Synthesis, characterization, and photoinduced antibacterial activity of porphyrin-type photosensitizers conjugated to the antimicrobial peptide apidaecin 1b

Article abstract of DOI:10.1021/jm301509n

Antimicrobial photodynamic therapy (aPDT) is an emerging treatment for bacterial infections that is becoming increasingly more attractive because of its effectiveness against multi-antibiotic-resistant strains and unlikelihood of inducing bacterial resistance. Among the strategies to enhance the efficacy of PDT against Gram-negative bacteria, the binding to a cationic antimicrobial peptide offers the attractive prospect for improving both the water solubilty and the localization of the photoactive drug in bacteria. In this work we have compared a number of free and apidaecin-conjugated photosensitizers (PSs) differing in structure and charge. Our results indicate that the conjugation of per se ineffective highly hydrophobic PSs to a cationic peptide produces a photosensitizing agent effective against Gram-negative bacteria. Apidaecin cannot improve the phototoxic activity of cationic PSs, which mainly depends on a very high yield of singlet oxygen production in the surroundings of the bacterial outer membrane. Apidaecin-PS conjugates appear most promising for treatment protocols requiring repeated washing after sensitizer delivery.

Full text of DOI:10.1021/jm301509n

Article
pubs.acs.org/jmc
Synthesis, Characterization, and Photoinduced Antibacterial Activity
of Porphyrin-Type Photosensitizers Conjugated to the Antimicrobial
Peptide Apidaecin 1b
Ryan Dosselli,†,∥ Cristiano Tampieri,‡ Ruben Ruiz-Gonzalez,§ Sonia De Munari,‡,⊥ Xavier Ragas,§
́
́
̀
David Sanchez-García,§ Montserrat Agut,§ Santi Nonell,§ Elena Reddi,† and Marina Gobbo*,‡
́
†Department of Biology, University of Padova, via U. Bassi 58/B, I-35121 Padova, Italy
‡Department of Chemical Sciences, University of Padova, via F. Marzolo 1, I-35131 Padova, Italy
§Institut Químic de Sarria,
̀
Universitat Ramon Llull, Vía Augusta 390, E-08017 Barcelona, Spain
S
* Supporting Information
ABSTRACT: Antimicrobial photodynamic therapy (aPDT) is
an emerging treatment for bacterial infections that is becoming
increasingly more attractive because of its effectiveness against
multi-antibiotic-resistant strains and unlikelihood of inducing
bacterial resistance. Among the strategies to enhance the
efficacy of PDT against Gram-negative bacteria, the binding to
a cationic antimicrobial peptide offers the attractive prospect
for improving both the water solubilty and the localization of
the photoactive drug in bacteria. In this work we have
compared a number of free and apidaecin-conjugated
photosensitizers (PSs) differing in structure and charge. Our
results indicate that the conjugation of per se ineffective highly hydrophobic PSs to a cationic peptide produces a photosensitizing
agent effective against Gram-negative bacteria. Apidaecin cannot improve the phototoxic activity of cationic PSs, which mainly
depends on a very high yield of singlet oxygen production in the surroundings of the bacterial outer membrane. Apidaecin−PS
conjugates appear most promising for treatment protocols requiring repeated washing after sensitizer delivery.
bacteria, are able to kill Gram-negative species.9−11 An
INTRODUCTION
■
alternative approach to improve the susceptibility of Gram-
negative bacteria to the photodynamic action of neutral
porphyrins involves the covalent attachment of the PS to a
polymer molecule containing basic amino groups12 or, as we
recently proposed, to a cationic antimicrobial peptide.13
Cationic antimicrobial peptides (CAMPs) are components of
the innate defense of many organisms, and they are being
considered a promising source of new antibiotics.14 In addition
to exerting direct antimicrobial effects, many studies have
documented their ability to affect various host cell activity and
to have a key modulatory role in the innate immune
response.44,45 Their overall positive charge ensures accumu-
lation at the polyanionic microbial cell surfaces that contain
acidic polymers, such as lipopolysaccharides, and wall-
associated teichoic acids in Gram-negative and Gram-positive
bacteria, respectively. Beyond the presence of several cationic
amino acids, a substantial proportion of hydrophobic amino
acid residues permit most of the CAMPs to fold into an
amphipathic structure, which allows them to insert into the
phospholipid bilayer of the cell membranes. After insertion,
antimicrobial peptides act by either disrupting the physical
The global diffusion of new antimicrobial infections as well as
the continuously increasing resistance of pathogens against
many of the commonly used antibiotics imposes a considerable
effort to develop alternative therapies to the use of classical
drugs. In this area antimicrobial photodynamic therapy (aPDT;
also termed photodynamic antimicrobial chemotherapy,
PACT) represents a very promising strategy, particularly for
the treatment of superficial and localized infectious diseases.1
The PDT concept comprises the action of three components: a
photosensitizer (PS), a light source of appropriate wavelength,
and oxygen. The interaction between light and the PS leads to
the generation of reactive oxygen species (ROS), e.g., singlet
oxygen, by two possible mechanisms involving either electron-
transfer (type I) or energy-transfer (type II) reactions.2 These
ROS are highly reactive and can damage a variety of cellular
components, e.g., proteins, nucleic acids, and lipids, resulting in
cytotoxicity.3−5 Advantages of aPDT over traditional antibiotics
include a broad-spectrum activity, also against antibiotic-
resistant species,6 and the lack of development of resistance
mechanisms due to the multitarget process.7,8
Porphyrins, commonly used as PSs in PDT, can efficiently
kill Gram-positive bacteria, whereas only cationic PSs, or
noncationic PSs in combination with agents that permeabilize
the highly organized outer membrane of Gram-negative
Received: October 17, 2012
Published: December 12, 2012
© 2012 American Chemical Society
1052
dx.doi.org/10.1021/jm301509n | J. Med. Chem. 2013, 56, 1052−1063

Journal of Medicinal Chemistry
Article
Figure 1. Structures of the photosensitizers and their peptide conjugates.
integrity of the membrane or translocating across the
membrane to hit internal bacterial targets. This multitarget
mode of action promises both low susceptibility to antibiotic
resistance and a broad spectrum of activity against a variety of
microorganisms. Among CAMPs, the family of short proline−
arginine-rich peptides attracts particular interest because of
some unique features, such as a higher activity against Gram-
negative bacteria, a relative stability against proteolysis, and a
very low toxicity against mammalian cells.15 Apidaecin 1b, an
insect 18-residue-long peptide belonging to this family, is
effective against a large number of Gram-negative bacteria and a
few Gram-positive bacteria,16 where it acts by a non-pore-
forming mechanism only partially elucidated.17,18 Mutagene-
sis19,20 and structure−activity relationship studies21−24 have
identified the C-terminal half of apidaecin as essential for its
antimicrobial activity, and several studies have shown that the
peptide is able to translocate a fluorescent tag into a bacterial
cell.22,24−26 In a preliminary communication we have shown
that the conjugation of apidaecin to 5-(4′-carboxyphenyl)-
10,15,20-triphenylporphyrin affords a new water-soluble PS (1c
in Figure 1) that, upon light activation, is able to kill both
Gram-positive and Gram-negative bacteria at concentrations at
which the antimicrobial peptide and, on Gram-negative
bacteria, the porphyrin alone are not effective.13 Herein we
report the synthesis, characterization, and phototoxicity studies
against Escherichia coli and methicillin-resistant Staphylococcus
aureus (MRSA) of new conjugates in which apidaecin and its C-
terminal octapeptide were modified at the N-terminus with
neutral or charged porphyrin and porphycene PSs.
RESULTS
■
Synthesis and Characterization of the Conjugates.
Porphyrins 1b and 3b, properly functionalized for the covalent
binding to the peptide, were prepared by established methods
using the Lindsey27 and Adler−Longo conditions,28 respec-
tively. We synthesized three new PS−peptide conjugates in
which the porphycene 2b or the cationic porphyrin 4b was
covalently linked to the N-terminal end of the antimicrobial
peptide apidaecin and 5-(4′-carboxyphenyl)-10,15,20-triphenyl-
porphyrin (1b) to its C-terminal segment (Figure 1). The
trimethylated porphyrin 4b was prepared from 3b by treatment
with an excess of methyl iodide in DMF, whereas alkaline
hydrolysis of the porphycene 2a released the functional group
from the subsequent conjugation. The synthesis of 2a has been
reported elsewhere.29 The side-chain-protected peptide se-
quences were automatically synthesized on a solid phase by the
standard Fmoc protocol,30 and after removal of the N-terminal
amino protecting group, the selected PS (1−4b, 2.5 equiv) was
coupled to the peptide chain using diisopropylcarbodiimide/N-
hydroxybenzotriazole as the activating agent. The yields of the
coupling reactions were in the range of 75−80% (based on
HPLC) except in the case of the cationic porphyrin 4b, which
apparently did not react with the peptide even using different
coupling reagents and reaction conditions. The conjugate
between apidaecin and this cationic PS was obtained from 3c,
side-chain-protected and still attached to the solid support, by
treatment with an excess of methyl iodide. Besides on pyridine
nitrogens, methylation also occurred on the π-nitrogen of the
histidine side chain, which usually remains unprotected during
1053
dx.doi.org/10.1021/jm301509n | J. Med. Chem. 2013, 56, 1052−1063

Journal of Medicinal Chemistry
Article
the peptide chain assembly. After cleavage and deprotection
from the solid support, the conjugates 1c, 1d, 2c, and 4e were
purified by reversed-phase HPLC and characterized by
analytical HPLC, electrospray mass spectrometry, and UV−
vis absorption spectroscopy.
Absorption and Fluorescence of the Conjugates. The
spectroscopic and photophysical properties of the PSs and their
conjugates were measured in aqueous and organic media and in
cell suspensions to assess the structural and environmental
effects as well as their correlation with antibacterial activity.
Methanol. Figures 2 and 3 show the absorption and
fluorescence spectra of the free PSs and of their peptide
conjugates in methanol. While the spectra of porphyrins are
essentially insensitive to conjugation (panels A, C, and D), clear
changes can be observed for the porphycene (panel B). In turn,
the fluorescence quantum yield does not change appreciably for
the porphyrins, while it drops by ca. 50% for the porphycene
(Table 1). Finally, all porphyrins show monoexponential
fluorescence decay kinetics, and conjugation does not change
the lifetime values either (Supporting Information, Figure
S1A,C,D). Again, the situation is different for the porphycenes
in that the conjugate 2c shows biexponential kinetics unlike the
free porphycene 2a (Figure S1B), and on average the singlet
state decays faster (Table 1).
Aqueous Solutions (PBS). Porphyrin 4b is water-soluble as a
consequence of its positively and negatively charged groups. It
remains water-soluble after conjugation (4e) with small but
clear shifts in the position of the Soret and Q absorption bands
and changes in their relative intensities (Figure 2E). The
fluorescence spectrum of 4b shows a single, structureless broad
band (Figure 3E), a behavior strikingly different from that in
methanol but in line with that of the related tetracationic meso-
tetrakis(N-methylpyridinium-yl)porphyrin (TMPyP).32 Conju-
gation to the peptide (4e) leads to partial recovery of the two
well-resolved fluorescence bands observed in methanol. The
fluorescence decay kinetics of 4b is monoexponential, albeit
with a lifetime much shorter than that in methanol. For the
conjugate 4e two decay components can be observed, whose
lifetimes are close to those of 4b in methanol and in PBS,
respectively (Figure S1E, Supporting Information, and Table
1). The kinetics is independent of concentration over 3 orders
of magnitude (Figure S2A and Table S1, Supporting
Information).
On the other hand, porphyrin 1b and porphycene 2a are
insoluble in water, and therefore, no fluorescence can be
recorded in this solvent. Conjugation to the peptides (1c, 1d,
and 2c, respectively) renders them water-soluble, but the
spectroscopic and photophysical properties change substantially
relative to those in methanol: the absorption spectra show
broadening of the Soret band and loss of structure in the Q
region (Figure 2F−H), and the fluorescence is dramatically
quenched. In addition, the fluorescence spectra of 1c and 1d
are slightly red-shifted (Figure 3G,H), and the decays are
biexponential (Figure S1G,H, Supporting Information, and
Table 1) and show a clear concentration trend (Figure S2B,C
and Table S1, Supporting Information).
Figure 2. Absorption spectra in methanol (A−D), PBS (E−H), and E.
coli suspensions (I−L), normalized to facilitate their comparison. The
concentration is 5 μM for all compounds. Methanol: (A) 4e (red) and
4b (blue); (B) 2c (red) and 2a (blue; dichloromethane as solvent);
(C) 1d (red) and 1b (blue); (D) 1c (red) and 1b (blue). PBS: (E) 4e
(red) and 4b (blue); (F) 2c (red) and 2a (blue); (G) 1d (red) and 1b
(blue); (H) 1c (red) and 1b (blue). E. coli suspensions (green): (I)
4e; (J) 2c; (K) 1d; (L) 1c. Spectra in methanol (red) and PBS (blue)
are given for comparison.
Finally, the porphycene conjugate 2c shows very similar
absorption spectra in PBS and in E. coli suspensions (Figure
2J). However, while we could record no fluorescence in PBS,
we were nevertheless able to observe extremely weak
biexponential fluorescence decay in the cells (Figure 3, Figure
S1J, Supporting Information, and Table 1).
E. coli Suspensions. The conjugate 4e shows in cell
suspensions the same absorption and fluorescence properties
as it does in PBS (Figures 2I and 3I and Table 1). On the other
hand, there are evident changes for conjugates 1c and 1d in cell
suspensions relative to PBS, particularly in the absorption
spectrum and in the fluorescence kinetics, which shows a third
decay component not present in PBS or in methanol (Table 1).
Singlet Oxygen Production and Decay. All free and
conjugated PSs were able to photosensitize the formation of
1O2 in methanol as evidenced by its phosphorescence at 1275
1
nm. The quantum yields of O2 production (ΦΔ) were in the
0.6−0.7 range for the porphyrins and the 0.1−0.3 range for the
1054
dx.doi.org/10.1021/jm301509n | J. Med. Chem. 2013, 56, 1052−1063

Journal of Medicinal Chemistry
Article
(Figure S3, Supporting Information) reveals that production of
singlet oxygen by conjugate 4e is as fast as that by the free
porphyrin 4b (Table 2), yet for the conjugates 1c, 1d, and 2c it
is 1 order of magnitude slower. On the other hand, the lifetime
1
of O2 for all conjugates in deuterated PBS is shorter than the
value expected in this solvent (67 μs),34 which is actually
observed only for the free porphyrin 4b. When E. coli
1
suspensions were studied, the O2 signals showed essentially
the same pattern as that in PBS solutions (Figure S4,
Supporting Information, and Table 2).
Circular Dichroism (CD) Studies. The conformational
properties of the conjugates were investigated by CD
spectroscopy in different environments, including water and
membrane-mimicking solvents, and compared to those of the
parent peptide. As a consequence of the high proline content (6
out of 18 residues), in water apidaecin assumes a prevalently
disordered, extended structure (Figure 4A), and in membrane-
mimicking environments it exists as a mixture of conformers
with a high percentage of nonrepetitive bent structures, most
probably β-turns (Figure 4C,D).23,25 The CD spectra of the
conjugates in water are characterized by a broad negative band
around 200 nm, much more intense for those with the cationic
porphyrin than for those with the neutral PS (Figure 4A).
Moreover, 1c and 2c showed a split Cotton effect in the Soret
band region (Figure 4B), indicating that porphyrins are chirally
oriented and close to one another in space.35 These results
suggest that while 4e can assume in water a fully extended
structure, conjugates 1c and 2c show the tendency to form
aggregates with the peptide chain probably folded over the
hydrophobic porphyrin platform36 and the porphyrins close to
one another to reduce the exposure to the solvent. Aggregation
was confirmed by a change in the UV spectra of these
conjugates moving from methanol to water. In organic solvent
(2,2,2-trifluoroethanol, TFE) and in the presence of SDS
micelles, the CD spectrum of the conjugates is similar to that of
the parent peptide, with a broad band at 202 nm, a shoulder at
220 nm, and comparable intensities (Figure 4C,D). This
suggests that in membrane-mimicking environments the
conformational preferences of the peptide are minimally
affected by the intramolecular interactions with the PS.
Photoinactivation of E. coli. To determine the photo-
sensitizing efficiency of our PSs and peptide−PS conjugates
against Gram-negative bacteria, E. coli suspensions were
incubated in the dark for 60 min with different concentrations
(1.5, 5, 10, and 15 μM) of the agents and then illuminated
(light dose of 36 J/cm2 of red light for 2a and 2c, 13.5 J/cm2 of
blue light for all other compounds) with or without washings.
At the concentrations used, apidaecin alone did not cause any
decrease of E. coli survival,13as did its C-terminal segment 1d, in
both the dark and light conditions (Table S3, Supporting
Information). In addition, neither 1b nor 2a caused any
bacteria photokilling in their unconjugated form. All conjugates
exhibited markedly concentration-dependent abilities to kill E.
coli under illumination with an efficiency that largely depended
on the type of conjugate (Figure 5A). Indeed, considering the
unwashed samples, 4b and 4e caused complete killing of
bacteria at 5 and 10 μM concentrations, respectively, while all
other agents were considerably less potent and at 15 μM
reduced the survival of E. coli by 4 log at maximum. Moreover,
it is interesting to notice that 1c was slightly more efficient than
1d and caused a bacterial mortality similar to that of 2c.
However, when illumination was carried out after three
washings of the cells to remove the unbound PS (Figure 5B;
Figure 3. Emission spectra in methanol (A−D), PBS (E−H), and E.
coli suspensions (I−L), normalized to facilitate their comparison. The
concentration is 5 μM for all compounds. Methanol: (A) 4e (red) and
4b (blue); (B) 2c (red) and 2a (blue; dichloromethane as solvent);
(C) 1d (red) and 1b (blue); (D) 1c (red) and 1b (blue). PBS: (E) 4e
(red) and 4b (blue); (F) 2c (red) and 2a (blue); (G) 1d (red) and 1b
(blue); (H) 1c (red) and 1b (blue). E. coli suspensions (green): (I)
4e; (J) 2c; (K) 1d; (L) 1c. Spectra in methanol (red) and PBS (blue)
are given for comparison.
porphycenes (Table 2). The kinetics of 1O2 production
matched the results of laser flash photolysis experiments for
the triplet PS decay (see the Supporting Information, Table
S2). Likewise, an excellent match was found between the
1
observed lifetime of O2 and the literature values. In PBS, only
porphyrins 4b and 4e retained their high ability to produce
1O2; all other compounds experienced a decrease of the ΦΔ
value by 20−30-fold (porphyrins 1c and 1d) or even 100-fold
1
(porphycene 2c; Table 2). Inspection of the kinetics of O2
1055
dx.doi.org/10.1021/jm301509n | J. Med. Chem. 2013, 56, 1052−1063

Journal of Medicinal Chemistry
Article
Table 1. Fluorescence Properties of the Peptide Conjugates and Model Compounds in Methanol, PBS, and E. coli Suspensions
(Fractional Amplitudes in Parentheses)
λF,max/nm
ΦF
τS/ns
a
b
compd
MeOH
PBS
E. coli
MeOH
PBS
ns
MeOH
10.1
PBS
E. coli
c
1b
1c
647
647
ns
0.040
0.050
ns
651
652
ns
651
0.006
9.8
10.7 (0.82)
3.1 (0.18)
10.5 (0.46)
7.5 (0.33)
3.1 (0.21)
5.9 (0.11)
4.9 (0.48)
2.5 (0.41)
1d
646
652
0.044
0.006
9.8
6.1 (0.61)
2.6 (0.39)
d
d
2a
2c
715
697
0.030
0.016
1.46
<0.0001
0.9 (0.94)
9.6 (0.06)
7.9
0.9 (0.75)
5.2 (0.25)
4b
4e
656
655
675
660
0.022
0.024
0.008
0.018
4.2
660
8.1
4.2 (0.49)
4.2 (0.48)
7.1 (0.52)
7.1 (0.51)
a
b
c
d
Cresyl violet as standard (ΦF(methanol) = 0.54).31 TMPyP as standard (ΦF(PBS) = 0.017).9 Not soluble. In toluene.
used for E. coli. Contrary to E. coli, both 1b and 4b were highly
photoactive against MRSA; they caused complete killing of
unwashed bacteria at, respectively, 50 nM and 0.5 μM
concentrations (Figure 6A). These findings are in agreement
with the well-known susceptibility of Gram-positive bacteria to
many cationic and neutral PSs. However, the photoactivities of
the two porphyrins were differently affected by the washing
treatments: 1b did not show any appreciable change, while the
potency of 4b was dramatically reduced (Figure S7A,
Supporting Information). Both 1b and 4b lost part of their
activity when conjugated to apidaecin or its C-terminal
segment. In particular, at a 0.5 μM concentration the
unconjugated porphyrins caused complete cell killing, while
4e, 1c, and 1d caused only about a 3 log reduction of survival
(Figure 6A). The washing of S. aureus before illumination did
not appreciably reduce the killing efficiency of 1c and 1d, while
it significantly affected that of 4e, as was observed with E. coli
(Figure S7A). 2a and its apidaecin conjugate 2c were less
effective than the porphyrin counterparts, being active in the
range of 1.5−15 μM, i.e., at 10-fold higher concentrations
(Figure 6B). However, conjugation with apidaecin strength-
ened the action of the PS (Figure S7B). As for E. coli, also with
S. aureus the photoinactivation experiments were performed as
a function of the incubation time, and no major differences in
the efficiency of S. aureus photoinactivation with conjugated
and unconjugated PSs were found by changing the incubation
times from 15 to 120 min (Figure S8, Supporting Information).
Uptake Experiments. Flow cytometry experiments were
carried out to evaluate the interaction of the PSs (free or
conjugated to apidaecin) following incubation with E. coli and
S. aureus cells under the same experimental conditions as in the
photoinactivation experiments. The measurements showed that
1b does not associate with E. coli cells very efficiently. As shown
in Figure 7, after incubation with 5 μM 1b, only a small fraction
of E. coli cells exhibited a fluorescence signal higher than the
basal value in both washed and unwashed cells. On the
contrary, MRSA cells incubated with 1b exhibited a
fluorescence signal whose intensity was orders of magnitude
higher than the background signal, suggesting efficient
porphyrin association/binding either before or after the
washings. The conjugation of 1b to apidaecin or its C-terminal
segment did not affect to a great extent the association of 1b
with MRSA, while it increased the association with E. coli, as
Table 2. Kinetics of Singlet Oxygen Production (ΦΔ) and
Decay (τΔ) of the Peptide Conjugates and Model
Compounds in Air-Saturated Methanol, PBS, and E. coli
Suspensions
ΦΔ
τΔ/μs
a
b
c
d
e
compd
MeOH
PBS
MeOH
PBS
E. coli
f
1b
1c
0.63
0.70
ns
9.8
9.8
ns
0.020
3.0
2.3
g
43
g
1d
2a
2c
4b
0.66
0.036
ns
9.5
45
h
0.26
ns
g
0.14
0.69
0.001
0.73
9.8
9.6
42
3.6
3.6
3.6
g
60
4e
0.67
0.89
9.6
3.6
g
36
a
b
TMPyP as standard (ΦΔ(methanol) = 0.74).3 TPPS as standard
(ΦΔ(water) = 0.69).3 Literature value 10.4 μs.33 Literature value 3.3
c
d
μs in PBS and 67 μs in D-PBS.33 In PBS. Not soluble. In D-PBS.
e
f
g
h
Toluene as solvent.
Figure S5, Supporting Information), only 1d retained an
efficacy similar to that observed in unwashed samples, while all
other PSs were considerably less potent. In particular, the loss
of efficacy of 4b and 4e in killing bacteria was drastic, with a
decreased killing ability of 6 log at the 10 μM concentration.
On the other hand, 1c and 2c, which were less photoactive
without washings, were less hampered by the washing
treatment. We also studied the photoinactivation of E. coli as
a function of the incubation times and using a concentration of
5 μM for 4b and 4e and 10 μM for all other compounds
(Figure S6, Supporting Information). The results showed that
the time of dark incubation of the cells with the PSs and their
conjugates before irradiation does not affect to a great extent
the photokilling efficiency of E. coli. Incubation times ranging
from 15 to 120 min gave very similar decreases of bacterial
survival when irradiation was carried out leaving the unbound
PS in solution. Only with 4b and 4e was the photoinactivation
ability slightly improved by increasing the incubation time, but
this effect was lost after the repeated washing treatment.
Photoinactivation of MRSA. The photoinactivation of
MRSA was studied by using the same experimental approach
1056
dx.doi.org/10.1021/jm301509n | J. Med. Chem. 2013, 56, 1052−1063

Journal of Medicinal Chemistry
Article
Figure 4. CD spectra of apidaecin and its conjugates in 10% methanol (A, amide region; B, Soret region), TFE (C), and 30 mM aqueous SDS (D).
The peptide concentration is 10 μM.
clearly shown by the higher fluorescence signals exhibited by a
large fraction of cells incubated with 1c and 1d. Both free and
apidaecin-conjugated 4b did not appreciably interact with E. coli
since only in unwashed cells was the measured mean
fluorescence intensity slightly higher than the background
signal. Similar results were observed with MRSA cells,
suggesting that 4b and its conjugate exhibit a very poor ability
to interact with both Gram-positive and Gram-negative
bacteria.
with components of the protein synthesis machinery, impairing
protein synthesis and folding.17,18 The full-length apidaecin
sequence is very important, and the C-terminal octapeptide
does not possess any antibacterial activity,21 nor is it able to
translocate a fluorescent cargo into bacterial cells.24 Thus, most
probably, the conjugate with this cationic peptide (1d) can
effectively bind to the bacterial cell wall without being able to
reach the cytosol.
Efficient PSs for PDT must have appropriate photophysical
properties, such as an intense red-light absorption band and a
high quantum yield of generation of both the long-lived excited
triplet state and cytotoxic ROS, in particular singlet oxygen,
1O2. To establish whether the peptide moiety negatively affects
the PS photosensitizing efficiency, the porphyrin−peptide
conjugates were submitted to a detailed photophysical
characterization. The fluorescence properties and the singlet
oxygen production ability of the peptide conjugates 1c−d and
4e in methanol (Tables 1 and 2) are essentially identical to
those of free PSs, demonstrating that the peptide moiety exerts
no influence on the PS photosensitizing efficiency in this
solvent. Only in the case of porphycene 2a did we observe
changes in the absorption spectrum and a reduction by about
DISCUSSION
■
In our previous paper13 we have already shown that the
conjugate 1c is endowed with antibacterial activity following
light activation. In this paper we have extended our
investigations to other PS−apidaecin conjugates, containing
either a neutral porphycene or a cationic porphyrin
(respectively 2a and 4b in Figure 1) to assess the effects of
structural modifications. Porphycene, a structural isomer of
porphyrin, was chosen for its larger absorption coefficients in
the red part of the spectrum, where light can deeply penetrate
into tissues.37 Because positively charged PSs are effective in
PDT against Gram-negative bacteria without the addition of
outer-membrane-disrupting agents,38 we hypothesized that a
conjugate between apidaecin and a cationic porphyrin could
further promote the uptake of the PS in Gram-negative
bacteria, thereby reducing the minimum effective dose.
Moreover, to establish whether the antimicrobial peptide is
able to direct the PS against specific bacterial targets, we also
synthesized a conjugate between porphyrin 1b and a short
cationic peptide (PRPPHPRL) corresponding to the C-
terminal segment of apidaecin. Although the mode of action
of apidaecin has not been determined in detail, several points of
evidence suggest that this peptide enters E. coli cells by a non-
pore-forming mechanism and, once inside the cell, interacts
1
50% of the fluorescence quantum yield of the PS and in O2
production after conjugation to the peptide, which may be
ascribed to interactions between the peptide and the macro-
cyclic core. The situation is different in an aqueous environ-
ment (PBS) where the photophysical data change considerably
relative to those in methanol for all conjugates and particularly
for those containing a neutral PS. The decrease in the
fluorescence quantum yield, the concentration dependence of
the fluorescence kinetics, and the slow kinetics of 1O2
production in 1c−d and 2c reveal the presence of
intermolecular interactions, as observed also by CD spectros-
1057
dx.doi.org/10.1021/jm301509n | J. Med. Chem. 2013, 56, 1052−1063

Journal of Medicinal Chemistry
Article
copy (Figure 4B). Such interactions account also for the 20-fold
lower production of 1O2 in PBS relative to methanol (Table 2).
Nevertheless, following illumination (with blue light for 1c−
d and red light for 2c), the conjugates were phototoxic against
E. coli cells, inducing a decrease of survival of 3−4 log at a 15
μM concentration (Figure 5). On the other hand, the
unconjugated PS 2a was completely ineffective toward E. coli
(data not shown), consistent with data previously reported for
1b13 and other neutral porphyrins, which are unable to diffuse
through the highly organized outer membrane of Gram-
negative bacteria.39 Peptide conjugates 1c−d and 2c associate
efficiently with E. coli cells, as suggested by the observation that
repetitive washing of bacteria treated with conjugates, before
illumination, caused only a moderate reduction of phototoxicity
(Figure 5B; Figure S5, Supporting Information). In the case of
1c−d this is also supported by the flow cytometry results
(Figure 7). Porphycene was not fluorescent enough for
cytometry studies, but the detection of fluorescence by time-
resolved techniques in the E. coli suspensions, but not in PBS
(Table 1), must be taken as proof of binding. A deeper
understanding of the type of binding/association of 1c−d to E.
coli cells was obtained by analyzing the fluorescence decay
kinetics: unlike PBS or methanol, three decay components were
observed in cell suspensions, which suggest multiple binding
sites (Table 1). The match between two of the three observed
lifetimes with those detected in PBS indicates that one binding
site is located in an aqueous-like environment. Thus, the third
decay component suggests that an additional binding site exists
where the conjugates experience a less hydrophilic environ-
ment.
However, the phototoxicity and the photophysical and flow
cytometry results for the apidaecin conjugate 1c and its
truncated analogue 1d are so similar that it is difficult to
propose a different localization of these conjugates in E. coli
cells. Most probably both conjugates can diffuse through the
outer membrane and localize in different environments, but
since the C-terminal apidaecin fragment is unable to translocate
a PS across the cytoplasmic membrane, we are led to conclude
that the apidaecin conjugate is also not able to reach the
bacterial targets of apidaecin in cytosol.
Figure 5. (A) E. coli photoinactivation with different concentrations
(1.5−15 μM) of 2c, 1c, 1d, 4b, and 4e. E. coli cells were incubated for
60 min with the PS and irradiated with a fixed light dose (36 J/cm2 of
600−750 nm red light for 2c, 13.5 J/cm2 of 390−460 nm blue light for
all other compounds). (B) E. coli photoinactivation with a 5 μM
concentration of apidaecin or the tested PSs. Irradiation with a fixed
light dose, as described above, was carried out without washing (black
bars) or after three washings with PBS (red bars); the asterisk means a
sterile solution.
The conjugate between the cationic porphyrin 4e and
apidaecin possesses a +6 net positive charge, well-distributed
along the whole molecule, that is expected to discourage the
aggregation phenomena observed in conjugates 1c and 2c. In
Figure 6. MRSA photoinactivation with different concentrations of (A) 1b, 1c, 1d, 4b, and 4e (1 nM to 1.5 μM) and (B) 2b and 2c (1.5−15 μM).
MRSA cells were incubated with the PS for 60 min and irradiated with a fixed light dose (36 J/cm2 of 600−750 nm red light for 2b and 2c, 13.5 J/
cm2 of 390−460 nm blue light for all other compounds).
1058
dx.doi.org/10.1021/jm301509n | J. Med. Chem. 2013, 56, 1052−1063

Journal of Medicinal Chemistry
Article
Figure 7. Uptake of the peptide conjugates and model PSs in E. coli and MRSA. The cells were incubated with the compounds for 60 min and then
analyzed by flow cytometry. Concentrations of the compounds: 5 μM for E. coli and 1.5 μM for MRSA.
confirmation of this, the far-UV CD spectrum of 4e in water is
very similar to that of the free peptide and no dichroic signal,
indicative of porphyrin−porphyrin interactions, was detected in
the Soret band region (Figure 4 A,B). Aggregation can also be
ruled out by the lack of concentration effects on the decay
kinetics (Figure S2A and Table S1, Supporting Information). It
can therefore be safely concluded that the major differences
observed between 4b and 4e (Table 1) are due to interactions
between apidaecin and the porphyrin within the conjugate.
Comparison of the fluorescence spectrum, quantum yield, and
kinetics for 4b and 4e reveals that two populations of
conjugates coexist, in which the porphyrin is either exposed
to water or shielded from it by the peptide. This conclusion is
consistent with the well-known solvent-polarity effects on the
fluorescence of tetrapyridinium porphyrins.32 Nevertheless, the
production of 1O2 was very high in PBS, comparable to or even
higher than that in methanol (Table 2). In fact, the conjugate
4e caused total photokilling of E. coli cells at a concentration
(10 μM) at which 1c induced a strong (4 log) but incomplete
reduction of cell survival (Figure 5A). The cationic porphyrin
4b proved to be even more potent than its apidaecin conjugate
4e (Figure 5A). The 1O2 lifetime data in Table 2, particularly in
1
deuterated PBS, indicate that apidaecin is able to quench O2.
1
Thus, because O2 molecules are generated in the vicinity of
apidaecin, some of them will be quenched by the peptide
during their lifetime rather than by cell components.
The washing of the cells before illumination, to remove the
unbound or weakly associated PS, caused a tremendous
reduction of photokilling of E. coli cells, and under these
conditions 4b became the least efficient PS (Figure 5B; Figure
S5, Supporting Information). Flow cytometry measurements
showed that only unwashed cells exhibited red fluorescence
slightly above the background after incubation with these
compounds (Figure 7). Thus, the results indicate that 4b and
4e associate very weakly with E. coli cells and the killing of
unwashed cells is caused mainly by singlet oxygen generated by
the PS molecules not associated with or loosely associated with
the bacteria. Several reports of PDT on Gram-negative bacteria
have pointed out that if singlet oxygen can be generated in
sufficient quantities near the bacterial outer membrane, it will
1059
dx.doi.org/10.1021/jm301509n | J. Med. Chem. 2013, 56, 1052−1063

Journal of Medicinal Chemistry
Article
a Dionex Summit dual-gradient HPLC instrument, equipped with a
four-channel UV−vis detector, using a Vydac 218TP54 column (250 ×
4.6 mm, 5 μm, flow rate at 1.5 mL/min). Mobile phases A (aqueous
0.1% trifluoroacetic acid (TFA)) and B (90% aqueous acetonitrile
containing 0.1% TFA) were used for preparing binary gradients. All
analyses were carried out under gradient conditions (10−50% B in 20
min, except as otherwise indicated). All crude peptides were purified to
95% or more homogeneity for analytical and other experimental
purposes. Semipreparative HPLC was carried out on a Shimadzu series
LC-6A chromatograph, equipped with two independent pump units, a
UV−vis detector, and a Vydac 218TP1022 column (250 × 22 mm, 10
μm, flow rate at 15 mL/min). Elutions were carried out by the same
mobile phases described above. All the purified peptides were analyzed
again by HPLC and HRMS. Mass spectral analyses were carried out on
a Mariner API-TOF workstation (PerSeptive Biosystems Inc.),
operating with ESI techniques in positive mode. NMR spectra were
recorded on a Varian Gemini 300 spectrometer (300 and 75.5 MHz
for 1H and 13C, respectively). UV−vis spectra were recorded at rt on a
Shimadzu UV-2501PC spectrophotometer or on a Lambda 5
spectrophotometer (Perkin-Elmer), in 0.1 or 1 cm quartz cells.
Chemical Synthesis. 19-Glutaramide-2,7,12,17-tetraphenylpor-
phycene (2b). A 44 mg (0.1 mmol) sample of 2a was dissolved in 14
mL of THF and combined with 14 mL of methanol, and 9 mL of 4 N
aqueous sodium hydroxide was added dropwise with stirring at rt
within 5 min. The reaction was stirred for an additional 45 min,
neutralized, and then precipitated under acidic conditions with the
complete addition of 50 mL of ice-cold 5% acetic acid. The flaky
precipitate was filtered, washed with water and then with water/
methanol (1:1), and dried. The title compound was obtained in the
be able to diffuse into the cell to inflict damage to the vital
structure.1,40 Our photophysical data thus support this
hypothesis.
Testing the above PS against MRSA reveals a number of
differences relative to E. coli. First, except for the porphycene,
the concentration needed to inactivate bacteria was 1 order of
magnitude lower. Interestingly, the porphycene conjugate was
almost equally active against both kinds of bacteria. The higher
susceptibility of Gram-positive bacteria to photodynamic
damage has long been known for a large number of PSs.41
Second, it is interesting to recall that washings did not
appreciably remove 1b and its conjugates (1c and 1d) from the
cells (Figure 7), nor did it decrease the photodynamic
inactivation of MRSA (Figure S7, Supporting Information).
In addition, the nonconjugated porphyrin 1b was the most
active PS. This probably can be related to its more hydrophobic
character, which helps in penetration of the cell wall and the
cytoplasmic membrane of Gram-positive bacteria. The
porphycene 2a was probably too hydrophobic, and its strong
aggregation in the external medium prevented cell entrance.
Concerning PSs 4b and 4e, the observations are similar to
those of E. coli: the nonconjugated PS 4b is more active than 4e
1
on account of its lower ability to quench O2, and both lose
their activity after washings due to the very poor binding to the
bacterial cells (Figure 7).
1
CONCLUSIONS
form of a dark green powder. Yield: 27 mg (70%). H NMR (400
■
MHz, DMSO-d6): δ = 12.10 (br s, 1H), 10.96 (s, 1H), 10.08 (s, 1H),
10.07 (s, 1H), 10.05 (s, 1H), 9.93 (d, J = 12, 1H), 9.91 (d, J = 12, 1H),
9.85 (s, 1H), 9.82 (s, 1H), 8.37 (m, 6H), 7.98 (d, J = 8, 2H), 7.88 (m,
5H), 7.74 (m, 5H), 7.60 (m, 2H), 4.29 (s, 1H), 3.93 (s, 1H), 2.25 (t, J
= 8, 2H), 2.16 (t, J = 8, 2H), 1.72 (q, J = 8, 2H). 13C NMR (100 MHz,
DMSO-d6): δ = 174.1, 171.9, 158.0, 144.7, 143.7, 142.8, 142.7, 141.3,
139.1, 137.7, 135.6, 135.4, 135.3, 134.8, 134.3, 133.3, 132.8, 131.1,
131.0, 130.3, 129.2, 129.2, 129.1, 128.0, 127.4, 127.1, 123.1, 125.0,
124.5, 116.0, 115.5, 113.9, 33.9, 33.06, 19.8. HRMS (ESI-TOF): m/z
calcd for C49H38N5O3 744.2969 [M + H]+, found 744.2970.
The search for more effective antimicrobial PSs has led to a
number of strategies. Among them, binding to an antimicrobial
peptide offers the attractive prospect of enhancing both the
water solubility of the PS and the efficiency of the PDT
treatments through a synergistic effect. In this work we have
compared a number of free and apidaecin-conjugated PSs
differing in structure and charge. Our results confirm previous
findings that the conjugation of per se ineffective highly
hydrophobic PSs to a cationic peptide produces a photo-
sensitizing agent effective against Gram-negative bacteria.
MRSA are even more susceptible to the action of conjugates,
which produce the same reduction of bacterial growth at one-
tenth the concentration. The apidaecin ability to penetrate
Gram-negative bacteria is unfortunately lost after conjugation
to a bulky PS, but the amphiphilic character conferred by the
peptide enforces the binding of the PS to the bacterial outer
membrane. Apidaecin−PS conjugates appear most promising
for treatment protocols requiring repetitive washing after
sensitizer delivery, where the most active cationic PSs, such
as 4b and its apidaecin conjugate 4e, are rapidly washed out.
On the other hand, apidaecin cannot improve the phototoxic
activity of the cationic porphyrin, which is mainly determined
by a very high yield of singlet oxygen production in the
surroundings of the bacterial outer membrane.
Synthesis of PS−Peptide Conjugates (General Procedure). The
peptide sequences were prepared on an automated Advanced
Chemtech 348Ω peptide synthesizer, on a 0.25 mmol scale, starting
from Fmoc-Leu-Wang (substitution 0.6 mmol/g of resin). The tert-
butyl group and 2,2,4,6,7-pentamethyldihydrobenzofuran-5-sulfonyl
group were used to protect tyrosine and arginine side chains,
respectively, and the trityl group to protect asparagine and histidine
side chains. Fmoc deprotection was achieved with 20% piperidine in
DMF (5 + 15 min). Couplings were performed in the presence of O-
(benzotriazol-1-yl)-N,N,N′,N′-tetramethyluronium hexafluorophos-
phate/N-hydroxybenzotriazole/N,N-diisopropylethylamine (reaction
time 45−60 min), using an excess of 4 equiv of the carboxyl
component. After the coupling of the last amino acid and removal of
the Fmoc group, the resin was washed with DMF and CH2Cl2 and
then dried under vacuum. The dried resins containing the protected
amino acid sequences were used in the coupling reaction to the
porphyrin derivatives. The H-peptide-resin (0.025 mmol) was swelled
in DMF for 1 h and then washed with DMF. To the peptidyl resin was
added 600 μL of a DMF−CH2Cl2 (1:1, v/v) solution containing 0.05
mmol of porphyrin (1b, 2b or 3b), 0.05 mmol of diisopropylcarbo-
diimide, and 0.05 mmol of 1-hydroxybenzotriazole. The reaction
mixture was shaken overnight and then filtered to remove the excess of
reagents. The resin was repeatedly washed with DMF and CH2Cl2,
until the filtrate was colorless, and then dried under vacuum. Cleavage
and deprotection was carried out by treatment of the resin with a
mixture of TFA−triisopropylsilane−water (95:2.5:2.5 by volume) for
1.5 h at rt. The resin was filtered and washed with TFA, and the
filtrates were combined and reduced under vacuum to a small volume.
Addition of cold ether yielded a green precipitate, which was
repeatedly washed with ether and dried under vacuum. The conjugates
EXPERIMENTAL SECTION
■
General Methods. All chemicals were commercial products of the
best grade available, and unless otherwise indicated, they were used
directly without further purification. The starting porphyrins, 5-(4-
carboxyphenyl)-10,15,20-triphenylporphyrin (1b),42 9-(glutaric meth-
ylesteramide)-2,7,12,17-tetraphenylporphycene (2a),29 and 5-(4-car-
boxyphenyl)-10,15,20-tris(4-pyridyl)porphyrin (3b) and its tris-N-
methylpyridinium iodide (4b),28 were prepared according to literature
procedures. 9-Fluorenylmethoxycarbonyl (Fmoc)-amino acids and all
other chemicals for the solid-phase synthesis were supplied by Sigma-
Aldrich. Fmoc-Leu-Wang resin was purchased from Novabiochem
(Merck Biosciences). Analytical HPLC separations were carried out on
1060
dx.doi.org/10.1021/jm301509n | J. Med. Chem. 2013, 56, 1052−1063

Journal of Medicinal Chemistry
Article
1−3c and 1d were purified by semipreparative HPLC (linear gradient
50−80% B in 20 min), and the fractions containing the conjugate were
collected and lyophilized to yield the pure compound. The purity was
≥95% as determined by analytical HPLC (isocratic 10% B for 3 min;
linear gradient 10−90% B in 30 min). The conjugates were
characterized as follows:
measurements in bacterial suspensions, bacteria were incubated in
the dark at the conditions employed for the photoinactivation
experiments, namely, 5 μM PS for 1 h. When required, the cells
were washed once and resuspended in PBS to a final concentration of
∼1 × 107 cfu mL−1, and 3 mL volumes of the suspensions were
irradiated with 3 million laser pulses at 532 nm under gentle stirring.
CD measurements were carried out on a Jasco-715 spectropolarimeter,
using a quartz cell of 0.1 cm path length. The spectra were recorded at
298 K and were the average of a series of six scans made at 0.1 nm
intervals in the 250−190 and 350−550 nm regions. Sample
concentrations in water (pH 7), TFE, and aqueous 30 mM SDS
were in the range of 10−13 μM. Ellipticity is reported as the mean
residue ellipticity [θ]R (deg cm2 dmol−1).
Bacteria Culture. E. coli ATCC 25922 and the methicillin-resistant
strain of S. aureus ATCC BAA-44 were purchased from LGC
Promochem (Teddington, U.K.). Cultures were maintained by two
weeks of subcultures in brain heart infusion (BHI; Difco, Detroit, MI)
agar. For spectroscopic measurements we used instead the E. coli ECT
strain, purchased from the Spanish-type cell culture collection.
Bacteria Photoinactivation. For the photoinactivation experi-
ments, the bacteria were grown overnight in BHI at 37 °C, harvested
by centrifugation, washed twice, and resuspended in PBS (10 mM
phosphate, 0.14 M NaCl, 2.7 mM KCl, pH 7.3) at a density of ∼2 ×
107 cells/mL. The cell density was evaluated by measuring the
turbidity of the suspension in a Perkin-Elmer spectrophotometer
(model Lambda 5). The bacteria used in the experiments were
collected from cultures in the stationary phase of growth.
The bacteria were incubated with different concentrations of the
PSs in the dark at rt for 60 min or with a fixed PS concentration for
different times. After incubation, the suspensions were (i) directly
exposed to light with the unbound PS left in the suspension (no
washing) or (ii) centrifuged (10000g for 5 min), the pellet
resuspended in 1 mL of PBS, and washed two additional times with
PBS before illumination (three washings). For illumination, aliquots of
cell samples obtained as described above were transferred into 96-well
plates (200 μL/well). Samples incubated with porphycene and its
conjugate with apidaecin 1b were irradiated with red light (600−750
nm) with the Waldmann PDT 1200 lamp supplied by Waldmann
MedizintechNnik; the cells were illuminated from the top of the plates
with a fluence rate at the level of the samples of 40 mW/cm2, as
measured with an International Light power meter, and for a total light
dose of 36 J/cm2. Bacteria incubated with porphyrins and their
conjugates were irradiated with blue light (390−460 nm, with a
maximum at 420 nm) emitted by a UV 236 lamp supplied by
Waldmann Eclairage SA; the cells were illuminated from the bottom of
the plates with a fluence rate of 15.2 mW/cm2, as measured with the
Waldmann UV-meter, and for a total light dose of 13.5 J/cm2. After
illumination, aliquots of bacteria suspensions were serially 10-fold-
diluted in PBS, and 50 μL volumes of the appropriate dilutions were
plated in duplicate onto BHI agar to determine colony-forming units
(cfu). Treated and untreated cells were incubated overnight at 37 °C
to allow colony formation. Suspensions of bacteria exposed to PSs but
kept in the dark and subjected to the same procedure applied to the
irradiated suspensions were also plated onto BHI agar after the
appropriate serial dilutions. Controls also included bacteria not
exposed to any agent and bacteria exposed to light or peptide only.
Each experiment was performed at least three times with independent
bacterial suspensions.
(1c) Yield: 75%. HPLC: tR = 22.0 min. UV−vis (methanol): λmax
/
nm (log ε/M−1 cm−1) 414 (5.48), 513 (4.05), 546 (3.70), 591 (3.45),
642 (3.35). HRMS: m/z calcd for C140H179N36O24 2748.38 [M + H]+,
found 2748.39.
(1d) Yield: 80%. HPLC: tR = 24.3 min. UV−vis (methanol): λmax
/
nm (log ε/M−1 cm−1) 413 (5.41), 512 (4.02), 548 (3.68), 589 (3.54),
644 (3.34). HRMS: m/z calcd for C91H104N21O11 1666.82 [M + H]+,
found 1666.84.
(2c) Yield: 50%. HPLC: tR = 30.3 min. UV−vis (methanol): λmax
/
nm (log ε/M−1 cm−1) 378 (4.99), 583 (4.40), 628 (4.53), 650 (4.54).
HRMS: m/z calcd for C144H186N37O25 2833.44 [M + H]+, found
2833.39.
(3c) Yield: 50%. HPLC: tR = 13.7 min. UV−vis (methanol): λmax
/
nm (log ε/M−1 cm−1) 413 (5.11), 510 (3.81), 544 (3.30), 586 (3.27),
642 (2.89). HRMS: m/z calcd for C137H176N39O24 2751.37 [M + H]+,
found 2751.39.
Conjugate 4e. The porphyrin−peptide conjugate 3c (0.02 mmol),
still attached to the solid support, was swelled in DMF, and a 10%
solution of methyl iodide in DMF (2 mL) was added to the peptidyl
resin. The reaction mixture was shaken overnight at rt and then filtered
to remove the excess of reagents. The resin was repeatedly washed
with DMF and CH2Cl2 and then dried under vacuum. After
detachment from the resin, as described above, the conjugate was
purified by semipreparative HPLC (linear gradient 17−30% B in 25
min). Yield: 50%. HPLC: tR = 12.4 min. UV−vis (methanol): λmax/nm
(log ε/M−1 cm−1) 424 (5.15), 515 (3.98), 556 (3.63), 590 (3.51), 647
(3.03). HRMS: m/z calcd for C141H187N39O24 702.615 [M4+], found
702.874.
Spectroscopic Measurements. Absorption spectra were re-
corded on a double-beam Cary 6000i spectrophotometer (Varian),
equipped with a 110 mm diameter integrating sphere and a high-
performance photomultiplier tube for diffuse transmittance measure-
ments. Absorption coefficients were derived from the slopes of
Lambert−Beer plots. Fluorescence emission spectra were recorded in
a Spex Fluoromax-4 spectrofluorometer. Fluorescence decays were
recorded with a time-correlated single-photon-counting system
(Fluotime 200) equipped with a red-sensitive photomultiplier.
Excitation was achieved by means of a 405 nm LED working at a
10 MHz repetition rate. The counting frequency was always below 1%.
Fluorescence decays were analyzed using the PicoQuant FluoFit 4.0
data analysis software. 1O2 phosphorescence was detected by means of
a customized PicoQuant Fluotime 200 system described in detail
elsewhere.43 Briefly, a diode-pumped pulsed Nd:YAG laser (FTSS355-
Q, Crystal Laser) working at a 10 kHz repetition rate at 532 nm (12
mW, 1.2 μJ per pulse) was used for excitation. A 1064 nm rugate notch
filter (Edmund Optics) was placed at the exit port of the laser to
remove any residual component of its fundamental emission in the
near-IR region. The luminescence exiting from the side of the sample
was filtered by a cold mirror (CVI Melles Griot) to remove any
scattered laser radiation and focused on the entrance slit of a Science
Tech 9055 dual-grating monochromator. A near-IR-sensitive photo-
multiplier tube assembly (H9170-45, Hamamatsu Photonics) was used
as the detector at the exit port of the monochromator. Photon
counting was achieved with a multichannel scaler (Becker&Hickl MSA
300 or PicoQuant’s Nanoharp 250). The time-resolved emission
signals were analyzed using the FluoFit software to extract lifetime
values. Laser flash photolysis measurements were carried out using a
Q-switched Nd:YAG laser (Surelite I-10, Continuum) with a right-
angle geometry and an analyzing beam produced by a Xe lamp (PTI,
75 W) in combination with a dual-grating monochromator (model
101, PTI) coupled to a photomultiplier (Hamamatsu R928). Kinetic
analysis of the individual transients was performed with software
developed in our laboratory. All spectroscopic measurements were
carried out in 1 cm quartz cuvettes (Hellma) at rt. For the
Flow Cytometry. The interaction of PSs with bacteria was
evaluated by flow cytometry. For these experiments, the bacteria were
subjected to the same treatments used for photoinactivation
experiments, but instead of being illuminated after incubation and
washings, they were analyzed with a FACSCanto II flow cytometer.
Samples were excited with the 488 nm laser, and fluorescence emission
signals were recorded at wavelengths higher than 670 nm. The bacteria
population was isolated from the instrument noise by setting
electronic gates on the dual-parameter dot plots of forward scatter
against side scatter. For each sample, 20 000 events were acquired and
analyzed with the FACSDiva software (BD). Samples not incubated
with the PSs were used to determine the cell background fluorescence.
1061
dx.doi.org/10.1021/jm301509n | J. Med. Chem. 2013, 56, 1052−1063

Journal of Medicinal Chemistry
Article
therapy: study of bacterial recovery viability and potential development
of resistance after treatment. Mar. Drugs 2010, 8, 91−105.
(9) Reddi, E.; Ceccon, M.; Valduga, G.; Jori, G.; Bommer, J.; Elisei,
F.; Latterini, L.; Mazzucato, U. Photophysical properties and
antibacterial activity of meso-substituted cationic porphyrins. Photo-
chem. Photobiol. 2002, 75, 462−470.
(10) Bertoloni, G.; Rossi, F.; Valduga, G.; Jori, G.; Vanlier, J.
Photosensitizing activity of water- and lipid-soluble phthalocyanines
on Escherichia coli. FEMS Microbiol. Lett. 1990, 71, 149−155.
(11) Minnock, A.; Vernon, D.; Schofield, J.; Griffiths, J.; Parish, J.;
Brown, S. Mechanism of uptake of a cationic water-soluble pyridinium
zinc phthalocyanine across the outer membrane of Escherichia coli.
Antimicrob. Agents Chemother. 2000, 44, 522−527.
(12) Tegos, G.; Anbe, M.; Yang, C.; Demidova, T.; Satti, M.; Mroz,
P.; Janjua, S.; Gad, F.; Hamblin, M. Protease-stable polycationic
photosensitizer conjugates between polyethyleneimine and chlorin-
(e6) for broad-spectrum antimicrobial photoinactivation. Antimicrob.
Agents Chemother. 2006, 50, 1402−1410.
(13) Dosselli, R.; Gobbo, M.; Bolognini, E.; Campestrini, S.; Reddi,
E. Porphyrin−apidaecin conjugate as a new broad spectrum
antibacterial agent. ACS Med. Chem. Lett. 2010, 1, 35−38.
(14) Hancock, R. E. W.; Sahl, H. Antimicrobial and host-defense
peptides as new anti-infective therapeutic strategies. Nat. Biotechnol.
2006, 24, 1551−1557.
(15) Otvos, L. The short proline-rich antibacterial peptide family.
Cell. Mol. Life Sci. 2002, 59, 1138−1150.
(16) Li, W.; Ma, G.; Zhou, X. Apidaecin-type peptides: biodiversity,
structure-function relationships and mode of action. Peptides 2006, 27,
2350−2359.
(17) Castle, M.; Nazarian, A.; Yi, S.; Tempst, P. Lethal effects of
apidaecin on Escherichia coli involve sequential molecular interactions
with diverse targets. J. Biol. Chem. 1999, 274, 32555−32564.
(18) Otvos, L.; O, I.; Rogers, M.; Consolvo, P.; Condie, B.; Lovas, S.;
Bulet, P.; Blaszczyk-Thurin, M. Interaction between heat shock
proteins and antimicrobial peptides. Biochemistry 2000, 39, 14150−
14159.
(19) Casteels, P.; Romagnolo, J.; Castle, M.; Casteels-Josson, K.;
Erdjumentbromage, H.; Tempst, P. Biodiversity of apidaecin-type
peptide antibioticsprospects of manipulating the antibacterial
spectrum and combating acquired resistance. J. Biol. Chem. 1994,
269, 26107−26115.
ASSOCIATED CONTENT
* Supporting Information
Fluorescence decay kinetics, singlet oxygen phosphorescence
kinetics, and bacteria photoinactivation after repetitive washings
and as a function of the incubation time. This material is
available free of charge via Internet at http://pubs.acs.org.
■
S
AUTHOR INFORMATION
Corresponding Author
*Phone: +39 0498275741. Fax: +39 0498275829. E-mail:
marina.gobbo@unipd.it.
■
Present Addresses
∥Centre for Integrative Bee Research (CIBER), ARC CoE in
Plant Energy Biology, MCS Building M316, The University of
Western Australia, 6009 Crawley, Australia.
⊥Chemistry Research Laboratory, Department of Chemistry,
University of Oxford, Mansfield Road, Oxford OX1 3TA,
United Kingdom.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
We thank Dr. Barbara Biondi for ESI-MS analysis. Financial
support for this research was obtained from the University of
Padova, Progetto di Ateneo 2009 CPDA090338, and from the
Spanish Ministerio de Economia y Competitividad through
Grants CTQ2010-20870-C03-01 and IT2009-0033. We thank
́
̀
the Italian Ministero dell’Istruzione dell’Universita e della
Ricerca for Mobility Grant IT10H5E13A. R.R.-G. thanks the
Generalitat de Catalunya (DURSI) and Fons Social Europeu
for a predoctoral fellowship.
ABBREVIATIONS USED
■
BHI, brain heart infusion; CAMP, cationic antimicrobial
peptide; cfu, colony-forming units; D-PBS, deuterated
phosphate-buffered saline; PDT, photodynamic therapy; PS,
photosensitizer; ROS, reactive oxygen species; TFE, 2,2,2-
trifluoroethanol; TMPyP, meso-tetrakis(N-methylpyridinium-
yl)porphyrin; TPPS, meso-tetrakis(4-sulfonatophenyl)-
porphyrin
(20) Taguchi, S.; Mita, K.; Ichinohe, K.; Hashimoto, S. Targeted
engineering of the antibacterial peptide apidaecin, based on an in vivo
monitoring assay system. Appl. Environ. Microbiol. 2009, 75, 1460−
1464.
(21) Dutta, R. C.; Nagpal, S.; Salunke, D. M. Functional mapping of
apidaecin through secondary structure correlation. Int. J. Biochem. Cell
Biol. 2008, 40, 1005−1015.
(22) Zhou, X.; Li, W.; Pan, Y. Functional and structural
characterization of apidaecin and its N-terminal and C-terminal
fragments. J. Pept. Sci. 2008, 14, 697−707.
(23) Gobbo, M.; Biondi, L.; Filira, F.; Gennaro, R.; Benincasa, M.;
Scolaro, B.; Rocchi, R. Antimicrobial peptides: synthesis and
antibacterial activity of linear and cyclic drosocin and apidaecin 1b
analogues. J. Med. Chem. 2002, 45, 4494−4504.
(24) Czihal, P.; Hoffmann, R. Mapping of apidaecin regions relevant
for antimicrobial activity and bacterial internalization. Int. J. Pept. Res.
Ther. 2009, 15, 157−164.
(25) Gobbo, M.; Benincasa, M.; Bertoloni, G.; Biondi, B.; Dosselli,
R.; Papini, E.; Reddi, E.; Rocchi, R.; Tavano, R.; Gennaro, R.
Substitution of the arginine/leucine residues in apidaecin Ib with
peptoid residues: effect on antimicrobial activity, cellular uptake, and
proteolytic degradation. J. Med. Chem. 2009, 52, 5197−5206.
(26) Matsumoto, K.; Orikasa, Y.; Ichinohe, K.; Hashimoto, S.; Ooi,
T.; Taguchi, S. Flow cytometric analysis of the contributing factors for
antimicrobial activity enhancement of cell-penetrating type peptides:
case study on engineered apidaecins. Biochem. Biophys. Res. Commun.
2010, 395, 7−10.
REFERENCES
■
(1) Hamblin, M. R.; Hasan, T. Photodynamic therapy: a new
antimicrobial approach to infectious disease? Photochem. Photobiol. Sci.
2004, 3, 436−450.
(2) Foote, C. S. Definition of type I and type II photosensitized
oxidation. Photochem. Photobiol. 1991, 54, 659−659.
(3) Redmond, R.; Gamlin, J. A compilation of singlet oxygen yields
from biologically relevant molecules. Photochem. Photobiol. 1999, 70,
391−475.
(4) Stark, G. Functional consequences of oxidative membrane
damage. J. Membr. Biol. 2005, 205, 1−16.
(5) Ravanat, J.; Di Mascio, P.; Martinez, G.; Medeiros, M.; Cadet, J.
Singlet oxygen induces oxidation of cellular DNA. J. Biol. Chem. 2000,
275, 40601−40604.
(6) Maisch, T. A New strategy to destroy antibiotic resistant
microorganisms: antimicrobial photodynamic treatment. Mini-Rev.
Med. Chem. 2009, 9, 974−983.
(7) Lambrechts, S.; Demidova, T.; Aalders, M.; Hasan, T.; Hamblin,
M. Photodynamic therapy for Staphylococcus aureus infected burn
wounds in mice. Photochem. Photobiol. Sci. 2005, 4, 503−509.
(8) Tavares, A.; Carvalho, C. M. B.; Faustino, M. A.; Neves, M. G. P.
M. S.; Tome, J. P. C.; Tome, A. C.; Cavaleiro, J. A. S.; Cunha, A.;
Gomes, N. C. M.; Alves, E.; Almeida, A. Antimicrobial photodynamic
1062
dx.doi.org/10.1021/jm301509n | J. Med. Chem. 2013, 56, 1052−1063

Journal of Medicinal Chemistry
Article
(27) Lindsey, J.; Schereiman, I.; Hsu, H.; Kearney, P.; Marguerettaz,
A. Rothemund and Adler−Longo reactions revisited: synthesis of
tetraphenylporphyrins under equilibrium conditions. J. Org. Chem.
1987, 52, 827−836.
(28) Ishikawa, Y.; Yamashita, A.; Uno, T. Efficient photocleavage of
DNA by cationic porphyrin-acridine hybrids with the effective length
of diamino alkyl linkage. Chem. Pharm. Bull. 2001, 49, 287−293.
̀ ́
(29) Anguera, G.; Llinas, M. C.; Batllori, X.; Sanchez-García, D. Aryl
nitroporphycenes and derivatives: first regioselective synthesis of
dinitroporphycenes. J. Porphyrins Phthalocyanines 2011, 15, 1−6.
(30) Chan, W. C.; White, P. D. Fmoc Solid Phase Peptide Synthesis;
Oxford University Press: New York, 2000; pp 41−74.
(31) Magde, D.; Brannon, J. H.; Cremers, T. L.; Olmsted, J. Absolute
luminescence yield of cresyl violet. A standard for the red. J. Phys.
Chem. 1979, 83, 696−699.
(32) Vergeldt, F. J.; Koehorst, R. B. M.; van Hoek, A.; Schaafsma, T.
J. Intramolecular interactions in the ground and excited states of
tetrakis(N-methylpyridyl)porphyrins. J. Phys. Chem. 1995, 99, 4397−
4405.
(33) Scaiano, J. C., Ed. CRC Handbook of Organic Photochemistry;
CRC Press: Boca Raton, FL, 1989; p 496.
(34) Ogilby, P. R.; Foote, C. S. Chemistry of singlet oxygen. 36.
Singlet molecular oxygen (1Δg) luminescence in solution following
pulsed laser excitation. Solvent deuterium isotope effects on the
lifetime of singlet oxygen. J. Am. Chem. Soc. 1982, 104, 2069−2070.
(35) Arai, T.; Inudo, M.; Ishimatsu, T.; Akamatsu, C.; Tokusaki, Y.;
Sasaki, T.; Nishino, N. Self-assembling of the porphyrin-linked acyclic
penta- and heptapeptides in aqueous trifluoroethanol. J. Org. Chem.
2003, 68, 5540−5549.
(36) Sibrian-Vazquez, M.; Jensen, T.; Fronczek, F.; Hammer, R.;
Vicente, M. Synthesis and characterization of positively charged
porphyrin-peptide conjugates. Bioconjugate Chem. 2005, 16, 852−863.
(37) Jori, G. Far-red-absorbing photosensitizerstheir use in the
photodynamic therapy of tumors. J. Photochem. Photobiol., A 1992, 62,
371−378.
(38) Valduga, G.; Breda, B.; Giacometti, G.; Jori, G.; Reddi, E.
Photosensitization of wild and mutant strains of Escherichia coli by
meso-tetra (N-methyl-4-pyridyl)porphine. Biochem. Biophys. Res.
Commun. 1999, 256, 84−88.
(39) George, S.; Hamblin, M. R.; Kishen, A. Uptake pathways of
anionic and cationic photosensitizers into bacteria. Photochem.
Photobiol. Sci. 2009, 8, 788−795.
(40) Dahl, T.; Midden, W.; Hartman, P. Pure singlet oxygen
cytotoxicity for bacteria. Photochem. Photobiol. 1987, 46, 345−352.
(41) Dai, T.; Huang, Y.; Hamblin, M. R. Photodynamic therapy for
localized infectionsstate of the art. Photodiagn. Photodyn. Ther. 2009,
6, 170−188.
(42) Swamy, N.; James, D.; Mohr, S.; Hanson, R.; Ray, R. An
estradiol-porphyrin conjugate selectively localizes into estrogen
receptor-positive breast cancer cells. Bioorg. Med. Chem. 2002, 10,
3237−3243.
(43) Jimenez-Banzo, A.; Ragas, X.; Kapusta, P.; Nonell, S. Time-
resolved methods in biophysics. 7. Photon counting vs. analog time-
resolved singlet oxygen phosphorescence detection. Photochem.
Photobiol. Sci. 2008, 7, 1003−1010.
(44) Brown, K. L.; Hancock, R. E. W. Cationic host defense
(antimicrobial) peptides. Curr. Opin. Immunol. 2006, 18, 24−30.
(45) Tavano, R.; Segat, D.; Gobbo, M.; Papini, E. The honeybee
antimicrobial peptide apidaecin differentially immunomodulates
human macrophages, monocytes and dendritic cells. J. Innate Immun.
2011, 3, 614−622.
1063
dx.doi.org/10.1021/jm301509n | J. Med. Chem. 2013, 56, 1052−1063