|Home | About | Journals | Submit | Contact Us | Français|
Photodynamic therapy (PDT) was initially discovered over 100 years ago by its ability to kill microorganisms, but its use to treat infections clinically has not been much developed. However, the present relentless increase in antibiotic resistance worldwide and the emergence of strains that are resistant to all known antibiotics has stimulated research into novel antimicrobial strategies such as PDT that are thought to be unlikely to lead to the development of resistance. In this chapter we will cover the use of PDT to kill pathogenic microbial cells in vitro and describe a mouse model of localized infection and its treatment by PDT without causing excessive damage to the host tissue.
It has been known since the first days of PDT early in the last century that certain microorganisms can be killed by the combination of non-toxic dyes, known as photosensitizers (PSs), and harmless visible light in vitro (1, 2). Throughout the years since those times there have been additional reports of bacteria, yeasts, fungi, and viruses being killed or inactivated by various combinations of PSs and light (3, 4). In the 1990s it was observed that there was a fundamental difference in susceptibility to PDT between Gram-positive and Gram-negative bacteria. It was found that in general neutral or anionic PS molecules are efficiently bound to and photodynamically inactivate Gram-positive bacterial and fungal cells, whereas Gram-negative bacterial cells are relatively resistant to these compounds (5). The high susceptibility of Gram-positive bacteria and fungi was explained by their physiology as their cytoplasmic membrane is surrounded by a relatively porous layer of peptidoglycan and lipoteichoic acid, or beta-glucan and chitin, respectively, and both these structures allow non-cationic PSs to cross (6). Then several groups of workers devised approaches that would allow PDI of Gram-negative species (5). These methods included using the polycationic peptide polymyxin B nonapeptide (5) and EDTA (7), which both increased the permeability of the Gram-negative outer membrane and allowed PSs that are normally excluded from the cell to penetrate to a location where the reactive oxygen species (ROS) generated on illumination to execute fatal damage. A second approach adopted by several groups is to use a PS molecule with an intrinsic positive charge. Wilson, Wainwright, and other groups have used the phenothiazinium salts such as toluidine blue O to carry out PDI of a large range of both Gram-positive and Gram-negative bacteria (8). The group in Italy led by Jori has used cationic porphyrins to photoinactivate Gram-negative species such as Vibrio anguillarum and E. coli (9). They found that washing the loosely bound PS from the cells before illumination decreased the killing and explained this finding by supposing that the first dose of light on PS bound to the outside of the outer membrane causes an initial limited photodamage that then allows further penetration of the PS (10). The group in Leeds, UK, led by Brown has used cationic phthalocyanines for PDI of Gram-negative bacteria (11). They investigated E. coli DH5a and in particular the mechanism of uptake. They found that incubation with PPC in the dark led to increased sensitivity of the bacteria to hydrophobic but not hydrophilic antibiotics. Incubation with PPC also led to increased uptake of radiolabeled protoporphyrin that was reversed in the presence of up to 50 mM Mg++ ions. These observations were consistent with the uptake of PPC proceeding through the self-promoted uptake pathway (12).
Our laboratory has introduced an approach to antimicrobial PDT in which an anionic PS (chlorin e6) is covalently conjugated to polymers with basic amino groups that can bear a cationic charge at biological pH values (13). These molecular constructs can be formed between poly-l-lysine chains (pL-ce6) or polyethylenimine polymers (either linear or branched; PEIce6). We have previously shown that these pL-ce6 and PEI-ce6 conjugates are highly effective in mediating the PDI of both Gram-positive and Gram-negative bacteria. Their positive charges help them to bind to the negatively charged bacteria and their polycationic nature enables them to penetrate the outer membrane of Gram-negative cells by disturbing the structure of the lipopolysaccharide layers. The macromolecular nature of these conjugates gives a temporal selectivity for bacteria over mammalian cells as the latter take them up by the time-dependent process of endocytosis, while they bind rapidly to bacteria.
Although PDI of bacteria has been known for over a 100 years (14), its use to treat infections has not been much developed (8). This may be partly due to the difficulty of monitoring the effectiveness of PDT in animal models of infection. The standard method of quantifying bacterial burdens in animal models of infection involves sacrificing the animal, removing tissue and homogenizing it, and carrying out serial dilutions to provide the number of CFU/g tissue. In order to improve this process of monitoring infection after PDT, we have developed a procedure that uses bioluminescent genetically engineered bacteria and a light-sensitive imaging system to allow real-time visualization of infections (15). The use of microbial cells that have been engineered to express luciferase and the imaging of their location and cell number has streamlined and refined studies involving PDT in animal infection models as the necessity to sacrifice the animals to acquire data on the progress of the infection has essentially been eliminated. Bacterial pathogenesis appeared to be unaffected by the presence of the luciferase genes, and bioluminescence can be detected throughout the study period in animals. Furthermore, the intensity of the bioluminescence measured from the living animal correlated well with the bacterial burden subsequently determined by standard protocols. When these bacteria are treated with PDT in vitro, the loss of luminescence parallels the loss of colony-forming ability.
We have developed several models of infections in wounds and soft-tissue abscesses in mice that can be followed by bioluminescence imaging. The size and intensity of the infection can be sequentially monitored in a non-invasive fashion in individual mice in real time. When photosensitizers are introduced into the infected tissue followed by illumination with red light, a light dose-dependent loss of luminescence is seen. If the bacterium is invasive, the loss of luminescence correlates with increased survival of the mice, while animals in control groups die of sepsis within 5 days. Healing of the PDT-treated wounds is not impaired and may actually be improved.
One problem that is evident when applying PDT for microbial infection is the fact that soon after cessation of illumination, the generation of antimicrobial ROS ceases and the lifetime of these ROS in tissue is very short. Therefore there is likely to be no reason why any microbial cell remaining alive cannot regrow without hindrance after illumination has finished. Bioluminescence imaging in fact demonstrates that in some circumstances this bacterial regrowth does occur. The hope for clinical application is that there exists some lower limit of infectious burden, so that if PDT is able to reduce the number of microbial cells beneath this limit, the host immune defense system will be able to “mop-up” the remaining microorganisms and cure will ensue. It may also be possible to repeat PDT for localized infections at defined time intervals.
The use of various animals as models for microbiological infections has been a fundamental part of infectious disease research for more than a century (16). Now, techniques of genetic alteration and manipulation have made possible the design of animals so as to be specifically applicable to the study of a myriad of diseases.
The intent for the use of animals as models of disease is to establish an infection that mimics that seen in humans. Ultimately, the goal is to seek means by which the infection can be thwarted. A key to developing an animal model is the selection of an animal whose physiology, reaction to an infection, and the nature of the infection itself all mirror as closely as possible the situation in humans. The data from animal models provide a means of indicating the potential of a treatment. Further study, involving humans, is always necessary before something such as a drug can be introduced for general use. Such human studies are subject to rigorous control.
The organisms described in this chapter are all classified as biosafety level two (BL2). This means that although these species are capable of causing disease in humans, they present no health hazard to laboratory personnel when standard universal precautions are taken in handling. These precautions include the use of personal protective equipment (gloves, laboratory coat, eye protection against splashes, and mask if aerosols are likely to be generated) and use of a Class 2 biosafety cabinet. All materials containing live microorganisms should be sterilized by autoclaving before disposal. Disinfectant sprays should be used to clean benches and hoods. UV light should be regularly used in biosafety cabinets to avoid contamination.
In this report we will describe the use of S. aureus and P. aeruginosa in vitro and infections caused by P. aeruginosa in vivo (see Note 1).
A convenient light source consists of a non-coherent incandescent lamp capable of delivering light into a fiber-optic probe (FOP) fitted with a band-pass filter (LumaCare, Newport Beach, CA). For MB we use a 660 ± 15 nm band-pass filter that provides approximately 1 W of light that can be focused into a spot of diameter between 1.5 and 5 cm depending on the distance from the end of the FOP.
The intensity of the light spot was measured as follows. A power meter (model DMM 199 with 201 standard head, Coherent, Santa Clara, CA) is used to measure the irradiance (power density in mW/cm2).
Serial dilutions use a considerable amount of disposable plasticware.
Mice are used as an animal model. Female BALB/c mice (Charles River Lab, Wilmington, MA) are obtained at 6 weeks of age when they weigh on average 20 g. Anesthesia is conveniently obtained by intraperitoneal injection of a mixture of ketamine/xylazine (100 mg/kg; 10:1 ratio). Mice are shaved on the back and the next day burns are created using brass blocks obtained from Small Parts, Inc. (Miami, FL).
All animal procedures must be approved by the Institutional Animal Care and Use Committee (IACUC) and must meet the guidelines of National Institutes of Health. The animals are housed one per cage (to prevent mice interfering with each other's wounds) and maintained on a 12-h light/dark cycle with access to food and water ad libitum. Mice receive buprenorphine (0.03 mg/kg SC BID) for 3 days after wounding for pain relief. Mice are euthanized according to protocol when their condition is assessed to be moribund.
1Sources and choice of microbiological strains. Defined strains of various species of microorganisms can be obtained from culture depositories or cell banks. ATCC (Manassas, VA) is the recognized vendor in the USA and many other worldwide collections are listed at http://www.bacterio.cict.fr/links.html. Stable bioluminescent bacteria are available from Caliper Life Sciences: http://www.caliperls.com/products/reagents/biolumine scent/light-producing-cells-and-microorganisms/micro organisms/
2Antimicrobial PS. The area of PS structure is probably where the largest variation is possible. There are a relatively large number of compounds (tens) that have been reported to be antimicrobial PS in the scientific literature. Many if not most of these PS molecules are positively charged. In other words they are cationic molecules that have one or more quaternized nitrogen atoms. These molecules may be tetrapyrroles that are based on porphyrin (22, 23), phthalocyanine (20, 24), or chlorin (25) backbones. A second large group of compounds is composed of cationic synthetic dyes such as phenothiazinium dyes or triarylmethane dyes. In this chapter we have chosen to describe the easily available MB in order to make it possible for most people to replicate the techniques described (26).
3PS quality control. It is known that solutions of most PS are not stable indefinitely even when stored at 4°C in the dark. The dyes that are used can aggregate in aqueous solution and this is frequently not visible to the naked eye (i.e., there is no visible precipitate). However, aggregation does mean that the dyes can substantially lose effectiveness in producing photokilling of microbial cells. Visible absorption spectra can help to monitor the activity of these dyes.
4Microbiological culture. The possibility of contamination is an ever-present danger in microbiology. Since rich growth media are used, stray microbial cells from the environment or from the laboratory can fairly easily grow in liquid media and, if care is not taken, can completely displace the desired species over time (in some cases without the experimenter realizing). Good aseptic technique will go a long way to avoid this occurrence. In addition it is useful to be able to recognize colony morphology of specific species on agar plates. A gram stain can be used to distinguish between Gram-positive and Gram-negative bacteria.
It is sometimes necessary to distinguish between log-phase and stationary-phase cultures.
5In vitro PDT. It is important to realize that there is a stoichiometric relationship between the number of microbial cells and the concentration of PS in the incubation mixture. In other words if the concentration of PS is increased then the fraction of bacteria killed will go up; similarly if the number of bacteria is decreased then the fraction killed will also go up. It is important to stir bacterial suspensions as the individual cells will settle to the bottom of the tube or well during illumination and when aliquots are withdrawn, the numbers of cells will steadily increase as the cell density rises as the total volume decreases. There is an upper limit to the concentration of PS that can be used if the compound is left in solution during the illumination. This is because of the self-shielding effect. This occurs when the dye in solution absorbs a significant proportion of the light falling on the bacterial suspension and prevents sufficient light reaching the PS-loaded cells. For many PSs this selfshielding happens when the concentration reaches about 300 μM.
If several unknown PSs are to be tested for effectiveness in antimicrobial PDT, it may be preferable to use a set of different PS concentrations (for instance 0.1, 0.3, 1, 3, 10, 30, and 100 μM) and a single light fluence (for instance 10 J/cm2 of the appropriate wavelength) instead of a single concentration and a set of fluences.
6Serial dilutions. Bacteria have a pronounced tendency to stick together in buffers. This means that unless care is taken the actual numbers of CFU counted in the successive serial dilutions may not reflect the calculated numbers. To avoid this effect the microcentrifuge tubes should each be individually vortexed for several seconds before the next dilution is made. In the worst case small concentrations of detergents may need to be added to the PBS to encourage bacterial dissociation form clumps.
7Animal models of infection. In order to create animal models of infection only certain microbial species (and even only certain strains within a single species) can be used. Bacteria are described as having varying degrees of virulence and pathogenicity. Moreover microbes that are highly pathogenic and virulent in humans may be markedly less so in mice (and vice versa). There have even been reports of differences in pathogenicity when a certain species and strain of microbe is tested for its ability to form an infection in two different strains of mice (27, 28).
8Creation of mouse burn infection. The susceptibilities of different areas in the BALB/c mouse body to bacterial infection are variable. For example, the lower back is usually more susceptible to infection than the upper back. To ensure that the extent of infection is relatively consistent in different mice, it is recommended that the burns be made consistently on the lower half of the mouse back. When applying the bacteria solution to the mouse burns, it is also important to smear the solution evenly on the whole burned area to ensure an even extent of infection within each burn. Mouse burns can also be made by exposing the dorsal surface of mice with a template (1 × 2 cm2 opening) to hot water bath (92–95°C) (29) or to the flame of an alcohol lamp (30). Bacteria can also be applied to the burn by subcutaneous injection of bacterial inoculum (31).
9Alternative mouse infection models. It is possible to use excisional wounds as a basis for bacterial infections (21, 32). In some cases subcutaneous abscesses can be used as mouse models. In this case the bacterial suspension is injected under the skin and into the muscle (e.g., thigh muscle). Depending on the species of bacteria employed, temporary immune suppression of the mice using cyclophosphamide injection may be necessary to allow the bacteria to become established (33).
10PDT of burn infections. The evaporation of the solvent (PBS) during PDT can cause non-PDT killing of bacteria. To prevent this artifact, it is recommended that aliquots of 10–15 μL PBS be added to the PDT-treated burns after each aliquot of light.
Photobleaching of MB can impair the effectiveness of PDT and is a commonly encountered problem in PDT. The effect of photobleaching can be eliminated or reduced by applying MB in aliquots to the burns.
Too much light (usually >350 J/cm2) can impair the selectivity of killing of bacteria and cause non-specific host tissue damage by PDT.
11Problem of regrowth. Bacterial regrowth after PDT is also a common problem. This is partially due to the fact that PDT is usually carried in the growing phase of the infection, which is usually from day 0 to day 3 after infection. For P. aeruginosa infection in mouse burns, bacterial regrowth can cause mortality of mice and subsequently failure of PDT. Repeated PDT (usually two to three times) or the combination of PDT with conventional antibiotics has shown to be a possible strategy to counteract the bacterial regrowth.
12Progress course of burn infection. The main determinants of the severity of the burn infection and whether the rodents develop sepsis and die are as follows: the virulence of the particular strain, the number of bacteria applied to the burn, the size of the burn expressed as % of TBSA, whether the bacteria are applied to the surface or injected into or beneath the burn, and the length of time the heated object or liquid is in contact with the mouse skin. Without PDT, P. aeruginosa usually can invade through the burn into mouse bloodstream and subsequently induce fatal infection. In addition, a third-degree burn and superimposed infection can cause a deficiency of the gut barrier (34, 35), which can subsequently promote bacterial translocation from the gut (36).