An automatic solid-phase synthesis of peptaibols

Article abstract of DOI:10.1021/jo802058x

An automated approach to peptaibols using microwave-assisted solid-phase peptide synthesis is demonstrated with a combination of HBTU and acid fluoride mediated couplings for normal and α,α-dialkylated amino acids, respectively. The method is utilized for

Full text of DOI:10.1021/jo802058x

acids is still not considered straightforward. One such class of
peptides is the peptaibols, which have received considerable
interest due to their membrane-disrupting abilities leading to
antibiotic activity.4,5 This holds, for example, for the intensely
studied peptaibol, alamethicin, which is considered an excellent
model for ion-channel behavior and trans-lipid pore formation.6
Peptaibols contain a high proportion of the R,R-dialkylated
amino acids, R-aminoisobutyric acid (Aib) and/or L-isovaline
(Iva), which, due to their increased steric bulk, couple to the
peptidyl resin in poor yields using standard SPPS coupling
reagents such as HBTU, HATU, or PyBOP.7,8 Because of these
difficulties, peptaibols or peptaibol fragments have been syn-
thesized using a variety of other methods such as solution-phase
segment condensations,9 expression,10 and SPPS approaches
including the solid-phase azirine/oxazolone method11 and the
amino acid fluoride approach.7,12 All these methods, however,
suffer from an intense consumption of man-hours due to their
use of noncommercially available reagents and in most cases a
lack of automation. Exactly this problem stimulated the
development of an automated procedure for peptaibol synthesis
presented here.
An Automatic Solid-Phase Synthesis of Peptaibols
Claudia U. Hjørringgaard, Jan M. Pedersen,*
Thomas Vosegaard, Niels Chr. Nielsen, and
Troels Skrydstrup*
Center for Insoluble Protein Structures, Department of
Chemistry and the Interdisciplinary Nanoscience Center,
Aarhus UniVersity, Langelandsgade 140,
8000 Aarhus C, Denmark
ts@chem.au.dk
ReceiVed September 22, 2008
To facilitate our labeling studies of alamethicin and other
peptaibols,13 we sought an effective, cheap, and automated solid-
phase synthesis of these peptides, using commercially available
reagents whenever possible and applying a minimum of
reprogramming of the peptide synthesizer. Our initial attempts
using standard HBTU-mediated Fmoc SPPS under microwave
irradiation failed. Likewise, in situ TFFH activation, which we
utilized previously for the semiautomatic synthesis of alame-
thicin,13 was not directly transferable to any of our fully
automatic synthesizers, although an automated synthesis of
An automated approach to peptaibols using microwave-
assisted solid-phase peptide synthesis is demonstrated with
a combination of HBTU and acid fluoride mediated couplings
for normal and R,R-dialkylated amino acids, respectively.
The method is utilized for the automated synthesis of several
full-length peptaibols, including alamethicin, tylopeptin,
ampullosporin, bergofungin, cervinin, trikoningin, trichogin,
and peptaibolin, reducing both synthesis time and costs
significantly as compared to other approaches. Furthermore,
the use of noncommercially available reagents is minimized.
(4) For special journal issues on peptaibols, see: Chem. BiodiVersity 2007,
4, 1021-1412; J. Pept. Sci. 2003, 9, 659-837.
(5) A peptaibol database exists: Whitmore, L.; Chugh, J. K.; Snook, C. F.;
Wallace, B. A. J. Pept. Sci. 2003, 9, 663–665. See also http://www.cryst.b-
bk.ac.uk/peptaibol.
(6) For a recent review, see: Leitgeb, B.; Szekeres, A.; Manczinger, L.;
Va´gvo¨lgyi, C.; Kredics, L. Chem. BiodiVersity 2007, 4, 1027–1051.
(7) (a) Wenschuh, H.; Beyermann, M.; Krause, E.; Brudel, M.; Winter, R.;
Schiimann, M.; Carpino, L. A.; Bienert, M. J. Org. Chem. 1994, 59, 3275–
3280. (b) Sapia, A. C.; Slomczynska, U.; Marshall, G. R. Lett. Pept. Sci. 1994,
1, 283–290.
(8) Selected examples: (a) Gisin, B. F.; Davis, D. G.; Borowska, Z. K.; Hall,
J. E.; Kobayashi, S. J. Am. Chem. Soc. 1981, 103, 6373–6377. (b) Augeven-
Bour, I.; Rebuffat, S.; Auvin, C.; Goulard, C.; Prigent, Y.; Bodo, B. J. Chem.
Soc., Perkin Trans. 1 1997, 1587–1594. (c) Peggion, C.; Coin, I.; Toniolo, C.
Pept. Sci. 2004, 76, 485–493. (d) Baldini, C.; Bellanda, M.; Peggion, C.; Djontu,
A. L.; Atagua, C.; Mammi, S.; Toniolo, C. Chem. BiodiVersity 2007, 4, 1129–
1143.
(9) Ovchinnikova, T. V.; Shenkarev, Z. O.; Yakimenko, Z. A.; Svishcheva,
N. V.; Tagaev, A. A.; Skladnev, D. A.; Arseniev, A. S. J. Pept. Sci. 2003, 9,
817–826.
(10) Stamm, S.; Heimgartner, H. Eur. J. Org. Chem. 2004, 3820–3827.
(11) Carpino, L. A.; Sadat-Aalaee, D.; Chao, H. G.; DeSelms, R. H. J. Am.
Chem. Soc. 1990, 112, 9651–9652.
Solid-phase peptide synthesis (SPPS) has become the pre-
ferred method for the synthesis of small peptides since its
implementation in the 1960s by Merrifield.1 The solid-phase
approach is especially attractive due to the possibility of full
automation of the synthesis process using robotic peptide
synthesizers. Two automatic SPPS techniques are dominant,
differing in the N-R-protecting group on the peptidyl-resin,
namely the Boc- and the Fmoc-approach,2,3 the latter becoming
increasingly popular as the use of hazardous hydrofluoric acid
is avoided.
Notwithstanding the success of SPPS, the synthesis of certain
peptides containing highly hindered nonproteinogenic amino
(1) Merrifield, R. B. J. Am. Chem. Soc. 1963, 85, 2149–2154.
(2) (a) Carpino, L. A. J. Am. Chem. Soc. 1957, 79, 4427–4431. (b) McKay,
F. C.; Albertson, N. F. J. Am. Chem. Soc. 1957, 79, 4686–4690. (c) Anderson,
G. W.; McGregor, A. C. J. Am. Chem. Soc. 1957, 79, 6180–6183.
(3) (a) Carpino, L. A.; Han, G. A. J. Org. Chem. 1972, 37, 3404–3409. (b)
Sheppard, R. J. Pept. Sci. 2003, 9, 545–552. (c) Chan, W. C., White, P. D., Ed.
Fmoc solid phase peptide synthesis: a practical approach; Oxford University
Press: Oxford, 2003.
(12) Wenschuh, H.; Beyermann, M.; Rothemund, S.; Carpino, L. A.; Bienert,
M. Tetrahedron Lett. 1995, 36, 1247–1250.
(13) (a) Bertelsen, K.; Pedersen, J. M.; Rasmussen, B. S.; Skrydstrup, T.;
Nielsen, N. C.; Vosegaard, T. J. Am. Chem. Soc. 2007, 129, 14717–14723. (b)
Vosegaard, T.; Bertelsen, K.; Pedersen, J. M.; Thøgersen, L.; Schiøtt, B.;
Tajkhorshid, E.; Skrydstrup, T.; Nielsen, N. C. J. Am. Chem. Soc. 2008, 130,
5028–5029.
10.1021/jo802058x CCC: $40.75
Published on Web 12/24/2008
2009 American Chemical Society
J. Org. Chem. 2009, 74, 1329–1332 1329

TABLE 1. Coupling Cycles Used during SPPS of Peptaibolsa
crude purity. Therefore, method B expands on method A by
using the standard double cycle on all proteinogenic amino acids
that are coupled onto an R,R-dialkylated amino acid. The
methods are exemplified for the synthesis of alamethicin F30
in Table 2 using the color-coding of the three different cycles
shown in Table 1.
Initially, we tested method A for the synthesis of short
peptaibols, namely peptaibolin (5 residues) and trichogin A IV
(11 residues), and we were pleased to see that this straightfor-
ward method gave the desired peptides in very good crude
purities of 94% and 96%, respectively, which enabled us to
isolate them in good yields using HPLC (Table 3, entries 1 and
2). Encouraged by these results, we proceeded to test the method
for the synthesis of the 20 residue peptaibol alamethicin.
However, in this case, method A proved inadequate as the crude
mixture contained a large proportion of deletion products (Table
3 entry 10). Especially apparent were products involving valine
(V) and glutamine (Q) deletions. The nature of the deletion
products led us to conclude that the limitations of method A
when synthesizing longer peptaibols were due to low yields of
peptide coupling of some of the proteinogenic amino acids onto
Aib. As we wanted to avoid the synthesis of further amino acid
fluorides, we tested method B (Table 2) on the alamethicin
synthesis, where couplings onto R,R-dialkylated amino acids
are performed using a double HBTU cycle. Gratifyingly, method
B gave a sufficiently pure crude product providing alamethicin
in a 24% yield after HPLC purification (Table 3, entry 11).
Furthermore, the result showed that when coupling two con-
secutive Aib residues onto the peptidyl resin using acid fluorides
the use of a double acid fluoride cycle is not required for the
second Aib coupling.
To further explore the scope of the methods, we gradually
increased the length of the peptaibol sequences from 5 to 20
while also synthesizing peptaibols containing isovaline and
hydroxyproline residues. The 11 residue peptaibol trikoningin
KB I was successfully synthesized in an isolated 21% yield
using method A, albeit with a relatively low crude purity of
35% (Table 3, entry 4). Employing method B to the synthesis
of trikoningin only increased the crude and isolated yields
slightly (Table 3, entry 5), indicating that some peptaibols
contain “difficult” sequences not related to low coupling yields,
which can normally be increased by employing method B.
Cervinin I containing 12 residues was synthesized in a 20%
yield employing method A (Table 3, entry 6), whereas the 14
peptaibol bergofungin D was isolated in a 11% yield using
method B (Table 3, entry 7). The synthesis of bergofungin
demonstrates that peptaibols containing hydroxyproline (O)
residues are also available applying the present approach. The
decreased yield of bergofungin D prompted us to use method
B for the longer peptaibols ampullosporin I and tylopeptin A.
Ampullosporin was isolated in a good yield of 22% eVen though
the sequence contains three consecutiVe Aib residues (Table 3,
entry 8). The successful synthesis of tylopeptin in an isolated
yield of 12% illustrates the ability of method B to also
incorporate the highly hindered isovaline residues (J), which
are common in peptaibols (Table 3, entry 9).17 Thus a variety
of peptaibols have been successfully synthesized using an
automated approach, incorporating Aib, Iva and hydroxyproline
residues, as well as other proteinogenic amino acids.
a For
a detailed description of the coupling conditions see the
Experimental Section and the Supporting Information.
TABLE 2. Methods Used for SPPS of Peptaibols (Exemplified by
Alamethicin F30)
alamethicin derivatives using TFFH activation with custom
programmed cycles has been reported once.14
Instead, we considered the acid fluoride method, utilizing
isolated and purified amino acid fluorides for each coupling.
However, to avoid the tedious synthesis of a range of these acid
fluorides we decided to attempt using them for the R,R-
dialkylated amino acids Aib and Iva only, while maintaining
standard HBTU conditions for the remaining couplings. The
acid fluorides of Aib or Iva are easily synthesized using the
method of Carpino and co-workers15 and are stable for months
in the freezer. They are also stable in a solution of DMF for
several hours.16
Modern peptide synthesizers, such as the one used in this
work coupled to a microwave, allow for easy customization of
the couplings for each individual amino acid. As depicted in
Table 1, we used three different coupling cycles for the
automated peptaibol synthesis, namely the factory-installed
standard cycle (black) and acid fluoride cycle (red), in which
an amino acid fluoride is added and the HBTU addition is
removed, and finally the factory installed double cycle (blue)
where the HBTU-mediated coupling process is repeated twice
prior to deprotection (Table 1). The use of HOBt as a additive
is not necessary when using microwave-assisted SPPS, as it has
no effect on racemization levels.
The cycles are applied to peptaibol synthesis in two ways,
method A and B, depending on the difficulty of the peptide
sequence (Table 2). Method A simply uses the standard cycle
for proteinogenic amino acids (including hydroxyproline) and
the acid fluoride cycle for the R,R-dialkylated amino acids (Aib
[U] and Iva [J]). For longer peptaibols, the yields of the coupling
of certain proteinogenic amino acids onto the R,R-dialkylated
amino acids are no longer high enough to achieve an acceptable
(14) Jung, G.; Redemann, T.; Kroll, K.; Meder, S.; Hirsch, A.; Boheim, G.
J. Pept. Sci. 2003, 9, 784–798.
(15) Carpino, L. A.; Sadat-Aalaee, D.; Chao, H. G.; DeSelms, R. H. J. Am.
Chem. Soc. 1990, 112, 9651–9652.
(16) Wenschuh, H.; Beyermann, M.; Rothemund, S.; Carpino, L. A.; Bienert,
M. Tetrahedron Lett. 1995, 36, 1247–1250.
(17) The lower yield obtained in the synthesis of tylopeptin A is due to the
fact that considerable amounts of the Iva deletion product were obtained. Two
peaks were observed for the purified tylopeptin A due to the presence of two
epimers of this peptide as racemic Iva was employed.
1330 J. Org. Chem. Vol. 74, No. 3, 2009

TABLE 3. Automated SPPS of Peptaibolsa
entry
peptaibol
sequenceb
method
synthesis timec (h)
crude purity (%)
isolated yielde (%)
1
2
3
4
5
6
7
8
peptaibolin
trichogin A IV
Ac-LULU-Phol
Oc-UGLUGGLUGI-Lol
A
A
Ad
A
B
A
B
B
B
A
B
4
6
16
6
7
6
8
9
9
11
12
94
96
89
35
46
29
24
27
28
71
20
24
21
24
20
11
22
12
trikoningin KB I
Oc-UGVUGGVUGI-Lol
cervinin 1
Ac-LUPULUPAUPV-Lol
bergofungin D
ampullosporin I
tylopeptin A
Ac-VUUVGLUUOQUOU-Phol
Ac-WAUULUQUUUQLUQ-Lol
Ac-WVUJAQAUSUALUQ-Lol
Ac-UPUAUAQUVUGLUPVUUEQ-Phol
9
10
11
alamethicin F30
<15
53
24
a All syntheses were performed on a CEM Liberty microwave-assisted peptide synthesizer on a 0.1 mmol scale using preloaded chlorotrityl resin. b U
) Aib, J ) (Rac)-Iva, O ) hydroxyproline, Phol ) L-phenylalaninol, Lol ) leucinol, Oc ) HO(CH2)7(CO)-. c Excluding acetylation and cleavage.
d The 5 min microwave coupling step was replaced by a 1 h room-temperature coupling step. e Purities of isolated peptides were all >90%.
It is evident from the results of Table 3 that the crude purity
of the peptides does not always correlate well with the isolated
yields (compare, for example, entries 1 and 2). We believe that
this discrepancy is mainly due to differences in the degree of
precipitation for both the peptaibols themselves and perhaps
more importantly the impurities of the crude peptide during
isolation.
Although microwave heating has been shown to improve
HBTU-mediated coupling of Aib residues in solution-phase
chemistry,18 we performed the synthesis of trichogin A IV
without the use of microwaves and obtained a similar result
(Table 3, entries 2 and 3). This implies that the method is easily
transferable to other peptide synthesizers without microwave
capabilities. However, it should be pointed out that the synthesis
time is much longer as coupling times required 1 h rather than
5 min under microwave heating.
Figure 1 illustrates the effect of employing method B when
method A proved insufficient in the synthesis of alamethicin
F30. When method A is employed the crude HPLC trace in
Figure 1a clearly shows a complex mixture with three main
products, none of which corresponds to alamethicin. On the
other hand, when method B was used the HPLC trace of the
crude mixture (Figure 1b) shows a main product corresponding
to the purified alamethicin illustrated in Figure 1c.
In summary, we have demonstrated a procedure for the fast
automated synthesis of peptaibols that utilizes a combination
of standard HBTU mediated solid phase peptide synthesis when
coupling proteinogenic amino acids, and acid fluoride peptide
synthesis when coupling R,R-dialkylated amino acids. The
method requires a minimum of synthetic effort prior to the SPPS
as only the syntheses of the Fmoc-amino acid fluorides of Aib
and/or Iva are required. The solid phase synthesis time was 12 h
or less, enabling overnight production of even the longest
peptaibols in good yields. Thus, this method is sufficiently
efficient and robust to allow for the expedient synthesis of eight
different full-length peptaibols as presented in this paper.
FIGURE 1. Analytical HPLC trace of (a) crude alamethicin using
method A, (b) crude alamethicin using method B, and (c) purified
alamethicin.
Experimental Section
in 10 mL of 1:1 DMF/CH2Cl2. The cycle used for each individual
amino acid of alamethicin is depicted in Tables 1 and 2. The
experimental conditions (see also the Supporting Infomation) of
the individual operations of the coupling cycles were as follows:
Fmoc Deprotection. 20% piperidine in DMF (7.0 mL) was
added, and microwave heating was employed for 30 s giving an
end temperature of 33 °C. After draining, 20% piperidine in DMF
(7 mL) was added again and the mixture microwave heated for 3
min reaching an end temperature of 75 °C. Nitrogen gas agitation
was used during heating.
Alamethicin F30 (Ac-UPUAUAQUVUGLUPVUUEQ-
Phol). The peptide was synthesized by applying method B on a
microwave-assisted peptide synthesizer using a 2-chlorotrityl
chloride resin preloaded with phenylalaninol (0.10 mmol). After
transfer to the reaction vessel, the resin was swelled using the
standard protocol of the SPPS machine involving 15 min immersion
(18) Santagada, V.; Fiorino, F.; Perissutti, E.; Severino, B.; De Filippis, V.;
Vivenzioa, B.; Caliendoa, G. Tetrahedron Lett. 2001, 42, 5171–5173.
J. Org. Chem. Vol. 74, No. 3, 2009 1331

Drain and Wash. Washings (4×) were performed using 7.0 mL
of DMF with nitrogen gas agitation.
Addition of Amino Acid. Amino acids were added as 0.20 M
tion in cold tert-butyl methyl ether. Repeated centrifugation,
decantation, and trituration (3 times) followed by lyophilization gave
the crude peptide (173 mg, 53% purity). The crude peptide was
dissolved in 1:1 MeCN/H2O and purified by semipreparative HPLC-
purification using a linear gradient from 40 to 60% of a solution of
0.1% TFA in MeCN in a solution of 0.1% TFA in H2O over 20
min with a flow rate of 5 mL/min, giving alamethicin F30 (47.8
mg, 24% yield, 95% purity). MALDI-TOF MS: C92H150N22O25
[M + Na+] calcd 1986.1, found 1986.6.
solutions in DMF (2.5 mL, 0.5 mmol).
Addition of Amino Acid Fluoride. Amino acid fluorides were
added as 0.20 M solutions in DMF (2.5 mL, 0.5 mmol).
Addition of HBTU. 2-(1H-Benzotriazole-1-yl)-1,1,3,3-tetram-
ethyluronium hexafluorophosphate (HBTU) was added as a 0.50
M solution in DMF (1.0 mL, 0.5 mmol).
Addition of DIPEA. Diisopropylamine was added as a 2.00 M
solution in N-methylpyrrolidone (0.5 mL, 1.0 mmol).
Microwave Heating. Microwave heating was employed for 5
min giving an end temperature of 80 °C. Nitrogen gas agitation
was used during heating.
After completion of the automatic synthesis following the final
Fmoc-deprotection cycle, the peptidyl-resin was washed twice with
10 mL of CH2Cl2 before being removed from the synthesizer.
N-Terminal acetylation was performed by treating the resin with
Ac2O (0.70 mmol) and DIPEA (1.40 mmol) in DMF (2 mL) for
45 min. Cleavage and deprotection of the peptide was performed
by treating the resin with 5 mL of TFA/CH2Cl2/H2O/TIPS 47:47:
4:2 (v/v) for 60 min. The cleavage mixture was concentrated to
approximately 1 mL under reduced pressure, followed by precipita-
Acknowledgment. We are deeply appreciative of generous
financial support from the Danish National Research Foundation,
the Danish Natural Science Research Council, iNANOSchool,
and Aarhus University.
Supporting Information Available: Experimental proce-
dures for all compounds, spectroscopic data for compounds 1-5,
and HPLC chromatograms for the synthesized peptides. This
material is available free of charge via the Internet at
http://pubs.acs.org.
JO802058X
1332 J. Org. Chem. Vol. 74, No. 3, 2009