Racemisation of N-Fmoc phenylglycine under mild microwave-SPPS and conventional stepwise SPPS conditions: Attempts to develop strategies for overcoming this

Article abstract of DOI:10.1002/psc.2398

We have been engaged in the microwave-solid phase peptide synthesis (SPPS) synthesis of the phenylglycine (Phg)-containing pentapeptide H-Ala-Val-Pro-Phg-Tyr-NH₂ (1) previously demonstrated to bind to the so-called BIR3 domain of the anti-apoptotic protein XIAP. Analysis of the target peptide by a combination of RP-HPLC, ESI-MS, and NMR revealed the presence of two diastereoisomers arising out of the racemisation of the Phg residue, with the percentage of the LLLDL component assessed as 49%. We performed the synthesis of peptide (1) using different microwave and conventional stepwise SPPS conditions in attempts to reduce the level of racemisation of the Phg residue and to determine at which part of the synthetic cycle the epimerization had occurred. We determined that racemisation occurred mainly during the Fmoc-group removal and, to a much lesser extent, during activation/coupling of the Fmoc-Phg-OH residue. We were able to obtain the desired peptide with a 71% diastereomeric purity (29% LLLDL as impurity) by utilizing microwave-assisted SPPS at 50°C and power 22Watts, when the triazine-derived coupling reagent DMTMM-BF₄ was used, together with NMM as an activator base, for the incorporation of this residue and 20% piperidine as an Fmoc-deprotection base. In contrast, the phenylalanine analogue of the above peptide, H-Ala-Val-Pro-Phe-Tyr-NH₂ (2), was always obtained as a single diastereoisomer by using a range of standard coupling and deprotection conditions. Our findings suggest that the racemisation of Fmoc-Phg-OH, under both microwave-SPPS and stepwise conventional SPPS syntheses conditions, is very facile but can be limited through the use of the above stated conditions.

Full text of DOI:10.1002/psc.2398

Research Article
Received: 6 September 2011
Revised: 7 December 2011
Accepted: 4 January 2012
Published online in Wiley Online Library: 26 March 2012
(wileyonlinelibrary.com) DOI 10.1002/psc.2398
Racemisation of N-Fmoc phenylglycine
under mild microwave-SPPS and
conventional stepwise SPPS conditions:
attempts to develop strategies for
overcoming this
Mohamed A. Elsawy,a* Chandralal Hewageb and Brian Walkera
We have been engaged in the microwave-solid phase peptide synthesis (SPPS) synthesis of the phenylglycine (Phg)-containing
pentapeptide H-Ala-Val-Pro-Phg-Tyr-NH2 (1) previously demonstrated to bind to the so-called BIR3 domain of the anti-apoptotic
protein XIAP. Analysis of the target peptide by a combination of RP-HPLC, ESI-MS, and NMR revealed the presence of two
diastereoisomers arising out of the racemisation of the Phg residue, with the percentage of the LLLDL component assessed as
49%. We performed the synthesis of peptide (1) using different microwave and conventional stepwise SPPS conditions in
attempts to reduce the level of racemisation of the Phg residue and to determine at which part of the synthetic cycle the epimer-
ization had occurred. We determined that racemisation occurred mainly during the Fmoc-group removal and, to a much lesser
extent, during activation/coupling of the Fmoc-Phg-OH residue. We were able to obtain the desired peptide with a 71% diastereo-
meric purity (29% LLLDL as impurity) by utilizing microwave-assisted SPPS at 50ꢀC and power 22Watts, when the triazine-derived
coupling reagent DMTMM-BF4 was used, together with NMM as an activator base, for the incorporation of this residue and 20%
piperidine as an Fmoc-deprotection base. In contrast, the phenylalanine analogue of the above peptide, H-Ala-Val-Pro-Phe-Tyr-
NH2 (2), was always obtained as a single diastereoisomer by using a range of standard coupling and deprotection conditions.
Our findings suggest that the racemisation of Fmoc-Phg-OH, under both microwave-SPPS and stepwise conventional SPPS syn-
theses conditions, is very facile but can be limited through the use of the above stated conditions. Copyright © 2012 European
Peptide Society and John Wiley & Sons, Ltd.
Supporting information may be found in the online version of this article.
Keywords: Fmoc-phenylglycine; racemisation; peptide diastereoisomers; DMTMM; super active esters; microwave-SPPS; conventional
stepwise SPPS; RP-HPLC; ESI-MS; (1D) and (2D) NMR
mechanism, whereas the oxazolone mechanism occurs in coupling
acyl-type amino protected amino acid residues and peptide
Introduction
9-Fluorenylmethoxycarbonyl (Fmoc) solid phase peptide synthesis
(SPPS) has become an important tool in peptide chemistry and
has revolutionized the preparation of synthetic peptides for use
in molecular biology, chemical biology, biochemistry, pharmacol-
ogy, drug discovery, and medicinal chemistry [1–5]. With the
increased demand for production of huge peptide libraries for high
throughput screening, the use of microwave synthesizers has
emerged as a complementary approach to conventional stepwise
SPPS for accelerating the synthesis procedure and decreasing side
products [6,7]. Although microwave-assisted synthesis can be
highly efficient in enhancing the rates of deprotection and cou-
pling steps, there are some reported cases of increased levels of
racemisation of some amino acid residues [8,9].
fragments. The degree and rate of racemisation depend on many
factors: inductive and mesomeric effects of amino acid side chain
residues, the nature of activators/coupling reagents used, the
basicity of tertiary amines used in activation/coupling steps and
Fmoc protecting group removal, reaction duration, temperature,
and synthesis mode whether conventional stepwise SPPS or
microwave-SPPS.
Because racemisation of a single chiral amino acid residue may
have a devastating effect on the biological action of a peptide,
approaches to minimize racemisation of vulnerable residues are
clearly essential.
In SPPS, the incidence of racemisation has been widely investi-
gated for many urethane-type amino protected amino acid
residues [8–14]. Racemisation mechanisms have been extensively
studied and fall into two main types: direct enolization and
oxazolone (azalactone) formation, which are both base-catalyzed
mechanisms [15]. In the case of urethane-type amino protected
amino acid residues, direct enolization is mostly the suggested
*
Correspondence to: Mohamed A. Elsawy, School of Pharmacy, Queen’s University
Belfast, 97 Lisburn Road, BT9 7BL, UK. E-mail: melsawy01@qub.ac.uk
a School of Pharmacy, Queen’s University of Belfast, BT9 7BL, UK
b Biochemistry Department, Conway Institute, University College of Dublin,
Dublin, Ireland
J. Pept. Sci. 2012; 18: 302–311
Copyright © 2012 European Peptide Society and John Wiley & Sons, Ltd.

RACEMISATION OF PHENYLGLYCINE IN MICROWAVE AND CONVENTIONAL SPPS
One such racemisation vulnerable residue is phenylglycine
(Phg). Although not encoded genetically, Phg has been incorpo-
rated into a broad-spectrum of synthetic biologically active
peptides such as anti-hepatitis c virus (HCV) genotype 1b replicon
inhibitors [16,17] and HCV NS3 protease inhibitors [18], b-lactam
antibiotics [19], gastrin-like peptides [20], platelet aggregation
inhibitors [21], anticoagulant D-phenylglycinamide derived inhibi-
tors of Factor Xa [22], vascularisation/angiogenesis inhibitors with
possible use in cancer therapy [23], neurokinin inhibitors [24],
D-phenylglycine-L-dopa conjugates that improved the bioavail-
ability of the neurotransmitter L-dopa [25,26], and pro-apoptotic
anticancer peptides [27].
In the course of our studies aimed at the synthesis of analogues
of the pro-apoptotic peptide H-AVPPhgY-NH2 (1), where the Phg
is a crucial residue for the binding of the peptide to its target
protein ‘BIR3 domain of XIAP’ [27], we have encountered extensive
racemisation of this crucial residue when using ‘standard’ solid
phase synthesis methods employing conventional and microwave-
assisted approaches. This paper reports on the comparative degrees
of racemisation engendered by using a range of synthesis condi-
tions and the substantial suppression of racemisation of this residue
by using the triazine-derived coupling reagent 4-(4,6-dimethoxy-
1,3,5-triazin-2-yl)-4-methylmorpholinium tetrafluoroborate (DMTMM-
BF4) with NMM as the activation base for the coupling steps and
20% piperidine/DMF for Fmoc-removal, under microwave-SPPS
mild conditions.
during synthesis. All reagents and amino acids are delivered
through metered sample loops for accurate dispensing.
Peptide Synthesis Protocols
H-AVPPhgY-NH2 (1), H-AVPFY-NH2 (2), Fmoc-PhgY-NH2 (3) and
H-PhgY-NH2 (4)
Peptides (1), (2), (3), and (4) were synthesized by the CEM Liberty
automated microwave peptide synthesizer, utilizing either con-
ventional or microwave-assisted modes. In essence, each peptide
was synthesized on a 0.1-mmol scale, using Rink amide MBHA
resin (0.154 g, substitution 0.65 meq/g). Fmoc removal was
performed using two repeat cycles employing (i) 2% DBU, (ii)
20% piperidine, (iii) 5% piperazine (iv) 20% morpholine, or (v)
20% triethanolamine as solutions in DMF. For the microwave-
assisted approach, a first deprotection cycle of 30 s at 50 Watts
was employed, followed by a second deprotection cycle of
3 min at 50 Watts. Both cycles were carried out at 75 ꢀC. The
conventional approach employed a 10-min exposure to the base,
followed by a 20-min exposure at room temperature, in the
absence of any microwave-induced heating. Coupling steps were
carried out by introducing each Fmoc-amino acid (0.2-M solution
in DMF) at a fivefold excess over resin loading, together with
various activation reagents and bases used in the molar ratios
indicated, including (i) HBTU/DIEA/AA (1/2/1), (ii) HBTU/NMM/
AA (1/2/1), (iii) DMTMM-Cl/NMM/AA (1/2/1), and (iv) DMTMM-
BF4/NMM/AA (1/2/1), using microwave-assisted synthesis or
HBTU/HOBT/DIEA/AA (1/1/2/1) for conventional synthesis. Each
coupling reaction was performed for 10 min at 22 Watts at a
temperature of either 50 or 75 ꢀC, for microwave-assisted synthe-
sis, or for 45 min at room temperature, in the absence of
microwave-induced heating for conventional synthesis. Finally,
cleavage of the peptide from the solid support and concomitant
removal of the side chain O-But protecting group of tyrosine was
performed manually at room temperature, using TFA/water/
triisopropylsilane (95/2.5/2.5, v/v/v) for two 1-h cycles, with washing
with dichloromethane after every cycle. The collected cleavage
reaction mixtures and washes were evaporated under vacuum,
at 30 ꢀC, cooled, and the products were precipitated by the
addition of cold diethylether. The precipitates were collected
by centrifugation (3–5 min at 2000–3000g), and the pellet was
washed thoroughly with diethylether. This process was repeated
two to three times, each time the solid was collected by centri-
fugation. After a brief drying in a vacuum to remove all traces of
diethyl ether, the peptide products were dissolved in a 10% TFA
aqueous solution and freeze dried.
Materials and Methods
Reagents
All amino acids were introduced as their Fmoc-protected deriva-
tives. Fmoc-Ala-OH, Fmoc-Pro-OH, and Fmoc-Phe-OH were
purchased from CEM (Buckingham, England, UK), whereas
Fmoc-Val-OH, Fmoc-phenylglycine (Fmoc-Phg-OH), and Fmoc-
Tyr-(tBu)-OH were obtained from Novabiochem (Darmstadt,
Germany). N-Hydroxybenzotriazole (HOBT) was also obtained
from CEM (Buckingham, England, UK). O-(Benzotriazol-1-yl)-N,
N,N’,N’-tetramethyluronium hexafluorophosphate (HBTU), 4-(4,6-
dimethoxy-1,3,5-triazin-2-yl)-4-methylmorpholinium tetrafluoroborate
(DMTMM-BF4), and Rink amide MBHA resin were obtained from
Iris Biotech GMBH (Marktredwitz,Germany). 4-(4,6-Dimethoxy-
1,3,5-triazin-2-yl)-4-methylmorpholinium chloride (DMTMM-Cl),
diisopropylethylamine (DIEA), N-methylmorpholine (NMM), mor-
pholine, piperidine, piperazine, triethanolamine, trifluoroacetic acid
(TFA), triisopropylsilane, N,N-dimethylformamide (DMF), dichloro-
methane (DCM), and diethylether were obtained from Sigma-
Aldrich (Gillingham, Dorset, England, UK). HPLC-grade methanol
and HPLC-grade trifluoroacetic acid were also supplied by Sigma-
Aldrich. Double-distilled, deionised water was used throughout.
RP-HPLC Analysis
Peptide purity was checked through RP-HPLC, using a Waters HPLC
system fitted with Waters 1525 binary HPLC pump and Waters 2489
UV/visible detector (l216 nm) (Waters, Milford, Massachusetts, USA)
and employing a Phenomenex Jupiter C12 column (250 Â 4.66 mm;
particle size, 10 mm). The runs were carried out on an analytical scale
with a flow rate of 1 ml/min. An elution gradient was utilized to re-
solve the product components that went from 90% solvent A
(0.05% TFA in H2O)/10% solvent B (0.05% TFA in CH3OH) to 10% sol-
vent A/90% solvent B, in 60 min for peptides (1) and (2). Another elu-
tion gradient was used for peptide (3), which went from 100% sol-
vent A to 10% solvent A/90% solvent B, in 35 min. Whereas, for
peptide (4), an elution gradient that went from 80% solvent A/20%
solvent B to 10% solvent A/90% solvent B in 45 min was applied.
Liberty™ Microwave Synthesizer
The Liberty™ system (CEM Microwave Technology Ltd., Buckingham,
England, UK) is an automated sequential peptide synthesizer that
utilizes Pepdrive© software for controlling each step of the synthesis
and can synthesize up to 12 different peptides in series. The Liberty™
contains a single-mode microwave (Discovery™) reactor in which a
reaction vessel (30-ml glass fritted for 0.025–1.0 mmol syntheses
scales, equipped with a spray head for reagent delivery and a fi-
ber-optic temperature probe for power delivery control) is placed.
The system is pressurized with nitrogen for conveying reagents, ag-
itating the reaction mixtures, and providing an inert environment
J. Pept. Sci. 2012; 18: 302–311 Copyright © 2012 European Peptide Society and John Wiley & Sons, Ltd. wileyonlinelibrary.com/journal/jpepsci

ELSAWY, HEWAGE AND WALKER
ESI-MS Spectroscopy
corresponds to Phg). We thought the latter explanation to be
the most likely, but we decided to perform 1D and 2D proton
NMR on samples of purified material from Peaks A and B to
confirm/establish the precise reason for the presence of the
two forms of the peptide.
Mass spectrometry was performed using a Thermo Finnigan LCQ
Deca™ instrument (Thermo Electron Corporation, Hertfordshire,
England, UK), which uses ElectroSpray Ionisation Ion-Trap Mass
Spectrometry, and was operated in nanospray mode. The pep-
tide samples were dissolved in 10% TFA in methanol to
perform mass spec analysis.
Figure 2 records the 1D proton, and the 2D TOCSY and ROESY
spectra for each component. All spectra showed clear chemical
shift dispersion and well-resolved peaks. Sequence-specific pro-
ton resonance assignment of the peptide sample was achieved
through the analysis of TOCSY and ROESY spectra. Almost all
the spin patterns were unambiguously assigned in the TOCSY
spectrum. The presence of aiH/Ni + 1H and side chain connectiv-
ity cross peaks in the ROESY spectra led to the complete
sequence-specific resonance assignment of the peptide. The
presence of connectivities of neighboring amino acid residues
supported the reliability of the resonance assignments (see
Figure 2). Sequence-specific resonance assignments were carried
out, which led to the full and unambiguous identification of all
the individual spin systems.
Each of the proline residues in the purified samples of both
peptides obtained from Peaks A and B exhibited an 8–9/1
trans-cis conformational ratio with no observable difference in
chemical shift, whereas chemical shift differences of the a proton
of the phenylglycine residue of each peptide showed a slight var-
iation of 0.15 ppm. In contrast, the amide proton of the tyrosine
residue of each peptide showed a remarkable chemical shift
difference of 0.33 ppm and small chemical shift variations for
a protons of 0.08 ppm. Aromatic protons of the phenylglycine
and tyrosine also showed noticeable changes for both peptides
(the chemical shifts are recorded in Table 1). This clearly indicates
the twist of the amide bond between the phenylglycine and tyro-
sine residues at the C-terminus of the sequence. In common with
many bioactive peptides, both peptide samples did not show any
secondary structural features in the DMSO medium.
NMR Spectroscopy
NMR experiments were performed using the 5-mm inverse probe
head on Bruker DRX 500 (Bruker BioSpin, Germany) operating at a
1H resonance frequency of 500.13 MHz at 298 K. (1 mg in 600ul
calculate, 0.35 mM). Peptide samples were prepared using a DMSO
solvent. NMR spectra were internally referenced for d-DMSO
resonating at 2.54 ppm, and the final volume of the peptide sample
was 600 ml.
One-dimensional (1D) proton NMR spectra were acquired with a
relaxation delay of 3.0 s and acquisition time of 2.34 s. Thirty-two
scans were acquired over a 7-kHz spectral width, and 32 768
data points were collected. The data sets were zero-filled to
65 536 data points and was multiplied by 0.2-Hz line broadening.
Two-dimensional (2D) total correlation spectroscopy (TOCSY) [28]
and rotating frame Overhauser enhancement spectroscopy
(ROESY) [29] data sets were acquired with a relaxation delay of
1.4s, acquisition time of 0.14 s, and a spectral width of 7 kHz. The
TOCSY data were acquired with an 80-ms mixing time, whereas
the ROESY data were acquired with a 200-ms spin-lock. The 2D
experiments were performed with 8 and 16 number of scans with
2048 (F2) and 1024 (F1) complex data points for the TOCSY and
NOESY data, respectively. 2D data were zero-filled to 2048 data
points in F1 prior to transformation. All data sets were apodised
in both dimensions by using a shifted squared sine bell window
function. All spectra were referenced internally to the residual 1-H
signal of the d-DMSO, resonating at 2.54 ppm. Data were processed
using the TOPSPIN program version 2.1 (Bruker BioSpin, Germany).
Consequently, the NMR analysis data are pinpointing that
racemisation of the Phg residue is the main cause for the
presence/generation of two diastereoisomers corresponding to
our designation of LLLLL (Peak A) and LLLDL (Peak B) in the
recorded RP-HPLC chromatogram (Figure 1(a)).
Results and Discussion
Our instinct was that the racemisation of the Phg residue
occurred during the synthesis of the target peptide and not
because of the optically impure starting material. However, we
had to establish formally that the latter could be excluded as a
possibility. Thus, the optical purity of Fmoc-Phg-OH was checked
by measuring its optical polarization, which showed an [a]20/D of
+85.4º (C = 1% in DMF), as reported in the supplier’s certificate of
analysis. This emphasizes the optical integrity of the starting
Fmoc-protected amino acid and points to the racemisation of this
residue occurring during the synthesis procedure.
Interestingly, when we synthesized the phenylalanine analogue,
H-AVPFY-NH2 (2), by using the same microwave-assisted conditions
and reagents as utilized for the synthesis of H-AVPPhgY-NH2, we
obtained a single product according to RP-HPLC analysis (Figure 3).
This product had a retention time of 29 min, and the ESI-MS m/z
profile revealed a charged species of m/z: 617.7 [M+ Na]+ of
relative abundance 100% (calcd m/z: 594) (Figure 3). This suggests
that the vulnerability of Phg towards racemisation is much more
pronounced under the same conditions than it is for Phe.
Microwave-SPPS synthesis of peptide H-AVPPhgY-NH2 (1) was
carried out under mild conditions, by employing an activation/
coupling cycle of 10 min at 75 ꢀC, at a power setting of 22 Watts.
In the first instance, HBTU was employed as an activator in
the presence of two equivalents of DIEA as an activator base.
Fmoc group removal was performed using 2% 1,8-diazabicyclo
[5.4.0]undec-7-ene (DBU) in DMF (Table 2, entry 1). The release
of the target peptide from the resin and the removal of the O-
But protecting group from tyrosine were carried out at room tem-
perature, and the crude peptide was checked with RP-HPLC, as
outlined above. The product was found to contain two major
peaks, A (percentage area 51%) and B (percentage peak area
49%) with retention times differing by 4 min (Figure 1(a)). The
ESI-MS analysis of Peak A (found m/z: 581.3 [M + 1]+, relative
abundance 100%) and Peak B (581.2 [M + 1]+, relative abundance
100%) revealed identical m/z profiles corresponding to the
calculated mass (580 Da) for the peptide.
This implies that Peaks A and B are isomers of a single
compound, arising either out of distinct cis–trans configuration
ratios of the proline residue or racemisation of, most likely, the
Phg residue, resulting in the formation of two diastereoisomers,
which we will designate as LLLLL and LLLDL (embolden residue
Bodanszky and Birkhimer have demonstrated previously that
the high racemisation tendency of aromatic amino acid deriva-
tives in the presence of tertiary amines can be attributed to the
abstraction of the a carbon proton, a mechanism know as direct
wileyonlinelibrary.com/journal/jpepsci Copyright © 2012 European Peptide Society and John Wiley & Sons, Ltd. J. Pept. Sci. 2012; 18: 302–311

RACEMISATION OF PHENYLGLYCINE IN MICROWAVE AND CONVENTIONAL SPPS
(a)
(b)
(c)
Figure 1. RP-HPLC profile of AVPPhgY-NH2 (1) synthesized by (a) microwave-SPPS with conditions and reagents detailed in (Table 2, entry 1), the
percentage of diastereoisomer LLLDL (Peak B) is 49% of the crude product; (b) microwave-SPPS with conditions and reagents detailed in (Table 2, entry
2), the percentage of diastereoisomer LLLDL (Peak B) is 43% of the crude product; (c) microwave-SPPS with NMM solution as activator base (Table 2,
entry 3), the percentage of diastereoisomer LLLDL (Peak B) is 37% of the crude; (d) conventional stepwise SPPS with conditions and reagents detailed
in (Table 2, entry 10), the percentage of diastereoisomer LLLDL (Peak B) is 36% of the crude product; (e) microwave-SPPS with conditions and reagents
detailed in (Table 2, entry 8), the percentage of diastereoisomer LLLDL (Peak B) is 31% of the crude product; (f) microwave-SPPS with conditions and
reagents detailed in (Table 2, entry 9), the percentage of diastereoisomer LLLDL (Peak B) is 29% of the crude product. The retention times for Peaks
A and B are 25 and 29 min, respectively.
enolization [28]. In the case of Phg, the a carbon is a benzilic car-
bon, thus the carbanion generated from the release of the acidic
a proton catalyzed by the tertiary amine is stabilized by the reso-
nance effect of the phenyl ring. The stabilized carbanion can then
be protonated from above or below the plane of the molecule. This
is a possible explanation for the high racemisation tendency of the
activated esters for such a residue (Figure 4). On the other hand,
racemisation through the oxazolone (azalactone) mechanism can-
not be a suggestion here, as the Phg residue is used as a ure-
thane-type amino protected residue.
According to the previously suggested mechanisms, Phg would
be susceptible to direct enolization at the benzilic a-proton either
during the coupling step when exposed to the activator tertiary
base N, N-diisopropylethylamine (DIEA) or during Fmoc removal
for which the non-nucleophilic base 1,8-diazabicyclo[5.4.0]undec-
7-ene (DBU) was used, initially, in our synthetic approach.
H-PhgY-NH2 (4) by using the same starting conditions used for
the synthesis of the pentapeptide H-AVPPhgY-NH2 (1) (see Ta-
ble 2, entry 1). The RP-HPLC profile of the Fmoc-protected
peptide (3) revealed only 3% DL diastereoisomer (Figure 5(a),
Peak E), whereas that of the de-protected peptide (4) showed it
contained 40% of the DL diastereoisomer (Figure 5(b), Peak G).
These results demonstrate that Fmoc-removal is the largest
contributing factor to Phg racemisation and that coupling/
activation proceeds is of minor importance.
Thus, our first attempts to reduce the Phg racemisation
focused on evaluating different bases for Fmoc-deprotection. In
the original synthesis of (1), 2% DBU was used as a deprotect
base (see Table 2, entry 1), which is one of the two most
commonly used bases for the removal of the Fmoc group in
Fmoc-based peptide synthesis [31–34], the other being the
nucleophilic base, piperidine [35,36].
To investigate at which part of the synthetic cycle the
epimerization had occurred, we synthesized the Fmoc-protected
dipeptide Fmoc-PhgY-NH2 (3) and its unprotected analogue
Because 2% DBU and 20% piperidine can be used interchange-
ably for the removal of Fmoc, we replaced the former with the
latter under exactly the same conditions to examine the effect
J. Pept. Sci. 2012; 18: 302–311 Copyright © 2012 European Peptide Society and John Wiley & Sons, Ltd. wileyonlinelibrary.com/journal/jpepsci

ELSAWY, HEWAGE AND WALKER
(d)
(e)
(f)
Figure 1. (continued)
Figure 2. Fingerprint region with sequential connectivities of the ROESY spectra of Y and Z in DMSO at 500.13 MHz.
of the deprotection base on Phg racemisation. We observed a
modest (6%) reduction in the percentage of LLLDL diastereoiso-
mer formed (43% as compared to 49%; see Table 2, entry 2) when
piperidine was used in place of DBU. The RP-HPLC traces of pep-
tide (1) synthesized using these two bases are recorded in Fig-
ures 1(a) (DBU) and 1(b) (piperidine). Because DBU has a pKa of
approximately 12 (in DMSO) and the weaker base piperidine
has a pKa 11.2 [37], the use of the latter would be expected to
result in a reduction of the abstraction of the Phg benzilic a-proton
and a decrease in the formation of the LLLDL diastereoisomer.
We then started using combinations of milder deprotection
and activation bases in an attempt at reducing racemisation in
both Fmoc-removal and coupling/activation steps, respectively.
When the switch in the use of piperidine for DBU for Fmoc group
removal was combined with the use of an activator base with
lower basicity, NMM (pKa 7.41) instead of DIEA (pKa 11), we
observed a further (6%) drop in the proportion of the LLLDL
diastereoisomer formed to 37% (Table 2, entry 3; RP-HPLC profile
of peptide (1) synthesized using these conditions is shown in
Figure 1(c)), regardless of the apparently lower contribution of
wileyonlinelibrary.com/journal/jpepsci Copyright © 2012 European Peptide Society and John Wiley & Sons, Ltd. J. Pept. Sci. 2012; 18: 302–311

RACEMISATION OF PHENYLGLYCINE IN MICROWAVE AND CONVENTIONAL SPPS
Table 1. Chemical shift values of peptides obtained from Peaks A and B (in brackets) with reference to d-DMSO at 2.54 ppm. Data were collected on
a Bruker 500.13 1H spectrometer at 25 C
Residue
NH
aH
bH
gCH2
dCH2
gCH3
Aromatic
1 Ala
2 Val
8.05 (8.05)
8.55 (8.55)
3.94 (3.94)
4.36 (4.36)
1.32 (1.32)
2.00 (2.01)
—
—
—
—
—
0.92 (0.92)
0.96 (0.97)
—
—
—
3 Pro
—
4.50 (4.51)
5.43 (5.58)
1.87, 2.05 (1.76, 2.09)
1.98, 1.98 (1.85, 1.91)
3.65, 3.74 (3.60, 3.77)
—
4 Phg
8.43 (8.46)
—
—
—
—
2,6 H; 7.70 (7.08)
3,5H; 7.37 (7.46)
4H; 7.31 (7.24)
2,6 H; 7.00 (6.91)
3,5H; 6.64 (6.56)
—
5 Tyr
8.23 (8.56)
9.19 (9.15)
4.40 (4.32)
2.75, 2.92 (2.67, 2.88)
—
—
—
—
—
—
Amide
—
—
Figure 3. RP-HPLC profile of AVPFY-NH2 synthesized by microwave-SPPS and by employing a 1-M solution of DIEA in DMF as an activator base. The
retention time for Peak C is 29 min.
[39] which reported on the reduction in the rate of racemisa-
tion of Boc-Phg-OH-activated esters in DMF that used NMM
in place of DIEA.
Given that the combination of the switch in ‘activator’ base and
‘Fmoc-deprotection’ base from DIEA + DBU to NMM + piperidine
resulted in a 12% reduction in the formation of the LLLDL diaste-
reoisomer (compare entries 1 and 3 in Table 2), we examined the
3
effect of retaining NMM as the ‘activator’ base but replacing piper-
1
idine with two even milder bases, piperazine (pKa 9.8) and morpho-
2
line (pKa 8.36) [37], to see if we could achieve even further reduc-
tions of this isomer. However, the amount of this isomer formed
with these combinations was not reduced any further (compare
entries 3, 5, and 6 in Table 2). Finally, when an even weaker base
triethanolamine (pKa 7.77) [33] was utilized, it proved ineffective
+
at catalyzing the removal of the Fmoc group, and no peptide prod-
uct whatsoever could be obtained (Table 2, entry 7).
To check the effect of microwave energy, reaction tempera-
ture, and coupling duration on the degree of Phg racemisation,
peptide (1) was synthesized manually through conventional
stepwise SPPS techniques at room temperature (Table 2, entry
10). The product purity was checked with RP-HPLC and, again,
two diastereoisomers were obtained. However, the proportion
Figure 4. Base-catalyzed racemisation of N-Fmoc-phenylglycine derivatives
of the LLLDL diastereoisomer formed was 36% (Figure 1(d)),
which is some 7% lower than that formed in the microwave-
assisted synthesis carried out at 75 ꢀC, using identical activation
reagents (HBTU + DIEA) and a deprotection base (piperidine)
(Table 2, compare entry 2 with 10). Increased levels of racemisa-
tion at elevated reaction temperatures have been reported
previously. For example, Bacsa et al. reported similar and
through the direct enolization mechanism.
the coupling step in the racemisation process as observed from
the Fmoc-PhgY-NH2 RP-HPLC trace (only 3% DL diastereoiso-
mer, Peak E in Figure 5(a)). This result is consistent with the
work of Anderson et al., [38] and Bodansky and Bodansky
J. Pept. Sci. 2012; 18: 302–311 Copyright © 2012 European Peptide Society and John Wiley & Sons, Ltd. wileyonlinelibrary.com/journal/jpepsci

ELSAWY, HEWAGE AND WALKER
Table 2. Conditions used for the synthesis of peptide (1), and the %LLLDL in the crude product as determined by RP-HPLC. NP, no product was
obtained using this base to remove the Fmoc group
Entry SPPS technique
Temp. (ꢀC)
Activator base Base used in Fmoc-deprotection
Activator
HBTU
Yield [%] Purity [%] %LLLDL
1
2
Microwave
75
DIEA
DIEA
NMM
DIEA
NMM
NMM
NMM
NMM
NMM
DIEA
2% DBU
53
57
51
51
47
48
—
58
65
43
96
98
98
97
95
95
—
86
97
99
49
43
37
36
36
35
NP
31
29
36
75
20% piperidine
20% piperidine
20% piperidine
5% piperidine
20% morpholine
20% triethanolamine
20% piperidine
20% piperidine
20% piperidine
HBTU
3
75
HBTU
4
50
HBTU
5
50
HBTU
6
50
HBTU
7
50
HBTU
8
50
50
DMTMM-Cl
DMTMM-BF4
HBTU/HOBT
9
10
Conventional
Room temp.
(a)
(b)
Figure 5. RP-HPLC profile of (a) Fmoc-PhgY-NH2 (3) synthesized by microwave-SPPS with conditions and reagents detailed in (Table 2, entry 1), the
percentage of diastereoisomer DL (Peak E) is 3% of the crude product. The retention times for Peaks D and E are 26.3 and 29.9 min, respectively. (b)
PhgY-NH2 (4) synthesized by microwave-SPPS with conditions and reagents detailed in (Table 2, entry 1), the percentage of diastereoisomer DL (Peak
G) is 40% of the crude product. The retention times for Peaks F and G are 6.6 and 9.77 min, respectively.
significant levels of racemisation of histidine and cysteine when
using both microwave and conventional SPPS at 86 ꢀC, indicating
that racemisation is a result of thermal effect, not the mode of
heat production [40]. Also, Palasek et al. observed that the micro-
wave-assisted coupling of the activated esters of the two residues
resulted in significant racemisation at 80 ꢀC, the extent of which
could be reduced substantially if the coupling step temperature
was reduced and conducted at 50 ꢀC [8]. In light of those
findings, we repeated the microwave synthesis of peptide (1) at
50 ꢀC (Table 2, entry 4) but using exactly the same activation
reagents, bases, and coupling times that had been employed at
75 ꢀC. In doing so, we achieved a 7% decrease in the amount of
LLLDL diastereoisomer formed (Table 2, compare entries 2 and
4), to a level (36%) identical to that formed during the conven-
tional synthesis conducted at room temperature (Table 2,
compare entries 4 and 10). So, it would appear that if the
activation and coupling steps are carried out at temperatures
up to 50 ꢀC, then there is no difference in the extent of racemisa-
tion of the Phg residue through microwave-assisted synthesis
and conventional SPPS approaches. Similar results have been
reported by Loffrendo et al. whose work has demonstrated that
virtually identical low levels of racemisation (<2%) were obtained
during the coupling of many vulnerable residues at 60 ꢀC (with
the exception of L-cysteine) for both microwave and conventional
stepwise SPPS [9]. Further confirmatory evidence that elevated
reaction temperatures do not necessarily lead to enhancement
of racemisation has been provided by Souza et al. who have in-
vestigated the utility of SPPS at elevated temperatures (SPPS-ET)
and reported that the (nonmicrowave-assisted) syntheses of three
simple model peptides at elevated temperature (55–75 ꢀC) did
not show significant increases in levels of racemisation compared
to syntheses carried out at room temperature [10].
wileyonlinelibrary.com/journal/jpepsci Copyright © 2012 European Peptide Society and John Wiley & Sons, Ltd. J. Pept. Sci. 2012; 18: 302–311

RACEMISATION OF PHENYLGLYCINE IN MICROWAVE AND CONVENTIONAL SPPS
Figure 6. DMTMM-mediated super active ester formation (faster Pathway A), base-catalyzed racemisation of N-Fmoc-phenylglycine derivatives
through the direct enolization mechanism (slower Pathway B) and N-(4,6-dimethoxy-1,3,5-triazin-2-yl) amino acid by product formation on the
resin (Pathway C).
J. Pept. Sci. 2012; 18: 302–311 Copyright © 2012 European Peptide Society and John Wiley & Sons, Ltd. wileyonlinelibrary.com/journal/jpepsci

ELSAWY, HEWAGE AND WALKER
In the case of the Phg residue, the thermal effect played a role in
increasing the degree of racemisation, leading to a 7% increase in
the level of the LLLDL diastereoisomers (Table 2, compare entries
2, 4 and 10). So, we recommend using lower temperatures for the
microwave-assisted SPPS of Phg-containing peptides (≤50 ꢀC) or
at room temperature for the conventional stepwise SPPS.
for both Fmoc-removal (piperidine, piperazine or morpholine)
and coupling/activation steps (NMM) are used, together with
DMTMM-BF4 as a coupling agent at 50 ꢀC may, to provide signif-
icant advantages in the synthesis of peptides containing Phg and
other amino acid residues with high propensity to undergo
racemisation. Although we were able to achieve a 20% reduction
in the racemisation level of Phg encountered, we believe that
further investigations are needed to find a suitable base that
can completely eliminate racemisation of this residue during
Fmoc-group removal while retaining a high rate of efficiency in
this deprotection step.
Despite the optimization in the selection of the bases used in
the activation and Fmoc group removal steps, and reduction in
reaction temperature, we were only able to achieve a modest
overall reduction (~13%) in the formation of the ‘undesired’
LLLDL diastereoisomer, rather than the complete suppression
we sought. Consequently, we decided to examine an alternative
activator for HBTU in the hope of enhancing the rate of coupling
of the Fmoc-Phg-OH residue in comparison to the rate of
racemisation of this activated building block during the coupling
step. Our attention was drawn to the triazine coupling reagent
4-(4,6-dimethoxy-1,3,5-triazin-2-yl)-4-methylmorpholinium chloride
(DMTMM-Cl) that was introduced by Kamiński et al. (1998) [41]
and which has been applied to the synthesis of esters [42],
amides [43], and simple peptides [44]. Of direct relevance to
the present study are the reports that DMTMM-Cl yields peptide
and peptidomimetic products with much lower levels of racemi-
sation (in some instances, with complete suppression of the
undesired formation of diastereoisomers) in comparison to
those produced using other activating agents such as 2-chloro-
4,6-dimethoxy-1, 3, 5-triazine (CDMT) and HBTU/HOBT [44–46].
When we utilized DMTMM-CI as the reagent of choice for all
of the coupling steps in the microwave-assisted synthesis of
H-AVPPhgY-NH2, at 50 ꢀC, using NMM and piperidine as the bases
for the activation and Fmoc group removal steps, respectively
(Table 2, entry 8), we were able to achieve an additional 5%
suppression for the formation of the LLLDL diastereoisomer
(see Figure 1(e) for the RP-HPLC trace for this synthesis and see
Table 2 to compare entries 4–6 and 10 with 8). We attribute this
to the known ability of DMTMM-Cl to form super active ester
derivatives of urethane-protected amino acids that are ideal
acylating agents for amines. The formation of these super active
esters was first postulated by Kamiński [47], and we suggest that
this mechanism of amide bond formation (Figure 6, pathway A)
proceeds much more rapidly than the direct epimerization of
the Phg a proton (Figure 6, pathway B), thus, suppressing the in-
corporation of the D-enantiomer into the target peptide
sequence. However, we noticed that the degree of purity for
the crude peptide was lower than that of the HBTU-based synthe-
sis (Figure 1(e), see also Table 2 to compare the crude peptide
product purity for entries 1–6 and 10 with 8). This could possibly
be due to the formation of resin-bound N-(4,6-dimethoxy-1,3,5-
triazin-2-yl) derivatives at every coupling step (Figure 6, pathway
C), a known side reaction to DMTMM-Cl. To overcome the
formation of those side products, we replaced the DMTMM-Cl
with its tetrafluoroborate salt analogue (DMTMM-BF4) which
was reported by Kamiński et al. to produce peptides with
relatively high yields and higher degrees of purity [48]. The
appearance of ‘side product’ peaks was substantially decreased
(Figure 1(f)), and the percentage purity of the crude peptide
was greatly improved (97%) compared to that of the crude
peptide produced by DMTMM-Cl (86%), as anticipated (Table 2,
compare entries 8 and 9).
Acknowledgements
This work was supported by the generous donation of Dr John
King (grant code: R9113PMY). Many thanks for Dr. Brendan Gilmore
and Dr. Wei Chen in the School of Pharmacy, Queen’s University of
Belfast for their help in training us in the use of the Liberty™ micro-
wave synthesis system, and to Adrienne Healy, School of Biological
Sciences, Queen’s University of Belfast, for performing the ESI-mass
spectrometry analysis.
References
1 Scanlon BD, Eefting MA, Lloyd CJ, Burgess WA, Simpson JR. Synthesis of
biologically-active human transforming growth factor-a by fluorenyl-
methoxycarbonyl solid phase peptide chemistry. J. Chem. Soc. Chem.
Commun. 1987; (7): 516–518.
2 Basak A, Boudreault A, Chen A, Chretien M, Seidah N, Lazure C. Appli-
cation of the multiple antigenic peptides (MAP) strategy to the
production of prohormone convertases antibodies: synthesis, charac-
terization and use of 8-branched immunogenic peptides. J. Pept. Sci.
1995; 1(6): 385–395.
3 Escher SE, Sticht H, Forssmann WG, Rosch P, Adermann K. Synthesis
and characterization of the human CC chemokine HCC-2. J. Pept.
Res. 1999; 54(6): 505–513.
4 Langlois C, Tessier S, Brkovic A, Fournier A. An improved procedure for
the preparation of the endothelin-B-receptor ligand sarafotoxin 6c. J.
Cardiovasc. Pharmacol. 2000; 36(5, Suppl. 1): S290–S291
5 Koyack JM, Cheng PR. Design and synthesis of b-peptides with biologi-
cal activity. Meth. Mol. Biol. (Totowa, NJ, United States) 2006; 340(Protein
Design): 95–109.
6 Bacsa B, Kappe CO. Rapid solid-phase synthesis of a calmodulin-bind-
ing peptide using controlled microwave irradiation. Nat. Protoc. 2007;
2(9): 2222–2227.
7 Sabatino G, Papini AM. Advances in automatic, manual and micro-
wave-assisted solid-phase peptide synthesis. Curr. Opin. Drug Discov.
Devel. 2008; 11(6): 762–770.
8 Palasek SA, Cox ZJ, Collins JM. Limiting racemisation and aspartamide
formation in micro-wave enhanced Fmoc solid phase peptide
synthesis. J. Pept. Sci. 2007; 13: 143–148.
9 Loffrendo C, Assunção NA, Gerhardt J, Miranda MTM. Microwave-
assisted solid-phase peptide synthesis at 60 ꢀC: alternative conditions
with low enantiomerization. J. Pept. Sci. 2009; 15: 808–817.
10 Souza MP, Tavares MFM, Miranda MTM. Racemisation in stepwise
solid-phase peptide synthesis at elevated temperatures. Tetrahedron
2004; 60: 4671–4681.
11 Han Y, Albericio F, Barany G. Occurrence and minimization of cysteine
racemisation during stepwise solid-phase peptide synthesis. J. Org.
Chem. 1997; 62(13): 4307–4312.
12 Angell Y, Alsina J, Albericio F, Barany G. Practical protocols for
stepwise solid-phase synthesis of cysteine-containing peptides. J. Pept.
Res. 2002; 60(5): 292–299.
13 Hogue RA, Bibbs L, Bonewald L, Fields G, Kelly J, McMurray J, Moore
W, Weintraub S. Techniques in Protein Chemistry VIII. Academic Press:
San Diego, CA, 1997; 875–890.
14 Kaiser T, Nicholson G, Kohlbau H, Voelter W. Racemisation studies of
Fmoc-Cys(Trt)-OH during stepwise Fmoc-solid phase peptide synthesis.
Tetrahedron Lett. 1996; 37(8): 1187–1190.
In conclusion, we have identified a set of synthetic conditions
that has enabled us to prepare our target peptide sequence
H-AVPPhgY-NH2, (1) with enhanced diastereomeric purity. We
are of the opinion that microwave-SPPS in which mild bases
wileyonlinelibrary.com/journal/jpepsci Copyright © 2012 European Peptide Society and John Wiley & Sons, Ltd. J. Pept. Sci. 2012; 18: 302–311

RACEMISATION OF PHENYLGLYCINE IN MICROWAVE AND CONVENTIONAL SPPS
15 Bodanszky M, Ondetti MA. Peptide synthesis Chapter VI. Interscience:
NY, 1966; 137.
16 Li L, Zhong M. Preparation of cycloalkylarayls, cycloalkylheteroaryls
and arylheteroaryls end-capped with amino acid or peptide deriva-
tives as inhibitors of HCV NSSA. PCT Int. Appl. WO 2010065681 A1
20100610 2010.
17 Li G, Fathi G, Yang Z, Liao Y, Zhu Q, Lam A, Sandrasagra A, Nawoschik
K, Cho H-J, Cao J, Ruoqiu W, Wobbe CR. Preparation of aryl substituted
pyrazoles, especially N-[4-[[1H-pyrazol-4-yl]ethynyl]phenyl]dipeptide
derivatives, as HCV antiviral agents and their use in the treatment of
hepatitis C infection. PCT Int. Appl. WO 2008048589 A2 20080424 2008.
18 Chen KX, Njoroge FG, Pichardo J, Prongay A, Butkiewicz N, Yao N,
Madison V, Girijavallabhan V. Potent 7-Hydroxy-1,2,3,4-tetrahydroiso-
quinoline-3-carboxylic acid-based macrocyclic inhibitors of Hepatitis
C virus NS3 protease. J. Med. Chem. 2006; 49(2): 567–574.
19 Hans M, Bovenberg RAL, Klaassen P, Van der Laan JM. Engineering of
isopenicillin N synthase and non-ribosomal peptide synthetase with
improved activity for production of b-lactam antibiotics. PCT Int. Appl.
WO 2008040731 A2 20080410 2008.
20 Schon I, Kifaludy L, Varro V, Varga L, Nafradi J. Phenylglycine-containing
peptides with gastrin action. Ger. Offen. DE 2754770 A1 19780706 1978.
21 Nishino N, Hayashida J, Arai T, Mihara H, Ueno Y, Kumagai H. Cyclo
(Àarginyl-sarcosyl-aspartyl-phenylglycyl-)2. Simple synthesis of an
RGD-related peptide with inhibitory activity for platelet aggregation.
J.Chem. Soc. Perkin Trans. 1: Organ.Bio-Organ.Chem. 1996; 9: 939–946.
22 Sheehan SM, Masters JJ, Wiley MR, Young SC, Liebeschuetz JW, Jones
SD, Murray CW, Franciskovich JB, Engel DB, Weber WW, Marimuthu J,
Kyle JA, Smallwood JK, Farmen MW, Smith GF. A four component
coupling strategy for the synthesis of D-phenylglycinamide-derived
non-covalent factor Xa inhibitors. Bioorg. Med. Chem. Lett. 2003; 13(14):
2255–2259.
23 Watabe S, Hayakawa S, Kinoshita N, Ooshima K, Yanagihara Y, Okada
T, Imaoka T, Tanaka M, Hirohashi M. Preparation of peptides containing
4-hydroxyphenylglycine and/or phenylglycine as vascularisation inhibi-
tors. Jpn. Kokai Tokkyo Koho JP 06107681 A 19940419 1994.
24 Caliendo G, Greco G, Grieco P, Perissutti E, Santagada V, Calignano A,
Mancuso F, Novellino E. Modification of the potent peptide FK888
with unusual amino acids: effects on activity on neurokinin receptors.
Farmaco 1996; 5(3)1: 197–201.
32 Wade JD, Bedford J, Sheppard RC, Tregear GW DBU as an Na-depro-
tecting reagent for the fluorenylmethoxycarbonyl group in continu-
ous flow solid-phase peptide synthesis. Pept. Res. 1991; 4(3): 194–199.
33 Kates SA, Sole NA, Beyermann M, Barany G, Albericio F. Optimized
preparation of Deca(L-alanyl)-L-valinamide by 9-fluorenylmethyloxyar-
bonyl (Fmoc) solid-phase synthesis on polyethylene glycol-polystyrene
(PEG-PS) graft supports, with 1,8-diazabicyclo[5.4.0]undec-7-ene (DBU)
deprotection. Pept. Res. 1996; 9(3): 106–113.
34 Szabo I, Schlosser G, Hudecz F, Mezo G. Disulfide bond rearrangement
during regioselective oxidation in PhS(O)Ph/CH3SiCl3 mixture for the
synthesis of a-conotoxin GI. Biopolymers (Pept. Sci.) 2007; 88(1): 20–28.
35 Balse PM, Kim HJ, Han G, Hruby VJ. Evaluation of new base-labile 2-(4-
nitrophenylsulfonyl) ethoxycarbonyl (Nsc)-amino acids for solid-phase
peptide synthesis. J. Pept. Res. 2000; 56(2): 70–79.
36 Di-Segni S, Giordano C, Rahimipour S, Ben-Aroya N, Koch Y, Fridkin M.
b-sulfonamido gonadotropin-releasing hormone analogs: synthesis
and evaluation of several parent hormone properties. J. Pept. Sci.
2005; 11(1): 45–52.
37 Hall Jr. HK. Correlation of the base strengths of amines. J. Am. Chem.
Soc. 1957; 79(20): 5441–5444.
38 Anderson GW, Zimmerman JE, Callahan FM. Racemisation control in
the synthesis of peptides by the mixed carbonic-carboxylic anhydride
method. J. Am. Chem. Soc. 1966; 88: 1338–1339.
39 Bodanszky M, Bodanszky A. Racemisation in peptide synthesis.
Mechanism-specific models. Chem. commun. (London) 1967; 12:
591–593.
40 Bacsa B, Horvati K, Bosze S, Andreae F, Kappe CO. Solid-phase synthe-
sis of difficult peptide sequences at elevated temperatures: a critical
comparison of microwave and conventional heating technologies. J.
Org. Chem. 2008; 73(19): 7532–7542.
41 Kamiński ZJ, Paneth P, Rudziński J. A Study on the activation of
carboxylic acids by means of 2-chloro-4,6-dimethoxy-1,3,5-triazine
and 2-chloro-4,6-diphenoxy-1,3,5-triazine. J. Org. Chem. 1998; 63(13):
4248–4255.
42 Kunishima M, Morita J, Kawachi C, Iwasaki F, Terao K, Tani S. Esterifica-
tion of carboxylic acids with alcohols by 4-(4,6-dimethoxy-1,3,5-triazin-
2-yl)-4-methylmorpholinium chloride (DMTMM). Synlett 1999; 1999(8):
1255–1256.
43 Kunishima M, Kawachi C, Iwasaki F, Terao K, Tani S. Synthesis and
characterization of 4-(4,6-dimethoxy-1,3,5-triazin-2-yl)-4-methylmor-
pholinium chloride. Tetrahedron Lett. 1999; 40(29): 5327–5330.
44 Falchi A, Giacomelli G, Porcheddu A, Taddei M. 4-(4,6-Dimethoxy
[1,3,5]triazin-2-yl)-4-methylmorpholinium chloride (DMTMM): a valu-
able alternative to PyBOP for solid-phase peptide synthesis. Synlett
2000; 2000(2): 275–277.
25 Wang C-L, Fan Y-B, Lu H-H, Tsai T-H, Tsai M-C, Wang H-P. Evidence of
D-phenylglycine as delivering tool for improving l-dopa absorption. J.
Biomed. Sci. London, United Kingdom 2010; 17: 71–78.
26 Wang H-P, Fan Y-B, Lu H-H, Hsu W-L. Preparation and intestinal
absorption of L-dopa-D-phenylglycine. Yaowu Shipin Fenxi 2002; 10(2):
81–87.
27 Oost TK, Sun C, Armstrong RC, Al-Assaad A, Betz SF, Deckweth TL,
Ding H, Elmore SW, Meadows RP, Olejniczak ET, Olesksijew A,
Oltersdorf T, Rosenberg SH, Shoemaker AR, Tomaselli KJ, Zou H, Fesik
SW. Discovery of potent antagonists of the antiapoptotic protein XIAP
for the treatment of cancer. J. Med. Chem. 2004; 47(18): 4417–4426.
28 Bax A, Davis DG. MLEV-17-based two-dimensional homonuclear
magnetization transfer spectroscopy. J. Magnetic Resonance (1969–1992)
1985; 65(2): 355–360.
45 Kunishima M, Kitao A, Kawachi C, Watanabe Y, Iguchi S, Hioki K, Tani S.
A racemisation test in peptide synthesis using 4-(4,6-dimethoxy-1,3,
5-triazin-2-yl)-4-methylmorpholinium chloride (DMT-MM). Chem. Pharm.
Bull. 2002; 50(4): 549–550.
46 Shieh W-C, Chen Z, Xue S, McKenna J, Wang R-M, Prasad K, Repič O.
Synthesis of sterically-hindered peptidomimetics using 4-(4,6-dimethoxy-
1,3,5-triazine-2-yl)-4-methyl-morpholinium chloride. Tetrahedron Lett.
2008; 49: 5359–5362.
29 Bothner-By AA, Stephens RL, Lee J, Warren CD, Jeanloz RW. Structure
determination of a tetrasaccharide: transient nuclear Overhauser
effects in the rotating frame. J. Am. Chem. Soc. 1984; 106(3): 811–813.
30 Bodanszky M, Birkhimer CA. Chimia (Switz.) 1960; 14: 368.
31 Carreno C, Mendez ME, Kim Y-D, Kim H-J, Kates SA, Andreu D,
Albericio F. Nsc and Fmoc Na-amino protection for solid-phase
peptide synthesis: a parallel study. J. Pept. Res. 2000; 56(2): 63–69.
47 Kamiński ZJ. The concept of superactive esters. Could peptide
synthesis be improved by inventing superactive esters? Int. J. Pept.
Protein Res. 1994; 43(3): 312–319.
48 Kamiński ZJ, Kolesińska B, Kolesińska J, Sabatino G, Chelli M, Rovero P,
Błaszczyk M, Głowka ML, Papini AM. N-Triazinylammonium tetrafluor-
oborates. A new generation of efficient coupling reagents useful for
peptide synthesis. J. Am. Chem. Soc. 2005; 127: 16912–16920.
J. Pept. Sci. 2012; 18: 302–311 Copyright © 2012 European Peptide Society and John Wiley & Sons, Ltd. wileyonlinelibrary.com/journal/jpepsci