Protease-Stable Polycationic Photosensitizer Conjugates between Polyethyleneimine and Chlorin(e6) for Broad-Spectrum Antimicrobial Photoinac
http://www.100md.com
《抗菌试剂及化学方法》
1.Wellman Center for Photomedicine, Massachusetts General Hospital, Boston, Massachusetts,2.Department of Dermatology, Harvard Medical School, Boston, Massachusetts,3.Department of Materials Science and Engineering, Graduate School of Engineering, University of Tokyo, Tokyo, Japan,4.Graduate Program in Cell Molecular and Developmental Biology, Sackler School of Graduate Biomedical Sciences, Tufts University School of Medicine, Boston, Massachusetts,5.Aga Khan Medical School, Karachi, Pakistan,6.Harvard-MIT Division of Health Sciences and Technology, Cambridge, Massachusetts
ABSTRACT
We previously showed that covalent conjugates between poly-L-lysine and chlorin(e6) were efficient photosensitizers (PS) of both gram-positive and gram-negative bacteria. The polycationic molecular constructs increased binding and penetration of the PS into impermeable gram-negative cells. We have now prepared a novel set of second-generation polycationic conjugates between chlorin(e6) and three molecular forms of polyethyleneimine (PEI): a small linear, a small cross-linked, and a large cross-linked molecule. The conjugates were characterized by high-pressure liquid chromatography and tested for their ability to kill a panel of pathogenic microorganisms, the gram-positive Staphylococcus aureus and Streptococcus pyogenes, the gram-negative Escherichia coli and Pseudomonas aeruginosa, and the yeast Candida albicans, after exposure to low levels of red light. The large cross-linked molecule efficiently killed all organisms, while the linear conjugate killed gram-positive bacteria and C. albicans. The small cross-linked conjugate was the least efficient antimicrobial PS and its remarkably low activity could not be explained by reduced photochemical quantum yield or reduced cellular uptake. In contrast to polylysine conjugates, the PEI conjugates were resistant to degradation by proteases such as trypsin that hydrolyze lysine-lysine peptide bonds, The advantage of protease stability combined with the ready availability of PEI suggests these molecules may be superior to polylysine-PS conjugates for photodynamic therapy of localized infections.
INTRODUCTION
Photodynamic therapy (PDT) employs a nontoxic dye termed a photosensitizer (PS) and low-intensity visible light, which in the presence of oxygen produce cytotoxic species (4-6). PDT has the advantage of dual selectivity in that the PS can be targeted to its destination cell or tissue, and in addition the illumination can be spatially directed to the lesion. PDT was originally discovered over 100 years ago by its effect on microorganisms (35) but since then has been principally developed as a treatment for cancer (11) and age-related macular degeneration (3).
Because of the well-known increase in multiantibiotic resistance among pathogenic microbes of all classes, PDT has attracted attention as a possible treatment for localized infections (9, 14, 46). It is known that gram-negative bacteria are resistant to PDT with many commonly used PS that will readily lead to phototoxicity for gram-positive species (29) and that PS bearing a cationic charge (33) or the use of agents that increase the permeability of the outer membrane will increase the efficacy of killing of gram-negative organisms (29).
In 1997 our laboratory formed the hypothesis that by covalently conjugating a suitable PS such as chlorin(e6) (ce6) to a poly-L-lysine (pL) chain, a bacterially targeted PS delivery vehicle could be constructed that would efficiently inactivate both gram-positive and gram-negative species (42). Because the resulting polycationic entity is a macromolecule, it is taken up by mammalian cells by the time-dependent process of endocytosis (although it binds rapidly to anionic sulfated chains of proteoglycans on the cell membrane). The polycationic conjugates could effectively bind to microbial cells after incubations as short as 1 minute, thus giving temporal selectivity for bacteria. We (16) compared the effectivenesses of pL-ce6 conjugates with chain lengths of either 8 or 37 lysines attached to precisely one ce6 molecule for bacterial photodynamic inactivation (PDI) and found the 37-lysine conjugate was able to efficiently mediate the photodestruction of both gram-positive and gram-negative species, while the eight-lysine conjugate and free ce6 were effective only against gram-positive bacteria.
A similar conjugate (composed of one ce6 molecule and a 5-amino-acid lysine chain) was subsequently used by another group to kill several oral pathogens in the presence of 25% whole blood (41). Polo and coworkers used conjugates between pL and porphycenes with significant phototoxic activity against gram-negative bacteria (40). In a subsequent report (25) these authors showed that several strains responsible for periodontal disease were efficiently inactivated by visible-light irradiation in the presence of porphycene-pL conjugates. Repeated photosensitization of surviving cells did not induce the selection of resistant bacterial strains or modify their sensitivity to antibiotic treatment. A recent report described the synthesis and antibacterial activities of conjugates between pL and meso-substituted porphyrins (44). We subsequently demonstrated the potency and selectivity of pL-ce6 conjugates and red light for treating actual infections by using the technology to successfully carry out PDT of infected wounds (15, 17), and deep established soft-tissue infections (13) in mice.
In the present report we describe the preparation, high-pressure liquid chromatography (HPLC) characterization, and broad-spectrum antimicrobial photoinactivation potential of a novel set of second-generation polycationic PS conjugates constructed using another commercially available polymer, polyethyleneimine (PEI). This macromolecular vehicle does not contain peptide bonds and therefore should be resistant to protease degradation, which could in theory be a limitation on the use of pL-PS conjugates to carry out in vivo PDT.
MATERIALS AND METHODS
Preparation of the PEI-ce6 conjugates. All reactions were carried out in the dark at room temperature. Linear PEI (80 mg; 189 μmol; average molecular weight, 423; degree of polymerization, 10; Aldrich Chemical Co, St. Louis, MO; catalog number 46,853-3) was dissolved in 3 ml dry dimethyl sulfoxide to which was added 120 mg (200 μmol) ce6 (Porphyrin Products, Logan, UT) and 80 mg (418 μmol) 1-ethyl-3-(3-dimethylaminopropyl)-carbodiimide hydrochloride (Sigma). Triethylamine (20 μl; Sigma) was then added and the mixture was stirred for 24 h. The progress of the reaction was checked by thin-layer chromatography (Polygram SIL G/UV254 plates; Macherey Nagel, Duren, Germany) in methylene chloride-methanol-ammonium hydroxide (8:2:0.5). After 24 h, when the reaction was complete, methanol (1 ml) and water (1 ml) were added and the mixture was evaporated to dryness under vacuum. The residue was dissolved in sodium acetate buffer (10 mM, pH 5.5, 3 ml), applied to a column of Sephadex G-25 (60 by 1 cm) and eluted with the same buffer at a flow rate of 3.3 ml per h; 3-ml fractions were collected, and fractions 14 to 21 were combined and evaporated to give the product PEI-ce6 lin. The substitution ratio was calculated from the absorption spectrum of the conjugate to be 1 ce6 per PEI chain, assuming the absorption coefficient of conjugated ce6 was the same as that of free ce6 (400 = 150,000 M–1 cm–1).
Using the methodology described above, conjugates were prepared between ce6 and two other PEI preparations (low molecular weight [LMW], 600 to 800, branched; Aldrich catalog number 40,871-9; high molecular weight [HMW], 10,000 to 25,000, branched; Aldrich catalog number 40,872-7). Using the same molar ratios of reactants, conjugates (PEI-ce6 LMW and PEI-ce6 HMW) with average substitution ratios of 1 ce6 per PEI chain were obtained.
A pL-ce6 conjugate that consisted of a pL chain of an average length of 164 lysines with an average of four ce6 molecules attached to each chain was used. This was prepared as previously described (15, 17).
HPLC characterization of conjugates. The HPLC system was from Shimadzu (Columbia, MD) and consisted of two LC-10ADvp solvent delivery modules for gradient and isocratic elution and an SPD-M10Avp UV-visible photodiode array detector. The operation and data acquisition of HPLC were conducted with a personal computer and SCl-10Avp system controller (software package CLASS-VP 7.1.1 SP1). Three types of HPLC columns were tested for the purpose of PEI conjugate characterization. These included an Agilent Hypersil BDS-C18 column, 4.6 by 250 mm, 5-μm particle size (Agilent Technologies, Palo Alto, CA); a Hamilton polymer reversed-phase PRP-1 column, 4.1 by 150 mm, 10-μm particle size (Hamilton Company, Reno, NV); and an Agilent Lichrospher 100 diol column, 250 by 4 mm, 5-μm particle size (Agilent Technologies).
Microbial strains and growth conditions. The following microbial strains were used: Staphylococcus aureus (strain 8325-4), Streptococcus pyogenes (strain 591, serotype M49), Escherichia coli (ATCC 25922), Pseudomonas aeruginosa (ATCC 19660), and Candida albicans (ATCC MYA574). Bacteria were routinely grown on brain heart infusion agar and broth and yeasts were routinely grown in yeast- mannitol (YM) broth or agar with shaking at 37°C. S. pyogenes was grown in the presence of 5% CO2. The next day, the cells were diluted 1/100 in fresh medium and left to grow until mid-log phase (optical density [OD], 0.6, equivalent to 108 cells/ml or 107 cells/ml for the yeast).
Photodynamic inactivation of microbial cells. Microbial suspensions in medium were centrifuged for 5 min and the supernatant was replaced with sterile phosphate-buffered saline (PBS); 3.5 ml of the resulting bacterial suspensions (at concentrations of 108/ml or 107 cells/ml for the yeast) were incubated in the dark at room temperature with either 1 or 10 μM of PEI-ce6 conjugates (lin, LMW, and HMW) or free ce6 for 10 min. Cells were incubated in the absence of PS to provide CFU values for an absolute control and to provide cells for the experiments with light alone. Some microbial cell/dye suspensions were centrifuged for 5 min and again the supernatant was replaced with same amount of PBS at the same cell concentration and 500-μl aliquots were then placed in the wells of 48-well plates and the remainder was used for uptake experiments (see below).
The wells were illuminated at room temperature using a 665-nm, 1-W diode laser (model BWF-665-1; B&W Tek, Newark, DE). The laser was coupled into a 1-mm optical fiber that delivered light into a lens. This assembly formed a uniform circular spot on the top of the 48-well plate which was 3 cm in diameter. Red light was delivered at an irradiance of 50 mW/cm2. The fluence range was from 0 to 16 J/cm2 (up to 5 min of illumination time) for PEI-ce6 lin and PEI-ce6 HMW and from 0 to 100 J/cm2 (up to 30 min of illumination time) for PEI-ce6 LMW.
At intervals during the illumination when the requisite fluences had been delivered, aliquots (100 μl) were taken from each well to determine CFU. Care was taken to ensure the contents of the wells were thoroughly mixed before sampling, as bacteria can settle to the bottom. The aliquots were serially diluted 10-fold in PBS to give dilutions of 10–1, 10–2, 10–3, 10–4, 10–5, and 10–6 times the original concentrations and 10-μl aliquots of each of the dilutions were streaked horizontally on square brain heart agar plates by the method of Jett et al. (21). Plates were streaked in triplicate and incubated for 24 h at 37°C in the dark (CO2 incubator for S. pyogenes). In general, three dilutions could be counted on each plate.
Controls were cells not treated with PS or light, cells incubated with PS but kept in 48-well plates covered with aluminum foil at room temperature for the duration of the illumination, and cells treated with light but not with PS. Survival fractions (SF) were routinely expressed as ratios of CFU of bacteria treated with light and of PS to CFU of microbes treated with neither. The SF at 0 J/cm2 gives a measure of the dark toxicity of the conjugates.
Trypsin degradation of conjugates. To compare the effects of trypsin we used a pL-ce6 conjugate prepared similarly to those previously described (15, 17). Conjugates (pL-ce6, PEI-ce6 lin, or PEI-ce6 HMW) were incubated with 1% trypsin solution (Sigma) in PBS without Ca and Mg overnight in the dark at 37°C and mixtures were then boiled at 100°C to inactivate the enzyme. The inactivated trypsin-conjugate mixtures were then used for PDI with E. coli and S. aureus in the same manner as unreacted conjugates. There was no effect of boiling on the solubility or PDI activity of the PEI-ce6 or pL-ce6 conjugate alone.
Photochemical tests. All chemicals were obtained from Aldrich (Milwaukee, WI) and used without further purification. The superoxide assay used nitro blue tetrazolium chloride (NBT) at 80 μM, NADH at 10 mM, and PS at 5 μM ce6 equivalent concentration, all dissolved in PBS (49). All ingredients were freshly prepared prior to the procedure and placed in a quartz cuvette and the spectrophotometer was blanked. Broad-band (400- to 700-nm) white light was delivered at a power of approximately 200 mW incident on the cuvette face, and measurements of the blue formazan product were taken every 10 seconds at 580 nm.
The singlet oxygen assay was based on the oxidation of 4-nitroso-N,N-dimethyl-aniline (RNO) (22). The cuvette contained L-histidine at 15 mM and PS at 5 μM ce6 equivalent concentration and was blanked. RNO was then added to achieve a measured absorbance of approximately 1 at 440 nm (13.3 μM concentration). Illumination was with broad-band white light at an incident power of approximately 3 W, and loss of absorbance was measured every 10 seconds at 440 nm.
Uptake of conjugates by microbial cells. Suspensions of microorganisms (108/ml for bacteria and 107/ml for C. albicans) were incubated for 10 min with PS in the dark at room temperature. Unbound PS was washed out by centrifugation of the mixture of dye and microorganisms for 6 min at 1,550 x g followed by resuspension of washed pellets in 3.5 ml PBS without Ca2+ and Mg2+. Aliquots (500 μl) of these suspensions were used for PDI experiments, and the remaining suspensions were centrifuged again and the pellets were dissolved in 6 ml 0.1 M NaOH-1% sodium dodecyl sulfate for at least 24 h. These suspensions were used for PS uptake measurement.
The fluorescence of dissolved pellets was measured on a spectrofluorimeter (FluoroMax3; SPEX Industries, Edison, NJ). The excitation wavelength was 400 nm and the emission spectra of the solution were recorded from 580 to 700 nm. The fluorescence was calculated from the height of the peaks recorded. Calibration curves were made from pure PS dissolved in NaOH-sodium dodecyl sulfate and used for determination of PS concentration in the suspension. Uptake values were obtained by dividing the number of nmol of PS in the dissolved pellet by the number of CFU obtained by serial dilutions and the number of PS molecules/cell was calculated by using Avogadro's number.
Statistics. Values are reported as the mean ± standard error of the mean (SEM). Differences between two means were assessed for significance by the two-tailed Student's t test, assuming equal or unequal variances of the standard deviations as appropriate. A value of P of <0.05 was considered significant.
RESULTS
Preparation and characterization of conjugates. Three types of HPLC column from different vendors were tested to develop a method to characterize the PEI conjugates (Fig. 1). To our knowledge, there have been no reports on HPLC analysis of PEI or its related compounds. Polyelectrolytes, such as PEI, are charged at most pHs. In PEI, the amino groups tend to be positively charged due to protonation at pH 7, which is usually the operational pH for most silica-based HPLC columns. These positively charged groups are problematic for HPLC analysis due to analyte stationary-phase interaction, which leads to partial or complete adsorption of analytes to the HPLC column, a phenomenon usually seen in protein and peptide HPLC analyses.
In this study, we observed complete adsorption of the PEI-ce6 conjugate to the Agilent C18 column. The polymer reversed-phase column (Hamilton PRP-1) was brought in for testing with the hope that the noncharged polymer stationary phase would have minimal electrostatic interaction with the PEI analytes. But the PRP column yielded broad peaks for all three PEI conjugates and hence is not the ideal column for PEI conjugate characterization. The diol column (Agilent Lichrospher 100 diol), on the other hand, has a stationary phase that combines a hydrophilic external surface (diol groups) and a hydrophobic internal surface (alkane linker) that can effectively preclude the previously described PEI-silanol group interaction. With the HPLC method developed in this study, the Lichrospher diol column showed different retention times and peak shapes for the three conjugates. The peaks obtained are shown in Fig. 2. The order of elution from the column is in order of increasing molecular weight (PEI-ce6 lin, 1,000; PEI-ce6 LMW, 1,300; and PEI-ce6 HMW, 18,000).
Dark toxicity of PEI and PEI conjugates. In order to compare the levels of dark toxicity exhibited by the PEI conjugates and by the unconjugated PEI preparations, we incubated the three PEI-ce6 conjugates and the corresponding PEI compounds at 10 μM in PBS with E. coli for 10 min. The survival fractions were as follows: PEI lin, 0.018 ± 0.0037, versus PEI-ce6 lin, 0.958 ± 0.13 (P < 0.0001); PEI LMW, 0.157 ± 0.058, versus PEI-ce6 LMW, 0.526 ± 0.225 (P = 0.18); and PEI HMW, 0.0067 ± 0.00011, versus PEI-ce6 HMW, 0.57 ± 0.267 (P < 0.0001). We conclude that the dark toxicity due to the polycationic nature of some PEI preparations is markedly reduced after conjugation to ce6, presumably because the anionic lipophilic dye molecule ameliorates the effect of multiple basic amino groups.
Photoinactivation of gram-positive species. We used two gram-positive species, S. aureus and S. pyogenes, which are both important human and animal pathogens. It rapidly became clear that the conjugate PEI-ce6 LMW had much lower activity than both the PEI-ce6 lin and PEI-ce6 HMW conjugates, with free ce6 having intermediate activity. Nevertheless, we used the same concentration of all four PS, i.e., 1 μM ce6 equivalent, and delivered additional light (up to 100 J/cm2 for PEI-ce6 LMW and up to 40 J/cm2 for free ce6) in an attempt to increase the extent of killing. As shown in Fig. 3A for S. aureus and Fig. 3B for S. pyogenes, there is a dramatic difference in the effectiveness of PDI between PEI-ce6 LMW (less than 2 logs of killing after 100 J/cm2) and the other two conjugates (5 to 6 logs of killing after 16 J/cm2). Free ce6 was intermediate in potency. The linear and HMW conjugates were similarly effective in mediating killing of S. pyogenes (Fig. 3B) but the HMW conjugate was significantly better (P < 0.05 at 8 and 16 J/cm2) than PEI-ce6 lin at mediating the killing of S. aureus (Fig. 3A).
Photoinactivation of gram-negative species. We studied two gram-negative species that are pathogenic to humans and animals, E. coli and P. aeruginosa. Because gram-negative species are more resistant to PDI than gram-positive bacteria (29), we increased the concentration of PS 10-fold over that used for gram positives (to 10 μM). Because of the relatively poor activity of PEI-ce6 LMW and ce6, we again delivered light in doses of up to 100 J/cm2 and 40 J/cm2 instead of the dose of 16 J/cm2 used for the other conjugates. Figure 4A shows the PDI of E. coli and Fig. 4B shows the PDI of P. aeruginosa. For E. coli the order of effectiveness is PEI-ce6 HMW > PEI-ce6 lin > ce6 > PEI-ce6 LMW. However, PEI-ce6 LMW is actually more effective against E. coli than against any of the other microbes tested. For P. aeruginosa the order of effectiveness is similar to that found for E. coli except there is little difference between PEI-ce6 lin and free ce6. Compared to E. coli, P. aeruginosa is more susceptible to PEI-ce6 HMW and less susceptible to PEI-ce6 LMW. The biggest difference observed between the gram-positive and gram-negative species is that PEI-ce6 lin is highly effective (5 logs of killing at 1 μM) against the former but almost ineffective (1 log of killing at 10 μM) against the latter.
Photoinactivation of C. albicans. We used C. albicans cells at a cell density 1/10 that of the bacterial species (107 instead of 108 cells/ml). We had previously showed that C. albicans cells are 10 times larger than bacterial cells by dry weight and therefore each cell needs to bind 10 times as much PS to allow equal killing (8). Figure 5 shows the killing of C. albicans mediated by the four PS incubated at 10 μM. In a pattern similar to that of the gram-positive bacteria, PEI-ce6 LMW is the least effective (2 logs at 100 J/cm2) and ce6 is next (2.5 logs at 40 J/cm2), while both PEI-ce6 lin and PEI-ce6 HMW are highly effective (4 to 6 logs at 16 J/cm2), although PEI-ce6 HMW is significantly better (P < 0.05) at killing the yeast than PEI-ce6 lin.
Effect of trypsin degradation of conjugates. We wished to test the hypothesis that pL-ce6 conjugates could be degraded by protease enzymes, thus destroying their polycationic character, while PEI-ce6 conjugates would be resistant to proteases because they do not contain peptide bonds. Figure 6A shows the PDI killing of E. coli obtained using PEI-ce6 HMW and pL-ce6 with and without trypsin treatment. While the killing produced by PEI-ce6 HMW with and without trypsin treatment was identical and similar to the killing produced by pL-ce6, the effect of trypsin treatment on the pL-ce6 conjugate was to significantly (P < 0.001) reduce the killing compared to that with the other three PS and gave greater than a 3-log difference in SF of E. coli at 16 J/cm2.
Figure 6B shows the effect of trypsin treatment of pL-ce6 and PEI-ce6 lin on the efficacy of PDI against S. aureus. PEI-ce6 lin was equally effective with and without trypsin treatment; however, pL-ce6 was significantly less effective (P < 0.01) than PEI-ce6 lin. Interestingly, when pL-ce6 was treated with trypsin it became much more effective at mediating killing (P < 0.0001), giving almost total killing after only 4 J/cm2.
Relative ROS production on illumination. In an attempt to explain the remarkably low PDI activity exhibited by PEI-ce6 LMW, we tested the hypothesis that its quantum yield of reactive oxygen species (ROS) was significantly lower than that of the other conjugates. Illumination of PS can proceed via type I or type II mechanisms (4). Type I frequently involves the light-mediated production of superoxide anion radical (O2– ·) by electron transfer to and from the triplet state in the presence of a reducing agent, while type II involves the production of singlet oxygen by energy transfer to ground-state oxygen from the PS triplet state. Both processes have been implicated in the PDI of microorganisms (7, 31, 32). The type I process is conveniently monitored in aqueous solution by the reduction of nitro blue tetrazolium in the presence of NADH (49), while the type II process can be monitored in aqueous solution by oxidation of RNO in the presence of histidine (22). Because the NBT reaction is a reduction and the RNO reaction is an oxidation there is no cross talk between the two reaction intermediates and their respective products.
As shown in Fig. 7A, there are no significant differences between the rates of NBT reduction caused by superoxide produced by any of the four PS tested. Figure 7B shows the relative rates of RNO photobleaching in the presence of histidine caused by singlet oxygen production by the four PS. As can be seen, the rates are in the order ce6 > PEI-ce6 LMW > PEI-ce6 lin > PEI-ce6 HMW. The differences between ce6 and PEI-ce6 LMW and between PEI-ce6 LMW and PEI-ce6 lin are significant (P < 0.05).
Uptake of conjugates by microbial cells. We then asked if the relative inactivity of PEI-ce6 LMW could be explained by markedly lower cell uptake of PS. Because we centrifuged the PS-incubated cells before illumination, only dye that actually bound to the cells was active in the PDI. Figure 8 depicts the uptake in terms of molecules of ce6 equivalent per cell after incubation of the gram-positive organisms with 1 μM and the other cells with a 10 μM equivalent concentration of ce6. As can be seen, the order of uptake of the four PS by both of the gram-positive species (S. aureus and S. pyogenes) was PEI-ce6 HMW > PEI-ce6-lin > ce6 > PEI-ce6 LMW. PEI-ce6 LMW was taken up significantly less (P < 0.01) than either PEI-ce6 HMW or PEI-ce6 lin. For the gram-negative species (E. coli and P. aeruginosa) the order was PEI-ce6 HMW > PEI-ce6-LMW > PEI-ce6 lin >> ce6. PEI-ce6 LMW uptake was significantly (P < 0.01) lower than that of PEI-ce6 HMW and higher than that of free ce6, but not significantly different from that of PEI-ce6 lin. C. albicans had the same order of PS uptake as the gram-negative organisms except that the uptake of ce6 was only very slightly lower than that of the conjugates.
DISCUSSION
This study has shown that covalent conjugates between PEI and ce6 can be potent broad-spectrum antimicrobial PS resistant to protease degradation and therefore constitute an alternative to the previously described pL-ce6 conjugates. The development of an HPLC method to characterize these cationic polymeric conjugates demonstrates that the preparations are free of nonconjugated ce6 and shows their retention times are proportional to their molecular weights.
The high susceptibility of the gram-positive organisms S. aureus and S. pyogenes to PDI was not unexpected. We used 1/10 the concentration of all the PS and obtained overall levels of killing roughly comparable to those obtained with gram-negative and yeast cells at a 10-fold-higher concentration. Many workers have shown that gram-positive bacteria are much more susceptible to PDI than other microbial cell types (29, 38, 46). We found that PEI-ce6 lin and PEI-ce6 HMW had roughly equal activities against S. aureus, and PEI-ce6 HMW was better against S. pyogenes. This was somewhat surprising since we had previously shown (16) a structure-function relationship in a pair of pL conjugates, where a conjugate that had an average of eight lysine monomers joined to one ce6 molecule performed better in killing S. aureus than a conjugate that had an average of 37 lysines joined to one ce6.
We proposed that the relatively porous outer coat of gram-positive species allows the smaller conjugate to pass through more efficiently, while the larger conjugate is bound to the negatively charged outer structures of the gram-positive cells due to its more pronounced cationic charge but cannot diffuse through the permeability barrier as efficiently due to its greater size. Presumably the efficiency of these polycationic conjugates in inactivating gram-positive species is partly a balance between molecular size (smaller is better) and cationic charge (more is better). This point is demonstrated in Fig. 7, where the pL-ce6 conjugate (MW 20,000) is less effective than PEI-ce6 lin and PEI-ce6 HMW but becomes much more effective after proteolysis.
The superiority of the large HMW conjugate in mediating the PDI of gram-negative species was also expected. In this case we propose that the larger, more polycationic structure can more effectively displace divalent cation from the lipopolysaccharide molecules in the outer membrane and is better able to increase outer membrane permeability, allowing access to targets such as the inner membrane, This process is known as self-promoted uptake (10, 34, 39) and is typified by the bactericidal mechanism of naturally occurring cationic antimicrobial peptides expressed by most life forms, such as defensins, cecropin, and gramicidin (18).
The yeast C. albicans behaved similarly to gram-positive bacteria in being effectively killed by the linear and HMW conjugates and moderately killed by free ce6 after exposure to modest levels of light. Each Candida cell is 10 times heavier than a Staphylococcus cell (8) and therefore 10 times more PS is needed to kill each cell (10 μM as opposed to 1 μM). Other workers have found that C. albicans, like gram-positive bacteria, is killed by PDI using noncationic PS (2, 24)
The remarkably low PDI activity of the LMW conjugate (especially compared to that of PEI-ce6 lin, which appears to have a very similar molecular structure in terms of size and charge) was surprising. It was also remarkable that PEI-ce6 LMW was in every case less active than free ce6. This might be possible to explain in the case of the gram-positive bacteria on the basis of increased diffusion of smaller compounds regardless of charge. We previously showed that free ce6 was more active against S. aureus than a pL-ce6 conjugate with 37 lysine residues. However, the greater activity of free ce6 than of PEI-ce6 LMW against both of the gram-negative species was particularly surprising.
We then asked if the photophysical properties of the photoactive dye had been adversely altered in the LMW conjugate compared to the other conjugates and free ce6. We therefore carried out two photochemical assays involving singlet oxygen-mediated oxidation of RNO in the presence of histidine and superoxide-mediated reduction of NBT in the presence of NADH to answer this question. The results showed that the rates of photochemical superoxide production were fairly similar among all four PS. However, we were surprised to observe that PEI-ce6 LMW had the second highest rate of photochemical singlet oxygen production (after free ce6, which was expected to be the highest), and PEI-ce6 lin and PEI-ce6 HMW, which were most effective at killing cells, actually had the lowest rates of 1O2 formation.
We then argued that the low activity of PEI-ce6 LMW must be because the molecule did not bind as effectively to the microbial cells. Therefore we examined uptake of the four PS by all five microbial species. The uptake of PEI-ce6 lin by the gram-positive cells was significantly the lowest of all four PS tested, but whether this observation represents a difference big enough to explain the very low PDI activity is debatable. The uptake of PEI-ce6 LMW by the gram-negative species and C. albicans was intermediate between that of PEI-ce6 lin and PEI-ce6 HMW. This was as we initially expected based on the increasing cationic character of the molecules. Low cellular binding then cannot explain the poor PDI efficacy of PEI-ce6 LMW against the gram-negative and fungal species. The low uptake of free ce6 by the gram-negative cells is due to the permeability barriers for anionic molecules exhibited by these species, and the high uptake of free ce6 by the yeast cells was expected, and this difference correlates with PDI killing.
It is possible that besides the effects of molecular size and overall cationic charge, another molecular property affects the microbial PDI activity of these conjugates. This may be due to changes in tertiary structure. It is known that naturally occurring antimicrobial peptides assume a helical structure with the cationic charges on one side of the helix and hydrophobic amino acid residues on the opposite side (12, 37, 48). Although PEI-ce6 lin and PEI-ce6 LMW are comparable in size and charge, the chief difference between the molecules is that one is a linear chain while the other is cross-linked. Therefore, the small cross-linked molecule may be unable to form the appropriate tertiary structure to increase microbial binding and penetration. Although the PEI-ce6 HMW conjugate is also cross-linked it is possible that its much greater content of cationic groups or its increased molecular flexibility due to its greater size could overcome the restriction seen with PEI-ce6 LMW. The finding that the dark toxicity of unconjugated PEI LMW is significantly lower than that of the other two PEI preparations also suggests that the PEI LMW structure has some fundamental difference from that of the other PEI compounds in how it binds or penetrates bacteria.
PEI and pL have been reported to act similarly on gram-negative bacteria (19, 20). Helander et al. demonstrated that PEI increased cell permeability to the fluorescent dye 1-N-phenylnaphthylamine (19). Several papers by Klibanov and colleagues (26, 27) have shown that alkylated derivatives of PEI can exert an antimicrobial effect when incorporated into surfaces and textiles. This was proposed to occur due to the fact that the alkyl chains were long enough to penetrate from the surface to the bacterial membrane (28).
Although pL-ce6 conjugates are highly efficient in mediating PDI of multiple microbial species in vitro (8, 16, 42) and in vivo (13, 15, 17), their structure is based on polypeptides composed of naturally occurring L-amino acids. This means that they are likely to be substrates for protease enzymes that are found in tissue. In particular, extracellular proteases derived from the host cells are found in areas of inflammation such as infections (1, 23) and proteases may also be produced by bacteria themselves as secreted virulence factors involved in tissue invasion (30, 47).
Many proteases have particular sequences of amino acids that are recognized and preferentially cleaved by the enzyme, but some are more general in their substrate specificity. Trypsin is a digestive proteinase from the pancreas that hydrolyzes peptide bonds in protein molecules that have carboxyl groups donated by the basic amino acids arginine and lysine, and therefore served as a proof-of-principle enzyme for the ability to cleave pL conjugates in solution. The effect of trypsin treatment of pL-ce6 on killing of E. coli was to significantly reduce the number of cells that were killed. This is presumably because the products are short oligopeptides, and if these peptides still bear a ce6 molecule they are not sufficiently polycationic to penetrate the outer membrane. By contrast, the effect of trypsin-treated pL-ce6 on the PDI of S. aureus was to dramatically increase killing. This observation can be explained by fact that the short peptides attached to PS molecules that are the products of proteolysis are better able to penetrate gram-positive cells than intact large molecules.
It is at present uncertain whether the peptide bond between the epsilon amino group of lysine and ce6 is hydrolyzed by trypsin, but it would seem unlikely. It should be pointed out that it is at present unclear if protease stability would be a real advantage in the use of these conjugates in PDT of actual infections. First, resistance to degradation would be an advantage only in combating gram-negative infections, and second, resistance to degradation might hinder removal of the remaining conjugate from the tissue after completion of the illumination.
In conclusion, second-generation polycationic conjugates between various PEI preparations and ce6 have been prepared and characterized. PEI has been widely studied as a DNA transfection agent for nonviral gene therapy and much is known about its properties and cytotoxicity (36, 43, 45). The broad-spectrum antimicrobial PDI activity of PEI-ce6 coupled with the possible advantage of protease stability and the ready availability of PEI suggest these molecules may be superior to pL-PS conjugates for PDT of localized infections.
ACKNOWLEDGMENTS
This work was funded by U.S. National Institutes of Health grant R01AI050875 to M. R. Hamblin. T. N. Demidova was supported by a Wellman Center of Photomedicine graduate student fellowship. M. Anbe was funded by the Ministry of Education, Culture, Sports, Science and Technology, Nano Bioengineering Education Program, University of Tokyo, Tokyo, Japan.
Deceased.
REFERENCES
Behera, A. K., E. Hildebrand, J. Scagliotti, A. C. Steere, and L. T. Hu. 2005. Induction of host matrix metalloproteinases by Borrelia burgdorferi differs in human and murine Lyme arthritis. Infect. Immun. 73:126-134.
Bliss, J. M., C. E. Bigelow, T. H. Foster, and C. G. Haidaris. 2004. Susceptibility of Candida species to photodynamic effects of photofrin. Antimicrob. Agents Chemother. 48:2000-2006.
Brown, S. B., and K. J. Mellish. 2001. Verteporfin: a milestone in ophthalmology and photodynamic therapy. Expert Opin. Pharmacother. 2:351-361.
Castano, A. P., T. N. Demidova, and M. R. Hamblin. 2004. Mechanisms in photodynamic therapy: part one—photosensitizers, photochemistry and cellular localization. Photodiagn. Photodyn. Ther. 1:279-293.
Castano, A. P., T. N. Demidova, and M. R. Hamblin. 2005. Mechanisms in photodynamic therapy: part three—photosensitizer pharmacokinetics, biodistribution, tumor localization and modes of tumor destruction. Photodiagn. Photodyn. Ther. 2:91-106.
Castano, A. P., T. N. Demidova, and M. R. Hamblin. 2005. Mechanisms in photodynamic therapy: part two—cellular signalling, cell metabolism and modes of cell death. Photodiagn. Photodyn. Ther. 2:1-23.
Dahl, T. A., W. R. Midden, and P. E. Hartman. 1989. Comparison of killing of gram-negative and gram-positive bacteria by pure singlet oxygen. J. Bacteriol. 171:2188-2194.
Demidova, T. N., and M. R. Hamblin. 2005. Effect of cell-photosensitizer binding and cell density on microbial photoinactivation. Antimicrob. Agents Chemother. 49:2329-2335.
Demidova, T. N., and M. R. Hamblin. 2004. Photodynamic therapy targeted to pathogens. Int. J. Immunopathol. Pharmacol. 17:245-254.
Devine, D. A., and R. E. Hancock. 2002. Cationic peptides: distribution and mechanisms of resistance. Curr. Pharm. Des. 8:703-714.
Dolmans, D. E., D. Fukumura, and R. K. Jain. 2003. Photodynamic therapy for cancer. Nat. Rev. Cancer 3:380-387.
Dykes, G. A., S. Aimoto, and J. W. Hastings. 1998. Modification of a synthetic antimicrobial peptide (ESF1) for improved inhibitory activity. Biochem. Biophys. Res. Commun. 248:268-272.
Gad, F., T. Zahra, K. P. Francis, T. Hasan, and M. R. Hamblin. 2004. Targeted photodynamic therapy of established soft-tissue infections in mice. Photochem. Photobiol. Sci. 3:451-458.
Hamblin, M. R., and T. Hasan. 2004. Photodynamic therapy: a new antimicrobial approach to infectious disease? Photochem. Photobiol. Sci. 3:436-450.
Hamblin, M. R., D. A. O'Donnell, N. Murthy, C. H. Contag, and T. Hasan. 2002. Rapid control of wound infections by targeted photodynamic therapy monitored by in vivo bioluminescence imaging. Photochem. Photobiol. 75:51-57.
Hamblin, M. R., D. A. O'Donnell, N. Murthy, K. Rajagopalan, N. Michaud, M. E. Sherwood, and T. Hasan. 2002. Polycationic photosensitizer conjugates: effects of chain length and Gram classification on the photodynamic inactivation of bacteria. J. Antimicrob. Chemother. 49:941-951.
Hamblin, M. R., T. Zahra, C. H. Contag, A. T. McManus, and T. Hasan. 2003. Optical monitoring and treatment of potentially lethal wound infections in vivo. J. Infect. Dis. 187:1717-1726.
Hancock, R. E., and M. G. Scott. 2000. The role of antimicrobial peptides in animal defenses. Proc. Natl. Acad. Sci. USA 97:8856-8861.
Helander, I. M., H. L. Alakomi, K. Latva-Kala, and P. Koski. 1997. Polyethyleneimine is an effective permeabilizer of gram-negative bacteria. Microbiology 143:3193-3199.
Helander, I. M., K. Latva-Kala, and K. Lounatmaa. 1998. Permeabilizing action of polyethyleneimine on Salmonella typhimurium involves disruption of the outer membrane and interactions with lipopolysaccharide. Microbiology 144:385-390.
Jett, B. D., K. L. Hatter, M. M. Huycke, and M. S. Gilmore. 1997. Simplified agar plate method for quantifying viable bacteria. BioTechniques 23:648-650.
Kralic, I., and S. El Mohsni. 1978. A new method for the detection of singlet oxygen in aqueous solution. Photochem. Photobiol. 28:577-581.
Lahteenmaki, K., S. Edelman, and T. K. Korhonen. 2005. Bacterial metastasis: the host plasminogen system in bacterial invasion. Trends Microbiol. 13:79-85.
Lambrechts, S. A., M. C. Aalders, and J. Van Marle. 2005. Mechanistic study of the photodynamic inactivation of Candida albicans by a cationic porphyrin. Antimicrob. Agents Chemother. 49:2026-2034.
Lauro, F. M., P. Pretto, L. Covolo, G. Jori, and G. Bertoloni. 2002. Photoinactivation of bacterial strains involved in periodontal diseases sensitized by porphycene-polylysine conjugates. Photochem. Photobiol. Sci. 1:468-470.
Lewis, K., and A. M. Klibanov. 2005. Surpassing nature: rational design of sterile-surface materials. Trends Biotechnol. 23:343-348.
Lin, J., S. Qiu, K. Lewis, and A. M. Klibanov. 2002. Bactericidal properties of flat surfaces and nanoparticles derivatized with alkylated polyethylenimines. Biotechnol. Prog. 18:1082-1086.
Lin, J., S. Qiu, K. Lewis, and A. M. Klibanov. 2003. Mechanism of bactericidal and fungicidal activities of textiles covalently modified with alkylated polyethyleneimine. Biotechnol. Bioeng. 83:168-172.
Malik, Z., H. Ladan, and Y. Nitzan. 1992. Photodynamic inactivation of Gram-negative bacteria: problems and possible solutions. J. Photochem. Photobiol. B 14:262-266.
Marquart, M. E., J. J. Dajcs, A. R. Caballero, B. A. Thibodeaux, and R. J. O'Callaghan. 2005. Calcium and magnesium enhance the production of Pseudomonas aeruginosa protease IV, a corneal virulence factor. Med. Microbiol. Immunol. 194:39-45.
Martin, J. P., and N. Logsdon. 1987. Oxygen radicals are generated by dye-mediated intracellular photooxidations: a role for superoxide in photodynamic effects. Arch. Biochem. Biophys. 256:39-49.
Martin, J. P., and N. Logsdon. 1987. The role of oxygen radicals in dye-mediated photodynamic effects in Escherichia coli B. J. Biol. Chem. 262:7213-7219.
Merchat, M., G. Bertolini, P. Giacomini, A. Villanueva, and G. Jori. 1996. Meso-substituted cationic porphyrins as efficient photosensitizers of gram-positive and gram-negative bacteria. J. Photochem. Photobiol. B. 32:153-157.
Minnock, A., D. I. Vernon, J. Schofield, J. Griffiths, J. H. Parish, and S. B. Brown. 2000. Mechanism of uptake of a cationic water-soluble pyridinium zinc phthalocyanine across the outer membrane of Escherichia coli. Antimicrob. Agents Chemother. 44:522-527.
Moan, J., and Q. Peng. 2003. An outline of the hundred-year history of PDT. Anticancer Res. 23:3591-3600.
Moghimi, S. M., P. Symonds, J. C. Murray, A. C. Hunter, G. Debska, and A. Szewczyk. 2005. A two-stage poly(ethylenimine)-mediated cytotoxicity: implications for gene transfer/therapy. Mol. Ther. 11:990-995.
Mor, A., and P. Nicolas. 1994. Isolation and structure of novel defensive peptides from frog skin. Eur. J. Biochem. 219:145-154.
Nitzan, Y., M. Gutterman, Z. Malik, and B. Ehrenberg. 1992. Inactivation of gram-negative bacteria by photosensitized porphyrins. Photochem. Photobiol. 55:89-96.
Piers, K. L., M. H. Brown, and R. E. Hancock. 1994. Improvement of outer membrane-permeabilizing and lipopolysaccharide-binding activities of an antimicrobial cationic peptide by C-terminal modification. Antimicrob. Agents Chemother. 38:2311-2316.
Polo, L., A. Segalla, G. Bertoloni, G. Jori, K. Schaffner, and E. Reddi. 2000. Polylysine-porphycene conjugates as efficient photosensitizers for the inactivation of microbial pathogens. J. Photochem. Photobiol. B 59:152-158.
Rovaldi, C. R., A. Pievsky, N. A. Sole, P. M. Friden, D. M. Rothstein, and P. Spacciapoli. 2000. Photoactive porphyrin derivative with broad-spectrum activity against oral pathogens in vitro. Antimicrob. Agents Chemother. 44:3364-3367.
Soukos, N. S., L. A. Ximenez-Fyvie, M. R. Hamblin, S. S. Socransky, and T. Hasan. 1998. Targeted antimicrobial photochemotherapy. Antimicrob. Agents Chemother. 42:2595-2601.
Thomas, M., Q. Ge, J. J. Lu, J. Chen, and A. M. Klibanov. 2005. Cross-linked small polyethylenimines: while still nontoxic, deliver DNA efficiently to mammalian cells in vitro and in vivo. Pharm. Res. 22:373-380.
Tome, J. P., M. G. Neves, A. C. Tome, J. A. Cavaleiro, M. Soncin, M. Magaraggia, S. Ferro, and G. Jori. 2004. Synthesis and antibacterial activity of new poly-S-lysine-porphyrin conjugates. J. Med. Chem. 47:6649-6652.
Varga, C. M., N. C. Tedford, M. Thomas, A. M. Klibanov, L. G. Griffith, and D. A. Lauffenburger. 2005. Quantitative comparison of polyethylenimine formulations and adenoviral vectors in terms of intracellular gene delivery processes. Gene Ther. 12:1023-1032.
Wainwright, M. 1998. Photodynamic antimicrobial chemotherapy (PACT). J. Antimicrob. Chemother. 42:13-28.
Watanabe, H., S. Miyoshi, T. Kawase, K. Tomochika, and S. Shinoda. 2004. High growing ability of Vibrio vulnificus biotype 1 is essential for production of a toxic metalloprotease causing systemic diseases in humans. Microb. Pathog. 36:117-123.
Wieprecht, T., O. Apostolov, M. Beyermann, and J. Seelig. 2000. Membrane binding and pore formation of the antibacterial peptide PGLa: thermodynamic and mechanistic aspects. Biochemistry 39:442-452.
Yamakoshi, Y., N. Umezawa, A. Ryu, K. Arakane, N. Miyata, Y. Goda, T. Masumizu, and T. Nagano. 2003. Active oxygen species generated from photoexcited fullerene (C60) as potential medicines: O2–* versus 1O2. J. Am. Chem. Soc. 125:12803-12809.(George P. Tegos, Masahiro)
ABSTRACT
We previously showed that covalent conjugates between poly-L-lysine and chlorin(e6) were efficient photosensitizers (PS) of both gram-positive and gram-negative bacteria. The polycationic molecular constructs increased binding and penetration of the PS into impermeable gram-negative cells. We have now prepared a novel set of second-generation polycationic conjugates between chlorin(e6) and three molecular forms of polyethyleneimine (PEI): a small linear, a small cross-linked, and a large cross-linked molecule. The conjugates were characterized by high-pressure liquid chromatography and tested for their ability to kill a panel of pathogenic microorganisms, the gram-positive Staphylococcus aureus and Streptococcus pyogenes, the gram-negative Escherichia coli and Pseudomonas aeruginosa, and the yeast Candida albicans, after exposure to low levels of red light. The large cross-linked molecule efficiently killed all organisms, while the linear conjugate killed gram-positive bacteria and C. albicans. The small cross-linked conjugate was the least efficient antimicrobial PS and its remarkably low activity could not be explained by reduced photochemical quantum yield or reduced cellular uptake. In contrast to polylysine conjugates, the PEI conjugates were resistant to degradation by proteases such as trypsin that hydrolyze lysine-lysine peptide bonds, The advantage of protease stability combined with the ready availability of PEI suggests these molecules may be superior to polylysine-PS conjugates for photodynamic therapy of localized infections.
INTRODUCTION
Photodynamic therapy (PDT) employs a nontoxic dye termed a photosensitizer (PS) and low-intensity visible light, which in the presence of oxygen produce cytotoxic species (4-6). PDT has the advantage of dual selectivity in that the PS can be targeted to its destination cell or tissue, and in addition the illumination can be spatially directed to the lesion. PDT was originally discovered over 100 years ago by its effect on microorganisms (35) but since then has been principally developed as a treatment for cancer (11) and age-related macular degeneration (3).
Because of the well-known increase in multiantibiotic resistance among pathogenic microbes of all classes, PDT has attracted attention as a possible treatment for localized infections (9, 14, 46). It is known that gram-negative bacteria are resistant to PDT with many commonly used PS that will readily lead to phototoxicity for gram-positive species (29) and that PS bearing a cationic charge (33) or the use of agents that increase the permeability of the outer membrane will increase the efficacy of killing of gram-negative organisms (29).
In 1997 our laboratory formed the hypothesis that by covalently conjugating a suitable PS such as chlorin(e6) (ce6) to a poly-L-lysine (pL) chain, a bacterially targeted PS delivery vehicle could be constructed that would efficiently inactivate both gram-positive and gram-negative species (42). Because the resulting polycationic entity is a macromolecule, it is taken up by mammalian cells by the time-dependent process of endocytosis (although it binds rapidly to anionic sulfated chains of proteoglycans on the cell membrane). The polycationic conjugates could effectively bind to microbial cells after incubations as short as 1 minute, thus giving temporal selectivity for bacteria. We (16) compared the effectivenesses of pL-ce6 conjugates with chain lengths of either 8 or 37 lysines attached to precisely one ce6 molecule for bacterial photodynamic inactivation (PDI) and found the 37-lysine conjugate was able to efficiently mediate the photodestruction of both gram-positive and gram-negative species, while the eight-lysine conjugate and free ce6 were effective only against gram-positive bacteria.
A similar conjugate (composed of one ce6 molecule and a 5-amino-acid lysine chain) was subsequently used by another group to kill several oral pathogens in the presence of 25% whole blood (41). Polo and coworkers used conjugates between pL and porphycenes with significant phototoxic activity against gram-negative bacteria (40). In a subsequent report (25) these authors showed that several strains responsible for periodontal disease were efficiently inactivated by visible-light irradiation in the presence of porphycene-pL conjugates. Repeated photosensitization of surviving cells did not induce the selection of resistant bacterial strains or modify their sensitivity to antibiotic treatment. A recent report described the synthesis and antibacterial activities of conjugates between pL and meso-substituted porphyrins (44). We subsequently demonstrated the potency and selectivity of pL-ce6 conjugates and red light for treating actual infections by using the technology to successfully carry out PDT of infected wounds (15, 17), and deep established soft-tissue infections (13) in mice.
In the present report we describe the preparation, high-pressure liquid chromatography (HPLC) characterization, and broad-spectrum antimicrobial photoinactivation potential of a novel set of second-generation polycationic PS conjugates constructed using another commercially available polymer, polyethyleneimine (PEI). This macromolecular vehicle does not contain peptide bonds and therefore should be resistant to protease degradation, which could in theory be a limitation on the use of pL-PS conjugates to carry out in vivo PDT.
MATERIALS AND METHODS
Preparation of the PEI-ce6 conjugates. All reactions were carried out in the dark at room temperature. Linear PEI (80 mg; 189 μmol; average molecular weight, 423; degree of polymerization, 10; Aldrich Chemical Co, St. Louis, MO; catalog number 46,853-3) was dissolved in 3 ml dry dimethyl sulfoxide to which was added 120 mg (200 μmol) ce6 (Porphyrin Products, Logan, UT) and 80 mg (418 μmol) 1-ethyl-3-(3-dimethylaminopropyl)-carbodiimide hydrochloride (Sigma). Triethylamine (20 μl; Sigma) was then added and the mixture was stirred for 24 h. The progress of the reaction was checked by thin-layer chromatography (Polygram SIL G/UV254 plates; Macherey Nagel, Duren, Germany) in methylene chloride-methanol-ammonium hydroxide (8:2:0.5). After 24 h, when the reaction was complete, methanol (1 ml) and water (1 ml) were added and the mixture was evaporated to dryness under vacuum. The residue was dissolved in sodium acetate buffer (10 mM, pH 5.5, 3 ml), applied to a column of Sephadex G-25 (60 by 1 cm) and eluted with the same buffer at a flow rate of 3.3 ml per h; 3-ml fractions were collected, and fractions 14 to 21 were combined and evaporated to give the product PEI-ce6 lin. The substitution ratio was calculated from the absorption spectrum of the conjugate to be 1 ce6 per PEI chain, assuming the absorption coefficient of conjugated ce6 was the same as that of free ce6 (400 = 150,000 M–1 cm–1).
Using the methodology described above, conjugates were prepared between ce6 and two other PEI preparations (low molecular weight [LMW], 600 to 800, branched; Aldrich catalog number 40,871-9; high molecular weight [HMW], 10,000 to 25,000, branched; Aldrich catalog number 40,872-7). Using the same molar ratios of reactants, conjugates (PEI-ce6 LMW and PEI-ce6 HMW) with average substitution ratios of 1 ce6 per PEI chain were obtained.
A pL-ce6 conjugate that consisted of a pL chain of an average length of 164 lysines with an average of four ce6 molecules attached to each chain was used. This was prepared as previously described (15, 17).
HPLC characterization of conjugates. The HPLC system was from Shimadzu (Columbia, MD) and consisted of two LC-10ADvp solvent delivery modules for gradient and isocratic elution and an SPD-M10Avp UV-visible photodiode array detector. The operation and data acquisition of HPLC were conducted with a personal computer and SCl-10Avp system controller (software package CLASS-VP 7.1.1 SP1). Three types of HPLC columns were tested for the purpose of PEI conjugate characterization. These included an Agilent Hypersil BDS-C18 column, 4.6 by 250 mm, 5-μm particle size (Agilent Technologies, Palo Alto, CA); a Hamilton polymer reversed-phase PRP-1 column, 4.1 by 150 mm, 10-μm particle size (Hamilton Company, Reno, NV); and an Agilent Lichrospher 100 diol column, 250 by 4 mm, 5-μm particle size (Agilent Technologies).
Microbial strains and growth conditions. The following microbial strains were used: Staphylococcus aureus (strain 8325-4), Streptococcus pyogenes (strain 591, serotype M49), Escherichia coli (ATCC 25922), Pseudomonas aeruginosa (ATCC 19660), and Candida albicans (ATCC MYA574). Bacteria were routinely grown on brain heart infusion agar and broth and yeasts were routinely grown in yeast- mannitol (YM) broth or agar with shaking at 37°C. S. pyogenes was grown in the presence of 5% CO2. The next day, the cells were diluted 1/100 in fresh medium and left to grow until mid-log phase (optical density [OD], 0.6, equivalent to 108 cells/ml or 107 cells/ml for the yeast).
Photodynamic inactivation of microbial cells. Microbial suspensions in medium were centrifuged for 5 min and the supernatant was replaced with sterile phosphate-buffered saline (PBS); 3.5 ml of the resulting bacterial suspensions (at concentrations of 108/ml or 107 cells/ml for the yeast) were incubated in the dark at room temperature with either 1 or 10 μM of PEI-ce6 conjugates (lin, LMW, and HMW) or free ce6 for 10 min. Cells were incubated in the absence of PS to provide CFU values for an absolute control and to provide cells for the experiments with light alone. Some microbial cell/dye suspensions were centrifuged for 5 min and again the supernatant was replaced with same amount of PBS at the same cell concentration and 500-μl aliquots were then placed in the wells of 48-well plates and the remainder was used for uptake experiments (see below).
The wells were illuminated at room temperature using a 665-nm, 1-W diode laser (model BWF-665-1; B&W Tek, Newark, DE). The laser was coupled into a 1-mm optical fiber that delivered light into a lens. This assembly formed a uniform circular spot on the top of the 48-well plate which was 3 cm in diameter. Red light was delivered at an irradiance of 50 mW/cm2. The fluence range was from 0 to 16 J/cm2 (up to 5 min of illumination time) for PEI-ce6 lin and PEI-ce6 HMW and from 0 to 100 J/cm2 (up to 30 min of illumination time) for PEI-ce6 LMW.
At intervals during the illumination when the requisite fluences had been delivered, aliquots (100 μl) were taken from each well to determine CFU. Care was taken to ensure the contents of the wells were thoroughly mixed before sampling, as bacteria can settle to the bottom. The aliquots were serially diluted 10-fold in PBS to give dilutions of 10–1, 10–2, 10–3, 10–4, 10–5, and 10–6 times the original concentrations and 10-μl aliquots of each of the dilutions were streaked horizontally on square brain heart agar plates by the method of Jett et al. (21). Plates were streaked in triplicate and incubated for 24 h at 37°C in the dark (CO2 incubator for S. pyogenes). In general, three dilutions could be counted on each plate.
Controls were cells not treated with PS or light, cells incubated with PS but kept in 48-well plates covered with aluminum foil at room temperature for the duration of the illumination, and cells treated with light but not with PS. Survival fractions (SF) were routinely expressed as ratios of CFU of bacteria treated with light and of PS to CFU of microbes treated with neither. The SF at 0 J/cm2 gives a measure of the dark toxicity of the conjugates.
Trypsin degradation of conjugates. To compare the effects of trypsin we used a pL-ce6 conjugate prepared similarly to those previously described (15, 17). Conjugates (pL-ce6, PEI-ce6 lin, or PEI-ce6 HMW) were incubated with 1% trypsin solution (Sigma) in PBS without Ca and Mg overnight in the dark at 37°C and mixtures were then boiled at 100°C to inactivate the enzyme. The inactivated trypsin-conjugate mixtures were then used for PDI with E. coli and S. aureus in the same manner as unreacted conjugates. There was no effect of boiling on the solubility or PDI activity of the PEI-ce6 or pL-ce6 conjugate alone.
Photochemical tests. All chemicals were obtained from Aldrich (Milwaukee, WI) and used without further purification. The superoxide assay used nitro blue tetrazolium chloride (NBT) at 80 μM, NADH at 10 mM, and PS at 5 μM ce6 equivalent concentration, all dissolved in PBS (49). All ingredients were freshly prepared prior to the procedure and placed in a quartz cuvette and the spectrophotometer was blanked. Broad-band (400- to 700-nm) white light was delivered at a power of approximately 200 mW incident on the cuvette face, and measurements of the blue formazan product were taken every 10 seconds at 580 nm.
The singlet oxygen assay was based on the oxidation of 4-nitroso-N,N-dimethyl-aniline (RNO) (22). The cuvette contained L-histidine at 15 mM and PS at 5 μM ce6 equivalent concentration and was blanked. RNO was then added to achieve a measured absorbance of approximately 1 at 440 nm (13.3 μM concentration). Illumination was with broad-band white light at an incident power of approximately 3 W, and loss of absorbance was measured every 10 seconds at 440 nm.
Uptake of conjugates by microbial cells. Suspensions of microorganisms (108/ml for bacteria and 107/ml for C. albicans) were incubated for 10 min with PS in the dark at room temperature. Unbound PS was washed out by centrifugation of the mixture of dye and microorganisms for 6 min at 1,550 x g followed by resuspension of washed pellets in 3.5 ml PBS without Ca2+ and Mg2+. Aliquots (500 μl) of these suspensions were used for PDI experiments, and the remaining suspensions were centrifuged again and the pellets were dissolved in 6 ml 0.1 M NaOH-1% sodium dodecyl sulfate for at least 24 h. These suspensions were used for PS uptake measurement.
The fluorescence of dissolved pellets was measured on a spectrofluorimeter (FluoroMax3; SPEX Industries, Edison, NJ). The excitation wavelength was 400 nm and the emission spectra of the solution were recorded from 580 to 700 nm. The fluorescence was calculated from the height of the peaks recorded. Calibration curves were made from pure PS dissolved in NaOH-sodium dodecyl sulfate and used for determination of PS concentration in the suspension. Uptake values were obtained by dividing the number of nmol of PS in the dissolved pellet by the number of CFU obtained by serial dilutions and the number of PS molecules/cell was calculated by using Avogadro's number.
Statistics. Values are reported as the mean ± standard error of the mean (SEM). Differences between two means were assessed for significance by the two-tailed Student's t test, assuming equal or unequal variances of the standard deviations as appropriate. A value of P of <0.05 was considered significant.
RESULTS
Preparation and characterization of conjugates. Three types of HPLC column from different vendors were tested to develop a method to characterize the PEI conjugates (Fig. 1). To our knowledge, there have been no reports on HPLC analysis of PEI or its related compounds. Polyelectrolytes, such as PEI, are charged at most pHs. In PEI, the amino groups tend to be positively charged due to protonation at pH 7, which is usually the operational pH for most silica-based HPLC columns. These positively charged groups are problematic for HPLC analysis due to analyte stationary-phase interaction, which leads to partial or complete adsorption of analytes to the HPLC column, a phenomenon usually seen in protein and peptide HPLC analyses.
In this study, we observed complete adsorption of the PEI-ce6 conjugate to the Agilent C18 column. The polymer reversed-phase column (Hamilton PRP-1) was brought in for testing with the hope that the noncharged polymer stationary phase would have minimal electrostatic interaction with the PEI analytes. But the PRP column yielded broad peaks for all three PEI conjugates and hence is not the ideal column for PEI conjugate characterization. The diol column (Agilent Lichrospher 100 diol), on the other hand, has a stationary phase that combines a hydrophilic external surface (diol groups) and a hydrophobic internal surface (alkane linker) that can effectively preclude the previously described PEI-silanol group interaction. With the HPLC method developed in this study, the Lichrospher diol column showed different retention times and peak shapes for the three conjugates. The peaks obtained are shown in Fig. 2. The order of elution from the column is in order of increasing molecular weight (PEI-ce6 lin, 1,000; PEI-ce6 LMW, 1,300; and PEI-ce6 HMW, 18,000).
Dark toxicity of PEI and PEI conjugates. In order to compare the levels of dark toxicity exhibited by the PEI conjugates and by the unconjugated PEI preparations, we incubated the three PEI-ce6 conjugates and the corresponding PEI compounds at 10 μM in PBS with E. coli for 10 min. The survival fractions were as follows: PEI lin, 0.018 ± 0.0037, versus PEI-ce6 lin, 0.958 ± 0.13 (P < 0.0001); PEI LMW, 0.157 ± 0.058, versus PEI-ce6 LMW, 0.526 ± 0.225 (P = 0.18); and PEI HMW, 0.0067 ± 0.00011, versus PEI-ce6 HMW, 0.57 ± 0.267 (P < 0.0001). We conclude that the dark toxicity due to the polycationic nature of some PEI preparations is markedly reduced after conjugation to ce6, presumably because the anionic lipophilic dye molecule ameliorates the effect of multiple basic amino groups.
Photoinactivation of gram-positive species. We used two gram-positive species, S. aureus and S. pyogenes, which are both important human and animal pathogens. It rapidly became clear that the conjugate PEI-ce6 LMW had much lower activity than both the PEI-ce6 lin and PEI-ce6 HMW conjugates, with free ce6 having intermediate activity. Nevertheless, we used the same concentration of all four PS, i.e., 1 μM ce6 equivalent, and delivered additional light (up to 100 J/cm2 for PEI-ce6 LMW and up to 40 J/cm2 for free ce6) in an attempt to increase the extent of killing. As shown in Fig. 3A for S. aureus and Fig. 3B for S. pyogenes, there is a dramatic difference in the effectiveness of PDI between PEI-ce6 LMW (less than 2 logs of killing after 100 J/cm2) and the other two conjugates (5 to 6 logs of killing after 16 J/cm2). Free ce6 was intermediate in potency. The linear and HMW conjugates were similarly effective in mediating killing of S. pyogenes (Fig. 3B) but the HMW conjugate was significantly better (P < 0.05 at 8 and 16 J/cm2) than PEI-ce6 lin at mediating the killing of S. aureus (Fig. 3A).
Photoinactivation of gram-negative species. We studied two gram-negative species that are pathogenic to humans and animals, E. coli and P. aeruginosa. Because gram-negative species are more resistant to PDI than gram-positive bacteria (29), we increased the concentration of PS 10-fold over that used for gram positives (to 10 μM). Because of the relatively poor activity of PEI-ce6 LMW and ce6, we again delivered light in doses of up to 100 J/cm2 and 40 J/cm2 instead of the dose of 16 J/cm2 used for the other conjugates. Figure 4A shows the PDI of E. coli and Fig. 4B shows the PDI of P. aeruginosa. For E. coli the order of effectiveness is PEI-ce6 HMW > PEI-ce6 lin > ce6 > PEI-ce6 LMW. However, PEI-ce6 LMW is actually more effective against E. coli than against any of the other microbes tested. For P. aeruginosa the order of effectiveness is similar to that found for E. coli except there is little difference between PEI-ce6 lin and free ce6. Compared to E. coli, P. aeruginosa is more susceptible to PEI-ce6 HMW and less susceptible to PEI-ce6 LMW. The biggest difference observed between the gram-positive and gram-negative species is that PEI-ce6 lin is highly effective (5 logs of killing at 1 μM) against the former but almost ineffective (1 log of killing at 10 μM) against the latter.
Photoinactivation of C. albicans. We used C. albicans cells at a cell density 1/10 that of the bacterial species (107 instead of 108 cells/ml). We had previously showed that C. albicans cells are 10 times larger than bacterial cells by dry weight and therefore each cell needs to bind 10 times as much PS to allow equal killing (8). Figure 5 shows the killing of C. albicans mediated by the four PS incubated at 10 μM. In a pattern similar to that of the gram-positive bacteria, PEI-ce6 LMW is the least effective (2 logs at 100 J/cm2) and ce6 is next (2.5 logs at 40 J/cm2), while both PEI-ce6 lin and PEI-ce6 HMW are highly effective (4 to 6 logs at 16 J/cm2), although PEI-ce6 HMW is significantly better (P < 0.05) at killing the yeast than PEI-ce6 lin.
Effect of trypsin degradation of conjugates. We wished to test the hypothesis that pL-ce6 conjugates could be degraded by protease enzymes, thus destroying their polycationic character, while PEI-ce6 conjugates would be resistant to proteases because they do not contain peptide bonds. Figure 6A shows the PDI killing of E. coli obtained using PEI-ce6 HMW and pL-ce6 with and without trypsin treatment. While the killing produced by PEI-ce6 HMW with and without trypsin treatment was identical and similar to the killing produced by pL-ce6, the effect of trypsin treatment on the pL-ce6 conjugate was to significantly (P < 0.001) reduce the killing compared to that with the other three PS and gave greater than a 3-log difference in SF of E. coli at 16 J/cm2.
Figure 6B shows the effect of trypsin treatment of pL-ce6 and PEI-ce6 lin on the efficacy of PDI against S. aureus. PEI-ce6 lin was equally effective with and without trypsin treatment; however, pL-ce6 was significantly less effective (P < 0.01) than PEI-ce6 lin. Interestingly, when pL-ce6 was treated with trypsin it became much more effective at mediating killing (P < 0.0001), giving almost total killing after only 4 J/cm2.
Relative ROS production on illumination. In an attempt to explain the remarkably low PDI activity exhibited by PEI-ce6 LMW, we tested the hypothesis that its quantum yield of reactive oxygen species (ROS) was significantly lower than that of the other conjugates. Illumination of PS can proceed via type I or type II mechanisms (4). Type I frequently involves the light-mediated production of superoxide anion radical (O2– ·) by electron transfer to and from the triplet state in the presence of a reducing agent, while type II involves the production of singlet oxygen by energy transfer to ground-state oxygen from the PS triplet state. Both processes have been implicated in the PDI of microorganisms (7, 31, 32). The type I process is conveniently monitored in aqueous solution by the reduction of nitro blue tetrazolium in the presence of NADH (49), while the type II process can be monitored in aqueous solution by oxidation of RNO in the presence of histidine (22). Because the NBT reaction is a reduction and the RNO reaction is an oxidation there is no cross talk between the two reaction intermediates and their respective products.
As shown in Fig. 7A, there are no significant differences between the rates of NBT reduction caused by superoxide produced by any of the four PS tested. Figure 7B shows the relative rates of RNO photobleaching in the presence of histidine caused by singlet oxygen production by the four PS. As can be seen, the rates are in the order ce6 > PEI-ce6 LMW > PEI-ce6 lin > PEI-ce6 HMW. The differences between ce6 and PEI-ce6 LMW and between PEI-ce6 LMW and PEI-ce6 lin are significant (P < 0.05).
Uptake of conjugates by microbial cells. We then asked if the relative inactivity of PEI-ce6 LMW could be explained by markedly lower cell uptake of PS. Because we centrifuged the PS-incubated cells before illumination, only dye that actually bound to the cells was active in the PDI. Figure 8 depicts the uptake in terms of molecules of ce6 equivalent per cell after incubation of the gram-positive organisms with 1 μM and the other cells with a 10 μM equivalent concentration of ce6. As can be seen, the order of uptake of the four PS by both of the gram-positive species (S. aureus and S. pyogenes) was PEI-ce6 HMW > PEI-ce6-lin > ce6 > PEI-ce6 LMW. PEI-ce6 LMW was taken up significantly less (P < 0.01) than either PEI-ce6 HMW or PEI-ce6 lin. For the gram-negative species (E. coli and P. aeruginosa) the order was PEI-ce6 HMW > PEI-ce6-LMW > PEI-ce6 lin >> ce6. PEI-ce6 LMW uptake was significantly (P < 0.01) lower than that of PEI-ce6 HMW and higher than that of free ce6, but not significantly different from that of PEI-ce6 lin. C. albicans had the same order of PS uptake as the gram-negative organisms except that the uptake of ce6 was only very slightly lower than that of the conjugates.
DISCUSSION
This study has shown that covalent conjugates between PEI and ce6 can be potent broad-spectrum antimicrobial PS resistant to protease degradation and therefore constitute an alternative to the previously described pL-ce6 conjugates. The development of an HPLC method to characterize these cationic polymeric conjugates demonstrates that the preparations are free of nonconjugated ce6 and shows their retention times are proportional to their molecular weights.
The high susceptibility of the gram-positive organisms S. aureus and S. pyogenes to PDI was not unexpected. We used 1/10 the concentration of all the PS and obtained overall levels of killing roughly comparable to those obtained with gram-negative and yeast cells at a 10-fold-higher concentration. Many workers have shown that gram-positive bacteria are much more susceptible to PDI than other microbial cell types (29, 38, 46). We found that PEI-ce6 lin and PEI-ce6 HMW had roughly equal activities against S. aureus, and PEI-ce6 HMW was better against S. pyogenes. This was somewhat surprising since we had previously shown (16) a structure-function relationship in a pair of pL conjugates, where a conjugate that had an average of eight lysine monomers joined to one ce6 molecule performed better in killing S. aureus than a conjugate that had an average of 37 lysines joined to one ce6.
We proposed that the relatively porous outer coat of gram-positive species allows the smaller conjugate to pass through more efficiently, while the larger conjugate is bound to the negatively charged outer structures of the gram-positive cells due to its more pronounced cationic charge but cannot diffuse through the permeability barrier as efficiently due to its greater size. Presumably the efficiency of these polycationic conjugates in inactivating gram-positive species is partly a balance between molecular size (smaller is better) and cationic charge (more is better). This point is demonstrated in Fig. 7, where the pL-ce6 conjugate (MW 20,000) is less effective than PEI-ce6 lin and PEI-ce6 HMW but becomes much more effective after proteolysis.
The superiority of the large HMW conjugate in mediating the PDI of gram-negative species was also expected. In this case we propose that the larger, more polycationic structure can more effectively displace divalent cation from the lipopolysaccharide molecules in the outer membrane and is better able to increase outer membrane permeability, allowing access to targets such as the inner membrane, This process is known as self-promoted uptake (10, 34, 39) and is typified by the bactericidal mechanism of naturally occurring cationic antimicrobial peptides expressed by most life forms, such as defensins, cecropin, and gramicidin (18).
The yeast C. albicans behaved similarly to gram-positive bacteria in being effectively killed by the linear and HMW conjugates and moderately killed by free ce6 after exposure to modest levels of light. Each Candida cell is 10 times heavier than a Staphylococcus cell (8) and therefore 10 times more PS is needed to kill each cell (10 μM as opposed to 1 μM). Other workers have found that C. albicans, like gram-positive bacteria, is killed by PDI using noncationic PS (2, 24)
The remarkably low PDI activity of the LMW conjugate (especially compared to that of PEI-ce6 lin, which appears to have a very similar molecular structure in terms of size and charge) was surprising. It was also remarkable that PEI-ce6 LMW was in every case less active than free ce6. This might be possible to explain in the case of the gram-positive bacteria on the basis of increased diffusion of smaller compounds regardless of charge. We previously showed that free ce6 was more active against S. aureus than a pL-ce6 conjugate with 37 lysine residues. However, the greater activity of free ce6 than of PEI-ce6 LMW against both of the gram-negative species was particularly surprising.
We then asked if the photophysical properties of the photoactive dye had been adversely altered in the LMW conjugate compared to the other conjugates and free ce6. We therefore carried out two photochemical assays involving singlet oxygen-mediated oxidation of RNO in the presence of histidine and superoxide-mediated reduction of NBT in the presence of NADH to answer this question. The results showed that the rates of photochemical superoxide production were fairly similar among all four PS. However, we were surprised to observe that PEI-ce6 LMW had the second highest rate of photochemical singlet oxygen production (after free ce6, which was expected to be the highest), and PEI-ce6 lin and PEI-ce6 HMW, which were most effective at killing cells, actually had the lowest rates of 1O2 formation.
We then argued that the low activity of PEI-ce6 LMW must be because the molecule did not bind as effectively to the microbial cells. Therefore we examined uptake of the four PS by all five microbial species. The uptake of PEI-ce6 lin by the gram-positive cells was significantly the lowest of all four PS tested, but whether this observation represents a difference big enough to explain the very low PDI activity is debatable. The uptake of PEI-ce6 LMW by the gram-negative species and C. albicans was intermediate between that of PEI-ce6 lin and PEI-ce6 HMW. This was as we initially expected based on the increasing cationic character of the molecules. Low cellular binding then cannot explain the poor PDI efficacy of PEI-ce6 LMW against the gram-negative and fungal species. The low uptake of free ce6 by the gram-negative cells is due to the permeability barriers for anionic molecules exhibited by these species, and the high uptake of free ce6 by the yeast cells was expected, and this difference correlates with PDI killing.
It is possible that besides the effects of molecular size and overall cationic charge, another molecular property affects the microbial PDI activity of these conjugates. This may be due to changes in tertiary structure. It is known that naturally occurring antimicrobial peptides assume a helical structure with the cationic charges on one side of the helix and hydrophobic amino acid residues on the opposite side (12, 37, 48). Although PEI-ce6 lin and PEI-ce6 LMW are comparable in size and charge, the chief difference between the molecules is that one is a linear chain while the other is cross-linked. Therefore, the small cross-linked molecule may be unable to form the appropriate tertiary structure to increase microbial binding and penetration. Although the PEI-ce6 HMW conjugate is also cross-linked it is possible that its much greater content of cationic groups or its increased molecular flexibility due to its greater size could overcome the restriction seen with PEI-ce6 LMW. The finding that the dark toxicity of unconjugated PEI LMW is significantly lower than that of the other two PEI preparations also suggests that the PEI LMW structure has some fundamental difference from that of the other PEI compounds in how it binds or penetrates bacteria.
PEI and pL have been reported to act similarly on gram-negative bacteria (19, 20). Helander et al. demonstrated that PEI increased cell permeability to the fluorescent dye 1-N-phenylnaphthylamine (19). Several papers by Klibanov and colleagues (26, 27) have shown that alkylated derivatives of PEI can exert an antimicrobial effect when incorporated into surfaces and textiles. This was proposed to occur due to the fact that the alkyl chains were long enough to penetrate from the surface to the bacterial membrane (28).
Although pL-ce6 conjugates are highly efficient in mediating PDI of multiple microbial species in vitro (8, 16, 42) and in vivo (13, 15, 17), their structure is based on polypeptides composed of naturally occurring L-amino acids. This means that they are likely to be substrates for protease enzymes that are found in tissue. In particular, extracellular proteases derived from the host cells are found in areas of inflammation such as infections (1, 23) and proteases may also be produced by bacteria themselves as secreted virulence factors involved in tissue invasion (30, 47).
Many proteases have particular sequences of amino acids that are recognized and preferentially cleaved by the enzyme, but some are more general in their substrate specificity. Trypsin is a digestive proteinase from the pancreas that hydrolyzes peptide bonds in protein molecules that have carboxyl groups donated by the basic amino acids arginine and lysine, and therefore served as a proof-of-principle enzyme for the ability to cleave pL conjugates in solution. The effect of trypsin treatment of pL-ce6 on killing of E. coli was to significantly reduce the number of cells that were killed. This is presumably because the products are short oligopeptides, and if these peptides still bear a ce6 molecule they are not sufficiently polycationic to penetrate the outer membrane. By contrast, the effect of trypsin-treated pL-ce6 on the PDI of S. aureus was to dramatically increase killing. This observation can be explained by fact that the short peptides attached to PS molecules that are the products of proteolysis are better able to penetrate gram-positive cells than intact large molecules.
It is at present uncertain whether the peptide bond between the epsilon amino group of lysine and ce6 is hydrolyzed by trypsin, but it would seem unlikely. It should be pointed out that it is at present unclear if protease stability would be a real advantage in the use of these conjugates in PDT of actual infections. First, resistance to degradation would be an advantage only in combating gram-negative infections, and second, resistance to degradation might hinder removal of the remaining conjugate from the tissue after completion of the illumination.
In conclusion, second-generation polycationic conjugates between various PEI preparations and ce6 have been prepared and characterized. PEI has been widely studied as a DNA transfection agent for nonviral gene therapy and much is known about its properties and cytotoxicity (36, 43, 45). The broad-spectrum antimicrobial PDI activity of PEI-ce6 coupled with the possible advantage of protease stability and the ready availability of PEI suggest these molecules may be superior to pL-PS conjugates for PDT of localized infections.
ACKNOWLEDGMENTS
This work was funded by U.S. National Institutes of Health grant R01AI050875 to M. R. Hamblin. T. N. Demidova was supported by a Wellman Center of Photomedicine graduate student fellowship. M. Anbe was funded by the Ministry of Education, Culture, Sports, Science and Technology, Nano Bioengineering Education Program, University of Tokyo, Tokyo, Japan.
Deceased.
REFERENCES
Behera, A. K., E. Hildebrand, J. Scagliotti, A. C. Steere, and L. T. Hu. 2005. Induction of host matrix metalloproteinases by Borrelia burgdorferi differs in human and murine Lyme arthritis. Infect. Immun. 73:126-134.
Bliss, J. M., C. E. Bigelow, T. H. Foster, and C. G. Haidaris. 2004. Susceptibility of Candida species to photodynamic effects of photofrin. Antimicrob. Agents Chemother. 48:2000-2006.
Brown, S. B., and K. J. Mellish. 2001. Verteporfin: a milestone in ophthalmology and photodynamic therapy. Expert Opin. Pharmacother. 2:351-361.
Castano, A. P., T. N. Demidova, and M. R. Hamblin. 2004. Mechanisms in photodynamic therapy: part one—photosensitizers, photochemistry and cellular localization. Photodiagn. Photodyn. Ther. 1:279-293.
Castano, A. P., T. N. Demidova, and M. R. Hamblin. 2005. Mechanisms in photodynamic therapy: part three—photosensitizer pharmacokinetics, biodistribution, tumor localization and modes of tumor destruction. Photodiagn. Photodyn. Ther. 2:91-106.
Castano, A. P., T. N. Demidova, and M. R. Hamblin. 2005. Mechanisms in photodynamic therapy: part two—cellular signalling, cell metabolism and modes of cell death. Photodiagn. Photodyn. Ther. 2:1-23.
Dahl, T. A., W. R. Midden, and P. E. Hartman. 1989. Comparison of killing of gram-negative and gram-positive bacteria by pure singlet oxygen. J. Bacteriol. 171:2188-2194.
Demidova, T. N., and M. R. Hamblin. 2005. Effect of cell-photosensitizer binding and cell density on microbial photoinactivation. Antimicrob. Agents Chemother. 49:2329-2335.
Demidova, T. N., and M. R. Hamblin. 2004. Photodynamic therapy targeted to pathogens. Int. J. Immunopathol. Pharmacol. 17:245-254.
Devine, D. A., and R. E. Hancock. 2002. Cationic peptides: distribution and mechanisms of resistance. Curr. Pharm. Des. 8:703-714.
Dolmans, D. E., D. Fukumura, and R. K. Jain. 2003. Photodynamic therapy for cancer. Nat. Rev. Cancer 3:380-387.
Dykes, G. A., S. Aimoto, and J. W. Hastings. 1998. Modification of a synthetic antimicrobial peptide (ESF1) for improved inhibitory activity. Biochem. Biophys. Res. Commun. 248:268-272.
Gad, F., T. Zahra, K. P. Francis, T. Hasan, and M. R. Hamblin. 2004. Targeted photodynamic therapy of established soft-tissue infections in mice. Photochem. Photobiol. Sci. 3:451-458.
Hamblin, M. R., and T. Hasan. 2004. Photodynamic therapy: a new antimicrobial approach to infectious disease? Photochem. Photobiol. Sci. 3:436-450.
Hamblin, M. R., D. A. O'Donnell, N. Murthy, C. H. Contag, and T. Hasan. 2002. Rapid control of wound infections by targeted photodynamic therapy monitored by in vivo bioluminescence imaging. Photochem. Photobiol. 75:51-57.
Hamblin, M. R., D. A. O'Donnell, N. Murthy, K. Rajagopalan, N. Michaud, M. E. Sherwood, and T. Hasan. 2002. Polycationic photosensitizer conjugates: effects of chain length and Gram classification on the photodynamic inactivation of bacteria. J. Antimicrob. Chemother. 49:941-951.
Hamblin, M. R., T. Zahra, C. H. Contag, A. T. McManus, and T. Hasan. 2003. Optical monitoring and treatment of potentially lethal wound infections in vivo. J. Infect. Dis. 187:1717-1726.
Hancock, R. E., and M. G. Scott. 2000. The role of antimicrobial peptides in animal defenses. Proc. Natl. Acad. Sci. USA 97:8856-8861.
Helander, I. M., H. L. Alakomi, K. Latva-Kala, and P. Koski. 1997. Polyethyleneimine is an effective permeabilizer of gram-negative bacteria. Microbiology 143:3193-3199.
Helander, I. M., K. Latva-Kala, and K. Lounatmaa. 1998. Permeabilizing action of polyethyleneimine on Salmonella typhimurium involves disruption of the outer membrane and interactions with lipopolysaccharide. Microbiology 144:385-390.
Jett, B. D., K. L. Hatter, M. M. Huycke, and M. S. Gilmore. 1997. Simplified agar plate method for quantifying viable bacteria. BioTechniques 23:648-650.
Kralic, I., and S. El Mohsni. 1978. A new method for the detection of singlet oxygen in aqueous solution. Photochem. Photobiol. 28:577-581.
Lahteenmaki, K., S. Edelman, and T. K. Korhonen. 2005. Bacterial metastasis: the host plasminogen system in bacterial invasion. Trends Microbiol. 13:79-85.
Lambrechts, S. A., M. C. Aalders, and J. Van Marle. 2005. Mechanistic study of the photodynamic inactivation of Candida albicans by a cationic porphyrin. Antimicrob. Agents Chemother. 49:2026-2034.
Lauro, F. M., P. Pretto, L. Covolo, G. Jori, and G. Bertoloni. 2002. Photoinactivation of bacterial strains involved in periodontal diseases sensitized by porphycene-polylysine conjugates. Photochem. Photobiol. Sci. 1:468-470.
Lewis, K., and A. M. Klibanov. 2005. Surpassing nature: rational design of sterile-surface materials. Trends Biotechnol. 23:343-348.
Lin, J., S. Qiu, K. Lewis, and A. M. Klibanov. 2002. Bactericidal properties of flat surfaces and nanoparticles derivatized with alkylated polyethylenimines. Biotechnol. Prog. 18:1082-1086.
Lin, J., S. Qiu, K. Lewis, and A. M. Klibanov. 2003. Mechanism of bactericidal and fungicidal activities of textiles covalently modified with alkylated polyethyleneimine. Biotechnol. Bioeng. 83:168-172.
Malik, Z., H. Ladan, and Y. Nitzan. 1992. Photodynamic inactivation of Gram-negative bacteria: problems and possible solutions. J. Photochem. Photobiol. B 14:262-266.
Marquart, M. E., J. J. Dajcs, A. R. Caballero, B. A. Thibodeaux, and R. J. O'Callaghan. 2005. Calcium and magnesium enhance the production of Pseudomonas aeruginosa protease IV, a corneal virulence factor. Med. Microbiol. Immunol. 194:39-45.
Martin, J. P., and N. Logsdon. 1987. Oxygen radicals are generated by dye-mediated intracellular photooxidations: a role for superoxide in photodynamic effects. Arch. Biochem. Biophys. 256:39-49.
Martin, J. P., and N. Logsdon. 1987. The role of oxygen radicals in dye-mediated photodynamic effects in Escherichia coli B. J. Biol. Chem. 262:7213-7219.
Merchat, M., G. Bertolini, P. Giacomini, A. Villanueva, and G. Jori. 1996. Meso-substituted cationic porphyrins as efficient photosensitizers of gram-positive and gram-negative bacteria. J. Photochem. Photobiol. B. 32:153-157.
Minnock, A., D. I. Vernon, J. Schofield, J. Griffiths, J. H. Parish, and S. B. Brown. 2000. Mechanism of uptake of a cationic water-soluble pyridinium zinc phthalocyanine across the outer membrane of Escherichia coli. Antimicrob. Agents Chemother. 44:522-527.
Moan, J., and Q. Peng. 2003. An outline of the hundred-year history of PDT. Anticancer Res. 23:3591-3600.
Moghimi, S. M., P. Symonds, J. C. Murray, A. C. Hunter, G. Debska, and A. Szewczyk. 2005. A two-stage poly(ethylenimine)-mediated cytotoxicity: implications for gene transfer/therapy. Mol. Ther. 11:990-995.
Mor, A., and P. Nicolas. 1994. Isolation and structure of novel defensive peptides from frog skin. Eur. J. Biochem. 219:145-154.
Nitzan, Y., M. Gutterman, Z. Malik, and B. Ehrenberg. 1992. Inactivation of gram-negative bacteria by photosensitized porphyrins. Photochem. Photobiol. 55:89-96.
Piers, K. L., M. H. Brown, and R. E. Hancock. 1994. Improvement of outer membrane-permeabilizing and lipopolysaccharide-binding activities of an antimicrobial cationic peptide by C-terminal modification. Antimicrob. Agents Chemother. 38:2311-2316.
Polo, L., A. Segalla, G. Bertoloni, G. Jori, K. Schaffner, and E. Reddi. 2000. Polylysine-porphycene conjugates as efficient photosensitizers for the inactivation of microbial pathogens. J. Photochem. Photobiol. B 59:152-158.
Rovaldi, C. R., A. Pievsky, N. A. Sole, P. M. Friden, D. M. Rothstein, and P. Spacciapoli. 2000. Photoactive porphyrin derivative with broad-spectrum activity against oral pathogens in vitro. Antimicrob. Agents Chemother. 44:3364-3367.
Soukos, N. S., L. A. Ximenez-Fyvie, M. R. Hamblin, S. S. Socransky, and T. Hasan. 1998. Targeted antimicrobial photochemotherapy. Antimicrob. Agents Chemother. 42:2595-2601.
Thomas, M., Q. Ge, J. J. Lu, J. Chen, and A. M. Klibanov. 2005. Cross-linked small polyethylenimines: while still nontoxic, deliver DNA efficiently to mammalian cells in vitro and in vivo. Pharm. Res. 22:373-380.
Tome, J. P., M. G. Neves, A. C. Tome, J. A. Cavaleiro, M. Soncin, M. Magaraggia, S. Ferro, and G. Jori. 2004. Synthesis and antibacterial activity of new poly-S-lysine-porphyrin conjugates. J. Med. Chem. 47:6649-6652.
Varga, C. M., N. C. Tedford, M. Thomas, A. M. Klibanov, L. G. Griffith, and D. A. Lauffenburger. 2005. Quantitative comparison of polyethylenimine formulations and adenoviral vectors in terms of intracellular gene delivery processes. Gene Ther. 12:1023-1032.
Wainwright, M. 1998. Photodynamic antimicrobial chemotherapy (PACT). J. Antimicrob. Chemother. 42:13-28.
Watanabe, H., S. Miyoshi, T. Kawase, K. Tomochika, and S. Shinoda. 2004. High growing ability of Vibrio vulnificus biotype 1 is essential for production of a toxic metalloprotease causing systemic diseases in humans. Microb. Pathog. 36:117-123.
Wieprecht, T., O. Apostolov, M. Beyermann, and J. Seelig. 2000. Membrane binding and pore formation of the antibacterial peptide PGLa: thermodynamic and mechanistic aspects. Biochemistry 39:442-452.
Yamakoshi, Y., N. Umezawa, A. Ryu, K. Arakane, N. Miyata, Y. Goda, T. Masumizu, and T. Nagano. 2003. Active oxygen species generated from photoexcited fullerene (C60) as potential medicines: O2–* versus 1O2. J. Am. Chem. Soc. 125:12803-12809.(George P. Tegos, Masahiro)