Automated 'X-Y' robot for peptide synthesis with microwave heating: Application to difficult peptide sequences and protein domains

Article abstract of DOI:10.1002/psc.1269

Precise microwave heating has emerged as a valuable method to aid solid-phase peptide synthesis (SPPS). New methods and reliable protocols, as well as their embodiment in automated instruments, are required to fully use this potential. Here we describe a new automated robotic instrument for SPPS with microwave heating, report protocols for its reliable use and report the application to the synthesis of long sequences, including the β-amyloid 1-42 peptide. The instrument is built around a valve-free robot originally developed for parallel peptide synthesis, where the robotic arm transports reagents instead of pumping reagents via valves. This is the first example of an 'X-Y' robotic microwave-assisted synthesizer developed for the assembly of long peptides. Although the instrument maintains its capability for parallel synthesis at room temperature, in this paper, we focus on sequential peptide synthesis with microwave heating. With this valve-free instrument and the protocols developed for its use, fast and efficient syntheses of long and difficult peptide sequences were achieved. Copyright

Full text of DOI:10.1002/psc.1269

Research Article
Received: 16 April 2010
Revised: 25 May 2010
Accepted: 2 June 2010
Published online in Wiley Interscience: 14 July 2010
(www.interscience.com) DOI 10.1002/psc.1269
Automated ‘X-Y’ robot for peptide
synthesis with microwave heating:
application to difficult peptide
sequences and protein domains
Leila Malik,‡ A. Pernille Tofteng,‡ Søren L. Pedersen, Kasper K. Sørensen
and Knud J. Jensen∗
Precise microwave heating has emerged as a valuable method to aid solid-phase peptide synthesis (SPPS). New methods and
reliable protocols, as well as their embodiment in automated instruments, are required to fully use this potential. Here we
describe a new automated robotic instrument for SPPS with microwave heating, report protocols for its reliable use and report
the application to the synthesis of long sequences, including the β-amyloid 1-42 peptide. The instrument is built around a
valve-free robot originally developed for parallel peptide synthesis, where the robotic arm transports reagents instead of
pumping reagents via valves. This is the first example of an ‘X-Y’ robotic microwave-assisted synthesizer developed for the
assembly of long peptides. Although the instrument maintains its capability for parallel synthesis at room temperature, in this
paper, we focus on sequential peptide synthesis with microwave heating. With this valve-free instrument and the protocols
c
developed for its use, fast and efficient syntheses of long and difficult peptide sequences were achieved. Copyright ꢀ 2010
European Peptide Society and John Wiley & Sons, Ltd.
Supporting information may be found in the online version of this article
Keywords: peptide synthesis; microwave heating; amyloid; LysM
Introduction
microwave-assisted SPPS by comparing it with SPPS using an
oil bath and concluded that the observed enhancement effect
was of thermal nature. Nonthermal effects were not evident,
which contradicts some previous suggestions [12]. Microwave
irradiation is an advantageous tool in peptide chemistry because
of the rapid and precise elevation of the temperature, and the
efficient temperature control during the synthesis. Furthermore,
if necessary, it can be combined with precise and reproducible
cooling after the reaction, e.g. by pressurized air.
In this paper, the use of a new fully automated robotic instru-
ment with precise microwave heating in SPPS is presented. The
instrument is built around a valve-free MultiSynTech instrument,
which incorporates an ‘X-Y’ robot equipped with digital syringe
pumps. The MultiSynTech instrument was originally developed
for parallel synthesis, where the robotic arm distributes reagents
(Figure S1, Supporting Information). Essentially, this mimics the
hand movements (i.e. it is chiro-mimic) in manual SPPS and is in
contrast to pumping reagents via valves, as in valve-based syn-
thesizers. The robot is combined with a modified Biotage Initiator
microwave instrument controlled by a single PC. The instrument
Solid-phase peptide synthesis (SPPS) has been tremendously
successful in the synthesis of short and medium length peptides
because of the development of efficient organic chemistry for
amide bond formation as well as for protection of amino groups
and side-chain functionalities. One of the current frontiers in
peptide science is the assembly of long peptides and small
proteins,whichhastoovercometheaccumulationofsidereactions
and the hurdles posed by so-called difficult sequences. In recent
years, precise and fast heating by microwave irradiation has
emerged as a new parameter to further optimize SPPS [1–6],
especially to enable the synthesis of long and difficult peptide
sequences or to improve their purity in general.
During the synthesis of difficult sequences, the peptide chain
most likely becomes partially inaccessible typically due to the
formation of secondary structures, especially β-sheets [7]. In
addition, steric hindrance from β-branched amino acids can
be a problem. Methods to suppress intramolecular aggregation
have been described and include pseudo-prolines [7], solvent
composition [8] and chaotropic salts [9,10]; however, their utility is
limited and the efficiency is variable. Intermolecular aggregation
could lead to poor solvation of the peptidyl-polymer, but it is less
pronounced with low-loading resins having a dynamic structure,
such as TentaGel. Heating is likely to reduce both the inter- and
intramolecular-derived aggregation and thereby decrease the
reaction time and improve the coupling efficiency of bulky and
β-branched amino acids, as well as Nα-deprotection of sterically
hinderedpeptidyl-polymers.Kappeet al.[11]haverecentlystudied
∗
Correspondence to: Knud J. Jensen, Faculty of Life Sciences, IGM, Section for
Bioorganic Chemistry, University of Copenhagen, Thorvaldsensvej 40, 1871
Frederiksberg C, Denmark. E-mail: kjj@life.ku.dk
Faculty of Life Sciences, IGM, Section for Bioorganic Chemistry, University of
Copenhagen, Thorvaldsensvej 40, 1871 Frederiksberg C, Denmark
‡
These authors contributed equally.
c
J. Pept. Sci. 2010; 16: 506–512
Copyright ꢀ 2010 European Peptide Society and John Wiley & Sons, Ltd.

AUTOMATED ‘X-Y’ ROBOT FOR PEPTIDE SYNTHESIS WITH MICROWAVE HEATING
maintains the capability for parallel synthesis at room tempera-
ture (RT); however, in this paper, we focus on sequential peptide
synthesis with microwave heating. The reaction vessel, a modified
plastic syringe with a filter bottom, is placed in the microwave
reactor for the duration of the synthesis (Figure S2). Mixing is
crucial during reactions with microwave heating in order to have
homogenous heat distribution in the reactor vessel, especially
when highly viscous reaction media are being used, as in SPPS.
Previously, mixing during microwave heating has been achieved
by magnetic stirring or bubbling with nitrogen, with the former
not being compatible with resins. However, vortexing has been
used extensively for SPPS at RT and has advantages over nitrogen
bubbling, as the latter carries a risk of blowing the resin to the
top of the reaction vessel. A novel feature of this new instrument
described here is that mixing is achieved by efficient vortexing
within the microwave cavity. Cooling is simply achieved as the
solvents used for washing are at RT.
the activated Fmoc amino acid in the coupling mixture was 0.18
M; however, a concentration of 0.15 M gave almost the same
results (data not shown). A complete cycle of deprotection, wash,
coupling and wash could be achieved in less than 30 min.
Protocol for Synthesis Using Conventionally RT Methods
Control peptides were prepared conventionally on a fully au-
tomated Syro Wave peptide synthesizer (Biotage AB) using
Fmoc-RAM-TG resin (0.433 g). Nα-Fmoc deprotection was per-
formed in two stages using solutions: (i) piperidine in DMF (2 : 3)
for 3 min and (ii) piperidine in DMF (1 : 4) for 10 min. Coupling
reactions were performed using Nα-Fmoc amino acids (5.2 equiv.,
0.5 M) which were dissolved in a 4 : 1 mixture of HOBt : HOAt (5.2
equiv., 0.5 M), HBTU (5 equiv., 0.43 M) and DIEA (10 equiv., 2 M)
in N-methyl-2-pyrrolidinone (NMP). After completing synthesis of
the peptide sequence, the resin was successively washed with
NMP (×3) and DCM (×3) and dried thoroughly. Peptide release
was performed using TFA–TES–H2O (95 : 2.5 : 2.5) for 2 h at RT.
The crude product was analyzed by LC-MS.
Materials and Methods
Materials
General Protocol for Synthesis of H-WFTTLISTIM-NH2 (1)
All organic solvents and Nα-Fmoc amino acids were obtained
from Iris Biotech GmbH and contained the following side-
chain-protecting groups: tert-butyl (OtBu, for Glu, Asp, Ser, Thr,
Tyr), 2,2,4,6,7-pentamethyl-dihydrobenzofuran-5-sulfonyl (Pbf, for
Arg), trityl (Trt, for Asn, Gln, His) and tert-butyloxycarbonyl (Boc for
Lys). All peptide amides were synthesized on a standard Fmoc-
RAM-TG resin from Rapp Polymere GmbH (loading 0.24 mmol/g).
Analytical HPLC was performed on a Dionex UltiMate 3000 with
Chromeleon 6.80SP3 software. Peptides were analyzed on a
Phenomenex Jupiter 300 Å C4 column (5 µm, 4.6 × 150 mm)
or a Phenomenex Gemini 110 Å C18 column (3 µm, 4.6 × 50 mm),
running a flow of 1.0 ml/min with a linear gradient with increasing
amount of buffer for 10–20 min (buffer A: 0.1% formic acid in
H2O; buffer B: 0.1% formic acid in CH3CN). Mass analyses were
performed using ESI-MS (MSQ Plus Mass Spectrometer, Thermo).
Peptide 1 was prepared on Fmoc-RAM-TG resin (0.433 g) (Tables 2
and 3, Figure 2). Deprotection was performed in two stages using
solutions: (i) piperidine in DMF (2 : 3) and (ii) piperidine in DMF
(1 : 4), with the first deprotection step at 60 ◦C for 2 min and then
followed by another deprotection step for 2 min at 60 ◦C. The
resin was then washed with NMP (3 × 45 s). Coupling reactions
were performed using Nα-Fmoc amino acids (5.2 equiv., 0.5 M),
which were dissolved in a 4 : 1 mixture of HOBt : HOAt (5.2 equiv.,
0.5 M), HBTU (5 equiv., 0.43 M) and DIEA (10 equiv., 2 M) in
NMP. Other coupling reagents were inserted as indicated in
Table 3. All couplings were performed for 5 min at 75 ◦C. Following
each coupling step, the resin was washed with NMP (×4). After
completing synthesis of the peptide sequence, the resin was
successivelywashedwithNMP(×3)andDCM(×3). Peptiderelease
was performed using TFA–TES–H2O (95 : 2.5 : 2.5) for 2 h at RT. The
crude product was analyzed by LC-MS (C18 column) via B gradient
elution (0–5 min: 5–37%, 5–8.5 min: 37–38% and 8.5–9.5 min:
38–100%) with an applied flow rate of 1.0 ml/min. Purity 44%.
ESI-MS, calculated monoisotopic composition for C58H90N12O14S,
1210.6 Da. Found: m/z 1211.6 [M + H]+.
Instrument for Robotic SPPS with Microwave Heating
Peptides were assembled on the fully automated solid-phase
microwave peptide synthesis system from Biotage AB (Syro
Wave). The instrument combines a Syro I synthesis robot from
MultiSynTech, with a Biotage Initiator microwave instrument,
incorporating a dedicated additional vortex mixer for the
microwave cavity (Figure S1). The microwave instrument has an
extended microwave guide and is attached on the left side of the
SyroI,whiletherightsidemaintainstheabilityforparallelsynthesis
atRT.Thereactionvesselisamodifiedplasticsyringe(withPTFEfrit)
which is vortexed within the microwave cavity during reactions.
The temperature was measured by a calibrated IR sensor. The
Initiator heated during coupling reactions and for some syntheses
also during Fmoc deprotections. The Initiator can deliver up to
200 W, but here it was programed to provide typically 40–160 W
during the ramp-up phase and a cap of max. 60 W was set during
the steady-state phase, where the temperature is maintained. For
a coupling temperature of 75 ◦C, the power required to maintain
a steady state was ∼40 W. The instrument was operated without
coolingduringmicrowaveirradiation.Thetotalvolumeofcoupling
reagents was 2.7 ml, which gave a volume of 3.3 ml with resin.
This volume proved efficient for reliable absorption of microwave
irradiation for a 10-ml reactor vial. The initial concentration of
Chiral Amino Acid Analysis of Peptide 1 Using COMU as
Coupling Reagent
Peptide1(Table 3, entry7, crudeproducts)wassubmittedtochiral
amino acid analysis at C.A.T. GmbH & Co. (Tu¨bingen, Germany)
to detect the level of D-enantiomer, which gave the following:
L-Threonine (>99.70%), D-Threonine (<0.10%), L-allo Threonine
(<0.10%), D-allo Threonine (<0.10%), L-Isoleucine (>99.70%), D-
Isoleucine, L-allo Isoleucine (<0.10%), D-allo Isoleucine (<0.10%)
Leucine (D-enatiomer 0.10%), Serine (D-enantiomer 0.10%), Me-
thionine (D-enantiomer 0.21%), Phenylalanine (D-enantiomer
0.21%) and Tryptophan (D-enantiomer 0.21%).
General Protocol for Synthesis of H-DAEFRHDSGYEVHHQKL
VFFAEDVGSNKGAIIGLMVGGVVIA-NH2 (2)
Peptide2waspreparedonFmoc-RAM-TGresin(0.433 g)(Figure 3).
Deprotection was performed in two stages using solutions:
(i) piperidine in DMF (2 : 3) and (ii) piperidine in DMF (1 : 4), with
c
J. Pept. Sci. 2010; 16: 506–512 Copyright ꢀ 2010 European Peptide Society and John Wiley & Sons, Ltd. www.interscience.com/journal/psc

MALIK ET AL.
a first deprotection step at 60 ◦C for 2 min and then followed
by another deprotection step for 2 min at 60 ◦C. Then washed
with NMP (3 × 45 s). Coupling reactions were performed using
Nα-Fmoc amino acids (5.2 equiv., 0.5 M), which were dissolved in
a 4 : 1 mixture of HOBt : HOAt (5.2 equiv., 0.5 M), HBTU (5 equiv.,
0.43 M) and DIEA (10 equiv., 2 M) in NMP. All couplings were
performed for 5 min at 75 ◦C. Following each coupling step, the
resin was washed with NMP (×4). After the incorporation of the
16th amino acid, a double-coupling protocol was performed using
the same method as above. After completing synthesis of the
peptide sequence, the resin was washed with NMP (×3) and DCM
(×3) and dried thoroughly. Peptide release was performed using
TFA–TES (97.5 : 2.5) (1 ml) for 2 h at RT, followed by addition of
EDT (16 µl) and TMSBr (13 µl) for 15 min. The crude product was
analyzed by LC-MS (C4 column) via B gradient elution (0–3.5 min:
5–30%, 3.5–15 min: 30–65% and 15–17 min: 65–100%) with an
applied flow rate of 1.0 ml/min. Purity 72%. ESI-MS, calculated
average isotopic composition for 4512.3 Da. Found: m/z 1129.0
The crude product was analyzed by LC-MS (C18 column) via B
gradient elution (0–9 min: 5%–100%) with an applied flow rate of
1.0 ml/min. The purity of peptide 4 was 75%. The purity of peptide
5 was ∼54%. ESI-MS, calculated average isotopic composition for
3682.97 Da. Found: m/z 1842.8 [M+2H]2+, 1228.8 [M+3H]3+, 921.8
[M+4H]4+
.
Results and Discussion
Couplings (reagents, temperature and time) and Nα-deprotection
(time and temperature) were optimized using peptide sequence
1 (Table 1), which originates from the C-terminus of the MuLV CTL
epitope [13]. Carpino et al. [14] have reported peptide 1 to be a
very difficult sequence. The optimized protocol was then used to
synthesize the following peptides (Table 1): human β-amyloid1-
42 peptide [15] (peptide 2), an analog of the peptide hormone
PYY3-36 [16,17] (peptide 3) and a region of the Nfr5 and LysM2
domain (Lotus japonicus sequence, peptide 4 and 5).
[M+4H]4+, 903.5 [M+5H]5+, 753.2 [M+6H]6+, 645.0 [M+7H]7+
.
Therefore, to evaluate the use of SPPS with microwave heating,
four peptides were assembled which in the literature have been
reported to be ‘difficult sequences’ (Table 1). The first sequence
investigated was the decapeptide derived from the MuLV CTL
epitope [13]. Jung and Redemann had previously reported that
the majority of deletion peptides come from the first nine C-
terminal amino acids, which prompted us to choose this short
sequence as a model for the evaluation of peptide synthesis,
in particular conventional at RT versus microwave heating. The
assemblyofpeptide1usingsemi-automatedsolid-phasesynthesis
haspreviouslybeenreportedtogiveaHPLCpurityof44%[18].Here
and in the following, yields indicate ‘crude purities’ after acidolytic
release of the peptides into solution. In the semi-automated
synthesis, the Nα-deprotections were performed at RT for 1 min,
followed by 2 min at 60 ◦C, whereas coupling reactions were at
80 ◦C for 10 min. Moreover, the semi-automated synthesis was
performed using a high-loaded PS-based resin. We build on these
results and aimed for a faster and more efficient route for fully
automated robotic peptide assembly with microwave heating.
We set out to test different reaction conditions using the
fully automated peptide synthesizer (Table 2). As a control, the
conventional synthesis of 1 (couplings for 45 min at RT and
deprotections at RT for 3 + 10 min) gave a HPLC purity of 35%
(Table 2, entry 1). Reducing the coupling time down to 5 min
further decreased the purity to 23% (Table 2, entry 2). Using
microwave irradiation during the coupling only (10 min at 75 ◦C)
and not the deprotections (3 + 10 min at RT) gave a purity of 40%
according to HPLC (Table 2, entry 3). We repeated the optimized
conditions from the semi-automated SPPS (couplings at 80 ◦C for
10 min and deprotections at 60 ◦C for 2 + 2 min), which afforded
a purity of 40% of 1 (Table 2, entry 4). To reduce cycle time further,
General Protocol for Synthesis of H-YLERELKKLERELKKLSPEE
LNRYYASLRHYLNLVTRQRY-NH2 (3)
Peptide 3 was prepared on TentaGel Rink amide resin (loading
0.24 mmol/g, 0.433 g). Deprotection was performed in two stages
using solutions: (i) piperidine in DMF (2 : 3) and (ii) piperidine
in DMF (1 : 4), With a first deprotection step at RT for 3 min
and then followed by another deprotection step for 10 min at
RT and washed with NMP (3 × 45 s). Coupling reactions were
performed using Nα-Fmoc amino acids (5.2 equiv., 0.5 M), which
were dissolved in a 4 : 1 mixture of HOBt : HOAt (5.2 equiv., 0.5
M), HBTU (5 equiv., 0.43M) and DIEA (10 equiv., 2 M) in NMP.
All couplings were performed for 10 min at 75 ◦C and double
couplings where applied in the entire sequence. Following each
coupling, the resin was washed with NMP (×4). After completing
synthesis of the peptide sequence, the resin was washed with
NMP (×3) and DCM (×3) and dried thoroughly. Peptide release
was performed using TFA–TES–H2O (95 : 2.5 : 2.5) for 2 h at RT.
The crude product was analyzed by LC-MS (C4 column) via B
gradient elution (0–14 min: 5–100%) with an applied flow rate
of 1.0 ml/min. Purity ∼25%. ESI-MS, calculated average isotopic
composition for C231H374N68O63, 5111.8 Da. Found: m/z 1279.2
[M+4H]4+, 1022.8 [M+5H]5+, 853.2 [M+6H]6+, 731.1 [M+7H]7+
.
General Protocol for Synthesis of H-LPERVKVVFPL-NH2 (4)
and H-YENLTNWNIVQAS-NPGVNPYLLPERVKVVFPL-NH2 (5)
Peptides 4 and 5 were prepared on TentaGel Rink amide resin
(loading 0.24 mmol/g, 0.433 g). Deprotection was performed in
two stages using solutions: (i) piperidine in DMF (2 : 3) and (ii)
piperidine in DMF (1 : 4), with a first deprotection step at RT for
3 min and then followed by another deprotection step for 10 min
at RT and washed with NMP (3 × 45 s). Coupling reactions were
performed using Nα-Fmoc amino acids (5.2 equiv., 0.5 M), which
were dissolved in a 4 : 1 mixture of HOBt : HOAt (5.2 equiv., 0.5
M), HBTU (5 equiv., 0.43 M) and DIEA (10 equiv., 2 M) in NMP.
All couplings were performed for 10 min at 75 ◦C and double
couplings were applied in the entire sequence. Following each
coupling, the resin was washed with NMP (×4). After completing
synthesis of the peptide sequence, the resin was washed with
NMP (×3) and DCM (×3) and dried thoroughly. Peptide release
was performed using TFA–TES–H2O (95 : 2.5 : 2.5) for 2 h at RT.
Table 1. The five difficult sequences
Peptide Sequences
1
2
3
4
5
H-WFTTLISTIM-NH2
H-DAEFRHDSGYEVHHQKLVFFAEDVGSNKGAIIGLMVGGVVIA-NH2
H-YLERELKKLERELKKLSPEELNRYYASLRHYLNLVVTRQRY-NH2
a
H-LPERVKVVFPL-NH2
H-YENLTNWNIVQASNPGVNPYLLPERVKVVFPL-NH2
a Peptide 4 is the first 11 amino acids from the C-terminal of peptide 5.
c
www.interscience.com/journal/psc Copyright ꢀ 2010 European Peptide Society and John Wiley & Sons, Ltd. J. Pept. Sci. 2010; 16: 506–512

AUTOMATED ‘X-Y’ ROBOT FOR PEPTIDE SYNTHESIS WITH MICROWAVE HEATING
Table 2. Different reaction conditions for the synthesis of peptide 1 (H-WFTTLISTIM-NH2)
Entry
Deprotection
Coupling reagent
Coupling temperature
Coupling time (min)
HPLC purity (%)
1
2
3
4
5
3 + 10 min, RT
3 + 10 min, RT
3 + 10 min,RT
2 + 2 min, 60 ◦C
2 + 2 min, 60 ◦C
HBTU/(HOBt : HOAt)
HBTU/(HOBt : HOAt)
HBTU/(HOBt : HOAt)
HBTU/(HOBt : HOAt)
HBTU/(HOBt : HOAt)
RT
1 × 45
1 × 5
35
23
40
40
46
RT
75 ◦C
80 ◦C
75 ◦C
1 × 10
1 × 10
1 × 5
Couplings reagent used was HBTU/(HOBt : HOAt).
Figure 1. HPLC chromatograms of protocol optimization for peptide 1. The purity of peptide 1 is given in the right side of each chromatogram. (1)
des-(Trp), (2) des-(Thr) and (3) [M+H].
we reduced the coupling time (5 min at 75 ◦C) and surprisingly
a noticeable better purity of 46% was obtained (Table 2, entry
Table 3. Different coupling reagents for the synthesis of peptide 1
(H-WFTTLISTIM-NH2)
5). The major impurities in all the runs were deletion peptides
Entry
Coupling reagent
HPLC purity (%)
des-(Trp) and des-(Thr) (Figure 1). We observed that using fresh
stock solutions of amino acid and especially new solutions of
Trp enhanced the purity by lowering, e.g., des-Trp byproduct
formation (data not shown). One of the concerns using elevated
temperature during SPPS is the possibility to form epimers. Here,
this question was addressed by comparison of previously reported
data for this decapeptide with our LC-MS analysis [14]. We saw
no indication of epimerization in any of our experiments, which
suggested that the heating does not cause significant epimers to
form during the synthesis of this peptide.
1
2
3
4
5
6
7
PyBOP/(HOBt: HOAt)
DIC/(HOBt : HOAt)
35
38
40
42
44
49
70
TSTU/(HOBt : HOAt)
HCTU/(HOBt : HOAt)
HBTU/(HOBt : HOAt)
HATU/(HOBt : HOAt)
COMU/(HOBt : HOAt)
In all the experiments, ‘‘deprotection was 2 + 2 min at 60 ◦C and the
coupling time was 5 min and the temperature 75 ◦C’’.
Thus, when comparing fully automated SPPS of decapeptide
1 under conventional conditions at RT with microwave-assisted
assembly,aclearadvantageofmicrowaveheatingemerged.When
applying microwave heating, the peptide was synthesized faster
(5 vs 45 min for coupling, 4 vs 13 min for deprotection). The main
advantage lies in the dramatic reduction in coupling time, whereas
the reduction in time for Fmoc deprotections is less pronounced.
Additionally, an increase in purity was observed from 35 to 46%
using microwave heating.
The overall time for a full cycle of coupling, deprotection and
washings is not solely determined by time used for couplings
c
J. Pept. Sci. 2010; 16: 506–512 Copyright ꢀ 2010 European Peptide Society and John Wiley & Sons, Ltd. www.interscience.com/journal/psc

MALIK ET AL.
and deprotections. The time used for washings and for reagent
transfer should also be included in the calculation. A cycle time
of 30 min for the difficult peptide 1 was obtained, which gives a
total synthesis time of 5.5 h (Table S1, Supporting Information).
These initial experiments clearly demonstrated that a peptide
synthesizer based on an ‘X-Y’ robot can indeed achieve fast
syntheses of peptides.
The peptide hormone PYY3-36 plays a central role in the
regulation of food intake and energy homeostasis [22]. The
optimized protocol described above was applied to the synthesis
of an analog of PYY3-36 (3) consisting of a helix-loop-helix
structure, with both helices being amphipathic (Table 1, entry
3). Although the C-terminal helix and the loop were maintained
from the native sequence, the N-terminal helix was inspired by
a structure reported by Vagt et al. [23]. In previous experiments,
we had observed that this sequence was difficult to synthesize
in high yield and purity. First, the synthesis of peptide 3 was
attempted using conventional conditions at RT (couplings for
45 min, deprotection for 3 + 12 min), but this gave no yield.
Increasing the coupling times at RT and repeating the couplings
gave a good yield and purity (coupling for 2 × 2 h, deprotection
for 3+12 min). Next, the sequence was synthesized on the robotic
synthesizer with microwave heating, which required optimization
of the protocol, as the initially used 5-min coupling at 75 ◦C
and deprotection 3 + 10 min at RT gave no product. However,
peptide 3 was successfully synthesized using double couplings
and microwave heating (2 × 10 min at 75 ◦C) and deprotections
withoutheating(3+10 minatRT).Thisaffordedpuritycomparable
with the much more time-consuming conventionally optimized
protocolabove. Thisclearlydemonstratedthatmicrowaveheating
appliedtothesynthesisofdifficultsequencescangivedramatically
shortened coupling times (coupling for 240 min conventionally vs
20 min using microwave heating).
The LysM domain was predicted to consist of two α-helices and
a two-stranded anti-parallel β-sheet in a β-α-α-β structure and has
been identified in Nfr5 [24] by sequence alignment of the crystal
structure with the LysM domain of Bacillus subtilis ykuD [25]. There
were significant synthetic challenges arising from the C-terminal
andtheN-terminalregions;thisispresumablyduetotheformation
of β-sheet-like structures, which is known to pose problems for
peptide chain assembly. Thus, peptide 4, which is derived from
the C-terminus of the LysM2 domain and contains several β-
branched and bulky amino acid residues, was assembled under
various conditions. When applying standard SPPS protocols at RT,
we were able to synthesize peptide 4 in very low yield, but when
using the very slow conventional protocol at RT, the peptide was
produced in reasonable yield (couplings for 2×2 h, deprotections
for 3 + 10 min). Then, we turned to the optimized protocol with
microwave heating for the synthesis of this sequence (couplings
for 2 × 10 min at 75 ◦C) and deprotections at RT (3 + 10 min),
this afforded an excellent HPLC purity of 75% compared with
previous work (Figure 4). Again, when comparing conventional
and microwave heating protocols, the immense advantage in
using microwave heating was seen in the reduction of coupling
times (20 min instead of 240 min) and the higher purity obtained
by avoiding the formation of several deletion peptides sequences
(data not shown). Notably, this was achieved without microwave
heating during the deprotections. Finally, we set out to synthesize
peptide 5 (continuing from peptide 4), which is the 32-mer from
the C-terminus of the LysM2. All couplings were achieved with
microwaveheatingandwiththeaboveoptimizedcondition,which
afforded the peptide in an acceptable yield 55%.
Next, weevaluatedtheuseofdifferentcouplingreagentsforthe
assembly of peptide 1 and their effect on purity. We maintained
the optimized microwave heating protocol described above and
HOBt/HOAtasnucleophilicadditive,butsystematicallysubstituted
HBTU with other coupling reagents, such as HATU, HCTU, TSTU,
DIC, PyBOP and COMU (Table 3). The result showed that TSTU,
HCTU and HBTU gave similar purities (Table 3, entry 3–5) and
that DIC and PyBOP gave reduced purities compared to HBTU
(Table 3, entry 1–2). HATU gave an increase in purity compared
with HBTU (Table 3, entry 6). COMU, a relatively new coupling
reagent, has been reported to give a significantly higher purity
of the Aib-analog of Leu-enkephalin pentapeptide (H-Tyr-Aib-Aib-
Phe-Leu-NH2) than standard coupling reagents, such as HATU and
HBTU [19]. We thus tested COMU as a coupling reagent for the
synthesis of peptide 1. This gave the highest HPLC purity (70%),
which was significantly better than for other coupling reagents.
Analysis of the LC-MS data revealed no epimerization (identical
masses for separate HPLC peaks); furthermore, a significant
reduction in deletion peptides was observed (Figure 2). Thus,
microwave heating with COMU as the coupling reagent gave the
best purity of decapeptide 1.
Peptide 1 (Table 3, entry 7) was taken as a representative
example and the crude peptide material was subjected to chiral
amino acid analysis. The crude peptide, rather than the purified
peptide, was chosen as this provided an analysis of all peptide
materialinthissynthesisandthusavoidingabias.Thisshowedonly
insignificant levels of epimerization, which provided additional
support for the absence of (or insignificant level of) epimerization
under these conditions.
The human β-amyloid1-42 peptide 2 is one of the main
constituentsinamyloidplaquesinthebrainofAlzheimer’spatients
and has been the target of significant attention in recent years;
however, its solid-phase assembly is non-trivial. The synthesis of
this peptide commenced with a reference experiment in which
peptide 2 was synthesized conventionally with an optimized,
but very time-consuming protocol at RT (couplings of the first
16 C-terminal residues for 45 min and the remaining residues
for 2 × 45 min, with Fmoc deprotection for 3 + 10 min). Analytical
HPLCrevealedapeakarisingfromtheMet(O)-35peptide,however,
acidolytic release of the peptide under reducing conditions [20],
afforded the crude product in a rewarding yield of 54%. The
application of microwave heating during the couplings (first 16
C-terminal residues for 5 min and the remaining for 2 × 5 min),
and the same deprotection condition at RT as above, provided
comparableyieldasforlongcouplingtimesatRT(datanotshown).
Next, microwave irradiation was used during coupling (as above)
and deprotection (2+2 min at 60 ◦C) this gave an improved purity
of 72% (Figure 3). In all syntheses of peptide 2, a peak arising from
the Met(O)-35 peptide was observed when no precautions were
taken, therefore all cleavages were performed using the above
reductive release protocol. The formation of Met(O)-35 peptide
in the synthesis of β-amyloid1-42 peptide has been observed by
others [21]. For this particular sequence, human β-amyloid1-42
peptide, heating during Fmoc removal proved advantageous.
Conclusion
In conclusion, we have developed reliable protocols for SPPS
of difficult and long sequences using a new automated robotic
instrument which combines the function of an ‘X-Y’ laboratory
c
www.interscience.com/journal/psc Copyright ꢀ 2010 European Peptide Society and John Wiley & Sons, Ltd. J. Pept. Sci. 2010; 16: 506–512

AUTOMATED ‘X-Y’ ROBOT FOR PEPTIDE SYNTHESIS WITH MICROWAVE HEATING
Figure 2. HPLC chromatograms of coupling reagent optimization experiments. The purity of peptide 1 is given in the right side of each chromatogram.
(1) des-(Trp) and (3) [M+H].
min
8.5
3.5
5.0
6.0
7.0
8.0
9.0
10.0
11.0
1.60
3.0
4.00
5.00
6.0
7.0
Figure 3. Crude analytical HPLC chromatogram of the β-amyloid1-42 (2)
after acidolytic release from the support under mild reducing conditions.
Figure 4. Crude analytical HPLC chromatogram of the C-terminal 11mer
(4) of the LysM2 domain.
robot with microwave heating. Difficult peptides as well as long
sequences including the β-amyloid 1–42 peptide and the LysM2
protein were successfully assembled in an excellent crude purity.
This study focused on the purities of the synthesized peptides;
future studies will also include analyses of the crude yields. In
general, syntheses were much faster than with conventional
c
J. Pept. Sci. 2010; 16: 506–512 Copyright ꢀ 2010 European Peptide Society and John Wiley & Sons, Ltd. www.interscience.com/journal/psc

MALIK ET AL.
protocols, where each cycle time with microwave heating often
could be efficiently performed in 30 min or less, instead of 3 h or
more required for difficult sequences at RT. Microwave heating
during Nα-Fmoc deprotections proved beneficial for the synthesis
of some peptides, but was not required in all cases. In general,
all standard coupling reagents gave comparable purities under
microwave conditions, with HATU slightly above average and DIC
as well as PyBOP slightly below.
9. Seebach D, Thaler A, Beck A. Solubilization of peptides in non-
polar organic solvents by the addition of inorganic salts: facts
and implications. Helv. Chim. Acta. 1989; 72: 857–867.
10. Thaler A, Seebach D, Cardinaux F. Lithium-salt effects in peptide
synthesis. Part II. Improvement of degree of resin swelling and of
efficiencyofcouplinginsolid-phasesynthesis. Helv.Chim.Acta. 1991;
74: 628–643.
`
`
11. Bacsa B, HorvaIti K, BoIfsze S, Andreae F, Kappe CO. Solid-phase
synthesis of difficult peptide sequences at elevated temperatures:
a critical comparison of microwave and conventional heating
technologies. J. Org. Chem. 2008; 73: 7532–7542.
12. Palasek S, Cox Z, Collins J. Limiting racemization and aspartimide
formation in microwave-enhanced Fmoc solid phase peptide
synthesis. J. Pept. Sci. 2007; 13: 143–148.
13. Redemann T, Jung G. In situ fluoride activation allows the
preparation of peptides not accessible by routine synthesis
protocols. Peptides (Proceedings of the 24th European Peptide
Symposium). 1996; 749–750.
14. Carpino LA, Krause E, Sferdean CD, Schu¨mann M, Fabian H,
Bienert M, Beyermann M. Synthesis of ‘difficult’ peptide sequences:
application of a depsipeptide technique to the Jung-Redemann 10-
and 26-mers and the amyloid peptide Ab(1–42). Tetrahedron Lett.
2004; 45: 7519–7523.
15. Tickler A,Clippingdale A,Wade J.Amyloid – asa‘‘DifficultSequence’’
insolidphasepeptidesynthesis.ProteinPept.Lett.2004;11:377–384.
16. Lerch M, Mayrhofer M, Zerbe O. Structural similarities of micelle-
bound peptide YY (PYY) and neuropeptide Y (NPY) are related to
their affinity profiles at the Y receptors. J. Mol. Biol. 2004; 339:
1153–1168.
17. Pedersen SL, Vrang N, Jensen KJ. Novel N-terminally modified PYY3-
36 analogs: design, structural analysis, affinity study and in vivo
evaluation. Biopolymers 2009; 4: 323.
18. Pedersen SL, Sørensen KK, Jensen KJ. Semi-automated microwave-
assisted SPPS: optimization of protocols and synthesis of difficult
sequences. Biopolymers (Peptide Science), 2010; 94: 206–212.
19. Subiros-Funosas R, Acosta GA, El-Faham A, Albericio F. Microwave
irradiation and COMU: a potent combination for solid-phase peptide
synthesis. Tetrahedron Lett. 2009; 50: 6200–6202.
Acknowledgements
We thank Dr Udo Treffer from MultiSynTech, and Mike Lally,
Anders Wikstro¨m, Ola Strandberg and Dr Amit Mehrotra from
Biotage for their essential contributions and their support as
well as Stig Jacobsen for his help in the laboratory. We also
acknowledge Andreas Hoel and Helmi Selin, previously at Biotage,
for their support. Financial support to KJJ from The Danish Council
for Strategic Research, the Danish National Research Foundation
(Centre for Carbohydrate Recognition and Signaling), the Danish
Council for Independent Research-Technology and Production
Sciences and the Villum-Kann-Rasmussen foundation is gratefully
acknowledged.
Supporting information
Supporting information may be found in the online version of this
article.
References
1. Bacsa B, Desai B, Dibo´ G, Kappe CO. Rapid solid-phase peptide
synthesisusingthermalandcontrolledmicrowaveirradiation.J. Pept.
Sci. 2006; 12: 633–638.
20. Beck W, Jung G. Convenient reduction of S-oxides in synthetic
peptides, lipopeptides and peptide libraries. Lett. Pept. Sci. 1994;
1: 31–37.
2. Bacsa B, Kappe C. Rapid solid-phase synthesis of
a
21. Quibell M, Turnell WG, Johnson T. Preparation and purification
of beta-amyloid (1–43) via soluble, amide backbone protected
intermediates. J. Org. Chem. 1994; 59: 1745–1750.
22. Batterham RL, Cowley MA, Small CJ, Herzog H, Cohen MA, Dakin CL,
Wren AM, Brynes AE, Low MJ, Ghatei MA, Cone RD, Bloom SR. Gut
hormone PYY3-36 physiologically inhibits food intake. Nature 2002;
418: 650–654.
calmodulinbinding peptide using controlled microwave irradiation.
Nat. Protoc. 2007; 2: 2222–2227.
3. Brandt M, Gammeltoft S, Jensen K. Microwave heating for solid-
phase peptide synthesis: general evaluation and application to
15-mer phosphopeptides. Int. J. Pept. Res. Ther. 2006; 12: 349–357.
4. Erde´lyi M, Gogoll A. Rapid microwave-assisted solid-phase peptide
synthesis. Synthesis 2002; 11: 1592–1596.
23. Vagt T, Zscho¨rnig O, Huster D, Koksch B. Membrane binding and
structure of de novo designed alpha-helical cationic coiled-coil-
forming peptides. ChemPhysChem 2006; 7: 1361–1371.
24. Madsen EB, Madsen LH, Radutoiu S, Olbryt M, Rakwalska M,
Szczyglowski K, Sato S, Kaneko T, Tabata S, Sandal N, Stougaard J.
5. Fara M, Díaz-Mocho´n J, Bradley M. Microwave-assisted coupling
with DIC/HOBt for the synthesis of difficult peptoids and
fluorescently labelled peptides – a gentle heat goes a long way.
Tetrahedron Lett. 2006; 47: 1011–1014.
6. Matsushita T, Hinou H, Kurogochi M, Shimizu H, Nishimura S. Rapid
microwave-assisted solid-phase glycopeptide synthesis. Org. Lett.
2005; 7: 877–880.
7. Wohr T, Wahl F, Nefzi A, Rohwedder B, Sato T, Sun X, Mutter M.
Pseudo-prolines as a solubilizing, structure-disrupting protection
technique in peptide synthesis. J. Am. Chem. Soc. 1996; 118:
9218–9227.
A
receptor kinase gene of the LysM type is involved in
legumeperception of rhizobial signals. Nature 2003; 425: 637–640.
25. Ohnuma T, Onaga S, Murata K, Taira T, Katoh E. LysM domains from
Pteris ryukyuensis chitinase-A. J. Biol. Chem. 2008; 283: 5178–5187.
8. Yamashiro D, Blake J, Hao Li C. The use of trifluoroethanol for
improved coupling in solid-phase peptide synthesis. Tetrahedron
Lett. 1976; 17: 1469–1472.
c
www.interscience.com/journal/psc Copyright ꢀ 2010 European Peptide Society and John Wiley & Sons, Ltd. J. Pept. Sci. 2010; 16: 506–512