Conformational behavior of Temporin L in aqueous solution: A computational/experimental study.

Full text

Page 1

Conformational Behavior of
Temporin A and Temporin L
in Aqueous Solution: A
Computational/Experimental
Study
M. D’Abramo1
A. C. Rinaldi2
A. Bozzi3
A. Amadei4
G. Mignogna5
A. Di Nola1
M. Aschi6
1Dipartimento di Chimica,
Universita ` di Roma ‘ ‘La
Sapienza,’ ’ Roma, Italia
2Dipartimento di Scienze e
Tecnologie Biomediche,
Sezione di Chimica Biologica e
Biotecnologie Biochimiche,
Universita ` di Cagliari,
Monserrato, Italia
3Dipartimento di Scienze e
Tecnologie Biomediche,
Universita ` de L’Aquila,
L’Aquila, Italia
Abstract:
carried out on aqueous temporin A and L, two short peptides belonging to an interesting class of
natural substances known to be active mainly against Gram-positive/negative bacteria and fungi.
Experimental results indicate the higher propensity of temporin L, with respect to temporin A, in
forming a-helical structures. These results were revisited by long-timescale MD simulations, in
which their a-helical propensity was investigated in the absence of trifluoroethanol. Results clearly
show the higher stability of a-helix conformations in temporin L; moreover, an interestingly strong
mechanical analogy emerges since both temporins show the same residue interval (from 7 to 10) as
the most energetically accessible for a-helix formation. Such studies provide some intriguing struc-
tural and mechanical evidence that may help in better understanding and rationalizing the confor-
mational behaviour of temporins in water solution and, ultimately, the inner principles of their
microbial targets selectivity and mechanism of action at the level of cell membranes.#2005 Wiley
Periodicals, Inc. Biopolymers 81: 215–224, 2006
Molecular dynamics (MD) simulations and circular dichroism (CD) experiments were
This article was originally published online as an accepted preprint. The ‘‘Published Online’’ date
corresponds to the preprint version. You can request a copy of the preprint by emailing the
Biopolymers editorial office at biopolymers@wiley.com
Keywords:
dichroism
temporins; antimicrobial peptides; molecular dynamics; free energy landscape; circular
4Dipartimento di Scienze e Tecnologie
Chimiche, Universita ` di Roma,
‘ ‘Tor Vergata,’ ’ Roma, Italia
5Dipartimento di Scienze Biochimiche ‘ ‘A.
Rossi Fanelli’ ’ and CNR, Universita ` ‘ ‘La
Sapienza,’ ’ Roma, Italia
6Dipartimento di Chimica, Ingegneria
Chimica e Materiali, Universita ` de L’Aquila,
L’Aquila, Italia
Received 11 July 2005;
revised 20 October 2005;
accepted 21 October 2005
Published online 31 October 2005 in Wiley InterScience
(www.interscience.wiley.com). DOI 10.1002/bip.20404
Correspondence to: M. Aschi; e-mail: aschi@caspur.it
Contract grant sponsor: Italian Ministry of Education, University
and Research.
Biopolymers, Vol. 81, 215–224 (2006)
#2005 Wiley Periodicals, Inc.
215

Page 2

INTRODUCTION
Gene-encoded antimicrobial peptides are key effec-
tors of the so-called innate immunity. An ever
increasing number of these molecules are being iso-
lated from a vast array of biological sources, either
prokaryotic and eukaryotic, including humans, which
they protect from the invasion of bacteria, protozoa,
fungi, and viruses.1,2Antimicrobial peptides display
an extreme diversity in their primary and secondary
structures, and usually have a rather large spectrum
of antibiotic activity. The low selectivity and the fast
killing of microbes are key features of the peptide-
based defenses that characterize its function as an
‘‘instant’’ immune system against microbial invaders,
as recently highlighted by Hans Boman.3This imme-
diate host response to infections plays an important
role not only in invertebrates, which exclusively
depend on it, but also in higher vertebrates, where it
comes into action before the adaptive immunity is
activated.4,5To face the challenge posed by the
spreading resistance of pathogenic microbial strains
to conventional antibiotics, the production of substi-
tute antibiotics with new activities and resistance-
avoiding properties has become an emergency.
Among the possible candidates, antimicrobial pepti-
des came recently under the spotlight as attractive
molecules to be potentially developed as therapeutic
anti-infective agents1,6and even as food preserva-
tives.7This spurred the initiation of studies aimed at
understanding their mode(s) of action.
A large body of evidence proves that killing of
microbes by antimicrobial peptides involves their ini-
tial interaction with the cytoplasmic membrane.8,9
The details of this interaction, and how this actually
leads to microbial death, however, are largely
unknown. Amphibians have proved to be an incredi-
bly rich source of antimicrobial peptides, stored in
skin granules destined for extracellular secretion.10,11
Temporins are a family of related antimicrobial pepti-
des first isolated from the skin of the European red
frog Rana temporaria.12Many other members of
this group, counting now over 40 peptides, have
later been found in several Rana species and also in
the venom of wasps.13–15Structurally, temporins are
characterized by being short (10–14 residues), by
bearing a net positive charge at neutral pH value, and
by their potential to adopt an amphipathic ?-helix
structure upon contact with membranes or when in
hydrophobic environments. The derived consensus
sequence for 36 frog-derived temporins is FLP-
LIASLLSKLL-NH2. Previously, temporins were found
to be active against Gram-positive/negative bacteria and
fungi, and to bind and permeate both artificial and
biological
information on the behavior of temporins, a wider
understanding of their modes of interaction with lipid
membranes and, more generally, of their antibacterial
mechanism, may well prove to be paradigmatic for
other short, naturally occurring peptides.
Many antimicrobial peptides, belonging to differ-
ent structural classes, present an unstructured confor-
mation in aqueous solution, and a marked increase in
secondary structure content with the assumption of an
amphipathic design usually takes place when these
molecules are transferred into a membrane-like envi-
ronment.19,20In this respect, however, a deep knowl-
edge of the structural–conformational features driv-
ing this drastic rearrangement to elicit the peptides’
antimicrobial potential is currently lacking. Computa-
tional methods may provide one of the most efficient
and reliable tools available nowadays to tackle this
important issue.21,22As a contribution, we therefore
decided to carry out molecular dynamics (MD) simu-
lations specifically addressing the behavior in water
solution of two antimicrobial peptides belonging to
the temporin family, namely temporin A (FLPL-
IGRVLSGIL) and temporin L (FVQWFSKFLGRIL).
Our approach is based on a joint application of exper-
imental (circular dihchoism, CD) measurements and
long-timescale MD simulations, with the precise aim
of evaluating the free energy conformational land-
scape of both peptides and their folding propensity in
water solution (i.e., in the absence of typical helical-
structure stabilizers such as trifluoroethanol, TFE),
looking for built-in conformational characteristics
that could plausibly rationalize the different spectrum
and level of activity on membrane-enveloped targets
recorded for temporins A and L. In particular, tempo-
rin A is preferentially active against Gram-positive
bacterial strains,13,16including some clinically impor-
tant antibiotic-resistant ones,13displays a moderately
lytic against human erythrocyte16and, as recently
shown, kills efficiently the human parasitic protozoan
Leishmania.23On the other side, temporin L has the
highest activity among all temporins studied to date
against human erythrocytes, fungi, and bacteria, in-
cluding Gram-negative strains.10,24
membranes.10,11,16–18
Besidesproviding
EXPERIMENTAL AND COMPUTATIONAL
METHODS
CD Measurements
CD measurements were carried out with a Jasco J710 spec-
tropolarimeter, equipped with a DP 520 processor, at 258C,
using a quartz cell of 2-mm path length. The peptide sam-
ples (65 ?M temporin L, 100 ?M temporin A) were pre-
216D’Abramo et al.
Biopolymers DOI 10.1002/bip

Page 3

pared in H2O–TFE solutions (0–80 % TFE, by volume).
For each sample, five spectra were recorded at the scan rate
of 20 nm/min and averaged.
MD Simulations
We performed two MD simulations of 290-ns time length
in the isochoric-isothermal (NVT) ensemble for the tempo-
rins A and L. Both the peptides were put initially in the ?-
helix conformation, at the centre of a box filled with the sin-
gle point charge (SPC) water model25at the typical water
density (55.32 mol/L). The first nanosecond was considered
as equilibration and then all the analysis included 289 ns. A
2-fs time step was used, the rototranslational motion was
removed,26the temperature was kept fixed at 300 K by the
isokinetic temperature coupling,27and the long-range elec-
trostatics was treated by means of the Particle Mesh Ewald
(PME) method.28A modified version of the Gromacs soft-
ware package29and the Gromos96 force field were used.
Note that during the simulations many unfolding/refolding
transitions occurred, possibly indicating that the initial con-
formation should not, or only poorly, influence the confor-
mational sampling of the system.
Conformational Analysis
The main difficulty arising when a conformational analysis of
a relatively large molecule is carried out is the proper defini-
tion of a ‘‘conformational coordinate,’’ i.e., a set of generalized
coordinates providing the directions in the phase space con-
necting the relevant conformational states. In this respect, a
powerful and rigorous approach is based on the essential
dynamics (ED) analysis. Although ED is widely described in
detail elsewhere,30we report some of its basic features.
Briefly, by diagonalizing either C-? or all-atoms positional
fluctuations covariance matrix, as provided by the MD simula-
tion, we obtain a set of eigenvectors and eigenvalues. The
eigenvectors represent the directions in configurational space
and the eigenvalues indicate the mean square fluctuations
along these axes. Sorting the eigenvectors by the size of the
corresponding eigenvalues, the configurational space can be
divided in a low dimensional (essential) subspace in which
most of the positional fluctuations are confined, and a high
dimensional subspace in which small and conformationally
irrelevant vibrations occur. The projection of the trajectory
onto the essential space may provide conformational free
energy (see next subsection) and, virtually, all the thermody-
namics of the peptide.
Thermodynamic Analysis
The free energy change for any transition from a reference
state ‘‘ref’’ to a generic state ‘‘i,’’ at constant volume and
temperature, can be calculated from the probabilities p
(obtained by the MD simulation) of finding the system in
both states ‘‘i’’ and ‘‘ref’’
?Aref;i¼ ?RT lnpi
pref
ð1Þ
where R is the ideal gas constant and T the (absolute) tem-
perature.
Moreover, combining the internal energy change Uref,i
¼ Ui? Uref(obtained averaging over the MD frames asso-
ciated to the ‘‘i’’ and ‘‘ref’’ states) we may also evaluate the
corresponding entropy variation via
?Sref;i¼?Uref;i? ?Aref;i
T
ð2Þ
In this article, the above equations have been used for eval-
uating the thermodynamics in the conformational space
(hereafter called essential plane) defined by the first two
essential eigenvectors obtained by the ED analysis. In this
case the reference (‘‘ref’’) condition was taken, for each
peptide, as the one corresponding to the absolute free-
energy minimum, i.e., it was evaluated a posteriori on the
basis of the resulting free energy landscape.
The same equations also provided the overall folding
thermodynamics, as obtained by MD simulation data. Thus,
we deliberately evaluated the global thermodynamic
changes due to the transitions from completely unfolded
(reference state) condition to each of the conformational
states defined by an increasing number n of residues in ?-
helix conformation (the minimum value of n was therefore
set at 4). Note that the choice of the reference state does not
obviously alter the thermodynamic picture. It is also impor-
tant to further stress that we are following the free energy
change for arranging at least four residues in the ?-helix
conformation. Therefore, for unfolded condition we indi-
cate the ensemble of structures in which such a condition is
not fulfilled. On the basis of the MD results (vide infra), we
found that the first two eigenvalues of the all-atoms cova-
riance matrix could account for the largest portion of the
phase space. For this reason, all the above analyses were
carried out on a bidimensional essential space hereafter
called ‘‘essential plane.’’
RESULTS AND DISCUSSION
Circular Dichroism
CD spectroscopy was used to determine the confor-
mation of synthetic temporin L and temporin A in
solution by recording the spectra in water and after
addition of TFE. Temporin A spectra were compara-
ble to those obtained for the same peptide by Wade
and colleagues under similar conditions, and pub-
lished elsewhere.13The experiments demonstrate that
for both peptides an increase in TFE concentration
caused a progressive change from a random coil to an
?-helical structure (Figure 1). In the case of temporin
L, the effect was almost complete at about 20% TFE
Temporin A and Temporin L in Aqueous Solution217
Biopolymers DOI 10.1002/bip

Page 4

(Figure 1a). This is somehow at variance from what
was seen with temporin A, which, although it dis-
played similar spectra with a minimum close to
200 nm, exhibited the maximal effect at 30% TFE
(Figure 1b). Thus, in temporin L the gradual change
from random coil to an ?-helical conformation was
significantly enhanced, and the peptide became
almost completely structured at a lower TFE/H2O
ratio. This higher propensity to adopt an ordered con-
formation even in a relatively poor hydrophobic sol-
vent, which clearly represents an intriguing difference
between the two temporins, prompted us to investi-
gate the thermodynamics as well as the energetics
associated to the dynamics of the peptides in solution.
In particular, MD simulations have been used in order
to evaluate the ‘‘intrinsic’’ ability of the two tempo-
rins in forming ?-helices—thus their folding propen-
sity even in the lack of TFE.
Structural Motions
In this section we show the results of the ED analysis
on the trajectories of temporins A and L, with the pre-
cise aim of identifying the main peptide internal
motions in water solution. The diagonalization of the
covariance matrix provides a set of eigenvectors and
eigenvalues, corresponding to generalized conforma-
tional coordinates and fluctuations. Among them,
only a small fraction is typically associated with sig-
nificant internal motions of the system, i.e., the corre-
sponding eigenvalues are significantly different from
zero. The all-atom eigenvalues spectrum of temporin
L (Figure 2) clearly shows that the first ten eigenvec-
tors are responsible for large part of the internal
motions (the same is observed for C-?). Similar
results were obtained for temporin A (data not
shown). The link between each eigenvector and the
relevant atomic structural motions can be studied by
analyzing the corresponding atomic components. The
results are reported in Figures 3 and 4.
In Figure 3a, the atom composition of the first two
(all-atoms) eigenvectors of temporin L shows that
these eigenvectors provide concerted motions mainly
involving the terminal residues (Phe1, Leu13) as well
as Gln3 and Lys7. Interestingly, the same analysis
conducted for the first two C-? eigenvectors (see Fig-
ure 3b) shows that Gln3 and Lys7 are associated to
almost zero components, whereas the terminal resi-
dues still correspond to high peaks. These results
indicate that the structural fluctuations involving
Gln3 and Lys7 in temporin L are mainly due to side-
chain motions. In the case of temporin A (Figure 4a),
the first two all-atoms eigenvectors largely involve
terminal motions (i.e., high values of the correspond-
ing components), but also with a significant compo-
nent associated with the Arg7 side chain (see Figure
4b). At variance with temporin L, the first two C-?
eigenvectors show the presence of four clearly sepa-
rated blocks, namely the terminal residues Leu4,
Ile5–Val8, and Leu9–Ser10 (see Figure 4b).
FIGURE 1
(b) in water and TFE.
CD spectra of temporin L (a) and temporin A
FIGURE 2
matrix from MD simulations of temporin L.
Eigenvalues of the all-atom covariance
218 D’Abramo et al.
Biopolymers DOI 10.1002/bip

Page 5

On the basis of these preliminary results, some
‘‘dynamical’’ differences between the two temporins are
clearly evident. Therefore, a more detailed thermody-
namic inspection on the essential plane—as described in
the methodological section—was carried out.
Thermodynamics
The 300 K Helmholtz free energy surface of temporin
A, as a function of the position in the essential plane,
shows the absolute minimum [which provides the
‘‘ref’’ conformation in Eq. (1)] in the centre of the
plane, and two local minima just 3–4 kJ/mol higher
(Figure 5). In Figure 6 we report the corresponding
300 K entropy surface as provided by Eq. (2). It is
interesting to observe that the two local free energy
minima are associated with entropy values signifi-
cantly lower than the absolute free energy minimum
one. At the same time, a large part of the accessible
plane, corresponding to a relatively high free energy,
is associated to higher entropy values. This finding
clearly indicates that temporin A free energy minima
are mainly determined by the internal energy (a phys-
ical condition apparently normal but not always pre-
dominant in biological molecules in solution; see
Ref. 31), and that the structures characterized by
higher entropy are not thermodynamically stable.
Differently from temporin A, temporin L shows a
behavior characterized by a ‘‘corrugated’’ free energy
surface with several minima (Figure 7). However,
similarly to temporin A, all these free energy minima
are associated to low entropy regions (Figure 8).
We show therefore that both temporins undergo a
rather typical internal energy-driven conformational
sampling in which high-entropy (highly disordered)
structures do not represent accessible states in these
conditions. In other words, temporins A and L both
FIGURE 3
all-atoms eigenvectors of temporin. (b) Absolute component
values of the first two C-? eigenvectors of temporin L.
(a) Absolute component values of the first two
FIGURE 4
all-atoms eigenvectors of temporin. (b) Absolute component
values of the first two C-? eigenvectors of temporin A.
(a) Absolute component values of the first two
Temporin A and Temporin L in Aqueous Solution219
Biopolymers DOI 10.1002/bip

Page 6

preferentially exist in partially organized (not fully
‘unfolded’) structures characterized by low internal
energy values but also relatively low entropy values.
This finding clearly required a more accurate inspec-
tion; we therefore calculated the global ?-helix forma-
tion free energy for an increasing number of residues
(from 4 to 13). The result is reported in Figure 9.
It is important to underline that a conformational
state is typically considered as ‘‘folded’’ when at least
six residues are organized in a helix structure. In Fig-
ure 9 it is evident that both temporins show in water
solution a positive ?-helix formation free energy (i.e.,
essentially a thermodynamic instability) when six or
more residues are involved. At the same time, it is
FIGURE 5
plane (nm).
Temporin A: 300 K Helmholtz free energy (kJ/mol) map on the all-atoms essential
FIGURE 6
Temporin A: 300 K entropy (J/mol K) map on the all-atoms essential plane (nm).
220D’Abramo et al.
Biopolymers DOI 10.1002/bip

Page 7

also evident from the same figure that temporin L is
in all cases characterized by an ?-helix formation free
energy systematically lower than temporin A, irre-
spective of the number of residues involved. Further-
more, when 4 or 5 residues are considered, temporin
L also shows a slightly negative helix formation free
energy. This finding may provide some additional
information on the actual conformational state of the
peptides. In particular, temporin L seems to exist
basically in a ‘‘semifolded’’ (low-entropy) state char-
acterized by a 4(5)-residue helix, which nicely con-
firms and strengthens the results shown in Figures 7
and 8. Moreover, the temporin L ‘‘dynamical’’ finger-
print reported in Figure 3b, characterized by a high
fluctuation almost exclusively confined within the ter-
minal, conceivably less ‘‘organized’’ (vide infra) resi-
dues, can also be explained at the light of these
results.
In conclusion, both peptides (in water at 300 K) do
not show any ?-helix conformation, even though
temporin L, consistent with the experimental evi-
dence obtained by CD spectroscopy (section 3.1),
FIGURE 7
plane.
Temporin L: 300 K Helmholtz free energy (kJ/mol) map on the all-atoms essential
FIGURE 8
Temporin L: 300 K entropy (J/mol K) map on the all-atoms essential plane.
Temporin A and Temporin L in Aqueous Solution 221
Biopolymers DOI 10.1002/bip

Page 8

shows a higher propensity to form stable ?-helices in
water compared to temporin A.
We also evaluated which residues specifically con-
tribute to the ?-helix formation. The interesting
aspect emerging from this analysis is that both pepti-
des are characterized by a topologically identical
region of ?-helix aggregation, as shown in Figure 10.
More specifically, the first structural organization,
that is the less work-intensive combination of 4 resi-
dues, involves the same positions along the sequence
even in the presence of different residues. This result
may provide, as far as we know for the first time, an
intriguing insight into the folding mechanism, charac-
terized by the presence of a core group of residues
triggering the folding (i.e., the residues from 7 to 10)
and whose thermodynamic accessibility drives the
global folding propensity and, possibly, the related
antimicrobial activity.
Indeed, it is conceivable that the greater antimicro-
bial and hemolytic activity exhibited by temporin L
could be ascribed to its higher propensity to assume a
folded conformation. In other words, the presence of
a partially folded structure in water solution may
plausibly facilitate, both thermodynamically and
kinetically, the peptide folding in the microbial mem-
brane. In addition, the higher positive charge (þ3)
possessed by temporin L could enhance its initial
binding to the negatively charged outer plasma mem-
brane of bacterial targets. Although both temporins
are too short to span the membrane bilayer and thus
form a simple transmembrane pore composed of an
helical cluster (the classic ‘‘barrel-stave’’ model), a
wealth of data indicates that temporins can insert
into and damage the cellular membrane as part of
their killing mechanism.10,11,16–18A number of other
models (e.g., the ‘‘carpet’’ and ‘‘sinking-raft’’ models)
have been proposed to account for peptide-induced
FIGURE 9
of temporin A (solid line) and temporin L (dashed line) as a
function of the number of folded residues.
The 300 K Helmholtz free energy of folding
FIGURE 10
90% of the trajectory are reported in bold, with the exception of the ones whose fraction is explic-
itly indicated) as emerged from MD simulations of temporin A (left) and L (right). The sequences
are reported according to the corresponding free energy of folding (see Figure 9).
The ?-helix forming residues (only the residues found to be present for more than
222 D’Abramo et al.
Biopolymers DOI 10.1002/bip

Page 9

membrane permeabilization/disruption, and envisage
at first the absorption of the peptide to the membrane
with the helices parallel to the surface, and a subse-
quent mechanism of membrane destabilization that
could in principle adapt also to temporins and other
short linear peptides.32,33Whereas it is not yet possi-
ble to indicate precisely the molecular mechanism of
interaction of temporins with lipid membranes, the
present investigation suggests that the ‘‘inherent’’
structural features, which ultimately depend on the
specific sequences, might be an essential determinant
of the biological activities of these antimicrobial pep-
tides.
CONCLUSIONS
CD spectroscopy was used to determine the confor-
mation of synthetic temporin L and temporin A in
solution. Analysis of the resulting patterns indicates
that temporin L displays a higher propensity to
acquire the ?-helix conformation. Long-timescale
MD simulations carried out on both temporins in
aqueous solutions confirm the experimental observa-
tions, clearly showing that, in the case of temporin L,
?-helix formation free energy is always lower than
that of temporin A. A more careful thermodynamic
analysis indicates that both peptides exist, in aqueous
solution, in a not completely random coil conforma-
tion, even if a positive Helmholtz free energy is found
for the structural rearrangement into an ?-helix con-
taining at least six residues. The greater ?-helix pro-
pensity, together with the higher net positive charge,
exhibited by temporin L may provide some quantita-
tive key aspects for proposing models of action plau-
sibly explaining its efficacy against selected micro-
bial targets.
AB wishes to acknowledge the financial support received
from the Italian Ministry of Education, University and
Research (PRIN 2004). The authors also acknowledge the
‘‘Consorzio per le Applicazioni di Supercalcolo per Univer-
sita ` e Ricerca (CASPUR-Rome)’’ for the computational
facilities.
REFERENCES
1. Zasloff, M. Nature 2002, 415, 389–395.
2. Bulet, P.; Stocklin, R.; Menin, L. Immunol Rev 2004,
198, 169–184.
3. Boman, H. G. Annu Rev Immunol 1995, 13, 61–92.
4. Boman, H. G. J Intern Med 2003, 254, 197–215.
5. Kimbrell, D. A.; Beutler, B. Nat Rev Genet 2001, 2,
256–267.
6. Hancock, R. E. W. Lancet Infect Dis 2001, 1, 156–
164.
7. Papagianni, M. Biotechnol Adv 2003, 21, 465–499.
8. McElhaney, R. N.; Prenner, E. J. Biochim Biophys
Acta 1999, 1462, 1–234.
9. Menestrina, G.; Dalla Serra, M.; Lazarovici, P., Eds.
Pore-Forming Peptides and Protein Toxins; Taylor
Francis: London and New York, 2003.
10. Rinaldi, A. C.; Mangoni, M. L.; Rufo, A.; Luzi, C.;
Barra, D.; Zhao, H.; Kinnunen, P. K. J.; Bozzi, A.; Di
Giulio, A.; Simmaco, M. Biochem J 2002, 368, 91–
100.
11. Rinaldi, A. C. Curr Opin Chem Biol 2002, 6, 799–804.
12. Simmaco, M.; Mignogna, G.; Canofeni, S.; Miele, R.;
Mangoni, M. L.; Barra, D. Eur J Biochem 1996, 242,
788–792.
13. Wade, D.; Silberring, J.; Soliymani, R.; Heikkinen, S.;
Kilpela ¨inen, I.; Lankinen, H.; Kuusela, P. FEBS Lett
2000, 479, 6–9.
14. Wade, D.; Silveira, A.; Silberring, J.; Kuusela, P.; Lan-
kinen, H. Protein Pept Lett 2000, 7, 349–357.
15. Wade, D. Internet J Chem 5, 2002, http://www.ijc.com/
articles/2002v5/5/
16. Mangoni, M. L.; Rinald, A. C.; Di Giulio, A.; Migno-
gna, G.; Bozzi, A.; Barra, D.; Simmaco, M. Eur J Bio-
chem 2000, 267, 1447–1454.
17. Rinaldi, A. C.; Di Giulio, A.; Liberi, M.; Gualtieri, G.;
Oratore, A.; Schinina `, M. E.; Simmaco, M.; Bozzi, A.
J Pept Res 2001, 58, 213–220.
18. Zhao, H.; Rinaldi, A. C.; Di Giulio, A.; Simmaco, M.;
Kinnunen, P. K. J. Biochemistry 2002, 41, 4425–
4436.
19. Vignal, E.; Chavanieu, A.; Roch, P.; Chiche, L.;
Grassy, G.; Calas, B.; Aumelas, A. Eur J Biochem
1998, 253, 221–228.
20. Raimondo, D.; Andreotti, G.; Saint, N.; Amodeo, P.;
Renzone, G.; Sanseverino, M.; Zocchi, I.; Molle, G.;
Motta, A.; Scaloni, A. Proc Natl Acad Sci USA 2005,
102, 6309–6314.
21. Roccatano, D.; Colombo, G.; Fioroni, M.; Mark, A. E.
Proc Natl Acad Sci USA 2002, 99, 12179–12184.
22. D’Alessandro, M.; Paci, M.; Amadei, A. Biopolymers
2004, 74, 448–456.
23. Mangoni, M. L.; Saugar, J. M.; Dellisanti, M.; Barra,
D.; Simmaco, M.; Rivas, L. J Biol Chem 2005, 280,
984–990.
24. Mangoni, M. L.; Papo, N.; Barra, D.; Simmaco, M.;
Bozzi, A.; Di Giulio, A.; Rinaldi, A. C. Biochem J
2004, 380, 859–865.
25. Amadei, A.; Chillemi, G.; Ceruso, M. A.; Grottesi, A.;
Di Nola, A. J Chem Phys 2000, 112, 9–23.
26. Evans, D. J.; Morriss, G. P. Statistical Mechanics of Non-
equilbrium Liquids; Academic Press: London, 1990.
27. Darden, T.; York, D.; Pedersen, L. J Chem Phys 1993,
98, 10089–10092.
28. Berendsen, H. J. C.; Postma, J. P. M.; van Gunsteren,
W. F.; Hermans, J. In Intermolecular Forces; Pullman,
B., Ed.; Reidel Publishing Company: Dordrecht, 1981;
pp 331–342.
Temporin A and Temporin L in Aqueous Solution 223
Biopolymers DOI 10.1002/bip

Page 10

29. van der Spoel, D.; Berendsen, H. J. C.; van Buuren, A. R.;
Apol, M. E. F.; Meulenhoff, P. J.; Sijbers, A. L. T. M.;
van Drunen, R. GROMACS User Manual. University
of Groningen, Groningen, The Nederlands, 1995.
30. Amadei, A.; Linssen, A. B. M.; Berendsen, H. J. C.
Proteins 1993, 17, 412–425.
31. Di Teodoro, C.; Aschi, M.; Amadei, A.; Roccatano, D.;
Malatesta, F.; Ottaviano, L. ChemPhysChem 2005, 6,
681–689.
32. Pokorny, A.; Almeida, P. F. F. Biochemistry 2004, 43,
8846–8857.
33. Lohner, K.; Blondelle, S. E. Comb Chem High
Throughput Screen 2005, 8, 241–256.
Reviewing Editor: Nancy Stellwagen
224D’Abramo et al.
Biopolymers DOI 10.1002/bip