Propargyl-Assisted Selective Amidation Applied in C-terminal Glycine Peptide Conjugation

Article abstract of DOI:10.1002/chem.201604247

Alkyl esters, such as propargyl esters, typically lack the electron-withdrawing inductive effects needed to participate in nucleophilic acyl substitution reactions. Herein, we report an unusual observation in which glycine propargyl ester derivatives displayed selective, base-independent reactivity towards linear alkylamines under mild, metal-free conditions. Through global reaction route mapping (GRRM) modeling calculations, it is predicted that these observations may be governed by factors related to hydrogen-bonding and intermolecular interactions, rather than electron-withdrawing inductive effects. Based on this concept of propargyl-assisted selective amidation, a direct application was made to develop a novel site-specific C-terminal glycine peptide bioconjugation technique as a proof-of-concept, which relies upon the selective reactivity of glycine propargyl esters over that of aspartate and glutamate side-chain-linked propargyl esters.

Full text of DOI:10.1002/chem.201604247

A Journal of
Accepted Article
Title: Propargyl-assisted Selective Amidation Applied in C-terminal
Glycine Peptide Conjugation
Authors: Kenward King Ho Vong, Satoshi Maeda, and Katsunori
Tanaka
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: Chem. Eur. J. 10.1002/chem.201604247
Link to VoR: http://dx.doi.org/10.1002/chem.201604247
Supported by

10.1002/chem.201604247
Chemistry - A European Journal
Propargyl-assisted Selective Amidation Applied in C-terminal
Glycine Peptide Conjugation
Kenward King Ho Vong1, Satoshi Maeda,2 and Katsunori Tanaka1,3,4
*
1Biofunctional Synthetic Chemistry Laboratory, RIKEN, 2-1 Hirosawa, Wako-shi, Saitama 351-0198,
Japan
2Department of Chemistry, Faculty of Science, Hokkaido University, Sapporo, 060-0810, Japan
3Biofunctional Chemistry Laboratory, A. Butlerov Institute of Chemistry, Kazan Federal University, 18
Kremlyovskaya Street, Kazan 420008, Russia
4JST, PRESTO, 2-1 Hirosawa, Wako-shi, Saitama 351-0198, Japan
* To whom correspondence should be addressed. E-mail: kotzenori@riken.jp
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/chem.2014xxxxx.
1
This article is protected by copyright. All rights reserved.

10.1002/chem.201604247
Chemistry - A European Journal
ABSTRACT
Alkyl esters, such as propargyl esters, typically lack the electron-withdrawing inductive effects needed
to participate in nucleophilic acyl substitution reactions. Herein, we report an unusual observation where
glycine propargyl ester derivatives displayed selective, base-independent reactivity towards linear
alkylamines under mild, metal-free conditions. Through global reaction route mapping (GRRM) model-
ing calculations, it is predicted that these observations may be governed by factors related to hydrogen
bonding and intermolecular interactions, rather than electron-withdrawing inductive effects. Based on
this concept of propargyl-assisted selective amidation, a direct application was made to develop a proof-
of-concept for a novel site-specific C-terminal glycine peptide bioconjugation technique, which relies
upon the selective reactivity of glycine propargyl esters over that of aspartate and glutamate side chain-
linked propargyl esters.
INTRODUCTION
Amide bond formation by nucleophilic acyl substitution is among one of the most well known and
practiced reactions in synthetic organic chemistry.[1] The principle driving force of this reaction is the
energetic stability of the leaving group after it has been displaced by nucleophilic addition. As such,
compounds with weak basic nature (ex/ halides, tosylates, and N-hydroxysuccinimide) are regarded as
the most effective leaving groups (Figure 1A).
Propargyl esters (propargyl alcohol pKa = 13.6), like most alkyl esters, are considered to be poor
leaving groups in amidation reactions (Figure 1B). To remove these groups, harsh conditions (high heat,
strong base) or the addition of catalysts (ex/ TiCl4,[2] Sb(OEt)3,[3] Zr(Ot-Bu)4-HOAt,[4] and a host of oth-
er inorganic and organic catalysts[5]) are required. Currently, propargyl esters have found limited usage
as acid and base stable protecting groups for carboxylates of amino acids.[6] As a side note, there is one
report of a propargyl ester derivative as a weak activating group in ester aminolysis.[7] However, con-
firmatory experiments suggest that this result may in fact be due to phthalimide ring opening (See Suppl.
Figure 1).
2
This article is protected by copyright. All rights reserved.

10.1002/chem.201604247
Chemistry - A European Journal
This paper largely describes the discovery of the unexpected selective aminolysis between glycine
propargyl esters and linear alkylamines, which has been observed to proceed under mild, metal-free
conditions (Figure 1C). Given that similar esters (comparable electron-withdrawing properties of sub-
stituents and leaving groups) are relatively unreactive, observed glycine propargyl ester aminolysis is
not believed to be mainly driven by electron-withdrawing inductive effects. GRRM-based computation-
al modeling suggests that hydrogen bonding/intermolecular interactions of substrate clusters may in-
stead play a more significant role in lowering the activation barrier for "propargyl-assisted selective
amidation". To address the utility of this reaction, a proof-of-concept for a novel site-specific C-terminal
glycine peptide bioconjugation method was designed and explored.
Figure 1. Ester aminolysis. A) Activated esters are known to be highly reactive in amidation reactions.
B) Alkyl esters, such as methyl and propargyl esters, are relatively stable and are not known to easily be
involved in amidation reactions. C) Glycine propargyl esters (ex/ 1a) were found to have unexpected
reactivity with linear alkylamines (ex/ hexylamine 3a).
RESULTS AND DISCUSSION
Initial experiments carried out in our laboratory discovered that Boc-glycine propargyl ester 1a and
hexylamine 3a formed an amide bond at room temperature, open to air, and in the absence of a base cat-
alyst (Figure 1C). Typically, yields range from 50-84% for 16 hr under various solvents that include
tetrahydrofuran, methanol, dioxane, as well as mixtures of dioxane in water (see Suppl Table 1). Exper-
iments also showed higher yields were obtained with longer reaction times and higher temperatures.
3
This article is protected by copyright. All rights reserved.

10.1002/chem.201604247
Chemistry - A European Journal
By comparing leaving groups of similar electron-withdrawing properties (pKa in the range of 13-14),
the importance of the propargyl alcohol leaving group is highlighted (Figure 2). Given that ester 1a pro-
duces significantly higher amidation yields than esters 1b-d, it can be speculated that in addition to be-
ing weakly electron-withdrawing, the propargyl moiety may also play a previously undescribed role that
accelerates aminolysis.
Figure 2. Comparison of the reactivities between different leaving groups of similar pKa in aminolysis
with hexylamine 3a. Leaving groups propargyl alcohol 1a, hydroxyacetone 1b, 1-(2-
hydroxyethyl)imidazole 1c, and 2-fluoroethanol 1d were used. Reactions were standardized to
0.25mmol of amine and ester in 1 ml of THF (250 mM) and stirred for 16 hours.
To gain better insight on the propargyl alcohol as a leaving group, various propargyl ester derivatives
were also tested (Table 1). In cases of C-1 carbon substitution (entries 2, 3) or increasing the carbon
backbone between the alkyne and ester moieties (entries 4, 5), a dramatic decrease in yields was ob-
served. In addition, the use of an allyl moiety (entry 6) or propargyl amide (entry 7) was also detri-
mental to reaction yields. With these results in mind, it can be suggested that the position and
derivatization surrounding the alkyne moiety plays a huge factor for reactivity. The only tolerable point
of derivatization appears to be the terminal alkyne position, where phenyl substitution (71% with 3a,
entry 8) yielded similar results to the underivatized one.
4
This article is protected by copyright. All rights reserved.

10.1002/chem.201604247
Chemistry - A European Journal
Table 1. Exploring propargyl alcohol leaving group
R
entry
1
substrate
yield of 41
conv2
80%
pKa structure
13.6
13.3
78%
1a
2
30%
35%
1e
3
4
5
6
7
8
9
13.3
14.2
14.8
14.4
19%
18%
4%
22%
20%
9%
1f
1g
1h
1i
11%
NR
15%
0%
1j
12.9
15.7
71%
NR
72%
0%
1k
1l
1Isolated yields, 2conversion calculated by isolation of remaining ester starting material; NR = no reac-
tion. Reactions were standardized to 0.25 mmol of amine and ester in 1 ml of solvent (250 mM).
In table 2, the next focal point was to address the significance of the ester substituent, with particular
emphasis on the role of the α-carbon amide bond of 1a as an electron-withdrawing group. Observations
show that by increasing the carbon chain length from the amide moiety, the reactivity of the ester is
greatly reduced (entries 2-3). Similar reductions in activities were also observed by replacing the amide
bond altogether with simple alkanes (entries 4-6). As expected, the importance of the α-carbon amide
bond as an electron-withdrawing group is confirmed by the restoration of activity with the α-methoxy-
containing 2f (77%, entry 7) derivative. In testing various glycine-based derivatives 2g-j, it was noted
that these derivatives were tolerated without significant compromise in reactivity (entries 8-11). Surpris-
ingly, however, replacing the glycine with other amino acid-based moieties (alanine, threonine, phenyl-
alanine, proline, aspatic acid, and glutamic acid, entries 12-17) lead to a dramatic reduction in reactivity.
These results suggest that substitution at the α-carbon disrupts reactivity, likely from steric hindrances.
5
This article is protected by copyright. All rights reserved.

10.1002/chem.201604247
Chemistry - A European Journal
Table 2. Exploring the ester substituents
entry
1
sub
R
prod
yield1
78%
conv2
80%
1a
4
2
3
4
5
7%
NR
NR
NR
10%
0%
0%
0%
2a
2b
2c
2d
5
-
-
-
6
7
-
NR
10%
80%
2e
2f
77%
6
Glycine-based
8
9
60%
78%
61%
80%
2g
2h
7
8
10
11
74%
47%
75%
52%
2i
2j
9
10
Other amino acid-based
12
13%
NR
19%
0%
2k
11
13
-
2l
14
15
-
-
NR
NR
0%
0%
2m
2n
16
-
NR
9%
2o
17
-
NR
0%
2p
1Isolated yields, 2conversion calculated by isolation of remaining ester starting material; NR = no reac-
tion. Reactions were standardized to 0.25 mmol of amine and ester in 1 ml of solvent (250 mM).
In table 3, various amine substrates were tested with 1a to identify the substrate scope. In general,
substrates with the best reactivity were linear alkylamines, such as compounds 3a-g (entries 1-7). Func-
tionalization of the alkyl chains appears to be tolerable to terminal alcohols, tertiary amines, and azides
(62-84% yields, entries 4-7). However, introduction of a terminal carboxylic acid (entries 8-9) appears
to abolish activity.
Evidence to support amine selectivity was collected by testing other nucleophilic groups. An alkyl
thiol 3j (entry 10) and an alkyl alcohol 3k (entry 11) displayed no observable reactivity. This is further
supported with the substrates 3d and 3e (entries 4-5), where reactivity was observed to be amine-
selective despite the presence of a terminal alcohol.
6
This article is protected by copyright. All rights reserved.

10.1002/chem.201604247
Chemistry - A European Journal
Table 3. Probing amine substrate scope
entry
amine
X─R
prod
yield1
conv2
Linear alkylamines
1
78%3
55%3
65%3
65%4
62%4
84%4
62%4
NR4
80%
55%
69%
70%
65%
95%
65%
0%
3a
4
2
3b
12
13
14
15
16
17
-
3
3c
4
3d
5
3e
6
3f
7
3g
8
3h
9
-
NR4
0%
3i
10
-
-
NR4
NR4
0%
0%
3j
11
3k
Branched/Bulky Amines
12
5%3
NR3
13%
0%
3l
18
13
-
3m
14
15
16
17
n = 0
n = 1
n = 2
n = 3
-
NR3
12%3
30%3
59%3
0%
3n
3o
3p
3q
15%
30%
63%
19
20
21
18
-
NR3
0%
3r
Biological Amines
R3
H
H
Ac
H
H
R4
H
Bn
R5
19
20
21
22
23
24
tBuO
tBuO
HO
MeO
MeO
MeO
2%3
NR3
NR4
NR4
NR4
NR4
2%
3%
0%
0%
0%
0%
3s
3t
3u
3v
3w
3x
22
-
-
-
-
(CH2)4NH2
(CH2)4NH2
CH2OH
CH2SH
H
-
25
26
glutathione
human serum albumin
-
-
NR4
NR5
0%
0%
3y
3z
3
1Isolated yields, 2conversion calculated by isolation of remaining ester starting material; THF; 450:50
dioxane/PBS Buffer pH 7.4; 5 overnight incubation at 37°C with 2 μmol of protein and ester in 200 μl of
5% DMSO in PBS Buffer pH 7.4 (10 μM); NR = no reaction. Generally, reactions were standardized to
0.25 mmol of amine and ester in 1 ml of solvent (250 mM).
7
This article is protected by copyright. All rights reserved.

10.1002/chem.201604247
Chemistry - A European Journal
To expand the scope of amine substrates, a series of branched and biological amines were also ex-
plored. Branched and bulky alkylamines 3l and 3m (entries 12-13) showed little to no reactivity, leading
us to believe that bulky substituents may sterically hinder reactivity. To further investigate, the phenyl
alkylamines 3n-q (entries 14-17) were tested, which showed a correlation between reactivity and in-
creasing chain length distance between the amine and phenyl substituents. Also tested was secondary
amine 3r (entry 18), which as expected, did not display any activity.
For biological amines, representative derivatives of selected amino acids (glycine, phenylalanine, ly-
sine, serine, and cysteine, entries 19-24) and the peptide glutathione 3y (entry 25) were shown to have
none or negligible reactivity with ester 1a, fueling speculations that this reaction would not proceed with
proteins. As expected, human serum albumin did not react with ester 1a following overnight incubation
at 37°C (entry 26).
As a means to explore the base-independency of this reaction, amines 3a, h, u were selected for a se-
ries of experiments (Table 4). In the first set of experiments, the reactivity of the amine substrates in the
absence of base was compared between propargyl ester 1a (entries 1, 4, 6) and N-hydroxysuccinimide
ester 1m (entries 8, 10, 12). Observations show that all amines reacted with ester 1m, whereas only
amine 3a reacted with ester 1a. As a complement to these experiments, 1 equivalent of a non-
nucleophilic base (triethylamine or pyridine) was added to the reactions. As expected, all yields with the
N-hydroxysuccinimide ester 1m (entries 9, 11, 13) increased. However, with propargyl ester 1a (entries
2, 3, 5, 7), 3h and 3u largely remained unreactive while yields with 3a surprisingly decreased. As a side
note, base addition also leads to the competing reaction: hydrolysis of the propargyl ester (entries 2, 5,
7). The discrepancy seen between starting material conversion and yield is largely attributed to this ef-
fect.
8
This article is protected by copyright. All rights reserved.

10.1002/chem.201604247
Chemistry - A European Journal
Table 4. Probing the role of base
ester (R)
amine (R)
prod entry
1
base3
yield1
covn2
-
83%
39%
41%
93%
94%
51%
2
3
NEt3
pyr
4
3a
4
5
-
NR
NR
0%
78%
-
NEt3
3h
1a
6
7
-
NR
NR
0%
62%
-
NEt3
3u
3a
8
9
-
88%
100%
100%
100%
4
NEt3
10
11
-
77%
93%
100%
100%
23
NEt3
3h
1m
12
13
-
59%
66%
100%
100%
24
NEt3
3u
1Isolated yields, 2conversion calculated by isolation of remaining ester starting material; 3either 1 equiv
of triethylamine (NEt3) or pyridine addition (pyr); NR = no reaction. Reactions were standardized to
0.25 mmol of amine and ester in 1 ml of solvent (250 mM).
Overall, given the contrasting observations against the rules of traditional, base-catalyzed
nucleophilic acyl substitution, a different approach to elucidate a possible reaction mechanism was
sought.
In the current state of computational chemistry, a majority of the modeling algorithms generally fol-
low a structure-optimization approach, which relies on predetermined coordinates and configurations of
a "rough" transition state to calculate an "optimal" transition state with the most stable energies and ge-
ometric structures. For chemical reactions where the transition state can be insinuated from similar reac-
tions, structure-optimization presents the most practical approach. However, in cases where the probable
reaction mechanism is unknown (ex/ glycine propargyl ester selectivity with linear alkylamines), struc-
ture-optimization calculations are burdened with the non-systematic bias of arbitrarily estimating the
"rough" transition state.
Global reaction route mapping (GRRM) is a recent computational approach that allows chemical re-
actions to be determined in a systematic and unbiased manner.[8] Based on anharmonic downward dis-
9
This article is protected by copyright. All rights reserved.

10.1002/chem.201604247
Chemistry - A European Journal
tortion following (ADDF) and artificial force induced reaction (AFIR) approaches,[9] quantum chemical
calculations can determine potential energy surfaces of all applicable isomers and isomerization path-
ways, which can be used to identify potential transition states and intermediates of the reaction path.
The utility and significance of this approach is highlighted in a few mechanistic studies.[10]
In this study, an AFIR-based unbiased search was done for the reaction between N-acetyl glycine
propargyl ester 2h and 1 or 2 molecular equivalents of N-methylamine (used as a hexylamine 3a re-
placement to shorten calculation times), where all the calculations were done using GRRM code com-
bined with Gaussian 09 program.[11] The obtained best path was further studied for reactions with differ-
ent reactant molecules (See Suppl. Figure 2-3).
One of the significant revelations from the transition state calculations is highlighted in figure 3A,
which calculated the bond lengths between the carbon of the ester and the oxygen of the leaving groups.
For the propargyl alcohol leaving group, the bond length is calculated to be about 1.54Å, which is
roughly 0.02-0.04Å longer when compared to other leaving groups (ex/ methanol, 2-fluoroethanol, 1-(2-
hydroxyethyl)imidazole, and hydroxyacetone). As a consequence of this added bond length, the C-1 hy-
drogen on the propargyl moiety appears to be better positioned and oriented to hydrogen bond with the
ester oxygen, potentially stabilizing the transition state. A possible reasoning for this effect may be due
to the planar nature and p orbitals of the alkyne group. With this theory in mind, the observed decrease
in reactivities for allyl ester 1i and derivatives 1g-h (entries 4-6, Table 2) can be justified due to the
suboptimal displacement of the alkyne moiety.
Since transition state stabilization likely also relies on other factors, additional hydrogen bonding in-
teractions were identified (Figure 3B). In calculations using 2:1 substrate ratio of amine to ester, one
hydrogen bonding interaction of note is found between the oxygen of the α-carbon amide hydrogen
bond and a hydrogen on a non-reacting amine molecule. Since calculations show that the activation bar-
rier for a 2:1 substrate reaction (ΔG = 71.6 kJ/mol for transition state) is significantly lower than a 1:1
substrate reaction (ΔG = 114.8 kJ/mol for transition state) as shown in supporting information (see
Supp. Figure 2-3), it can be strongly suggested that non-reacting amines may play a substantial role in
10
This article is protected by copyright. All rights reserved.

10.1002/chem.201604247
Chemistry - A European Journal
stabilizing the transition state through a hydrogen bonding network. However, experimental data also
shows that replacement of the amide to a ketone largely abolishes activity (entry 6, Table 2). This leads
us to believe that the α-carbon amide may also have an additional role as a weak electron withdrawing
group to pull electron density from the ester, thereby increasing reactivity.[12]
With the theory that multiple amines are involved in a hydrogen bonding network, it would be rea-
sonable to assume that displacing or interrupting their formation will have detrimental effects for reac-
tivity. As such, the observed decrease following base addition (entries 1-3, Table 4) may likely be the
result of solvent ionization (ex/ H2O) or displacement of relatively weak hydrogen bonds. This further
explains why even the addition of the weak base pyridine (pKa of 5.25) could reduce yields, despite the
greater intrinsic basicity of hexylamine (pKa 10.56).
Another important consideration for multiple amine substrates is the potential influence of intermo-
lecular interactions. To address the observed selectivity for linear alkylamines, it can be hypothesized
that when multiple substrates hydrogen bond with the ester, a tightly-packed amine cluster is formed.
Intermolecular interactions may potentially stabilize or destabilize these clusters depending on the type
of amine substrate. For example, when amines are linear and largely nonpolar (Figure 3C), hydrophobic
interactions between linear substituents may promote amine cluster stabilization. However, this en-
hanced stabilization from intermolecular interactions would likely be offset with bulky amines (ex/ 3l-
r), as steric hindrances will likely disrupt ideal hydrogen bonding distances. Also, in the case of sub-
strates containing carboxylic acids (ex/ 3h-i) or esters (ex/ 3s-y), factors such as charge repulsion or in-
tramolecular interactions are likely to affect amine substrate cluster formation.
11
This article is protected by copyright. All rights reserved.

10.1002/chem.201604247
Chemistry - A European Journal
Figure 3. Modeling study analysis. A) Comparison of the C-O bond lengths (between carbon of the es-
ter and oxygen of the leaving groups) and hydrogen bonding distances (between oxygen of the ester and
hydrogen of the C-1 propargyl moiety) for calculated TS2/3 states of various esters. Image is a close-up
for the TS2/3 state of ester 2h. B) Speculated key hydrogen bonding interactions for the 2:1 (amine to
ester) substrate TS2/3 state of ester 2h. C) Representation of hypothesized intermolecular interactions
between linear and non-polar substrate substituents and their potential role in stabilizing amine clusters.
Given the recent advances in N-terminal protein bioconjugation,[13] the need to further develop C-
terminal bioconjugation techniques is highlighted. Unlike the N-terminus, it is widely acknowledged
that the C-terminus of proteins has greater flexibility (due to the C-terminal tail) and exposure to solvent.
Targeting these sites, in effect, will decrease the likelihood of interfering with biological activity.
Currently, literature examples of chemical-based strategies for C-terminal bioconjugation are either
based upon either thioester- or ester-mediated nucleophilic acyl substitutions.[14] For thioester-based
strategies, numerous techniques can be used to displace the corresponding leaving groups, which in-
cludes transthioesterification followed by S,N-acyl shift by either cysteine[15] or synthetic probes,[16]
nucleophilic addition by oxyamino-based compounds,[17] traceless staudinger ligation-based amide for-
mation.[18] In addition, research has also shown the use of thioacids as an viable alternative.[19]
With ester-based strategies, on the other hand, the only available methods are through carbodiimide-
induced activated esters and salicylaldehyde ester-induced ligation.[20] One of the advantages of working
with ester-ligations over thioester-ligations is the ease of synthesis from unmodified, native pep-
12
This article is protected by copyright. All rights reserved.

10.1002/chem.201604247
Chemistry - A European Journal
tides/proteins. However, side chain reactivity with aspartate or glutamate residues is one of the biggest
drawbacks, as none of these methods are site-specific for the carboxylic acid of the C-terminus. In addi-
tion, once the activated ester is formed, reactivity with nucleophilic side chains (ex/ lysine ε-NH2 group)
becomes plausible.
Based on the observed reactivity of glycine propargyl esters with linear amines, the design and de-
velopment of a C-terminal glycine-specific peptide bioconjugation technique was explored. A series of
short peptides 27a-f (Table 5) were propargylated using standard DIC/DMAP coupling protocols. In or-
der to shorten reaction times to 1 hr, the propargylated peptides were incubated with 20 equivalents of
the azide-containing amine 3g under buffer conditions. Subsequent purification gave the desired C-
azido linked peptides 28a-f in excellent yields (87-97%).
Table 5. List of propargylated peptides conjugated at C-terminus
entry
peptide
27a
27b
27c
27d
27e
sequence
prod
28a
28b
28c
28d
28e
28f
yield1
97%
86%
89%
87%
89%
93%
1
2
3
4
5
6
Ala-Lys-Ile-Gly-OPgyl
Gly-Thr-Phe-Gly-Val-Gly-OPgyl
Z-Gln-Gly-OPgyl
Z-Met-Tyr-Gly-OPgyl
Z-Gly-Pro-Leu-Gly-OPgyl
Z-Phe-His-Ile-Gly-OPgyl
27f
1Isolated yields; pept = peptide. Generally, peptides 27a-f (18-92 mM final concentration) were incu-
bated with amine 3g (0.3-1.8 M final concentration) for 1 hour in (1:1) dioxane/PBS Buffer pH 7.4 (200
μl).
Previously, hexylamine 3a showed no observable reactivity with the β- and γ-propargyl esters of as-
partate 2o and glutamate 2p derivatives, respectively (Table 2, entries 16-17). Given these results, it was
reasoned that site-selectivity for C-terminal glycines may be achieved over aspartate and glutamate side
chain residues. During development, however, propargylated aspartate and glutamate side chain residues
were found to be prone to aspartimide and glutarimide side product formation.[21] This issue was suc-
cessfully bypassed with the use of a bulkier propargyl ester (2-methyl propargyl) in peptides 29a,b. Un-
der 20 equivalents of the azide-containing amine 3g for 5 hours, desired C terminal-azido linked pep-
tides 30a-b were obtained in 75-79% yields (Figure 4). HPLC purification cleanly gave desired product
13
This article is protected by copyright. All rights reserved.

10.1002/chem.201604247
Chemistry - A European Journal
without significant side product formation. In depth NMR analysis was also done to verify that reactivi-
ty was site-specific to the C-terminal glycine end (see Supp. Figure 4-6). As a proof of concept, the re-
maining aspartate side chain ester of peptide 30b was deprotected with palladium under mild condi-
tions,[22] followed by click chemistry to fluorescent probe 31 to give 32b.
Figure 4. Scheme for proposed C-terminal glycine peptide conjugation: 2-Methyl-propargylated pep-
tides 29a,b (100 mM) were incubated with amine 3g (2 M) for 5 hours in solvent (10% DMSO in PBS
buffer for 29a, 50% Dioxane in PBS buffer for 29b). Allylpalladium (II) chloride dimer was used as the
palladium species for ester hydrolysis.
Overall, site-selectivity of the C-terminal glycine propargyl ester was achieved over that of the β- and
γ-propargyl esters of aspartate and glutamate, respectively. In addition, no interference was found with
peptide residues with nucleophilic side chains (ex/ Lys) nor with the exposed N-terminus (ex/ 27a-b).
So far in these studies, millimolar to molar concentrations were used to provide adequate NMR analysis
to show selective C-terminal Gly conjugation (over that of Asp and Glu side chains). In a practical bio-
logical context, however, micromolar concentrations are typically used. To prove reactivity is main-
tained at smaller concentrations, peptide 29a (100 μM and 1 mM) was incubated with 20 equivalents of
amine 3g and the reaction mixture was directly analyzed by HPLC (Figure 5). As confirmed by MS, the
major peak indicates single amidation product, while the secondary peak is identified as unreacted
propargyl ester starting material. In terms of yields, observations show that desired product 30a pro-
duced under 1 mM conditions were as expected (~75% from HPLC) following 5 hours. On the other
14
This article is protected by copyright. All rights reserved.

10.1002/chem.201604247
Chemistry - A European Journal
hand, under 100 μM conditions, longer reaction times were needed (~12 hours) to reach the 75% yield
threshold.
Figure 5. HPLC trace of reaction mixture between amine 3g (20 equivalents) and A) 1 mM of peptide
29a for 5 hours, or B) 100 μM of peptide 29a for 12 hours. Mixtures were incubated at room tempera-
ture in 10% DMSO in PBS Buffer pH 7.4. HPLC analysis followed method 10 as written in the SI data.
Single amidation product 30a is highlighted in green with the retention time of 13.1 min while starting
material 29a is highlighted in orange with the retention time of 14.8min.
CONCLUSIONS
While more and more research efforts are oriented towards exploring complex, metal catalyzed reac-
tions, many simple, non-metal catalyzed reactions have become overlooked. With this work, we have
uncovered the unique reactivity of glycine propargyl esters to undergo "propargyl-assisted selective
amidation" with linear alkylamines. Rather than being dependent on the electron withdrawing inductive
effects (of the leaving group or ester substituents) like most ester aminolysis reactions, it is hypothesized
that this reaction is largely driven by collective hydrogen bonding/intermolecular interactions of sub-
strate clusters. If the proposed mechanism holds true, this reaction exhibits an interesting example where
hydrogen-bonding plays a central role in the previously ineffective reactivity of propargyl esters. Given
the need to further understand these observations, it is hoped that this work can further spark interest
and research efforts in propargyl group chemistry, which may open the door for similar reactions to be
discovered in the future.
Currently, about 6% of proteins naturally have a C-terminal glycine (32,519 out of 549,835 proteins
from the UniProtKB/Swiss-Prot database). Despite this, we believe there is potential for glycine
propargyl ester-based bioconjugation to be adapted for a wider scope of protein targets; mainly due to
15
This article is protected by copyright. All rights reserved.

10.1002/chem.201604247
Chemistry - A European Journal
current ease of site-directed mutagenesis techniques and the limited biological impact of C-terminal tail
modification. One potential strategy that can be envisioned for the facile preparation and utilization of
this bioconjugation technique is based upon spin column purification. Following each major steps; 1)
EDC-based propargyl ester formation, 2) glycine propargyl ester amidation, and 3) palladium-catalyzed
propargyl ester removal; spin column filtration can in theory be used to separate excess reagents from
the protein. In addition, EDC-based carbodiimide crosslinking chemistry with proteins is commercially
available as a kit from ThermoFisher. If successful, this approach offers significant advantages over cur-
rent ester-based approaches (ex/ selective against glutamate/aspartate), is compatible with unprotected
peptides, and would be a viable alternative to thioester-based bioconjugation in thiol-sensitive condi-
tions. Efforts in our research groups are already underway to develop a robust bioconjugation method
aimed at C-terminal glycine-containing proteins.
EXPERIMENTAL
The general synthetic procedure of compounds 4-24, 27a-f, 29a-b goes by the following: corresponding
amines (1 equiv for 4-24, 20 equiv for 27a-f, 29a-b) and propargylic esters (1 equiv, 0.1-0.3 mmol)
were dissolved in the corresponding solvent (THF, DCM, Dioxane, MeOH, H2O, PBS Buffer pH 7.4;
200-1000 μl) under specific temperatures (rt, 60°C) and times (1, 6, 16 hr). To workup, the solvents
were removed by vacuum and the desired products was purified by flash column chromatography or
HPLC purification. Detailed spectra and HPLC methods can be found in the supplementary information.
ACKNOWLEDGMENT
This work was supported by JSPS KAKENHI Grant Numbers JP16H03287, JP16K13104, and
JP15H05843 in Middle Molecular Strategy; K.K.H.V. was supported by the JSPS Postdoctoral Fellow-
ship for Research Abroad; A part of this study was done under the support of JST CREST; and by a sub-
sidy from the Russian Government “Program of Competitive Growth of Kazan Federal University
among World’s Leading Academic Centers”.
16
This article is protected by copyright. All rights reserved.

10.1002/chem.201604247
Chemistry - A European Journal
REFERENCES
[1]
a) A. C. Satterthwait, W. P. Jencks, J. Am. Chem. Soc. 1974, 96, 7018-7031; b) F. M. Menger, J.
H. Smith, J. Am. Chem. Soc. 1972, 94, 3824-3829.
[2]
[3]
C. A. G. N. Montalbetti, V. Falque, Tetrahedron 2005, 61, 10827-10852.
a) K. Ishihara, Y. Kuroki, N. Hanaki, S. Ohara, H. Yamamoto, J. Am. Chem. Soc. 1996, 118,
1569-1570; b) Y. Kuroki, K. Ishihara, N. Hanaki, S. Ohara, H. Yamamoto, Bull. Chem. Soc. Jpn.
1998, 71, 1221-1230.
[4]
[5]
C. Han, J. P. Lee, E. Lobkovsky, J. A. Porco, J. Am. Chem. Soc. 2005, 127, 10039-10044.
a) H. Morimoto, R. Fujiwara, Y. Shimizu, K. Morisaki, T. Ohshima, Org. Lett. 2014, 16, 2018-
2021; b) B. Gnanaprakasam, D. Milstein, J. Am. Chem. Soc. 2011, 133, 1682-1685; c) J. Lee, S.
Muthaiah, S. H. Hong, Adv. Synth. Catal. 2014, 356, 2653-2660; d) N. Caldwell, C. Jamieson, I.
Simpson, A. J. B. Watson, Chem. Commun. 2015, 51, 9495-9498; e) T. Ohshima, Y. Hayashi, K.
Agura, Y. Fujii, A. Yoshiyama, K. Mashima, Chem. Commun. 2012, 48, 5434-5436; f) X. Yang,
V. B. Birman, Org. Lett. 2009, 11, 1499-1502; g) K. E. Price, C. Larrivée-Aboussafy, B. M.
Lillie, R. W. McLaughlin, J. Mustakis, K. W. Hettenbach, J. M. Hawkins, R. Vaidyanathan, Org.
Lett. 2009, 11, 2003-2006; h) M. Movassaghi, M. A. Schmidt, Org. Lett. 2005, 7, 2453-2456; i)
C. Sabot, K. A. Kumar, S. Meunier, C. Mioskowski, Tetrahedron Lett. 2007, 48, 3863-3866; j)
F. J. Weiberth, Y. Yu, W. Subotkowski, C. Pemberton, Org. Process Res. Dev. 2012, 16, 1967-
1969.
[6]
a) P. Ilankumaran, N. Manoj, S. Chandrasekaran, Chem. Commun. 1996, 1957-1958; b) R.
Ramapanicker, R. Gupta, R. Megha, S. Chandrasekaran, Int. J. Pept. 2011, 2011, Article ID
854952.
[7]
[8]
[9]
M. Bodanszky, Chem. Ind. 1957, 524.
S. Maeda, K. Ohno, K. Morokuma, Phys. Chem. Chem. Phys. 2013, 15, 3683-3701.
a) S. Maeda, T. Taketsugu, K. Morokuma, K. Ohno, Bull. Chem. Soc. Jpn. 2014, 87, 1315-1334;
b) S. Maeda, Y. Harabuchi, M. Takagi, T. Taketsugu, K. Morokuma, Chem. Rec. 2016, DOI:
10.1002/tcr.201600043.
[10] W. M. C. Sameera, S. Maeda, K. Morokuma, Acc. Chem. Res. 2016, 49, 763-773.
[11] a) S. Maeda, et al., GRRM, a developmental version, 2016, Hokkaido University; b) M. J.
Frisch, et al., Gaussian 09, Revision D.01, 2009, Gaussian, Inc., Wallingford CT.
[12] a) A. G. M. Barrett, in Comprehensive Organic Synthesis (Eds.: B. M. Trost, I. Fleming),
Pergamon, Oxford, 1991, pp. 235-257; b) K. Tanaka, M. Pimentel, N. Berova, K. Nakanishi,
Bull. Chem. Soc. Jpn. 2005, 78, 1843-1850.
[13] a) J. I. MacDonald, H. K. Munch, T. Moore, M. B. Francis, Nat. Chem. Biol. 2015, 326-331; b)
M. Raj, H. Wu, S. L. Blosser, M. A. Vittoria, P. S. Arora, J. Am. Chem. Soc. 2015, 137, 6932-
6940.
[14] E. Basle, N. Joubert, M. Pucheault, Chem. Biol. 2010, 17, 213-227.
[15] P. E. Dawson, T. W. Muir, I. Clark-Lewis, S. B. H. Kent, Science 1994, 266, 776-779.
[16] a) C. T. T. Wong, C. L. Tung, X. Li, Mol. BioSyst. 2013, 9, 826-833; b) C. Sun, G. Luo, S.
Neravetla, S. S. Ghosh, B. Forood, Bioorg. Med. Chem. Lett. 2013, 23, 5203-5208.
[17] L. Yi, H. Sun, Y. W. Wu, G. Triola, H. Waldmann, R. S. Goody, Angew. Chem., Int. Ed. 2010,
49, 9417-9421.
[18] a) S. S. van Berkel, M. B. van Eldijk, J. C. M. van Hest, Angew. Chem., Int. Ed. 2011, 50, 8806-
8827; b) E. Saxon, J. I. Armstrong, C. R. Bertozzi, Org. Lett. 2000, 2, 2141-2143.
[19] a) X. Zhang, F. Li, X. W. Lu, C. F. Liu, Bioconjugate Chem. 2009, 20, 197-200; b) D. Crich, K.
Sana, S. Guo, Org. Lett. 2007, 9, 4423-4426; c) D. Crich, I. Sharma, Angew. Chem., Int. Ed.
2009, 48, 7591-7594; d) P. Karmakar, R. S. Talan, S. J. Sucheck, Org. Lett. 2011, 13, 5298-
5301.
[20] a) X. Li, H. Y. Lam, Y. Zhang, C. K. Chan, Org. Lett. 2010, 12, 1724-1727; b) Y. Zhang, C. Xu,
H. Y. Lam, C. L. Lee, X. Li, Proc. Natl. Acad. Sci. U. S. A. 2013, 110, 6657-6662.
[21] R. Subirós-Funosas, A. El-Faham, F. Albericio, Tetrahedron 2011, 67, 8595-8606.
17
This article is protected by copyright. All rights reserved.

10.1002/chem.201604247
Chemistry - A European Journal
[22] a) M. Pal, K. Parasuraman, K. R. Yeleswarapu, Org. Lett. 2003, 5, 349-352; b) J. Li, J. Yu, J.
Zhao, J. Wang, S. Zheng, S. Lin, L. Chen, M. Yang, S. Jia, X. Zhang, P. R. Chen, Nat. Chem.
2014, 6, 352-361.
18
This article is protected by copyright. All rights reserved.

10.1002/chem.201604247
Chemistry - A European Journal
Table of contents entry
Glycine propargyl esters show selective amidation activity with linear alkylamines, even in the absence
of a base catalyst. Calculations suggest hydrogen-bonding and intermolecular interactions play a central
role for reactivity. Using these findings, applications to a site-specific C-terminal peptide conjugation
technique is presented.
Keywords: propargyl esters, selective amidation, peptides, hydrogen bonds, amides
19
This article is protected by copyright. All rights reserved.