Penetratin and derivatives acting as antifungal agents

Article abstract of DOI:10.1016/j.ejmech.2008.02.019

The synthesis, in vitro evaluation, and conformational study of RQIKIWFQNRRMKWKK-NH₂ (penetratin) and related derivatives acting as antifungal agents are reported. Penetratin and some of its derivatives displayed antifungal activity against the

Full text of DOI:10.1016/j.ejmech.2008.02.019

Available online at www.sciencedirect.com
European Journal of Medicinal Chemistry 44 (2009) 212e228
http://www.elsevier.com/locate/ejmech
Original article
Penetratin and derivatives acting as antifungal agents
a,b
d
e
Marcelo F. Masman a,b,c, Ana M. Rodrıguez , Marcela Raimondi , Susana A. Zacchino ,
´
Paul G.M. Luiten c, Csaba Somlai f, Tamas Kortvelyesi g,
Botond Penke f, Ricardo D. Enriz a,b,
*
a Facultad de Qu´ımica, Bioqu´ımica y Farmacia, Universidad Nacional de San Luis, Chacabuco 915, 5700 San Luis, Argentina
b IMBIO-SL, CONICET, UNSL, Chacabuco 915, 5700 San Luis, Argentina
c Department of Molecular Neurobiology, Centre for Behaviour and Neurosciences, University of Groningen, Kerklaan 30, 9751 NN Haren, The Netherlands
d Microbiolog´ıa, Facultad de Medicina, Universidad Nacional de Rosario, Santa Fe 3100, 2000 Rosario, Argentina
e Farmacognosia, Facultad de Ciencias Bioqu´ımicas y Farmace´uticas, Universidad Nacional de Rosario, Suipacha 531, 2000 Rosario, Argentina
f Department of Medical Chemistry, University of Szeged, Do´m ter 8, H-6720 Szeged, Hungary
g Department of Physical Chemistry, University of Szeged, Rerrich Sq. 1, H-6720 Szeged, Hungary
Received 26 December 2007; received in revised form 8 February 2008; accepted 11 February 2008
Available online 29 February 2008
Abstract
The synthesis, in vitro evaluation, and conformational study of RQIKIWFQNRRMKWKKeNH2 (penetratin) and related derivatives acting
as antifungal agents are reported. Penetratin and some of its derivatives displayed antifungal activity against the human opportunistic pathogenic
standardized ATCC strains Candida albicans and Cryptococcus neoformans as well as clinical isolates of C. neoformans. Among the compounds
tested, penetratin along with the nonapeptide RKWRRKWKKeNH2 and the tetrapeptide RQKKeNH2 exhibited significant antifungal activities
against the Cryptococcus species. A comprehensive conformational analysis on the peptides reported here using three different approaches,
molecular mechanics, simulated annealing and molecular dynamics simulations, was carried out. The experimental and theoretical results allow
us to identify a topographical template which may provide a guide for the design of new compounds with antifungal characteristics against
C. neoformans.
Ó 2008 Elsevier Masson SAS. All rights reserved.
Keywords: Penetratin; Cationic peptides; Cryptococcosis; Candida infections; Molecular dynamics; Conformational analysis
1. Introduction
patients [1,2]. Invasive fungal infections as well as dermato-
mycoses produced by fungal organisms with even low viru-
lence can be life threatening [3] for individuals with
increased vulnerability such as neonates, cancer patients re-
ceiving chemotherapy, organ transplant patients, and burns pa-
tients, apart from those with acquired immunodeficiency
syndrome (AIDS). Other risk factors include corticosteroid
and antibiotic treatments, diabetes, lesions of epidermis and
dermis, malnutrition, neutropenia and surgery [2]. Many fun-
gal infections are caused by opportunistic pathogens that
may be endogenous or acquired from the environment (Can-
dida, Cryptococcus, Aspergillus infections). Patients with sig-
nificant immunosuppression frequently develop Candida
esophagitis. Cryptococcosis, caused by the encapsulated yeast
Cryptococcus neoformans, has been the cause of fungal
Fungal infections are a persistent major health problem,
which especially affect and threaten immunocompromised
Abbreviations: AIDS, acquired immunodeficiency syndrome; CPP, cell-
penetrating peptide; SA, simulated annealing; MD, molecular dynamics;
MIC, minimum inhibitory concentration; MFC, minimum fungicide concen-
tration; EDMC, electrostatically driven Monte Carlo; RMSD, route mean
square deviation; Rg, radius of gyration; RMSF, root mean square fluctuation;
SASA, solvent accessible surface area; TFEd2, trifluoroethanol-d2; MEPs, mo-
lecular electrostatic potentials.
´
´
* Corresponding author. Facultad de Quımica, Bioquımica y Farmacia, Uni-
versidad Nacional de San Luis, Chacabuco 915, 5700 San Luis, Argentina.
Tel.: þ54 2652 423789; fax: þ54 2652 431301.
E-mail address: denriz@unsl.edu.ar (R.D. Enriz).
0223-5234/$ - see front matter Ó 2008 Elsevier Masson SAS. All rights reserved.
doi:10.1016/j.ejmech.2008.02.019

M.F. Masman et al. / European Journal of Medicinal Chemistry 44 (2009) 212e228
213
mortality among HIV-infected patients. This organism has
a predilection for the central nervous system and leads to se-
vere, life-threatening meningitis. In addition, an increasing
number of normal individuals, including children in third-
world nations [4] that suffer deficient sanitation and education,
are prone to fungal infections, especially those involving the
skin and mucosal surfaces.
amino acid homeodomain of the Antennapedia protein of
Drosophila was able to translocate over cell membranes. In
order to understand the driving force for the internalization,
the homeodomain was modified by site-directed mutagenesis
leading to the discovery that its third helix was necessary
and also sufficient for membrane translocation, which re-
sulted in the development of a 16 amino acid-long CPP
called penetratin (1) [19]. Thus, peptide 1, a synthetic 16
amino acid peptide from the third helix of Antennapedia ho-
meodomain [19,20], is a cationic amphipathic peptide which
might penetrate cell membranes via a proposed ‘‘inverted
micelle’’ pathway. However, the mechanism of membrane
translocation is not well known. The question is whether
the peptide is internalized via endocytosis which is energy-
dependent or via direct transport, while the latter mechanism
is scarcely known at present, it is believed that the process is
non-receptor mediated [20,21]. In addition, we previously
provide evidence on the energy-dependent and lipid raft-me-
diated endocytic uptake of penetratin [22]. Peptide 1 has
been proposed as a universal intracellular delivery vehicle
[23]. Since 1 possesses 16 amino acids and a charge of
þ8, it might be included in the general classification of ‘‘cat-
ionic antimicrobial peptide’’. To the best of our knowledge
this is the first study reporting the antifungal activity of 1
and structurally related derivatives.
The aim of the present investigation is exploring the
antifungal potential of 1 and its derivatives against Candida al-
bicans and C. neoformans. To better characterize the struc-
tureeantifungal activity relationship of peptide 1 and related
compounds under study the present analysis explored influ-
ences of amino acid substitutions and deletions on its antifun-
gal activity. In addition, an extensive conformational analysis
of 1 and its derivatives was carried out using three different ap-
proaches: molecular mechanics, simulated annealing (SA) and
molecular dynamics (MD) simulations. The ability of each
method to obtain the different conformations is tested and
compared. Conformational and electronic studies were carried
out in order to identify a topographical and/or a sub-structural
template, which may be the starting structure for the design of
new antifungal compounds.
Although it appears that many drugs are available for the
treatment of systemic and superficial mycoses, there are in
fact only a limited number of effective antifungal drugs [1].
Many of the antifungal compounds currently available have
undesirable effects or are very toxic (amphotericin B); are fun-
gistatic and not fungicidal (azoles), or lead to the development
of resistance, as with 5-fluorocytosine [5]. Amphotericin B,
developed in the 1950s, still remains as a widely used antifun-
gal drug, most recently gaining renewed applications through
lipid based formulations. According to Polak [6] ideal drugs to
cure fungal infections have not been discovered yet. In the
meantime, resistance to currently available antifungal agents
continues to grow [7]. Although combination therapy has
emerged as a good alternative to bypass these disadvantages
[6,8], there is an urgent need for a next generation of safer
and more potent antifungal agents [1,8]. These explorations
have resulted in the identification of novel molecules, which
could prove promising for further future development. Among
them, some natural peptides were recently described as
antifungal compounds, inhibiting a broad spectrum of fungi
[9e11]. It has also been reported that a group of cationic an-
timicrobial peptides are major players in the innate immune
response [12,13]. These peptides are very ancient elements
of the immune response of all species of animal and plant
life, and the induction pathways for these compounds in verte-
brates, insects and plants [12e14] are highly conserved.
Furthermore, it is becoming increasingly clear that cationic
antimicrobial peptides play many potential roles in inflamma-
tory responses, which represent an orchestration of the mech-
anisms of innate immunity.
Small cationic peptides [15,16] are abundant in nature and
have been described as ‘‘nature’s antibiotics’’ or ‘‘cationic an-
timicrobial peptides’’. These peptides are 12e50 amino acids
long with a net positive charge of þ2 or þ9, which is due to an
excess of basic arginine and lysine residues, and approxi-
mately 50% hydrophobic amino acids [15]. These molecules
are also folded in three dimensions so that they have both a hy-
drophobic face comprising non-polar amino acid side chains,
and a hydrophilic face of polar and positively charged resi-
dues: these molecules are amphipathic. Despite these two sim-
ilarities these compounds vary considerably in length, amino
acid sequence and secondary structure. The different spatial
orderings include small b-sheets stabilized by disulphide
bridges, amphipathic a-helices and, less commonly, extended
and loop structures.
2. Results and discussion
2.1. Antifungal activity
To evaluate the antifungal potential, concentrations of pep-
tides up to 200 mM were incorporated in the growth media
according to previously reported procedures [11,24,25]. Com-
pounds producing no inhibition of fungal growth at 200 mM
were considered inactive. Table 1 gives the antifungal activity
obtained for peptide 1 against C. albicans and C. neoformans.
Peptide 1 displayed a significant antifungal activity against
both fungi being C. neoformans the more sensitive species.
It is interesting to note that 1 displayed a significant degree
of inhibition against C. neoformans even at low concentrations
(90% of inhibition was observed at 12.5 mM and 100% at
25 mM). The inhibitory effect observed against C. albicans
Cell-penetrating peptides (CPPs) are defined as peptides
with a maximum of 30 amino acids, which are able to enter
cells in a seemingly energy-independent manner, thus being
able to translocate across membranes in a non-endocytotic
fashion [17]. In 1991, Joliot et al. [18] reported that the 60

214
M.F. Masman et al. / European Journal of Medicinal Chemistry 44 (2009) 212e228
was slightly lower than that obtained for C. neoformans at
12.5 mM but similar at 25e200 mM.
In order to gain insight into the spectrum of activity, pep-
tide 1 was tested against a panel of clinical isolates of
C. neoformans. The minimum inhibitory concentration (MIC)
values of 1 were determined against this new panel by using
three endpoints: MIC100, MIC80 and MIC50 (the minimum con-
centration of compounds that inhibit 100, 80 and 50% of growth,
respectively). The application of a less stringent end point such
as MIC80 and MIC50 has been shown to consistently represent
the invitro activity of compounds and many times provide a bet-
ter correlation with other measurements of antifungal activity
such as the minimum fungicide concentration (MFC) [26]. In
addition to MIC determinations, the evaluation of MFC of 1
against this panel was accomplished by sub-culturing a sample
of media from MIC tubes showing no growth, onto drug-free
agar plates. So, peptide 1 was tested against 10 clinical isolates
´
of C. neoformans, all provided by the Malbran Institute (Buenos
Aires). These results are shown in Table 2 and the activity was
similar to, or lower than, those obtained against the standard
strain (ATCC 32264).
Peptide 2 possesses the same amino acids of peptide 1 but
located in a completely different sequence. In fact, the se-
quence of this peptide was randomly generated. This peptide
showed no antifungal activity against both fungi tested at
12.5 mM but inhibited 96 and 60% of the growth of C. neofor-
mans and of C. albicans at 200 mM, respectively. It is clear
that the antifungal activity of 1 and 2 is markedly different,
whereas 1 showed a significant antifungal activity against
both C. albicans and C. neoformans; peptide 2 was practically
inactive. On the basis of these results it can be concluded that
the sequence as well as the different spatial orderings of the
cationic, polar and hydrophobic residues are important deter-
minants for the antifungal activity. In contrast, the positive
charge (þ8) of 1 appears to be a necessary requirement but
not by itself sufficient to produce the antifungal response.
To study the structureeantifungal activity relationship on
these peptides, the effects of structural changes in the
Table 2
Minimum inhibitory concentrations (MIC100, MIC80 and MIC50) and mini-
mum fungicidal concentration (MFC) of penetratin (1) against clinical isolates
of C. neoformans
Voucher
specimen
MIC100 MIC80 MIC50 MFC Amph. B Itz. MIC100
MIC100
IM 983040
IM 972724
IM 042074
IM 983036
IM 00319
31.2
62.5
62.5
62.5
62.5
62.5
62.5
31.2
62.5
62.5
62.5
62.5
31.2
31.2
31.2
31.2
31.2
31.2
31.2
31.2
31.2
31.2
31.2
16
31.2 0.13
62.5 0.06
62.5 0.25
62.5 0.13
62.5 0.25
62.5 0.25
62.5 0.13
<0.015
0.25
<0.015
<0.015
<0.015
<0.015
0.25
IM 972751
IM 031631
IM 031706 125
IM 961951 125
IM 052470 125
16
16
125
0.25
0.06
0.50
0.50
<0.015
<0.015
125
125
31.2
MIC100, MIC80 and MIC50: concentration of a compound that caused 100, 80
or 50% reduction of the growth control, respectively. Within voucher speci-
men: IM ¼ Malbran Institute (Buenos Aires, Argentina); Amph.
B ¼ amphotericin B; Itz. ¼ itraconazole.

M.F. Masman et al. / European Journal of Medicinal Chemistry 44 (2009) 212e228
215
sequence of peptide 1 were considered. Our principal goal was
to synthesize shorter derivatives of 1 trying to maintain the an-
tifungal activity as much as possible. In the first step we syn-
thesized compound 3, a nonapeptide possessing the last four
amino acids of 1. In this peptide we maintain the same num-
bers of cationic amino acids (Arg and Lys) deleting Gln-2,
Ile-3, Ile-5, Phe-7, Gln-8, Asn-9 and Asn-12. Peptide 3 dis-
played a lower antifungal activity with respect to 1. The anti-
fungal activity against C. neoformans and C. albicans was
moderately effective but still significant. We decided to per-
form changes on peptide 3 and then we synthesized peptides
4e6. In peptide 4 we replaced the two Trp residues of peptide
3 (Trp-3 and Trp-7) by Phe. This structural change yielded a re-
duction of antifungal activity (compare the % of inhibition of
3 and 4 in Table 1), which is not an unexpected result; a role
for Trp as translocation determinant of peptides has been pro-
posed [27] and mutation of both tryptophans in peptide 1 was
found to abolish internalization [19]. In addition, it has previ-
ously been reported that tryptophans are poorly replaceable by
phenylalanine in 1 and derivatives when tested for their pene-
trating properties [28]. Our results lend support to previously
reported findings, but in addition demonstrate the antifungal
activity of these peptides. Octapeptide 5 was obtained by de-
leting Trp-3 from peptide 3; in turn heptapeptide 6 was ob-
tained by deleting Trp-3 and Trp-7 from peptide 3. Whereas
octapeptide 5 displayed only a marginal antifungal activity,
peptide 6 was practically ineffective in comparison to their
congeners.
peptide 1 is of paramount importance. Linear peptides are
highly flexible and therefore to determine the biologically rel-
evant conformations is a matter of high complexity, which
requires an exhaustive conformational analysis of these struc-
tures. Consequently, we carried out calculations using three
different approaches: electrostatically driven Monte Carlo
(EDMC) calculations implemented in the ECCEPAK [29]
package, SA calculations using the Tinker Molecular Model-
ling package [30] and MD simulations from the GROMACS
program [31,32].
2.2.1. EDMC results
EDMC results are summarized in Table 3 and Fig. 1 and
more
details are given in Tables SIeSVIII in Supplementary
material. Calculations yielded a large set of conformational
families for each peptide studied. The total number of confor-
mations generated for each peptide varied between 47 391 and
129 922, and the number of those accepted was 5000 for all the
cases. In the clustering procedure, an RMSD (root meꢁan1
˚
square deviation) of 0.75 A and a cutoff of 30 kcal mol
were used. The number of families after clustering varied be-
tween 137 and 1001. The total number of families accepted
with a relative population higher than 0.20% varied between
11 and 86. Their populations sum up to ca. 80% of all confor-
mations in each case (see Table 3).
All low-energy conformers of the peptides studied here
were then compared to each other. The comparison involved
the spatial arrangements, relative energy and populations.
A total of 639 different families were found for peptide 1.
However, 82.92% of total population of this peptide corre-
sponded to only 11 families (Table 3). It is interesting to
note that the energetically preferred family comprises
70.48% of the entire population. Thus, this family which
adopts an a-helix structure is the most representative form
of this molecule. This conformation is characterized by stabi-
lizing hydrogen bonds between the carbonylic oxygen (residue
i) and the NH group (residue i þ 4). The first and the last res-
idues do not present a stable structure. A spatial view of this
conformation is shown in Fig. 1a. The second most populated
family (7.46%) corresponds to a structure possessing the first
In order to further understand the above experimental re-
sults, we performed a conformational study of the peptides re-
ported here using different approaches.
2.2. Conformational study of peptide 1 and derivatives
A large number of studies have been performed in order to
shed light on the structural aspects and mechanism of action
for translocation of 1. However, compared to these mechanis-
tic properties, the conformational intricacies of this compound
have received relatively little attention. It is, however, obvious
that a better understanding of the conformational behavior of
Table 3
Selected conformational search and clustering results for peptides 1e8 optimized at the EDMC/SRFOPT/ECCEP/3 level of theory
d
Peptide
Generateda
Acceptedb
# Fc
# F0.20%
% Pe
Electrostatical
Random
Thermal
Total
Electrostatical
Random
Thermal
Total
1
2
3
4
5
6
7
8
a
8973
9200
8050
7490
7905
7191
3007
3939
119 535
120 229
107 710
102 380
106 150
98 483
575
493
304
245
294
213
32
129 083
129 922
116 064
110 115
114 349
105 887
47 391
1431
1349
1372
1121
1176
1192
508
3245
3373
3412
3697
3606
3635
4466
4379
324
278
216
182
218
173
26
5000
5000
5000
5000
5000
5000
5000
5000
639
703
270
288
242
137
505
481
11
19
6
82.92
81.92
88.94
88.72
89.76
91.02
82.82
83.34
11
11
6
44 352
54 133
74
45
54
58 126
579
42
Number of conformations generated electrostatically, randomly and thermally during the conformational search.
Number of conformations accepted from those generated electrostatically, randomly and thermally during the conformational search.
# F represents the total number of conformational families as result of the clustering run.
b
c
d
e
# F0.20% represents the number of conformational families with populations above 0.20%.
% P represents the sum of the percent relative population of # F0.20%
.

216
M.F. Masman et al. / European Journal of Medicinal Chemistry 44 (2009) 212e228
To better characterize the peptide spatial orientations, we
plotted Edmundson wheel representations of peptides 1e6
(Fig. 2). The representation obtained for peptide 1 displays
two clearly differentiated faces: the ‘‘charged face’’ (denoted
in dashed blue line in Fig. 2) and a more extended ‘‘non-
charged face’’ (denoted in full green line). The first face iden-
tifies residues R11, K4, K5 and R1 as those accounting for the
mutual coulombic binding. The first three residues are located
on the same side of the helical peptide and hence we desig-
nated it as the ‘‘charged face’’. These positively charged resi-
dues are able to produce salt bridges with the hydrophilic part
of the lipids. The non-charged face is more extensive and is
formed by six hydrophobic (M12, I5, W6, I3, W14 and F7)
and two polar residues (N9 and Q2). However, it should be
noted that 1 displays a homogeneously distributed remainder
of the positively charged residues. Thus, residues K16, K13
and K10 are strategically intercalated along the ‘‘non-charged
face’’. This is a striking difference with respect to peptide 2,
which displays two ‘‘charged faces’’ where all the cationic res-
idues are concentrated. In peptide 2 there are two non-charged
faces; however, it should be noted that even adding the two
non-charged faces of peptide 2, this portion is markedly lower
with respect to the only non-charged face obtained for 1. The
Edmundson wheel representations obtained for peptides 3 and
4 are very similar displaying a very extensive charged face and
a markedly reduced non-charged face. Peptide 5 in turn gives
only a minimal non-charged zone corresponding to the W6
residue and the rest is ‘‘charged face’’. Obviously, peptide 6
displays a completely charged face because only cationic res-
idues form it. Previously Lensink et al. [33] reported that a ho-
mogeneous distribution of positively charged residues along
the axis of the helical peptide, and especially K4, R5, and
K11 contribute to the association of peptide 1 with lipids.
Our EDMC results are in a complete agreement with those
MD simulations. In addition a very good correlation between
the antifungal activities and the potential penetrating proper-
ties of these peptides are particularly striking.
Fig. 1. Spatial view of the preferred forms obtained for peptide 1. (a) The
global minimum (a-helix structure) and (b) the second more populated
conformation.
four residues without a stable structure; residues 5e7 in a turn
structure and residues 10e15 with a typical a-helix structure.
The residues 8 and 9 present a bend structure connecting the
turn moiety with the a-helix portion. The last residue does
not display a stable form (Fig. 1b). However, this family has
an energy gap of 22.61 kcal molꢁ1 with respect to the global
minimum.
For peptide 2 a total of 703 different families were ob-
tained, from which 19 families comprise 81.92% of the total
population. The most populated family (61.06%) presents an
a-helix conformation from residues 3 to 15. This conformation
has an energy gap of 1.27 kcal molꢁ1 with respect to the global
minimum, which has 6.04% of population. The lowest energy
conformation possesses the following structure: from residues
2 to 5 in a turn structure; residue 6 in a bend form and from
residues 7 to 15 in a typical a-helix structure. Residues 1
and 16 do not show any stable structure. In general the confor-
mational behavior of 1 and 2 is comparable. However, in pep-
tide 1 the most populated family (a typical a-helix structure) is
also the energetically preferred one. In contrast, for peptide 2,
the fully a-helix structure is not the most preferred form.
Compounds 3e6 display a closely related conformational
behavior preferring a helical structure for the most populated
families. Peptides 3e6 are somewhat more rigid with respect
to 1. This fact might be appreciated comparing the total num-
ber of conformational families obtained for each compound
(Table 3).
2.2.2. SA results
The initial structure of peptides 1e6 was extended. The
secondary structures of the lowest energy conformers calcu-
lated by DSSP program [34] are summarized in Table 4.
The best structure of peptide 1 contains bends (I3eW6), 310-
helix (F7eN9) and b-turn (R10eR11, K13eK15) and bends
(F7eN9, R11eK13) in AMBER99 and OPLS-AA force field
calculations, respectively. Hierarchical clustering shows bend-
ing and helical backbone structures for peptide 1 for the most
representative clusters. OPLS-AA calculations, as in peptides
2e6, predict H-bonds in other positions than AMBER99.
Four H-bonds were formed between O(i) and HeN( j ), two
H-bonds between O(i) and HeN(i þ 3), one H-bond between
O(i) and HeN(i þ 2), and O(i) and HeN(i þ 4) (AMBER99
results). The results of OPLS-AA calculations predict eight
H-bonds formed between O(i) and HeN( j ), in peptide
1 four H-bonds between O(i) and HeN(i þ 2) and two H-
bonds between O(i) and HeN(i þ 4), one between O(i) and
HeN(i þ 3). In peptide 2, bend (Q3eK4, I6, Q12eK13) and

M.F. Masman et al. / European Journal of Medicinal Chemistry 44 (2009) 212e228
217
CF
CF
NCF
NCF
(1)
(2)
NCF
CF
NCF
NCF
(4)
(3)
CF
CF
CF
CF
(5)
(6)
NCF
Fig. 2. Edmundson wheel representations of peptides 1e6. The number in the center of the wheel corresponds to the peptide number. The ‘‘charged’’ (CF) and
‘‘non-charged’’ (NCF) faces are shown in blue dashed lines and full green lines, respectively. Positively charged amino acids are denoted with blue dots, the polar
ones with orange, and the hydrophobic ones with yellow. (For interpretation of the references to colour in this figure legend, the reader is referred to the web
version of this article.)
b-turn (W8eF10, M14eI15) alternate. Almost all of the bend
structures remained (Q3eN5, F10, K13) and a-helix formed
(I6eR9) which includes H-bonds. Peptides 3e6 contain a-he-
lix, 310-helix and b-turn in the central residues in AMBER99
results. OPLS-AA predicts bend structures almost at the same
residues.
in Supplementary material). The root mean square fluctuation
(RMSF) of the backbone atoms (Fig. S5 in Supplementary
material) and the hydrophilic and hydrophobic solvent acces-
sible surface areas (SASAs) were also calculated. The second-
ary structures of peptides were analyzed by sampling
trajectories every 10 ps with the DSSP program [34].
In all peptides simulated here, the initial 310-helix was de-
stroyed in the first 50e100 ps. The RMSD and the RMSF of
the backbone during the simulation characterize this change
in their secondary structure. The relative small change in the
RMSD of the peptides in the trajectory is evidence for the sta-
bilization of the backbone structure. RMSD for simulations
2.2.3. MD simulations
In the trajectory analysis of peptides 1e6, the total and po-
tential energies, radius of gyration (Rg) and the RMSD of the
backbone (NeCaeC(carbonyl)) atoms related to the structure at
the end of equilibration (100 ps) were calculated (Figs. S1eS4

218
M.F. Masman et al. / European Journal of Medicinal Chemistry 44 (2009) 212e228
Table 4
initial conformation of 310-helix was destroyed and a-helix,
b-turn/bend and a stable random meander structures at the
N- and C-terminal regions fluctuated during the whole simula-
tion. The residues 5e10 have shown the highest preference of
310-helix conformation. Here, in peptide 2, the b-turn and
bend conformations were mainly formed at residues 2e4
and 11e15. Also in this simulation the initial and final resi-
dues appear to have random coil structure because of the flex-
ibility of these residues. A mixture of coil, bend and b-turn
conformations was formed after 86 ns of simulation. This mix-
ture was observed until the end of the simulation. This is a dif-
ferent result with respect to the simulation performed on
peptide 1.
Secondary structures of the best conformation of peptides 1e8 obtained from
simulated annealing calculations by using AMBER99 and OPLS-AA force
fields (FFs)
Peptidea FF
Secondary structureb
Residue number
1 2
3
4
5
6
7 8 9 10 11 12 13 14 15 16
1
2
3
4
5
6
7
8
AMBER99 _ _
OPLS-AA _ _
AMBER99 _ _
OPLS-AA _ _
S
_
S
S
S
_
S
S
S
_
_
S
S
_
S
G G G T
T
S
_
_
_
S
S
_
T
S
S
S
T
_
T
_
_
_
_
_
S
_
S
S
_
T
T
T
T
_
T
_
H H H H S
AMBER99 _ _ G G G T
T
_
T
_
_
_
_
_
_
_
OPLS-AA _ _
AMBER99 _ _
OPLS-AA _ _
AMBER99 _ _
OPLS-AA _ _
_
T
_
_
_
S
T
S
S
T
S
S
T
S
T
S
In simulations 3e6, all derivatives adopt a helix-like con-
formation. However, while peptide 4 displays both a-helical
and 310-helical features, the structure of peptide 3 is predom-
inantly a 310-helix.
H H H H _
S
S
_
_
_
_
_
AMBER99 _ _ G G G _
OPLS-AA _ _
AMBER99 _ T T
OPLS-AA _ _
AMBER99 _ T T
OPLS-AA _ _
S
S
_
_
_
_
S S
It is interesting to note that the vectors of the dipole mo-
ment of the solvent molecules had no definite direction in
the periodic box. This indicates that the solvent had no electro-
static directional effect on the peptide structure. The largest
deviation of the dipole moment at the wall of the periodic
box was ꢀ0.05 D.
_
_
a
Peptide codes used in Table 1.
The secondary structure code obtained from DSSP program. H: 4-helix (a-
b
helix); S: bend; G: 3-helix (310-helix); T: H-bonded turns; _: loops or irregular
elements.
In summary our MD simulations indicate that peptides 1e6
adopt a helix-like conformation. However, whereas peptide 1
displays a marked preference for an a-helix structure, peptide
2 shows a mixture of beta turn, bend and 310-helix, being the
preferred form the 310-helix features.
1e6 is shown in Fig. S4 in Supplementary material. RMSD in-
creased to 0.2e0.8 nm in all cases and remained almost con-
stant in simulations 3e6. The fluctuation in RMSD is
attributed to slight changes in structure. In simulations 1 and
2 the conformations fluctuated between helical and turn/bend
structures (Figs. 3 and 4).
2.3. Comparison of theoretical results obtained from
different approaches
The N- and C-terminal residues in all simulations appear to
have a large flexibility as indicated by the change of RMSF
values during the simulations. The change at central residues
is moderate in simulations 1 and 2 (about 0.25 nm). The sim-
ulations 3e5 have shown a 0.15e0.20 nm change at residues
2e4 and 7. Also, the simulations 3 and 5 in the region which
seems to be the most sensitive in the conformational change
have a w0.35 nm change at residues 5 and 4, respectively.
In all simulations, Rgs remained almost constant, except in
1 and 2, where the values fluctuated, but the secondary struc-
tures might be considered stable (Fig. S3).
The conformational changes in simulation 1 are shown in
Fig. 3. The initial conformation returned and remained stable
in simulation 1, suggesting that the starting helical structure
was destroyed to form a mixture of a-helix, b-turn and bend
in the structure at residues 2e15. Such a conformational be-
havior was observed until the end of the simulation. The res-
idues 4e6 have shown the highest preference for a-helix
conformation. The initial and final residues appear to have
a random coil structure because of the flexibility of these res-
idues. For the same peptide in water Czajlik et al. [35] found
a significant amount of helix-like conformation, even in
a membrane-mimetic solvent system (TFEd2/water ¼ 9:1), by
1H NMR. Thus, our MD simulation is in very good agreement
with the experimental results. In simulation 2 (Fig. 4), the
The energetically preferred cluster obtained with EDMC
calculations for peptide 1 (70.48% of the total population)
contains an a-helix structure, the second most populated clus-
ter (7.46% of the total population) displays a-helix and b-turn
structures. On the other hand, the energetically most stable
form obtained in SA with minimization showed bend, 310-
helix and b-turn structures using AMBER99 force field, and
only bend and coil features using OPLS-AA force fields (see
Table 4). The energetically preferred form of peptide 2 showed
bend and b-turn structures for AMBER99 force field, and bend
and a-helix structures for OPLS-AA force field. Peptides 3e6
showed structures with helical or consecutive turn secondary
structures (AMBER99 force field) and bend with coil features
(OPLS-AA force field). These peptides have shown a higher
flexibility than peptides 1e2 due to their smaller size. In sum-
mary, AMBER99 force field has shown a slight preference for
helical structures. Thus, this force field has a better correlation
with ECCEP/3 force field than OPLS-AA force field. The
OPLS-AA results differ significantly showing some differ-
ences in 4 and j angle values and H-bond positions. Despite
this fact, all force fields used here predict a helix-like structure
for peptides 1e6. The N- and C-terminal residues have shown
a high flexibility, since no regular stable structure could be
observed.

M.F. Masman et al. / European Journal of Medicinal Chemistry 44 (2009) 212e228
219
Fig. 3. Change in the secondary structure during molecular dynamics simulation for peptide 1.
These results support the use of the MD simulations for
these peptides. It is clear, however, that in order to obtain a rel-
atively complete picture about the conformational intricacies
of peptide 1 and derivatives at least 100 ns of simulation ap-
pears to be necessary. Such simulations can provide useful in-
formation about the conformational preferences and molecular
flexibility of 1 and derivatives, which might be useful to get
a more profound understanding of the biological response of
these peptides.
Comparing the results obtained for the conformational
analysis using the different approaches, we can conclude
that, in general, these methods predict a helix-like structure
for peptide 1 and derivatives. These results are also in agree-
ment with the experimental results obtained from NMR [35].
to generate stereoelectrostatic forces. Thus, the electronic
study of peptides was performed using MEPs [36]. MEPs
have been shown to provide reliable information, both on
the interaction sites of molecules with point charges and on
the comparative reactivities of those sites [36,37]. More posi-
tive potentials reflect nucleus predominance, while less posi-
tive values represent rearrangements of electronic charges
and lone pairs of electrons. The fundamental application of
this study is the analysis of non-covalent interactions [37],
mainly by investigating the electronic distribution in the mol-
ecule. Thus, this methodology was used to evaluate the elec-
tronic distribution around molecular surface for peptides
reported here. The MEPs of peptide 1 are shown in Fig. 5
and the MEPs of peptides 3 and 6 are plotted in Fig. 6. We
evaluated the MEPs of all peptides tested but we show here
only the MEPs obtained for the three peptides, which dis-
played a significant antifungal activity.
2.4. Molecular electrostatic potentials (MEPs)
Knowledge of the stereoelectronic attributes and properties
of peptide 1 and derivatives will contribute significantly to the
elucidation of the molecular mechanism involved in the anti-
fungal activity. Molecular electrostatic fields and molecular
electrostatic potentials (MEPs), which are their visualisation,
offer an informative description of the capacity of peptides
To better appreciate the electronic behavior of 1 and con-
sidering that two different faces were signalled in Fig. 2, we
present the MEPs of 1, showing the two faces of this peptide
(Fig. 5). Fig. 5a gives the ‘‘charged face’’ (CF) characterized
by the presence of four cationic residues (R1, K4, R11 and
K15). Although it is possible to visualise residue K16 near

220
M.F. Masman et al. / European Journal of Medicinal Chemistry 44 (2009) 212e228
Fig. 4. Change in the secondary structure during molecular dynamics simulation for peptide 2.
to this face, in fact this residue is somewhat shifted in the di-
rection of the non-charged face. It has been previously re-
ported that peptideelipid association occurs through
formation of salt bridges between the positively charged resi-
dues K4, R11 and K15 and the lipid phosphate groups [33]. In
addition, tryptophan fluorescence studies previously showed
the importance of peptide with positively charged residues
for the initial binding to negatively charged vesicles, since
double R/K / A mutations involving the residues K4/R10/
R11/K13/K15 significantly decreased the binding affinity
[38]. The MEPs of 1 suggest that the above-mentioned resi-
dues (R1, K4, R11 and K15) could be responsible for the ini-
tial binding. The previously reported [33] p-stacking
interaction between F7 and R11 residues might be also appre-
ciated on this face. Although the main positive potentials (V(r)
ranging from 0.60 to 0.43 el auꢁ3) are concentrated on this
face, it should be noted that there is a relatively homogeneous
distribution of positively charged residues along the entire
structure. Thus, residues R10, K16 and K13 are strategically
located in an alternated fashion within the non-charged face.
Since the non-charged moiety is too large, two different
views of the MEPs were plotted in order to better visualise
this face (Fig. 5b and c). Fig. 5b displays four hydrophobic
residues I3, W6, F7 and W14. It appears that a kind of p-
stacking cluster through W6/R10/W14 occurs in this portion
of 1. Lensink et al. [33] reported that these residues could pro-
tect the peptide from the water phase. A clear hydrophobic in-
teraction between I3 and W6 might be also appreciated in this
figure. Fig. 5c displays a more polar face in comparison to
Fig. 5b, since it possesses the three polar residues of 1 (Q2,
Q8 and N9). Interestingly, I5 is located in an intercalated po-
sition with respect to polar residues and therefore there are no
interactions between them. These results suggest that these
polar residues could be highly solvated. Mutation of either
tryptophan decreases internalization, whereas double substitu-
tion completely inhibits peptide internalization [19,22,39].
These results indicate that peptide 1 is not sufficiently hydro-
phobic to insert deeply into phospholipid model membranes
[21,40]. Therefore charge neutralization is required for
a deeper insertion of the peptide into the hydrophobic core
of the membrane. The extended non-charged face alternating
cationic residues among the hydrophobic and polar ones ob-
served in the MEPs of 1 appears to be operational in this
sense.
Fig. 6a gives the MEPs obtained for peptide 3. This surface
shows a more extended positively charged face with respect to

M.F. Masman et al. / European Journal of Medicinal Chemistry 44 (2009) 212e228
221
Fig. 5. Electrostatic potential-encoded electron density surfaces of peptide 1. (a) ‘‘Charged face’’, (b) and (c) two different views of the ‘‘non-charged face’’. The
surfaces were generated with Gaussian 03 using RHF/6-31G single point calculations. The coloring represents electrostatic potential with red indicating the stron-
gest attraction to a positive point charge and blue indicating the strongest repulsion. The electrostatic potential is the energy of interaction of the positive point
charge with the nuclei and electrons of a molecule. It provides a representative measure of overall molecular charge distribution. The colour-coded is shown at the
left. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
1. In this portion are located five cationic residues R4, K8, R1,
R5 and K9. The MEPs obtained for peptide 6 are shown in
Fig. 6b. A very extensive deep blue zone with potential values
in the order of 0.60 el auꢁ3 clearly dominates this surface.
Previously reported MD simulations indicated that the aro-
matic residues do not contribute to the initial binding, but
rather to the subsequent insertion of peptide 1 between the bi-
layer head groups, where they shield the peptide from the
aqueous phase [33]. The importance of hydrophobic residues
seems to be crucial for the antifungal activity as well. Compar-
ing the MEPs obtained for peptides 1, 3 and 6 it is evident that
they correlate very well with their antifungal activities: the
most active peptide of this series, intermediate activity and
completely inactive molecule, respectively. These results
lend additional support for the key role of the hydrophobic res-
idues in these peptides acting as antifungal compounds.

222
M.F. Masman et al. / European Journal of Medicinal Chemistry 44 (2009) 212e228
Fig. 6. Electrostatic potential-encoded electron density surfaces of peptides 3 (a) and 6 (b). The colour-coded is shown at the left. (For interpretation of the ref-
erences to colour in this figure legend, the reader is referred to the web version of this article.)
2.5. Small size antifungal peptides
length of the backbone of these peptides they cannot achieve
a stable structure. Although both tetrapeptides display a statis-
tical coil [41] structure, their conformational preferences are
markedly different. Whereas peptide 7 displays a quasi-p-he-
lix form (j1 ¼ ꢁ62.9, v2 ¼ ꢁ88.1, j2 ¼ 81.8, v3 ¼ ꢁ86.4,
j3 ¼ 72.3 and v4 ¼ ꢁ82.7), peptide 8 prefers a quasi-a-helix
structure (j1 ¼ ꢁ18.7, v2 ¼ ꢁ73.9, j2 ¼ ꢁ25.5, v3 ¼ ꢁ82.3,
j3 ¼ ꢁ27.6 and v4 ¼ ꢁ78.4). This is a striking difference ob-
tained for these tetrapeptides. SA calculations indicate that
these peptides are more flexible than peptides 1e6 (Table 4).
The small size of these peptides could explain the irregular
structures obtained for peptides 7e8 (OPLS-AA force field),
whereas turn structures were preferred for the central residues
of these peptides (AMBER99). MD calculations indicate that
peptides 7 and 8 show a statistical coil structure during the
whole simulation time because of their high flexibility. How-
ever, from MD a clear difference obtained for these tetrapep-
tides is that peptide 8 displays a tendency to adopt a turn
The consideration, that a peptide-based antifungal agent
should be as short as possible in order to reduce costs, promp-
ted us to synthesize even shorter derivatives of 1. Thus, with
the aim to predict the potential antifungal effect of small
size peptides, we performed a molecular modeling study in
a series of tetrapeptides. Several different tetrapeptides were
modeled; however, we report here only the results obtained
for two of them (KWKKeNH2 (7) and RQKKeNH2 (8)), be-
ing representative of the entire series.
Tetrapeptides 7 and 8 displayed a completely different con-
formational and electronic behavior. EDMC, SA and MD re-
sults obtained for these peptides are summarized in Tables 3
and 4 and Tables SVIIeSVIII and Figs. S1eS5 in Supplemen-
tary material, respectively. EDMC results obtained for tetra-
peptides 7 and 8 predict a higher molecular flexibility in
comparison to the longer peptides. Also, considering the

M.F. Masman et al. / European Journal of Medicinal Chemistry 44 (2009) 212e228
223
conformation, whereas peptide 7 has the tendency to achieve
extended conformations. The spatial ordering adopted by pep-
tide 8 is closely related to that obtained for the first four res-
idues of peptide 1 (see Fig. 7c right hand).
Fig. 7 gives the MEPs obtained for compounds 1, 3 and 8 in
a front view. The similar MEP values obtained for peptides 1
and 8 can be appreciated in this figure. In contrast a more pos-
itively charged MEP was obtained for compound 3 as it was
previously discussed.
On the basis of the stereoelectronic similarity obtained for
peptides 1 and 8, it is reasonable to think that tetrapeptide 8
could present some antifungal activity, but not peptide 7.
Thus, to confirm our hypothesis, we decided to test the anti-
fungal effects of tetrapeptides 7 and 8. As expected, peptide
7 did not show any antifungal activity with the concentrations
reported here. In contrast, peptide 8 displayed a moderate but
significant antifungal effect against C. neoformans (Table 1). It
must be pointed out that peptide 8 displays antifungal activity
Fig. 7. Electrostatic potential-encoded electron density surfaces obtained for peptides 1 (a), 3 (b) and 8 (c) in a front view of the N-terminal portion (left hand). Also
this figure shows at the right hand: (a) a frontal spatial view of 1, (b) an overlapped stereoview of peptide 1 (white) with peptide 3 (yellow), and (c) an overlapped
stereoview of peptide 1 (white) with peptide 8 (green). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version
of this article.)

224
M.F. Masman et al. / European Journal of Medicinal Chemistry 44 (2009) 212e228
only against C. neoformans; in contrast this compound was not
active against C. albicans. This is a striking difference with re-
spect to 1. It is clear, however, that the antifungal potency
found for 8 against C. neoformans is very interesting particu-
larly considering its small size.
solubility and bioavailability, or by minimizing toxicity and
overcoming drug resistance.
4. Experimental section
Inspection of Fig. 7 led us to appreciate that although pep-
tides 1, 3 and 8 are of different size, they display some stereo-
electronic similarities primarily near the N-terminal portion,
implicating a likely overlap on these moieties. Cellular uptake
experiments previously demonstrated the crucial role of argi-
nines in cell-penetrating peptides [42]. Also, MD simulations
indicated that arginine residues could be responsible for the
initial electrostatic binding forming bidentate hydrogen bonds
with lipid phosphate groups. It should be noted that the first
residue of peptides 1, 3 and 8 is arginine. In fact there are var-
ious ways in which 8 may produce its antifungal activity, on
which we can only speculate. In that respect the similar stereo-
electronic behaviors observed between 8 and the N-terminal
portions of 1 and 3 are particularly noteworthy.
4.1. Synthetic methods
Solid phase synthesis of 1, RQIKIWFQNRRMKWKKe
NH2, was carried out manually on a p-methylbenzhydrylamine
resin (MBHA, 0.39 mmol/g) with standard methodology using
Boc-strategy. Side chain protecting groups were as follows:
Tos for Arg and 2-chloro-Z for Lys. All protected amino acids
were coupled in CH2Cl2 or CH2Cl2/DMF solvent mixture in
a ratio of 1:1 (for Gln and Asn) using DCC (2.5 equiv.) and
HOBt (2.5 equiv.). Amino acid incorporation was monitored
by Kaiser [45] ninhydrin test. After coupling of the amino
acid, Boc-deprotection was effected by using TFA/CH2Cl2
(1:1) for 5 min first then repeated for 25 min. The completed
peptide resin was treated with liquid HF/dimethyl sulfide/ani-
sole/indole (86:6:4:2) at 5 ꢂC for 1 h. HF was removed and the
resulted free peptide was solubilized in 10% aqueous acetic
acid and lyophilized. The crude peptide was purified by semi-
preparative RP HPLC on a Lichrosorb C-18 column
(16 ꢃ 250 mm, 7 mm) with a linear gradient of acetonitrile
30e100%, 0.1% TFA in 70 min, 4 ml/min flow. The appropri-
ate fractions were pooled and lyophilized. The purified peptide
was characterized by HPLC and mass spectrometry using
a Finnigan TSQ 7000 tandem quadrupole electrospray mass
spectrometer. M þ H calc.: 2244.3, found: 2244.5. Analytical
HPLC conditions: column: Luna C-18, 5 mm, 4.6 ꢃ 250 mm;
gradient: 30e50% AcN, 0.1% TFA in 10 min, 1.2 ml/min
flow, 220 nm. Rt: 4.280 min. Same synthesis procedures and
analytical HPLC conditions were applied for the peptide ana-
logues (2e8) as follows:
3. Conclusions
In the present paper, we report the design, synthesis and an-
tifungal effects of peptide 1 and derivatives, a new series of
antifungal peptides. Among the peptides tested, RQI-
KIWFQNRRMKWKKeNH2 (1), RKWRRKWKKeNH2 (3)
and RQKKeNH2 (8) displayed the most powerful inhibitory
effect against C. neoformans. A detailed conformational and
electronic study supported by theoretical calculations helped
us to identify a possible ‘‘biologically relevant conformation’’
for these peptides. A particular combination of cationic and
hydrophobic residues adopting a definite spatial ordering ap-
pears to be the key parameter for the transition from hydro-
philic to hydrophobic phase, which could be a necessary
step to produce the antifungal activity. We believe that our re-
sults could contribute to an understanding of the minimal
structural requirements for the antifungal potential of peptides
reported here and to the design of novel structurally related
agents. These results are very encouraging in that they show
a great potential for peptides 1 and 8 as a truly novel class
of antifungal compounds particularly against the yeast
C. neoformans. Thus, the antifungal activity reported in this
paper for peptides 1 and 8 opens promising ways for the devel-
opment of a new antifungal agent for treatment of cryptococ-
cosis. Since C. neoformans remains an important life-
threatening complication for immunocompromised hosts, par-
ticularly for patients who have undergone transplantation of
solid organs and therefore, new compounds effectively acting
against this fungus are highly needed [43,44].
Retention
factor (min)
Gradient
elution (%)
RQIKIWFQNRRMKWKKeNH2 (1)
WKQKNIKWRFRQKMIReNH2 (2)
RKWRRKWKKeNH2 (3)
RKFRRKFKKeNH2 (4)
RKRRKWKKeNH2 (5)
RKRRKKKeNH2 (6)
4.280
11.200
13.200
9.387
9.339
4.576
9.012
3.553
30e50 (10 min)
5e80 (25 min)
5e35 (15 min)
5e30 (10 min)
5e35 (15 min)
5e35 (15 min)
5e35 (15 min)
5e20 (15 min)
KWKKeNH2 (7)
RQKKeNH2 (8)
Mass spectrometric analysis of peptides 1e8 is given in
Supplementary material.
4.2. Antifungal evaluation
Peptide 1 has been used as a non-viral, non-toxic and
highly efficient vector for delivering bioactive substances,
that by themselves are membrane-impermeable, to the cyto-
plasm or nucleus of cells [18,19]. The antifungal activity
found for 1 is very interesting by itself but it is also important
considering its potential use as a carrier for other known anti-
fungal drugs. Thus, this system may improve the pharmaceu-
tical properties of the drugs by, for example, improving
4.2.1. Microorganisms and media
For the antifungal evaluation, standardized strains from the
American Type Culture Collection (ATCC), Rockville, MD,
USA, were used (C. albicans ATCC 10231 and C. neoformans
ATCC 32264). Active compounds were tested against clinical
´
isolates from the Malbran Institute [(M), Av. Velez Sarsfield
563, Buenos Aires]. The isolates included 10 strains of C. neo-
formans which voucher specimens are presented in Table 2.

M.F. Masman et al. / European Journal of Medicinal Chemistry 44 (2009) 212e228
225
Strains were grown on Sabouraudechloramphenicol agar
slants for 48 h at 30 ꢂC, maintained on slopes of Sabour-
audedextrose agar (SDA, Oxoid) and sub-cultured every 15
days to prevent pleomorphic transformations. Inocula of cells
were obtained according to reported procedures and adjusted
to 1e5 ꢃ 103 cells with colony forming units (CFU)/ml [46].
spp. or Cryptococcus sp., respectively. Microplates were read
in a VERSA Max microplate reader (Molecular Devices, Sun-
nyvale, CA, USA). Amphotericin B was used as positive con-
trol (100% inhibition). Tests were performed in duplicates.
Reduction of growth for each peptide concentration was calcu-
lated as follows: % of inhibition ¼ 100 ꢁ (OD405 PTW ꢁ
OD405 SCW)/OD405 GCW ꢁ OD405 SCW.
4.3. Antifungal susceptibility testing
Statistical analysis: Data were statistical analyzed by one-
way analysis of variance. A p value of <0.05 was considered
significant.
4.3.1. MIC determinations
MIC of each extract or compound was determined by using
broth microdilution techniques according to the guidelines of
the Clinical and Laboratory Standards Institute (CLSI), for-
merly National Committee for Clinical and Laboratory Stan-
dards for yeasts (M27-A2) [46]. MIC values were
determined in RPMI-1640 (Sigma, St. Louis, MO, USA) buff-
ered to pH 7.0 with MOPS. Microtiter trays were incubated at
35 ꢂC in a moist, dark chamber, and MICs were visually re-
corded at 48 h. For the assay, stock solutions of pure com-
pounds were twofold diluted with RPMI from 256 to
0.98 mg/ml (final volume ¼ 100 ml) and a final DMSO concen-
tration ꢄ1%. A volume of 100 ml of inoculum suspension was
added to each well with the exception of the sterility control
where sterile water was added to the well instead. Ketocona-
zole, amphotericin B and itraconazole (all from Sigma Chem-
ical Co., St. Louis, MO, USA) were used as positive controls.
Endpoints were defined as the lowest concentration of drug re-
sulting in total inhibition (MIC100) of visual growth compared
to the growth in the control wells containing no antifungal.
MIC80 and MIC50 were defined as the lowest concentration
of a compound that caused 80 and 50% reduction of the
growth control, respectively (culture media with the microor-
ganism but without the addition of any compound) and were
determined spectrophotometrically with the aid of a VERSA
Max microplate reader (Molecular Devices, USA). The MFC
of peptide 1 against each isolate was also determined as fol-
lows: after determining the MIC, an aliquot of 5 ml sample
was withdrawn from each clear well of the microtiter tray
and plated onto a 150 mm RPMI-1640 agar plate buffered
with MOPS (Remel, Lenexa, KS). Inoculated plates were in-
cubated at 30 ꢂC and MFC was recorded after 48 h. The
MFC was defined as the lowest concentration of each com-
pound that resulted in total inhibition of visible growth. All
tests were performed in duplicates.
4.4. Computational methods
4.4.1. EDMC calculations
The conformational space of each peptide was explored us-
ing the method previously employed by Liwo et al. [47] that
included the electrostatically driven Monte Carlo (EDMC)
method [29]. Conformational energy was evaluated using the
ECEPP/3 force field [48]. This force field employs rigid va-
lence geometry. Hydration energy was evaluated using a hydra-
˚
tion-shell model with a solvent sphere radius of 1.4 A and
atomic hydration parameters that have been optimized using
nonpeptide data (SRFOPT) [49]. In this model, in addition
to a sum of electrostatic, non-bonding, hydrogen bond and tor-
sional energy terms, the total conformational energy includes
terms accounting for loop closing and peptide solvation. The
conformation with minimized energy was subsequently per-
turbed by changing its torsional f and J angles using the
Monte Carlo method [50]. Piela’s algorithm [51], which was
also applied at this stage, greatly improves the acceptance co-
efficient. In this algorithm f and J angles are changed in
a manner which allow the corresponding peptide group to
find the most proper orientation in the electrostatic field of
the rest of the peptide chain. The energy of the new conforma-
tion is minimized, compared to the previous one and may be
accepted or discarded on the basis of energy and/or geometry.
If the new energy-minimized conformation is similar in shape
and in energy to the starting conformation, it is discarded. Oth-
erwise, the energy of the new conformation is compared to the
energy of the parent conformation. If the new energy is lower,
the new conformation is accepted unconditionally, otherwise
the Metropolis criterion [52] is applied in order to accept or
reject the new conformation. If the new conformation is ac-
cepted, it replaces the starting one; otherwise another perturba-
tion of the parent conformation is tried. A temperature jump
may be included if the perturbation is not successful for an ar-
bitrarily chosen number of iterations. The process is iterated
until a sufficient number of conformations have been accepted.
The detailed procedure is described in Ref. [53].
4.3.2. Determination of percentage of inhibition
The test was performed in 96-well microplates. Peptide test
wells (PTWs) were prepared with stock solutions of each pep-
tide in DMSO (maximum concentration ꢄ 2%), diluted with
RPMI-1640 to final concentrations 200e12.5 mM. Inoculum
suspension (100 ml) was added to each well (final volume in
the well ¼ 200 ml). A growth control well (GCW) (containing
medium, inoculum, the same amount of DMSO used in PTW,
but compound-free) and a sterility control well (SCW) (sam-
ple, medium and sterile water instead of inoculum) were in-
cluded for each strain tested. Microtiter trays were incubated
in a moist, dark chamber at 35 ꢂC, 24 or 48 h for Candida
In order to explore the conformational space extensively,
we carried out 10 different runs, each of them with a different
random number, for each peptide studied. Since the EDMC
procedure uses random numbers, there is a need to initialize
the random number generator by providing an integer. There-
fore, we collected a total of 5000 accepted conformations for
each peptide studied. Each EDMC run was terminated after
500 energy-minimized conformations had been accepted.

226
M.F. Masman et al. / European Journal of Medicinal Chemistry 44 (2009) 212e228
The parameters controlling the runs were the following: a tem-
perature of 298.15 K was used for the simulations. A temper-
ature jump of 50 000 K was used and the maximum number of
allowed repetitions of the same minimum was 50. The maxi-
mum number of electrostatically predicted conformations per
iteration was 400; the maximum number of random-generated
conformations per iteration was 100; the fraction of random/
electrostatically predicted conformations was 0.30. The maxi-
mum number of steps at one increased temperature was 20 and
the maximum number of rejected conformations until a tem-
perature jump executed was 100. Only trans peptide bonds
(u y 180ꢂ) were considered.
i þ 2, i þ 3, i þ 4, respectively. Residues without recognized
secondary structures are assigned as irregular or loop (random
coil) structure. For clustering 5000 structures a perl script in
MMTSB [56] was used.
4.4.3. MD simulations
Molecular dynamics (MD) simulations were performed us-
ing the GROMACS 3.2.1 package of programs [31,32], with
the OPLS-AA force field [57e59]. The calculations were car-
ried out using
a
standard helix 310 (f ¼ ꢁ49.00ꢂ,
j ¼ ꢁ26.00ꢂ) as a starting structure. The peptides 1e8 were
embedded in a box containing the SPC water model [60]
˚
that extended to at least 10 A between the solutes and the
The ensemble of obtained conformations was then clus-
tered into families using the program ANALYZE [54], which
applies the minimal-tree clustering algorithm for separation,
edge of the box. The total number of water molecules was be-
tween 1389 and 2791. Then Clꢁ ions were added to the sys-
tems by replacing water molecules in random positions, thus
making the whole system neutral. For details see Table SIX in
Supplementary material. Prior to dynamics simulation, internal
constraints were relaxed by energy minimization. Following the
minimization, an MD equilibration run was performed under
position restraints for 20 ps. An unrestrained run was then initi-
ated. During the MD runs, the LINCS algorithm [61] was used to
constrain the lengths of hydrogen containing bonds; the waters
were restrained using the SETTLE algorithm [62]. The time
step for the simulations was 0.002 ps. The simulations were
run under NPT conditions, using Berendsen’s coupling algo-
rithm [63] for keeping the temperature and pressure constant
(P ¼ 1 bar, tP ¼ 0.5 ps; T ¼ 310 K, tT ¼ 0.1 ps). The compress-
ibility was 4.8 ꢃ 10ꢁ5 barꢁ1. Van der Waals forces were treated
using all heavy atoms, energy threshold of 30 kcal molꢁ1
,
˚
and RMSD of 0.75 A as separation criteria for all peptides
studied here. Molecules containing seven, eight and 16 amino
acids residues were clustered using the same method but in-
stead of using all heavy atoms it used only the backbone atoms
(Ca, N and C(carbonyl)). This procedure allows for substantial
reduction of the number of conformations and eliminates
repetitions.
4.4.2. SA calculations
To find the best structures of peptides studied, 5000 struc-
tures were generated by simulated annealing (SA) imple-
mented in Tinker Version 4.2 [55]. Peptides were generated
by the software and minimized with a charge and van der
Waals cutoff of 1.8 and 1.4 nm, respectively, with OPLS-AA
and AMBER99 force fields. A taper of 0.8 was applied to
smooth the cutoff to zero in the calculations. The solvation
was simulated by GB/SA (generalized Born/surface area)
˚
using a 12 A cutoff. Long-range electrostatic forces were treated
using the particle mesh Ewald method (PME) [64]. The coordi-
nates were saved every 10 ps. The total simulation time was
100 ns for every peptide. The analysis of the simulations was
performed using the analysis tools provided in the GROMACS
package.
model. The tolerance of the minimization was
0.001 kcal molꢁ1
ꢁ1. The following protocol was used in
˚
A
the simulated annealing: 5000 times were repeated 2000 steps
equilibration in 1000 K and 2000 steps cooling to 50 K expo-
nentially with 1 fs step-size. After cooling the structures were
minimized with the previous method. The minimized structure
was applied for the next step. The secondary structures of the
peptides were analyzed by DSSP [34]. The algorithm mea-
sures the geometry and the hydrogen bonding in peptides.
The secondary structures are assigned by the following way:
turns within 3, 4 and 5 residues, respectively, are assigned
on the basis of hydrogen bonds between i to i þ 1 (three resi-
due turn), i to i þ 1 or i þ 2 (four residue turn), i to i þ 1 or
i þ 2 or i þ 3 (five residue turn). A b-bridge is assigned
when two non-overlapping stretches of three residues each,
i ꢁ 1, i, i þ 1 and j ꢁ 1, j, j þ 1 form H-bonding parallel or
antiparallel pattern. Two or more consecutive b-bridge struc-
tures are assigned as b-sheet structures. A bend is defined as
a five residue turn without H-bonds with a curvature of at least
70ꢂ between the first three residues (i ꢁ 2, i ꢁ 1 and i) and the
last three residues (i, i þ 1 and i þ 2). The position of the bend
is marked at i. Two consecutive turns at position i ꢁ 1 and i
form a helix. 310-Helix is marked at positions i, i þ 1 and
i þ 2. a-Helix and p-helix at i, i þ 1, i þ 3 and i, i þ 1,
4.4.4. Molecular electrostatic potentials
Quantum mechanics calculations were carried out using the
Gaussian 03 program [65]. We use the most populated confor-
mations of peptides 1e8 obtained from EDMC calculations.
Subsequently, single point ab initio (RHF/6-31G(d)) calcula-
tions were carried out. The electronic study was carried out
using molecular electrostatic potentials (MEPs) [36]. These
MEPs were calculated using RHF/6-31G(d) wave functions
and MEP graphical presentations were created using the
MOLEKEL program [66].
Acknowledgements
This work is part of the Hungarian-Argentine Intergovern-
mental S&T Cooperation Programme and was supported by
the Research and Technological Innovation Foundation and
by the SECyT, as well as by grants from Universidad Nacional
de San Luis and grants to S.A.Z. and R.D.E. from the Agencia
´
´
´
de Promocion Cientıfica y Tecnologica de la Argentina (ANP-
CyT) PICT R 260. It is part of the Iberoamerican Program of
Science and Technology for the Development (CYTED,

M.F. Masman et al. / European Journal of Medicinal Chemistry 44 (2009) 212e228
227
[26] E.J. Ernst, E.E. Roling, R. Petzold, D.J. Keele, M.E. Klepser,
Antimicrob. Agents Chemother. 46 (2002) 3846e3853.
[27] M. Schiffer, C.-H. Chang, F.J. Stevens, Protein Eng. 5 (1992) 213e214.
´
[28] T. Letoha, S. Gaal, C. Somlai, A. Czajlik, A. Perczel, B. Penke, J. Mol.
Recognit. 16 (2003) 272e279.
[29] D.R. Ripoll, H.A. Scheraga, Biopolymers 27 (1988) 1283e1303.
[30] J.W. Ponder, TINKER e Software Tools for Molecular Design, User’s
Guide for Version 4.2, Washington University School of Medicine,
2004.
Network RIBIOFAR). T.K. and B.P. thank the Hungarian
Research Fund OTKA K61577. R.D.E. is a member of the
CONICET (Argentina) staff. M.F.M is a fellow of CONICET
(Argentina). M.F.M. is a recipient of a Bernouilli Fellowship
from the University of Groningen. The authors thank E. Meny-
´
hart for the technical assistance of peptide synthesis.
[31] E. Lindahl, B. Hess, D. van der Spoel, J. Mol. Model. 7 (2001) 306e
317.
Appendix. Supplementary material
[32] D. van Der Spoel, A.R. van Buuren, E. Apol, P.J. Meulenhoff,
D.P. Tieleman, A.L.T.M. Sijbers, B. Hess, K.A. Feenstra, E. Lindahl,
R. van Drunen, H.J.C. Berendsen, GROningen Machine for Chemical
Simulations, Department of Biophysical Chemistry, University of
Groningen, Groningen, The Netherlands, 2002.
Supplementary data associated with this article can
be found in the online version, at doi:10.1016/j.ejmech.2008.
02.019.
[33] M.F. Lensink, B. Christiaens, J. Vandekerckhove, A. Prochiantz,
M. Rosseneu, Biophys. J. 88 (2005) 939e952.
[34] W. Kabsch, C. Sander, Biopolymers 22 (1983) 2577e2637.
References
[1] T.J. Walsh, A. Groll, J. Hiemenz, R. Fleming, E. Roilides, E. Anaissie,
Clin. Microbiol. Infect. 10 (2004) 48e66.
´
[35] A. Czajlik, E. Mesko, B. Penke, A. Perczel, J. Pept. Sci. 8 (2002)
151e171.
[36] P. Politzer, D.G. Truhlar (Eds.), Chemical Applications of Atomic and
Molecular Electrostatic Potentials, Plenum Press, New York, 1991.
[2] N. Georgopapadakou, J. Tkacz, Trends Microbiol. 3 (1995) 98e104.
[3] M. Nagiec, E. Nagiec, J. Baltisberger, G. Wells, R. Lester, R. Dickson,
J. Biol. Chem. 272 (1997) 9809e9817.
[4] S. Ablordeppey, P. Fan, I. Ablordeppey, L. Mardenborough, Curr. Med.
Chem. 6 (1999) 1151e1195.
´
´
[37] G. Naray-Szabo, G.G. Ferenczy, Chem. Rev. 95 (1995) 829e847.
[38] B. Christiaens, J. Grooten, M. Reusens, A. Joliot, M. Goethals,
J. Vandekerckhove, A. Prochiantz, M. Rosseneu, Eur. J. Biochem. 271
(2004) 1187e1197.
[5] T. White, K. Marr, R. Bowden, Clin. Microbiol. Rev. 11 (1998)
382e402.
[6] A. Polak, Mycoses 42 (1999) 355e370.
[39] M.H. Lindberg, H. Biverstahl, A. Graslund, L. Maler, Eur. J. Biochem.
270 (2003) 3055e3063.
´
[40] C.E.B. Brattwall, P. Lincoln, B. Norden, J. Am. Chem. Soc. 125 (2003)
´
[7] J. Bartroli, E. Turmo, M. Alguero, E. Boncompte, M. Vericat, L. Conte,
´
J. Ramis, M. Merlos, J. Garcıa-Rafanell, J. Forn, J. Med. Chem. 41
14214e14215.
[41] J.A. Vila, D.R. Ripoll, H.A. Baldoni, H.A. Scheraga, J. Biomol. NMR 24
(2002) 245e262.
(1998) 1855e1868.
[8] J. Bartroli, E. Turmo, M. Alguero, E. Boncompte, M. Vericat, L. Conte,
´
´
J. Ramis, M. Merlos, J. Garcıa-Rafanell, J. Med. Chem. 41 (1998)
´
´
[42] P.E.G. Thoren, D. Persson, P. Isakson, M. Goksor, A. Onfelt, B. Norden,
Biochem. Biophys. Res. Commun. 307 (2003) 100e107.
[43] D.P. Kontoyiannis, E. Mantadakis, G. Samonis, J. Hosp. Infect. 53 (2003)
243e258.
1869e1882.
[9] R.E.W. Hancock, R. Leher, Trends Biotechnol. 16 (1998) 82e88.
[10] P. Bulet, R. Stocklin, Protein Pept. Lett. 12 (2005) 3e11.
¨
´
[11] M.F. Masman, A.M. Rodrıguez, L. Svetaz, S.A. Zacchino, C. Somlai,
[44] N. Singh, Lancet Infect. Dis. 3 (2003) 703e708.
[45] E. Kaiser, R.L. Colescott, C.D. Bossiuger, P.I. Cook, Anal. Biochem. 34
(1970) 595e598.
[46] Wayne (Ed.), second ed.NCCLS, National Committee for Clinical
and Laboratory Standards, Method M27-A2, vol. 22, 2002, pp. 1e
29.
I.G. Csizmadia, B. Penke, R.D. Enriz, Bioorg. Med. Chem. 14 (2006)
7604e7614.
[12] M. Zasloff, Curr. Opin. Immunol. 4 (1992) 3e7.
[13] T. Ganz, R.I. Lehrer, Pharmacol. Ther. 66 (1995) 191e205.
[14] J.A. Hoffman, F.C. Kafatos, C.A. Janeway Jr., R.A.B. Ezekowitz,
Science 284 (1999) 1313e1318.
[47] A. Liwo, A. Tempczyk, S. Oldziej, M.D. Shenderovich, V.J. Hruby,
S. Talluri, J. Ciarkowski, S. Kasprzykowski, L. Lankiewicz,
Z. Grzonka, Biopolymers 38 (1996) 157e175.
[15] R.E.W. Hancock, Lancet 349 (1997) 418e422.
[16] R.E.W. Hancock, A. Patrzykat, Curr. Drug Targets Infect. Disord. 2
(2002) 79e83.
´
[48] G. Nemethy, K.D. Gibson, K.A. Palmer, C.N. Yoon, G. Paterlini,
[17] U. Langel (Ed.), Cell-penetrating Peptides: Processes and Applications,
CRC Press, Boca Raton, FL, 2002.
A. Zagari, S. Rumsey, H.A. Scheraga, J. Phys. Chem. 96 (1992)
6472e6484.
[49] R.L. Williams, J. Vila, G. Perrot, H.A. Scheraga, Proteins: Struct., Funct.,
Genet. 14 (1992) 110e119.
[18] A. Joliot, C. Pernelle, H. Deagostini-Bazin, A. Prochiantz, Proc. Natl.
Acad. Sci. U.S.A. 88 (1991) 1864e1868.
[19] D. Derossi, A.H. Joliot, G. Chassaing, A. Prochiantz, J. Biol. Chem. 269
(1994) 10444e10450.
[20] D. Derossi, S. Calvet, A. Trembleau, A. Brunissen, G. Chassaing,
A. Prochiantz, J. Biol. Chem. 271 (1996) 18188e18193.
[21] G. Drin, M. Mazel, P. Clair, D. Mathieu, M. Kaczorek, J. Temsamani,
Eur. J. Biochem. 268 (2001) 1304e1314.
[50] Z. Li, H.A. Scheraga, Proc. Natl. Acad. Sci. U.S.A. 84 (1987)
6611e6615.
[51] L. Piela, H.A. Scheraga, Biopolymers 26 (1987) S33eS58.
[52] N. Metropolis, A.W. Rosenbluth, M.N. Rosenbluth, A.H. Teller,
E.J. Teller, Chem. Phys. 21 (1953) 1087e1091.
´
[53] M. Sidor, J. Wojcik, D. Pawlak, J. Izdebski, Acta Biochim. Pol. 46 (1999)
´
[22] T. Letoha, S. Gaal, C. Somlai, Z. Venkei, H. Glavinas, E. Kusz, E. Duda,
641e650.
[54] H.A. Scheraga, D.R. Ripoll, A. Liwo, C. Czaplewski, User Guide
ECEPPAK and ANALYZE Programs, 1983.
´
A. Czajlik, F. Petak, B. Penke, J. Pept. Sci. 11 (2005) 805e811.
[23] E. Dupont, A.H. Joliot, A. Prochiantz, in: U. Langel (Ed.), Cell-penetrating
Peptides: Processes and Applications, CRC Press, Boca Raton, FL, 2002,
pp. 23e51.
[24] S.A. Zachino, R.A. Yunes, V. Cechinel Filho, R.D. Enriz, V. Kouznetsov,
J.C. Ribas, in: M. Rai, D. Mares (Eds.), Plant-derived Antimycotics:
Current Trends and Future Prospects, Haworth Press, New York, 2003,
pp. 1e47.
[55] C.E. Kundrot, J.W. Ponder, F.M. Richards, J. Comput. Chem. 12 (1991)
402e409.
[56] M. Feig, J. Karanicolas, C.L. Brooks III, MMTSB Tool Set, MMTSB
NIH Research Resource, The Scripps Research Institute, 2001. Available
from:<http://feig.bch.msu.edu/mmtsb/Main_Page>.
[57] W.L. Jorgensen, D.S. Maxwell, J. Tirado-Rives, J. Am. Chem. Soc. 118
(1996) 11225e11236.
´
´
[25] M. Sortino, P. Delgado, S. Juarez, J. Quiroga, R. Abonıa, B. Insuasty,
M. Nogueras, L. Rodero, F.M. Garibotto, R.D. Enriz, S.A. Zacchino,
Bioorg. Med. Chem. 15 (2007) 484e494.
[58] E.K. Watkins, W.L. Jorgensen, J. Phys. Chem. A 105 (2001) 4118e4125.

228
M.F. Masman et al. / European Journal of Medicinal Chemistry 44 (2009) 212e228
[59] G.A. Kaminski, R.A. Friesner, J. Tirado-Rives, J. Phys. Chem. B 105
(2001) 6474e6487.
H. Nakatsuji, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa,
M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai, M. Klene,
X. Li, J.E. Knox, H.P. Hratchian, J.B. Cross, C. Adamo, J. Jaramillo,
R. Gomperts, R.E. Stratmann, O. Yazyev, A.J. Austin, R. Cammi,
C. Pomelli, J.W. Ochterski, P.Y. Ayala, K. Morokuma, G.A. Voth,
P. Salvador, J.J. Dannenberg, V.G. Zakrzewski, S. Dapprich,
A.D. Daniels, M.C. Strain, O. Farkas, D.K. Malick, A.D. Rabuck,
K. Raghavachari, J.B. Foresman, J.V. Ortiz, Q. Cui, A.G. Baboul,
S. Clifford, J. Cioslowski, B.B. Stefanov, G. Liu, A. Liashenko,
P. Piskorz, I. Komaromi, R.L. Martin, D.J. Fox, T. Keith, M.A. Al-La-
ham, C.Y. Peng, A. Nanayakkara, M. Challacombe, P.M.W. Gill,
B. Johnson, W. Chen, M.W. Wong, C. Gonzalez, J.A. Pople, Gaussian
03, Revision B.05, Gaussian, Inc., Pittsburgh, PA, 2003.
[60] H.J.C. Berendsen, J.P.M. Postma, W.F. van Gunsteren, J. Hermans, in:
B. Pullman (Ed.), Intermolecular Forces, Reidel, Dordrecht, 1981, pp.
331e342.
[61] B. Hess, H. Bekker, H.J.C. Berendsen, J.G.E.M. Fraaije, J. Comput.
Chem. 18 (1997) 1463e1472.
[62] S. Miyamoto, P.A. Kollman, J. Comput. Chem. 13 (1992) 952e962.
[63] H.J.C. Berendsen, J.P.M. Postma, W.F. van Gunsteren, A. DiNola,
J.R. Haak, J. Chem. Phys. 81 (1984) 3684e3690.
[64] T. Darden, D. York, L. Pedersen, J. Chem. Phys. 98 (1993) 10089e10092.
[65] M.J. Frisch, G.W. Trucks, H.B. Schlegel, G.E. Scuseria, M.A. Robb,
J.R. Cheeseman, J.A. Montgomery Jr., T. Vreven, K.N. Kudin,
J.C. Burant, J.M. Millam, S.S. Iyengar, J. Tomasi, V. Barone,
B. Mennucci, M. Cossi, G. Scalmani, N. Rega, G.A. Petersson,
[66] P. Flukiger, H.P. Luthi, S. Portmann, J. Weber, MOLEKEL 4.0, Swiss
¨
Center for Scientific Computing, Manno, Switzerland, 2000.
¨