Improving the biological activity of the antimicrobial peptide anoplin by membrane anchoring through a lipophilic amino acid derivative

Article abstract of DOI:10.1016/j.bmcl.2013.05.002

The lipophilic amino acid, (S)-2-aminoundecanoic acid, was synthesized and incorporated at a number of specific positions within the peptide sequence of anoplin. These lipophilic anoplin analogs showed to be more active against Escherichia coli and Staphy

Full text of DOI:10.1016/j.bmcl.2013.05.002

Bioorganic & Medicinal Chemistry Letters 23 (2013) 3749–3752
Contents lists available at SciVerse ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Improving the biological activity of the antimicrobial peptide anoplin
by membrane anchoring through a lipophilic amino acid derivative
Jack C. Slootweg a, Timo B. van Schaik a, H. (Linda) C. Quarles van Ufford a, Eefjan Breukink b,
Rob M. J. Liskamp a,c, Dirk T. S. Rijkers a,
⇑
a Medicinal Chemistry and Chemical Biology, Utrecht Institute for Pharmaceutical Sciences, Department of Pharmaceutical Sciences, Faculty of Science, Utrecht University,
P.O. Box 80082, 3508 TB Utrecht, The Netherlands
b Membrane Biochemistry and Biophysics, Bijvoet Center for Biomolecular Research, Department of Chemistry, Faculty of Science, Utrecht University, Utrecht, The Netherlands
c Chemical Biology and Medicinal Chemistry, School of Chemistry, University of Glasgow, Glasgow, G12 8QQ Scotland, United Kingdom
a r t i c l e i n f o
a b s t r a c t
Article history:
The lipophilic amino acid, (S)-2-aminoundecanoic acid, was synthesized and incorporated at a number of
Received 16 April 2013
Revised 29 April 2013
Accepted 1 May 2013
Available online 9 May 2013
specific positions within the peptide sequence of anoplin. These lipophilic anoplin analogs showed to be
more active against Escherichia coli and Staphylococcus aureus compared to native anoplin, while the EC50
-
value of hemolysis was at least one order of magnitude lower than the MIC values. This was in sharp con-
trast to the N-acylated anoplin derivative, where a gain in activity also led to a complete loss of selectiv-
ity. Thus, the incorporation of a lipophilic amino acid residue into anoplin enhanced the antimicrobial
activity, while selectivity towards microbial membranes was retained.
Keywords:
Anoplin
Antimicrobial peptides
Lipoamino acids
Lipopeptides
Ó 2013 Elsevier Ltd. All rights reserved.
Membrane interactions
Peptide pharmaceuticals
Peptide synthesis
There is a growing need for the development of novel antibio-
tics, or for improving upon existing ones, as established antibiotic
compounds continue to lose ground in the struggle against resis-
tant bacteria. Of particular concern are the Methicillin resistant
Staphylococcus aureus (MRSA) and vancomycin resistant Enterococ-
cus faecium (VRE) infections that continue to rise. Clearly
compounds active against these resistant classes are urgently
required.1
Host-defensive antimicrobial peptides (AMPs), have consider-
able promise as novel antibacterial agents.2 Their mode of action
is (target-unspecific) permeabilization of bacterial membranes
which induces little stable resistance, since it is very difficult for
a bacterium to change its membrane composition in order to coun-
teract the activity of AMPs. This might explain the great potential
of AMPs as lead compounds for the development of the next gen-
eration peptide-based antibiotic drug molecules, like the FDA-
approved lipopeptides Caspofungin 13 and Daptomycin 2,4 which
are active against fungal and bacterial infections, respectively
(Fig. 1).
Lipopeptides form a subclass of antimicrobial peptides in which
a lipophilic alkyl chain acts as a membrane anchor. The length of
the alkyl chain is highly important for the bioactivity, since analogs
with a truncated alkyl chain show a dramatic decrease in antimi-
crobial activity.5,6 Acylation of the N-terminal
a-amino functional-
ity is a well-known approach to increase membrane affinity.7–12
However, such N-acylation results in a non-charged amino termi-
nus, while a positively charged N-terminus, in combination with
a peptide sequence that is rich in arginine and lysine residues, is
often important for activity and selectivity.10 This is especially
the case when the peptide is meant to interact with negatively
charged bacterial membranes, and not with overall neutral mam-
malian membranes.
To increase the membrane affinity for membrane-acting anti-
microbial peptides, without sacrificing these important positively
charged backbone/side chain functionalities, a lipophilic amino
acid derivative ((S)-2-aminoundecanoic acid) has been designed
and synthesized that can be incorporated at any position of the
peptide sequence. Herein, we report that the antimicrobial deca-
peptide anoplin, H-Gly1-Leu2-Leu3-Lys4-Arg5-Ile6-Lys7-Thr8-Leu9-
Leu10-NH2, was modified at residues Leu2, Ile6, and Leu10, with
the lipophilic amino acid residue, and that these anoplin deriva-
tives were found to be ten times more active compared to native
⇑
Corresponding author. Tel.: +31 (0)6 2026 0572/(0)30 253 7307; fax: +31 (0)30
253 6655.
E-mail address: D.T.S.Rijkers@uu.nl (D.T.S. Rijkers).
0960-894X/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.bmcl.2013.05.002

3750
J. C. Slootweg et al. / Bioorg. Med. Chem. Lett. 23 (2013) 3749–3752
OH
the temperature to 0 °C), however, it turned out that for a complete
HO
O
OH
H
N
conversion of the starting materials, reaction at reflux temperature
was required. In the next step, hydrogenation of the double bond
and the simultaneous removal of the benzylic protecting groups
of compound 8 was initially attempted with 10%-Pd on activated
charcoal in a hydrogen atmosphere in tert-BuOH/H2O as the sol-
vent. Under these conditions the reaction did not go to completion,
even at 55 psi (3.8 bar) hydrogen pressure. Gratifyingly, hydroge-
nation in the presence of Pearlman’s catalyst (10%-Pd(OH)2 on acti-
vated carbon) in MeOH, a complete conversion was obtained. The
desired (S)-2-aminoundecanoic acid was difficult to isolate due
to the amphiphatic character in its zwitterionic form. Therefore,
N
N
H
O
O
O
HO
O
O
NH
Caspofungin 1
O
H2N
OH
N
H
N
H
NH OH
NH2
N
HO
HO
H2N
O
O
H
N
N
H
N
H
HN
NH
O
O
NH
O
O
protection of the
(after removal of the catalyst by filtration) in the presence of
Fmoc-ONSu and DIPEA as base. Finally, N- -(9-fluorenylmethyl-
a-amino group was performed in MeOH/CH2Cl2
CONH2
O
O
HO2C
HO2C
H
N
O
N
H
N
H
a
NH
O
O
O
HO
H
N
NH
oxycarbonyl)-(S)-2-aminoundecanoic acid (Fmoc-Laa-OH) 9 was
obtained in a modest yield of 38% (62% per step) after aqueous
work-up and purification by column chromatography.18
A previously reported structure–activity relationship study of
anoplin performed by Hansen and co-workers19 indicated that any
substitution of the lysine or arginine residues by alanine or modifi-
CO2H
O
N
H
N
O
H
O
Daptomycin 2
O
NH2
HO
O
Figure 1. Structural formula of Caspofungin 1 and Daptomycin 2.
cation of the a-amino terminus resulted in anoplin derivatives with
an increased hemolytic activity. Furthermore, antimicrobial activity
was completely lost when the leucine residues at position 2 and 10,
or the isoleucine at position 6, were replaced by alanine (see Fig. 2A
for amino acid numbering). These data indicate that the overall
charge of +4 is essential for selectivity, while hydrophobicity is re-
quired for activity. Therefore, we hypothesized that incorporation
of (S)-2-aminoundecanoic acid on position 2, 6, or 10 might increase
activity while selectivity remained unaffected. The rationale for this
design strategy is illustrated in Figure 2B, since the sites of modifica-
tion form a hydrophobic patch as indicated by the helical wheel rep-
resentation of the anoplin sequence. To test this hypothesis, anoplin
derivatives 11–13 were synthesized, and their activity/selectivity
profile was analyzed and compared to anoplin 10 and N-decanoylat-
ed anoplin 14 (Fig. 2).20
The peptides 10–14 were tested for their potency to inhibit bac-
terial growth, expressed as their minimal inhibition concentration
(MIC), of a Gram-negative (E. coli) as well as a Gram-positive
(S. aureus) bacterium. Selectivity against bacterial membranes
was measured by exposing red blood cells to the peptides in a
hemoglobin leakage assay (EC50). Anoplin 10 was active against
anoplin, while their selectivity towards microbial membranes re-
mained unaffected.
The lipophilic amino acid derivative (S)-2-aminoundecanoic
acid was synthetically accessible from L-glutamic acid 3 in only
four steps, as shown in Scheme 1.13,14 To this end, 3 was treated
with benzyl bromide under basic conditions15 to obtain the
N,N,O,O-perbenzylated Bzl-N(Bzl)Glu(OBzl)-OBzl derivative 4 in a
good yield of 73%. Then, diester 4 was subjected to a controlled
DIBAL-H mediated reduction, according to the method of Martín
and co-workers,16 to afford aldehyde 5, which was used without
further purification, in the subsequent Wittig reaction.17 In a sepa-
rate flask, hexyltriphenylphosphonium bromide 6 was treated with
hexamethyldisilazane potassium salt (KHMDS) in toluene to gen-
erate in situ ylide 7 and after 15 min at 0 °C, aldehyde 5 was added,
and the reaction mixture was heated at reflux temperature for 2 h
to give alkene 8 in an overall yield of 54% after purification by col-
umn chromatography. In the original literature procedure,16 the
Wittig reaction was performed at À78 °C (and gradually raising
i. DIBAL-H
HO
O
O
O
O
Et2O/hexane
-78oC, 15 min
ii. aq. work-up
i. NaOH (2 equiv) in H2O
ii. BnBr (4 equiv)/K2CO3 (4 equiv)
reflux, 3 h (73%)
OH
O
O
H2N
N
N
O
O
O
3
5
4
KHMDS
in toluene
0oC, 15 min
5 in toluene
0oC to reflux, 2h
Br
(54%; 4 into 8)
(in situ)
Ph
Ph
Ph
O
P
P
N
Ph
Ph
Ph
O
6
7
8
1. Pd(OH)2/C/H2
in MeOH, rt, 16 h
2. Fmoc-ONSu/DIPEA
in MeOH/CH2Cl2, rt, 16 h
(overall yield: 38%)
O
OH
O
N
H
O
9
Fmoc-Laa-OH
Scheme 1. Synthesis of N-
a-(9-fluorenylmethyloxycarbonyl)-(S)-2-aminoundecanoic acid (Fmoc-Laa-OH) 9 from L-glutamic acid 3.

J. C. Slootweg et al. / Bioorg. Med. Chem. Lett. 23 (2013) 3749–3752
3751
(A)
(B)
1
2
3
4
5
6
7
8
9
10
7
3
H2N
H2N
H2N
H2N
Lys
Lys
Anoplin (10)
Leu Leu
Arg Ile
Leu Leu
Gly
Thr
NH2
NH2
NH2
NH2
NH2
Leu Lys
10
Leu
4
Lys
Thr
Gly
6
8
Ile
Leu
Ano-Laa02 (11)
Ano-Laa06 (12)
Ano-Laa10 (13)
C9-Anoplin (14)
Lys
Lys
Laa Leu
Arg Ile
Leu Leu
Thr
Gly
Gly
Gly
Gly
2
1
Leu Arg
5
9
Leu Leu Lys Arg
Laa Lys
Lys
Thr Leu Leu
Thr Leu Laa
Thr Leu Leu
Lys
Lys
Leu Leu
Leu Leu
Arg Ile
H
N
Lys
Arg Ile
O
Figure 2. (A) Schematic representation of anoplin and the four newly synthesized lipophilic analogs. Herein, Laa ((S)-2-aminoundecanoic acid) has been incorporated at
positions 2 (11), 6 (12), and 10 (13), respectively. As a control, an N- -decanoyl-anoplin derivative (14) has been prepared. (B) Helical wheel representation of anoplin that
indicates the amphiphilic character of this antimicrobial peptide, since the membrane-acting amino acid residues form a hydrophobic patch.
a
both strains with a MIC value of 41 and 20.6
and highly selective against bacterial membranes since no hemoly-
sis was observed up to 500 g/mL (Table 1 and Fig. 3). The lipo-
lg/mL, respectively,
l
philic anoplin analogs 11–13 were 4 to 8 times more active
against E. coli and even 10 times more active against S. aureus
compared to anoplin (Table 1). As can be seen from Figure 3, the
hemolytic activity of analogs 11–13 was higher than anoplin
(39–108 lg/mL versus >500 lg/mL), however, the concentration
at which 50% lysis will occur (EC50) was even one order of magni-
tude higher than the respective MIC values, an indication that
these lipophilic anoplin derivatives were still selective towards
bacterial membranes (Table 1). The specificity of antimicrobial
agents toward bacterial membranes is expressed as the therapeutic
index (TI), and is calculated by the ratio of hemolytic activity (EC50
Figure 3. Hemolytic activity of anoplin compared with its hydrophobic analogs.
in lg/mL) and antimicrobial activity (MIC in l
g/mL).21 Thus a lar-
ger value in therapeutic index implies an increased antimicrobial
specificity. The TI values are shown in Table 1, and especially ano-
plin derivative 11 was found to be very promising regarding its
antimicrobial specificity compared to anoplin. The N-decanoylated
anoplin derivative 14 is an example of a non-selective antimicro-
bial peptide since its EC50 value was found to be in the same con-
with anoplin 10 (at 50
l
g/mL) and ꢀ41% of CF leakage was ob-
served (as shown in Fig. 4A). Membrane lysis induced by anoplin
derivatives 11–13 was found to be in the same order (47%, 45%,
and 48%, respectively), while CF leakage induced by the N-deca-
noylated anoplin derivative 14 was increased to approximately
62% (Fig. 4A). As a model for an abundant bacterial cell membrane,
LUVs consisting of an equimolar amount DOPC and the anionic li-
pid 1,2-dioleoyl-sn-glycero-3-phosphoglycerol (DOPG) were used.
The interaction of anoplin with DOPC/DOPG vesicles resulted in
approximately 35% CF leakage, while membrane lysis induced by
anoplin derivatives 11–13 was increased (48 to 52%), as shown
in Fig. 4B. Interaction of the N-decanoylated anoplin derivative
14 with DOPC/DOPG vesicles resulted in almost 100% CF leakage,
an indication that this peptide was highly unspecific, since it had
also the highest leakage-% of DOPC vesicles (Fig. 4). Based on these
centration range as its MIC value: 5 versus 2.5 lg/mL, while its TI
value was rather low (2.0) as shown in Table 1 and Figure 3.
In a second biochemical assay, peptides 10–14 were tested for
their interaction with model membrane systems.22 For this pur-
pose, large unilamellar vesicles (LUVs) were loaded with carboxy-
fluorescein (CF) as the fluorophore, and the ability of the peptides
to induce membrane permeability was measured by monitoring
the release of CF by fluorescence spectroscopy. LUVs composed
of 1,2-dioleoyl-sn-glycero-3-phosphocholine (DOPC), a zwitter-
ionic lipid mimicking mammalian cell membranes, was treated
Table 1
Physico-chemical properties and biological activities of anoplin and its derivatives
MICb
(l
g/mL)
MICb
(l
g/mL)
EC50
(
lg/mL)
c
Peptide
Mw (Da)
Rta (min)
E. coli
TId
S. aureus
TId
Anoplin (10)
1152.8
1222.9
1222.9
1222.9
1306.9
18.3
20.3
20.3
20.2
22.2
41
>12.2
20.6
2.5
2.5
2.5
2.5
>24.3
43.2
15.6
30.8
2.0
>500
ꢀ108
ꢀ39
ꢀ77
ꢀ5
Ano-Laa02 (11)
Ano-Laa06 (12)
Ano-Laa10 (13)
C9-Anoplin (14)
5.2–10.3
10.3–20.6
10.3–20.6
2.5
10.5–20.8
1.9–3.8
3.7–7.5
2.0
a
Retention times were determined on a C8 column using a gradient of buffer B (0.1% TFA in CH3CN/H2O 95:5 v/v) in buffer A (0.1% TFA in CH3CN/H2O 5:95 v/v) for 20 min
with a flow rate of 1.0 mL/min.23
b
Antimicrobial activity is expressed as the minimal inhibitory concentration (MIC).
c
Hemolytic activity is expressed as the concentration of the peptide that induces 50% hemolysis (EC50).
d
Therapeutic index = EC50 (in lg/mL)/MIC (in l
g/mL); larger values indicate an improved antimicrobial specificity.21

3752
J. C. Slootweg et al. / Bioorg. Med. Chem. Lett. 23 (2013) 3749–3752
NMR spectra (8 and 9), and copies of the analytical HPLC and ESI
MS data (10–14)) associated with this article can be found, in the
(A)
online version, at http://dx.doi.org/10.1016/j.bmcl.2013.05.002.
References and notes
1. Editorial ‘The antibiotic alarm’, in: Nature 2013, 495, 141 (issue March 14,
2013).
2. For selected reviews on AMPs, see (a) Zasloff, M. Nature 2002, 415, 389; (b)
Steinstraesser, L.; Kraneburg, U.; Jacobsen, F.; Al-Benna, S. Immunobiology 2011,
216, 322; (c) Giuliani, A.; Rinaldi, A. C. Cell. Mol. Life Sci. 2011, 68, 2255; (d)
Zucca, M.; Scutera, S.; Savoia, D. Mini Rev. Med. Chem. 2011, 11, 888; (e)
Brogden, N. K.; Brogden, K. A. Int. J. Antimicrob. Agents 2011, 38, 217; (f) Engler,
A. C.; Wiradharma, N.; Ong, Z. Y.; Coady, D. J.; Hedrick, J. L.; Yang, Y.-Y. Nano
Today 2012, 7, 201; (g) Pasupuleti, M.; Schmidtchen, A.; Malmsten, M. Crit. Rev.
Biotechnol. 2012, 32, 143.
3. Deresinski, S. C.; Stevens, D. A. Clin. Infect. Dis. 2003, 36, 1445.
4. Raja, A.; LaBonte, J.; Lebbos, J.; Kirkpatrick, P. Nat. Rev. Drug Disc. 2003, 2, 943.
5. Counter, F. T.; Allen, N. E.; Fukuda, D. S.; Hobbs, J. N.; Ott, J.; Ensminger, P. W.;
Mynderse, J. S.; Preston, D. A.; Wu, C. Y. E.; Lilly, T.; Lilly, E. J. Antibiot. 1990, 43,
616.
(B)
6. Taft, C. S.; Selitrennikoff, C. P. J. Antibiot. 1990, 43, 433.
7. Chu-kung, A. F.; Bozzelli, K. N.; Lockwood, N. A.; Haseman, J. R.; Mayo, K. H.;
Tirrell, M. V. Bioconjugate Chem. 2004, 15, 530.
8. Malina, A.; Shai, Y. Biochem. J. 2005, 390, 695.
9. Majerle, A.; Kidric, J.; Jerala, R. J. Antimicrob. Chemother. 2003, 51, 1159.
10. Zweytick, D.; Pabst, G.; Abuja, P. M.; Jilek, A.; Blondelle, S. E.; Andrä, J.; Jerala,
R.; Monreal, D.; Martinez de Tejada, G.; Lohner, K. Biochim. Biophys. Acta 2006,
1758, 1426.
11. Mak, P.; Pohl, J.; Dubin, A.; Reed, M. S.; Bowers, S. E.; Fallon, M. T.; Shafer, W. M.
Int. J. Antimicrob. Agents 2003, 21, 13.
12. Avrahami, D.; Shai, Y. J. Biol. Chem. 2004, 279, 12277.
13. For the synthesis of lipophilic amino acids, among others (S)-2-
aminoundecanoic acid, several procedures have been published previously:
(a) Padrón, J. M.; Kokotos, G.; Martín, T.; Markidis, T.; Gibbons, W. A.; Martín, V.
S. Tetrahedron: Asymmetry 1998, 9, 3381; (b) Papini, A. M.; Nardi, E.; Nuti, F.;
Uziel, J.; Ginanneschi, M.; Chelli, M.; Brandi, A. Eur. J. Org. Chem. 2002, 2736; (c)
Blanchfield, J. T.; Dutton, J. L.; Hogg, R. C.; Gallagher, O. P.; Craik, D. J.; Jones, A.;
Adams, D. J.; Lewis, R. J.; Alewood, P. F.; Toth, I. J. Med. Chem. 2003, 46, 1266; (d)
Woon, E. C. Y.; Arcieri, M.; Wilderspin, A. F.; Malkinson, J. P.; Searcey, M. J. Org.
Chem. 2007, 72, 5146.
Figure 4. Lysis of large unilamellar vesicles expressed in a percentage of leakage
induced by anoplin and its hydrophobic analogs as measured on neutral DOPC
vesicles (A) and negatively charged DOPC/DOPG vesicles (B), at a peptide concen-
tration of 50 lg/mL.
14. For a review on the synthesis and application of lipoamino acid derivatives,
see: Ziora, Z. M.; Blaskovich, M. A.; Toth, I.; Cooper, M. A. Curr. Top. Med. Chem.
2012, 12, 1562.
15. Gmeiner, P.; Kärtner, A.; Junge, D. Tetrahedron Lett. 1993, 34, 4325.
16. Kokotos, G.; Padrón, J. M.; Martín, T.; Gibbons, W. A.; Martín, V. S. J. Org. Chem.
1998, 63, 3741.
17. To check the presence of the aldehyde functionality, the crude reaction product
was analyzed by 1H NMR (CDCl3, 300 MHz) and the aldehyde peak could be
identified at d 9.58 ppm.
leakage experiments it became clear that the lipophilic anoplin-de-
rived peptides 11–13 displayed an increased permeabilization
activity towards anionic lipid containing membranes most likely
due to an increased affinity for these membranes, while N-acylation
increased membrane affinity but led to a loss of selectivity.
In conclusion, an efficient five-step synthesis of N-
nylmethyloxycarbonyl)-(S)-2-aminoundecanoic acid starting from
-glutamic acid is reported, including its application as building
a-(9-fluore-
18. N-
9: Rf 0.16 (hexane/EtOAc 1:1 v/v), [
CDCl3) d = 0.87 (t (J = 6.7 Hz), 3H, CH3), 1.34 (broad s, 14H, CH2), 1.69 (m, 1H,
bCH2), 1.89 (m, 1H, bCH2), 4.22 (t (J = 6.9 Hz), 1H, CH Fmoc), 4.44 (m, 3H, CH
a
-(9-fluorenylmethyloxycarbonyl)-(S)-2-aminoundecanoic acid (Fmoc-Laa-OH)
L
a
]
D
+4.8 (c 1 in CHCl3); 1H NMR (300 MHz,
block in solid phase peptide synthesis. The incorporation of this
lipophilic amino acid in an antimicrobial peptide increases the
affinity of this peptide toward bacterial cell membranes, likely by
acting as a membrane anchor. Using this approach, antimicrobial
specificity was preserved and in one particular case (anoplin deriv-
ative 11) highly promising compared to native anoplin. These data
may imply that amino acid substitution by a lipophilic residue
could be a general approach to improve drug-like properties of
membrane-interacting antimicrobial peptides.
a
(1H)/CH2 Fmoc (2H)), 5.23 (d (J = 8.2 Hz), 1H, NH), 7.28–7.77 (m, 8H, arom CH
Fmoc); 13C NMR (75 MHz, CDCl3) d = 14.1, 22.7, 25.2, 29.1, 29.3, 29.4, 29.5, 31.9,
32.3, 47.2, 53.8, 67.1, 120.0, 125.1, 127.1, 127.7, 141.3, 143.7, 143.8, 156.1,
177.4.
19. Ifrah, D.; Doisy, X.; Ryge, T. S.; Hansen, P. R. J. Pept. Sci. 2005, 11, 113.
20. The peptides were synthesized on a Tentagel S-RAM resin using the Fmoc/tBu
SPPS protocol with BOP/DIPEA as coupling reagents. Crude peptides 10–14
were purified by preparative HPLC and characterized by analytical HPLC and
mass spectrometry. See the Supplementary data section for experimental
details.
21. Chen, Y.; Mant, C. T.; Farmer, S. W.; Hancock, R. E. W.; Vasil, M. L.; Hodges, R. S.
J. Biol. Chem. 2005, 280, 12316.
Acknowledgments
22. Breukink, E.; Van Kraaij, C.; Demel, R. A.; Siezen, R. J.; Kuipers, O. P.; De Kruijff,
B. Biochemistry 1997, 36, 6968.
23. The analytical HPLC run time was as follows: 0–5 min: 100% buffer A; 5–
25 min: from 100% buffer A to 100% buffer B; 25–30 min: 100% buffer B; 30–
35 min: from 100% buffer B to 100% buffer A; 35–40 min: 100% buffer A.
Dr. Johan Kemmink is acknowledged for his help to use
MathematicaÒ for calculating the EC50 values. This research
was financed by the Dutch Organization for Scientific Research—
Chemical Sciences (NWO-CW) as an ECHO-Grant 700.57.014.
Supplementary data
Supplementary data (details of the synthetic procedures and
compound analysis (4–9), biological assays, copies of the 1H, 13C