
grade and used as supplied unless otherwise stated. Where a
solvent was described as dry it was purchased as anhydrous
grade. All resins and Fmoc amino acids for peptide synthesis
were purchased from Merck Biosciences with the exception of
FmocHis(Me)-OH which was obtained from Bachem.
Peptide thioester synthesis
Synthesis and thioesterification of model peptides 1(a/b), 2(a/b),
4(a, b), 6(a, b) and 7(a, b). Synthesis of these compounds
was performed as previously described.15 10 mL aliquots of
each 1.0 mL reaction were taken at 24, 48, and 96 h intervals
and analysed by HPLC. Conversion to product was determined
through comparison of peak area at 230 nm. The identity of the
starting materials and products were confirmed at t = 0 and t =
96 h by LC-MS.
Native chemical ligation. A recombinant EPO fragment com-
prising residues 33–166 (approx 1.0 mg)17,24 was dissolved in 6 M
guanidine hydrochloride containing 300 mM Na phosphate buffer,
pH 7.0 (0.25 mL), containing 20 mM TCEP and 50 mM MPAA
and added to thioester 11 (5 mg and still containing about 50%
10). The reaction was shaken at room temperature for 24 h. A
sample (3.0 mL) was taken for LC-MS analysis after which time the
reaction appeared complete. Further TCEP (10 mM) was added
and the reaction shaken at room temperature for a further 1 h.
Synthesis of EPO residues 1–33 (8). The peptide con-
taining a C-terminal HC-motif was prepared on 0.05 mmol
scale using standard Fmoc SPPS procedures. Briefly, com-
This journal is
The Royal Society of Chemistry 2010
Org. Biomol. Chem., 2010, 8, 1351–1360 | 1357
©

View Article Online
The neoglycoprotein was then partially purified by dilution with
water (10 volumes) followed by incubation on ice for 0.5 h. The
precipitated protein was collected by centrifugation at maximum
speed on a microcentrifuge for 5 min. (the synthetic fragment is
highly water soluble whereas the protein is not). The protein pellet
was resolubilised in 6 M guanidine hydrochloride, treated with
dithiothreitol to a final concentration of 10 mM then incubated
at 37 ◦C for 0.5 h. A single repetition of the precipitation and
resolubilisation process (as above) afforded a protein sample that
was largely free of all synthetic components (10, or 11) though
the process can be repeated further (see supporting information
for more details). Since the aim of the experiment was merely
to verify thioester formation through NCL the product was not
purified further nor refolded.
equiv. DIPEA for 16 h. The process was repeated. Afterwards the
resin was exhaustively washed with DCM and DMF.
Cleavage from the solid support. 17 was cleaved from the resin
by treatment with the cleavage cocktail (95% TFA, 2.5% v/v water
and 2.5% v/v EDT) for 3 h. After 3 h the resin was filtered off
and treated once again with fresh cleavage cocktail for another
3 h. The combined TFA filtrates containing peracetylated 12 were
precipitated with cold diethyl ether and collected by centrifugation
at 3000 rpm for 10 min. The precipitate was dried by flushing with
nitrogen for 30 min. The precipitate was resuspended in a water–
methanol solution (10 : 1) containing 5% v/v hydrazine hydrate
(6.0 ml) and stirred at room temperature for 24 h. 12 was then
was purified by reverse phase semi preparative HPLC (water–
acetonitrile 5–95% over 40 min, 0.1% TFA, tR = 14.2 min) and
obtained as a fluffy white solid (36 mg, 47% yield based on initial
resin loading) after lyophilisation. Analytical data of compound
12: C56H93N13O32S2. MW: 1524.79. Found: (M + 1) 1525.8; (M +
2) 763.5.
Synthesis of oligosaccharide mimic 12
Solid-phase peptide synthesis. The synthesis of compound 12
was conducted on 0.10 mmol scale on Rink Amide MBHA
resin. The Fmoc group was removed by treatment with 20%
(v/v) piperidine in DMF (2 ¥ 5 min), and similarly between
each coupling step. Coupling reactions with Fmoc amino acids
were conducted using 0.50 mmol (5 equiv.) of each amino acid,
HBTU/HOBt and DIPEA for 3 h. The reaction progress was
monitored using the Kaiser ninhydrin test. After each coupling
and deprotection the resin was washed with DCM and DMF
(5 min each).
Synthesis of compound 18
The acetylene-modified peptide (3.60 mg, 2.67 ¥ 10-3 mmol) and 12
(14.30 mg, 9.38 ¥ 10-3 mmol) were dissolved in 0.1 M PBS buffer
solution; pH 8 (1.90 mL). Sodium ascorbate (13.00 mg, 6.66 ¥
10-2 mmol) and tris-benzyltriazolylmethylamine (TBTA) (35.0 mg,
6.66 ¥ 10-2 mmol) were added to the solution. The final suspension
was sonicated for 5 min in a cold water bath. Then, from a
stock solution in PBS, CuSO4·5H2O (0.67 mg, 2.67 ¥ 10-3 mmol,
26.7 mM, 100 mL) was added. The reaction was finally shaken for
Cys-StBu deprotection. DTT (200 mg) was dissolved in dry
DMF (1.8 mL) and 2.5% v/v DIPEA was added. After stirring
for 5 min the solution was transferred to a peptide synthesis vessel
containing resin-bound SStBu protected peptide (13). After 16 h
the resin was filtered off and washed exhaustively with DMF, then
DMF–H2O (1 : 1), DMF and finally DCM. This procedure was
then repeated.
◦
24 h at 20 C and 1200 rpm in an Eppendorf thermomixer. The
reaction was next centrifuged at 14000 rpm for 5 min, and filtered
through celite plug in a Pasteur pipette. The filtrate was purified
by reverse phase semi-preparative HPLC (water–acetonitrile 5%–
95% over 40 min, 0.1% TFA). Fractions containing the product
were lyophilised to afford a fluffy white solid 18 (tR = 16.1 min,
8.10 mg, 69%), and a white solid 12 (tR = 14.2 min) of recovered
starting material (5 mg). C164H268N40O84S8. MW: 4400.6. Found:
(M + 3) 1468.6, (M + 4 (+K)) 1111.5, (M + 4) 1101.7, (M + 5)
881.7. Deconvoluted mass = 4402.8 Da.
N-Acetyl lactosaminyl bromoacetamide couplings. Bromoac-
etamide 14 (2 eq. per thiol: 0.10 mmol ¥ 2 thiols ¥ 2 eq. = 302 mg,
0.40 mmol) was dissolved in DMF (1.0 mL) and NEt3 (84 mL,
0.6 mmol) and transferred to a peptide synthesis vessel containing
resin-bound SStBu deprotected peptide. The reaction was allowed
to proceed for 24 h. After this time, the resin was filtered and
washed exhaustively with DMF and then DCM.
Ligation between green fluorescent protein (GFP)- and EPO
(residues 22–32) containing two oligosaccharide mimetics
Alloc deprotection. Pd(PPh3)4 (58 mg, 0.05 mmol) was dis-
solved in a solution of 1 : 1 DMF–CHCl3 (1.5 mL) and 1,3-
dimethylbarbituric acid (156 mg, 1 mmol, 10 eq) was added. The
yellow solution was transferred to the reaction vessel containing
the resin and shaken at room temperature for 2 h. The resin (now
red in colour) was then washed sequentially with the following
solutions: 0.5% v/v DIPEA in DMF (4 ¥ 2.0 mL); DMF (6 ¥
2.00 mL); 0.5% w/v sodium diethyldithiocarbamate trihydrate in
DMF (4 ¥ 2.0 mL); followed by a final wash with DMF (6 ¥
2.0 mL). The washes returned the resin to its original colour.
The GFP gene was PCR amplified from the commercially available
cloning vector pcDNA3.1/NT-GFP-TOPO and subsequently
cloned into expression vector pTYB-1 using the NdeI/EcoRI
(after silencing of the internal NdeI restriction site) restriction sites.
Positive colonies were fully sequenced across the coding region.
The resulting plasmid was expressed as described as below.
Translated sequence
MASKGEELFTGVVPILVELDGDVNGHKFSVSGEGEGD-
ATYGKLTLKFICTTGKLPVPWPTLVTTFSYGVQCFSRYP-
DHMKRHDFFKSAMPEGYVQERTISFKDDGNYKTRAE-
VKFEGDTLVNRIELKGIDFKEDGNILGHKLEYNYNSH-
NVYITADKQKNGIKANFKIRHNIEDGSVQLADHYQQN-
TPIGDGPVLLPDNHYLSTQSALSKDPNEKRDHMVLLE-
FVTAAGITHGMDELYKEFLEGSS (bold residues are vector
Coupling of 2-acetamido-2-deoxy-3,6-di-O-acetyl-4-carboxy-
methyl-b-D-glucopyranosyl azide. The resin was washed with
DMF (2.0 mL). Coupling reactions with 2-acetamido-2-deoxy-
3,6-di-O-acetyl-4-carboxymethyl-b-D-glucopyranosyl azide, 16
(97 mg, 0.25 mmol, 2.5 equiv.), 2.5 equiv. PyBOP/HOBt and 5
1358 | Org. Biomol. Chem., 2010, 8, 1351–1360
This journal is
The Royal Society of Chemistry 2010
©

View Article Online
derived amino acids) Calculated molecular weight
27616.9 Da.
=
50 ml of conjugate or cleaved GFP at 4 ◦C for 48 h. Beads
were washed again as described and bound GFP-glycopeptide
visualised by exposure to U.V. radiation, using a 4 W UV lamp,
Expression of GFP-Sce intein–chitin binding domain (CBD) fu-
sion protein. Escherichia coli strain BL21(DE3) was transformed
with pTYB1-GFP. A single transformed colony was inoculated
into 10.0 ml of LB medium supplemented with ampicillin to a
final concentration of 100 mg ml-1 and grown to saturation at
37 ◦C with shaking. This culture was used to inoculate 2 ¥ 500 ml
of fresh LB ampicillin medium. Cultures were grown at 37 ◦C
with shaking until A600 reached 0.6. Expression of the GFP-Sce
intein-CBD fusion was induced by addition of isopropyl-thio-b-D-
galactopyranoside (IPTG) to each culture at a final concentration
of 1 mM. Induction was allowed to proceed for 5 h at 30 ◦C with
shaking. Bacteria were harvested by centrifugation at 8000 rpm
in a JA 10.500 rotor (Beckman) for 15 min at 4 ◦C and stored at
-20 ◦C until required.
R
model UVGL-25 Mineralightꢀ Lamp at 254 nm in a darkened
cabinet. To demonstrate the carbohydrate-specific nature of the
observed lectin interaction, bound glycoprotein–GFP conjugate
was competed off the lectin beads by mixing with a 100 mM
solution of D-(+)-galactose in Dulbecco’s PBS, after which beads
were washed and visualised as described above.
SDS PAGE analysis. For GFP ligation studies, samples were
mixed with sodium dodecyl sulfate polyacrylamide gel elec-
trophoresis (SDS PAGE) loading buffer, boiled for 5 min and
resolved by SDS PAGE on pre-cast 12% polyacrylamide gels using
the Protean III system (BioRad). Preparations of erythropoietin
were precipitated by addition of 20 sample volumes of methanol
and acetone solution (1 : 1 v/v), incubated at -20 ◦C for 30 min and
collected by centrifugation. Samples were resolubilised in 20 ml of
8 M urea and boiled with loading buffer as described. Proteins were
resolved by SDS PAGE on pre-cast 18% polyacrylamide gels as
described. Electrophoresed proteins were visualised by Coomassie
staining.
Purification of GFP-Sce intein-CBD fusion protein. Pelleted
cells were resuspended in 10.0 ml of column buffer (20 mM Tris-
HCl; pH 8.0, 500 mM NaCl, 1 mM EDTA, 1 mM PMSF) and
lysed by sonication on ice (20 cycles of 30 s sonication separated by
30 s cooling on ice). The crude lysate was cleared by centrifugation
at 12000 rpm in a JA 25.50 rotor (Beckman) for 15 min at 4 ◦C. The
supernatant was loaded onto a column containing 1.5 ml of chitin
beads (NEB) pre-equilibrated with 20 ml of column buffer and
allowed to drain by gravity flow. Unbound proteins were washed
from the beads with 50 ml of column buffer prior to native chemical
ligation. For control experiments, on-column cleavage of GFP was
performed by addition of 10 ml of cleavage buffer (column buffer
containing 50 mM DTT). Cleavage buffer was allowed to drain
after which the column was immediately sealed and incubated at
4 ◦C for 18 h. Cleaved GFP was eluted in 15 ml of cleavage buffer
containing 1 mM PMSF.
Acknowledgements
The authors acknowledge financial support from BBSRC, The
Wellcome Trust and The Royal Society.
Notes and references
1 (a) P. E. Dawson, T. W. Muir, I. Clark-Lewis and S. B. Kent, Science,
1994, 266; (b) D. Macmillan, Angew. Chem., Int. Ed., 2006, 45, 7668;
(c) D. Rauh and H. Waldmann, Angew. Chem., Int. Ed., 2007, 46, 826;
(d) J.-P. Pellois and T. W. Muir, Curr. Opin. Chem. Biol., 2006, 10, 487;
(e) V. Muralidharan and T. W. Muir, Nat. Methods, 2006, 3, 429; (f) C.
Chatterjee, R. K. McGinty, J.-P. Pellois and T. W. Muir, Angew. Chem.,
Int. Ed., 2007, 46, 2814; (g) S. B. H. Kent, Chem. Soc. Rev., 2009, 38,
338–351; (h) R. J. Payne and C.-H. Wong, Chem. Commun., 2010, 46,
21–43.
2 (a) D. Macmillan and C. R. Bertozzi, Tetrahedron, 2000, 56, 9515;
(b) T. J. W. Tolbert and C.-H. Wong, J. Am. Chem. Soc., 2000, 122,
5421; (c) D. Macmillan and C. R. Bertozzi, Angew. Chem., Int. Ed.,
2004, 43, 1355; (d) B. Wu, J. Chen, J. D. Warren, G. Chen, Z. Hua
and S. J. Danishefsky, Angew. Chem., Int. Ed., 2006, 45, 4116; (e) Y.-Y.
Yang, S. Ficht, A. Brik and C.-H. Wong, J. Am. Chem. Soc., 2007,
129, 7690; (f) N. Yamamoto, Y. Tanabe, R. Okamoto, P. E. Dawson
and Y. Kajihara, J. Am. Chem. Soc., 2008, 130, 501; (g) C. Piontek,
D. V. Silva, C. Heinlein, C. Po¨hner, S. Mezzato, P. Ring, A. Martin,
F. X. Schmid and C. Unverzagt, Angew. Chem., Int. Ed., 2009, 48, 1941;
(h) C. Piontek, P. Ring, O. Harjes, C. Heinlein, S. Mezzato, N. Lombana,
C. Po¨hner, M. Pu¨ttner, D. V. Silva, A. Martin, F. X. Schmid and C.
Unverzagt, Angew. Chem., Int. Ed., 2009, 48, 1936; (i) P. Nagorny, B.
Fasching, X. Li, G. Chen, B. Aussedat and S. J. Danishefsky, J. Am.
Chem. Soc., 2009, 131, 5792.
Native chemical ligation of GFP to unnatural glycopeptide. For
native chemical ligation with 18, 5 ml of ligation buffer (200 mM
sodium phosphate pH 8.0, 100 mM NaCl, 5 mM TCEP, 2%
(w/v) MESNa) was added to the chitin bead-bound GFP and
allowed to drain by gravity flow after which the column was
immediately sealed. 18 (4.2 mg) was resuspended in 100 ml of
ligation buffer and added to the chitin beads. This reaction
was mixed at intervals at room temperature for 48 h. Proteins
were eluted from the column in 2 ml of ligation buffer (without
MESNa). Ligation buffer was removed from eluted proteins and
replaced with Dulbecco’s phosphate buffered saline (PBS) pH 7.4
by centrifugation through a 30 kDa molecular weight cut-off
filter (VivaScience) during which procedure any residual (un-
ligated) glycopeptide was removed from the ligation product.
Following buffer exchange, the GFP-glycopeptide conjugate was
resuspended in 200 ml of Dulbecco’s PBS.
3 (a) P. H. Seeberger, Chem. Soc. Rev., 2008, 37, 19; (b) X. Zhu and R. R.
Schmidt, Angew. Chem., Int. Ed., 2009, 48, 1900.
4 (a) P. M. St Hilaire and M. Meldal, Angew. Chem., Int. Ed., 2000, 39,
1162; (b) L. L. Kiessling, J. E. Gestwicki and L. E. Strong, Angew.
Chem., Int. Ed., 2006, 45, 2348; (c) S.-Y. Chen, S. Cressman, F. Mao,
H. Shao, D. W. Low, H. S. Beilan, E. N. Cagle, M. Carnevali, V.
Gueriguian, P. J. Keogh, H. Porter, S. M. Stratton, M. C. Wiedeke,
L. Savatski, J. W. Adamson, C. E. Bozzini, A. Kung, S. B. H. Kent,
J. A. Bradburne and G. G. Kochendoerfer, Chem. Biol., 2005, 12, 371;
(d) S. I. van Kasteren, H. B. Kramer, H. H. Jensen, S. J. Campbell, J.
Kirkpatrick, N. J. Oldham, D. C. Anthony and B. G. Davis, Nature,
2007, 446, 1105; (e) S. I. v. Kasteren, S. J. Campbell, S. Serres, D. C.
Anthony, N. R. Sibsonc and B. G. Davis, Proc. Natl. Acad. Sci. U. S. A.,
2009, 106, 18.
Lectin interaction studies. To demonstrate carbohydrate–
lectin interaction, 100 ml of cross-linked 4% beaded agarose dis-
playing lectin from Ricinus communis (Sigma) was pre-equilibrated
in Dulbecco’s PBS and added to 50 ml of GFP–glycopeptide
conjugate or DTT-hydrolysed GFP alone as negative control.
Solutions were gently mixed for 1 h at room temperature. Beads
were washed 5 times with 1 ml of Dulbecco’s PBS to remove
unbound conjugate. Residual buffer was carefully removed after
the final wash and beads were further incubated with an additional
This journal is
The Royal Society of Chemistry 2010
Org. Biomol. Chem., 2010, 8, 1351–1360 | 1359
©

View Article Online
5 (a) M. R. Pratt and C. R. Bertozzi, J. Am. Chem. Soc., 2003, 125, 6149;
(b) I. S. Carrico, B. L. Carlson and C. R. Bertozzi, Nat. Chem. Biol.,
2007, 3, 321; (c) E. M. Sletten and C. R. Bertozzi, Angew. Chem., Int.
Ed., 2009, 48, 6974.
6 (a) R. M. Bill and S. L. Flitsch, Chem. Biol., 1996, 3, 145; (b) T. Buskas,
S. Ingale and G.-J. Boons, Glycobiology, 2006, 16, 113R; (c) H. Hojo
and Y. Nakahara, Peptide Sci, 2007, 88, 308; (d) D. P. Gamblin, E. M.
Scanlan and B. G. Davis, Chem. Rev., 2009, 109, 131.
7 (a) B. Li, H. Song, S. Hauser and L. X. Wang, Org. Lett., 2006, 8,
3081; (b) W. Huang, H. Ochiai, X. Zhang and L.-X. Wang, Carbohydr.
Res., 2008, 343, 2903; (c) H. Ochiai, W. Huang and L.-X. Wang, J. Am.
Chem. Soc., 2008, 130, 13790; (d) T. B. Parsons, J. W. B. Moir and A. J.
Fairbanks, Org. Biomol. Chem., 2009, 7, 3128.
8 (a) B. L. Pentelute, Z. P. Gates, V. Tereshko, J. L. Dashnau, J. M.
Vanderkooi, A. A. Kossiakoff and S. B. H. Kent, J. Am. Chem. Soc.,
2008, 130, 9695; (b) K. Mandal, B. L. Pentelute, V. Tereshko, A. A.
Kossiakoff and S. B. H. Kent, J. Am. Chem. Soc., 2009, 131, 1362;
(c) G. G. Kochendoerfer, S.-Y. Chen, F. Mao, S. Cressman, S. Traviglia,
H. Shao, C. L. Hunter, D. W. Low, E. N. Cagle, M. Carnevali, V.
Gueriguian, P. J. Keogh, H. Porter, S. M. Stratton, M. C. Wiedeke, J.
Wilken, J. Tang, J. J. Levy, L. P. Miranda, M. M. Crnogorac, S. Kalbag,
P. Botti, J. Schindler-Horvat, L. Savatski, J. W. Adamson, A. Kung,
S. B. H. Kent and J. A. Bradburne, Science, 2003, 299, 884.
9 J. Xiao and T. J. Tolbert, Org. Lett., 2009, 11, 4144.
15 J. Kang and D. Macmillan, Org. Biomol. Chem., 2009, 7, 4918.
16 (a) T. Murase, T. Tsuji and Y. Kajihara, Carbohydr. Res., 2009, 344,
762; (b) K. Hirano, D. Macmillan, K. Tezuka, T. Tsuji and Y. Kajihara,
Angew. Chem., Int. Ed., 2009, 48, 9557–9560.
17 See supporting information for further details.
18 Note that 7 (Fig. 1) is not a thioester though conversion is plotted on
the same graph for comparison.
19 (a) Z. Du, P. T. Shemella, Y. Liu, S. A. McCallum, B. Pereira, S. K.
Nayak, G. Belfort, M. Belfort and C. Wang, J. Am. Chem. Soc., 2009,
131, 11581; (b) C. Ludwig, D. Schwarzer and H. D. Mootz, J. Biol.
Chem., 2008, 283, 25264.
20 N. Suree, C. K. Liew, V. A. Villareal, W. Thieu, E. A. Fadeev, J. J.
Clemens, M. E. Jung and R. T. Clubb, J. Biol. Chem., 2009, 284, 24465.
21 Coupling of cysteine to hydroxy-functionalised resins employing
HBTU/HOBt may well lead to racemisation of the C-terminal cysteine
though this should be of little consequence since the cysteine is excised.
Cysteine racemisation can also be minimised by using a trityl resin, e.g.
NovaSYN TGT.
22 S. Elliott, D. Chang, E. Delorme, T. Eris and T. Lorenzini, J. Biol.
Chem., 2004, 279, 16854.
23 The D8E mutation is non-essential as hydrolysis at Asp is slow relative
to thioesterification but the conservative mutation is unlikely to affect
bioactivity as this region does not bind the EPO receptor.
24 D. Macmillan and L. Arham, J. Am. Chem. Soc., 2004, 126,
9530.
10 (a) N. J. Davis and S. L. Flitsch, Tetrahedron Lett., 1991, 32, 6793; (b) D.
Macmillan, A. M. Daines, M. Bayrhuber and S. L. Flitsch, Org. Lett.,
2002, 4, 1467.
11 T. R. Chan, R. Hilgraf, K. B. Sharpless and V. V. Fokin, Org. Lett.,
2004, 6, 2853.
25 E. C. B. Johnson and S. B. H. Kent, J. Am. Chem. Soc., 2006, 128, 6640.
26 C. Unverzagt and J. Seifert, Tetrahedron Lett., 2000, 41, 4549.
27 A. Bendele, J. Seely, C. Richey, G. Sennello and G. Shopp, Toxicol. Sci.,
1998, 42, 152.
12 J. P. Richardson and D. Macmillan, Org. Biomol. Chem., 2008, 6, 3977.
13 Y. Shin, K. A. Winans, B. J. Backes, S. B. H. Kent, J. A. Ellman and
C. R. Bertozzi, J. Am. Chem. Soc., 1999, 121, 11684.
14 J. Kang, J. P. Richardson and D. Macmillan, Chem. Commun., 2009,
407.
28 D. Macmillan, J. Blanc, J. PCT, Int. Appl., 2008, WO2008001109.
29 (a) S. T. Anisfeld and P. T. Lansbury, J. Org. Chem., 1990, 55, 5560;
(b) J. S. Miller, V. Y. Dudkin, G. J. Lyon, T. W. Muir and S. J.
Danishefsky, Angew. Chem., Int. Ed., 2003, 42, 431.
30 www.neb.com.
1360 | Org. Biomol. Chem., 2010, 8, 1351–1360
This journal is
The Royal Society of Chemistry 2010
©