If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
The potential of phage therapy in cystic fibrosis: Essential human-bacterial-phage interactions and delivery considerations for use in Pseudomonas aeruginosa-infected airways
Telethon Kids Institute, University of Western Australia, Nedlands 6009, Western Australia, AustraliaSchool of Paediatrics and Child Health, University of Western Australia, Nedlands 6009, Western Australia, Australia
Telethon Kids Institute, University of Western Australia, Nedlands 6009, Western Australia, AustraliaSchool of Paediatrics and Child Health, University of Western Australia, Nedlands 6009, Western Australia, AustraliaDepartment of Respiratory Medicine, Princess Margaret Hospital for Children, Perth 6001, Western Australia, AustraliaCentre for Cell Therapy and Regenerative Medicine, School of Medicine and Pharmacology, University of Western Australia, Nedlands 6009, Western Australia, AustraliaSchool of Public Health, Curtin University, Bentley 6102, Western Australia, Australia
1 The full membership of the Australian Respiratory Early Surveillance Team for Cystic Fibrosis (AREST CF) is available at www.arestcf.org.
AREST CF
Footnotes
1 The full membership of the Australian Respiratory Early Surveillance Team for Cystic Fibrosis (AREST CF) is available at www.arestcf.org.
Affiliations
Telethon Kids Institute, University of Western Australia, Nedlands 6009, Western Australia, AustraliaDepartment of Respiratory Medicine, Princess Margaret Hospital for Children, Perth 6001, Western Australia, AustraliaMurdoch Childrens Research Institute, Parkville, 3052 Melbourne, Victoria, AustraliaDepartment of Paediatrics, University of Melbourne, Parkville, 3052 Melbourne, Victoria, Australia
As antimicrobial-resistant microbes become increasingly common and a significant global issue, novel approaches to treating these infections particularly in those at high risk are required. This is evident in people with cystic fibrosis (CF), who suffer from chronic airway infection caused by antibiotic resistant bacteria, typically Pseudomonas aeruginosa. One option is bacteriophage (phage) therapy, which utilises the natural predation of phage viruses upon their host bacteria. This review summarises the essential and unique aspects of the phage-microbe-human lung interactions in CF that must be addressed to successfully develop and deliver phage to CF airways. The current evidence regarding phage biology, phage-bacterial interactions, potential airway immune responses to phages, previous use of phages in humans and method of phage delivery to the lung are also summarised.
Health care service costs for treating patients with cystic fibrosis (CF) and non-CF bronchiectasis are estimated at approximately 25,000–50,000 USD per patient per year [
]. A large proportion of these costs are associated with recurrent hospital visits as a result of exacerbated chronic bacterial infections, of which Pseudomonas aeruginosa is the most prevalent in older children and adults [
]. While P. aeruginosa is ubiquitous in the environment, certain strains evolve to thrive in conditions available in diseased lungs, and once established, difficult to eradicate [
Antibiotic treatment is currently the most widely used therapy for P. aeruginosa infections in the lung. Consequently, antibiotic resistance has been observed in transmissible epidemic strains of P. aeruginosa [
], and there is an urgent need for changes to current practice to protect patients from and eradicate antibiotic-resistant bacteria, which have been identified in hospitals around the world [
]. Antimicrobial resistance in the so-called “ESKAPE” pathogens including P. aeruginosa is one of the greatest risks to human health today. Recent expert reviews suggest that deaths due to antimicrobial resistant infections will cause more deaths than cancer in coming decades [
] and are a dire warning to the world that we must increase our capacity to produce new classes of antimicrobial therapies. In CF, the resistance of P. aeruginosa isolates to traditional antimicrobials is increasing in both the United States and Australia [
]. There are several novel treatment options for P. aeruginosa and other bacterial infections both in CF and in wider bacterial disease contexts, which are at various stages of investigation. Amongst these, bacteriophage (phage) therapy was recently included in a list of the top ten novel antibacterial therapies warranting further investigation, alongside others such as antimicrobial peptides, pathogen-specific antibodies and vaccination [
]. Alternative strategies that show promise require funding and investigation urgently, before complete antibiotic resistance occurs in epidemic bacterial strains affecting CF populations across the globe.
Phage therapy has many attractive features in this disease setting, including a long history of human use with limited side effects. As a naturally-occurring antibacterial easily obtained in the biosphere, it appears to be a viable option for development. P. aeruginosa infections are contained within the lung; however, the lung is a highly specialised organ, containing structures and cells with functions including both respiratory and immunological properties as an epithelial layer. This isolated microbiological and immunological environment provides a unique opportunity for targeted inhaled aerosol therapy development. Therefore, this review will investigate the utilisation of inhaled phage therapies in the context of the special CF lung environment to which phage therapies would be introduced, investigating the mode of delivery to the CF lung and the potential first-line innate immunological responses of the airway epithelium of the lung airway based on current scientific evidence. We aim to identify gaps in knowledge that currently constrain the use of phages in CF.
2. Phage biology and life cycle
Bacteriophages are viruses that infect and replicate within bacteria, and like other viruses, may have double stranded or single stranded DNA or RNA genomes. Phages are generally specific to strains within a species due to their method of replication, which involves binding to a specific cell surface receptor on the host bacterial cell. The prototypical phage will attach to the relevant bacterial receptor, inject their genetic material, overtake the cellular machinery to instruct the cell to produce new phage virions, and then burst the host cell, releasing progeny. This mechanism of replication can be exploited to provide an advantageous antimicrobial effect in humans via phage therapy.
Modern technologies have allowed researchers to study the proteomic and genomic changes in the host cell upon phage infection, increasing our understanding of the biology of phage replication [
]. Despite this, detailed biological studies of phages are limited to a few well-studied examples, and there are vast numbers of phages that have never been described. For many that have been isolated and described, the cellular activity and genome remain unstudied. This is part of what makes phages an exciting possibility for future therapy; that a vast untapped biological resource exists from which the next chapter of medicinal therapies may come. There is a long history of phage therapy, which has been well-described in previous reviews [
There are several families of viruses that bacteriophages may belong to; unlike bacteria, which are increasingly classified by their genetic relatedness, phage classification is mostly based on morphology and host range (due to the limited number of phages with sequenced genomes). More generally, phages may also be classified into two groups based on their behavioural characteristics within their host. These are (i) the lytic phages, which are virulent and obligate pathogens of their bacterial host, continually entering a cycle of attachment and invasion, then multiplication and lysis of the host cell, and (ii) the lysogenic phages, which, although they can enter the lytic cycle when stimulated, can also become a passenger in the host DNA by integrating into the genome. The latter group are not generally considered good candidates for phage therapy because they may have the ability to transfer virulence genes between hosts, also being limited in their antibacterial activity relative to their lytic counterparts [
]. Therefore, to exclude the presence of integrase genes, which indicate that the phage is a lysogen, phages for therapeutic treatment should have their genomes sequenced and annotated [
Phage-host pharmacodynamics are complex, and dependent on the replication rate of phages at the site of infection, as well as the density of susceptible bacterial hosts and the growth rate of that host in human tissues [
]. Whether antimicrobial therapies should be used prophylactically or therapeutically in times of exacerbation is yet to be determined, as well as which species of bacteria in the lung to target. With these limitations in mind, there are many potential benefits to the use of phages as a clinical intervention in CF. However, there are three potential causes of failure of phage therapy in the CF airway context; (i) the inability of the phage to make physical contact with the target bacterial cells, (ii) bacterial strains having or developing resistance to a phage through mutation and natural selection, and (iii) presence of large numbers of dead bacteria in a chronic infection, causing phage applied to bind to cells in which they cannot replicate.
Phages must enter their host by attachment to specific receptor molecules on the surface of the bacterial host cell to cause successful infection. Cells that lack the correct receptor are inherently resistant to infection with a particular phage, and this can limit the usefulness of an individual phage in the therapeutic context. At the microscopic level, pathogens such as P. aeruginosa grow as discrete colonies within the thick anaerobic mucus of CF patients, and once the inhaled therapy has reach the site of action, the phage must be able to penetrate the mucus to infect and lyse the bacterial cells [
]. This is potentially difficult in diseased lungs, with thick mucus providing a physical barrier and development of bacterial biofilms common, and this can be a mechanism by which a phage is blocked from binding to its host receptor, preventing infection. While bacteria such as P. aeruginosa may protect themselves against phage infection through production of biofilms, phages such as F116 may have developed strategies to overcome this issue, inducing bacterial production of enzymes that degrade these biofilms [
]. Additionally, infection of P. aeruginosa with phage DMS3 prevents the bacteria from forming biofilms, mediated by the hijacking of the bacterium's own immune response to the phage [
]. Selection of phages that can penetrate biofilms may increase the success of phage therapy in the CF lung. As a contingency strategy, combination therapy with other treatments designed to break up mucus or biofilms in CF may increase effectiveness of phage therapies. Reassuringly, recent evidence from Escherichia coli phage T4 suggests that it is possible for phages to reproduce in bacterial cells in stationary phase of growth, as is likely to occur in CF biofilms [
]. It will be important to address whether phages can reach the site of diseased lung if significant congestion and occlusion of airways is present in clinical trials.
Bacteria can possess or acquire a range of adaptations to prevent infection with phage, which were recently reviewed [
], identified in a small percentage of sequenced bacteria (~10% of ~1500) in the NCBI genome database of bacteria. These systems suggest that some pathogens treated with phage may possess defence mechanisms that make treatment of infection with phage difficult. However, there is likely to be a fitness cost to pathogens for having and expressing these bacterial defences, which limits their prevalence [
Pre-adapting parasitic phages to a pathogen leads to increased pathogen clearance and lowered resistance evolution with Pseudomonas aeruginosa cystic fibrosis bacterial isolates.
]. Screening of bacterial isolates for resistance to therapeutic phage strains prior to phage administration could be used to personalise therapy in people with CF. The possibility of phage resistance in bacteria should not prevent its development as a therapeutic tool but appropriate clinical management guidelines for patients with phage resistant bacterial infections should be developed by hospitals employing this therapy, as is currently performed for antibiotics.
The final hurdle identified to use of phage therapy in the CF lung is that phages might bind to the many dead microbes rather than live and potentially threatening microbes. Therefore, the success of phage therapy is dependent on phage virions outnumbering bacterial cells either by delivery at high concentration, or replication to high concentrations in the lung. The former depends largely on the formulation and method of delivery, as will be discussed further below, and the latter on complex interactions between the host and phage that are difficult to predict on a case to case basis. This is an issue that will need to be considered for all novel therapies by conducting trials in the target CF population.
4. Human immunological responses to phage therapy
The possible progression of phage therapy from laboratory use to medicinal interventions arouses concern on behalf of patients that phages, as non-self antigens, will stimulate inflammatory responses resulting in detrimental effects for recipients. Therefore, studies investigating the immunogenicity of phages on the human immune response are of importance to progress research to clinical use. As shown in this section, many studies have explored the effects of the ubiquitous T4 phage, and research into other phages related to CF pathogens is needed.
Dabrowska et al. recently demonstrated the antigenicity of Escherichia coli T4-phage proteins for humans and mice, showing that most individuals tested had phage-specific antibodies in circulation, which may reduce the functionality of these phages in vivo in combination with the complement system in the blood [
]. E. coli and T4-like phages are abundant in humans and the environment, which may explain the abundance of antibodies against these phages. In humans treated with different phages for various infections, the level of anti-phage antibodies detected are dependent on the route of administration, but this also appears to differ based on the phage strain. Moreover, individual responses to the same phage strain may differ depending on prior exposure [
], which has implications for repeated therapy with the same phage. The production of anti-phage antibody may be most clinically relevant in its capacity to interfere with phage therapy in intravenous studies, yet the presence of neutralising antibody does not preclude therapeutic benefits [
]. There is some evidence from mouse experiments that pre-existing inflammation may increase the rate of phagocytic clearance of phage from the body during therapy [
]. Phages do indeed generate antibody responses of varying strength in serum, but whether specific antibodies against phages are produced in the lung environment and would have significant impact on phage therapy in the lung is unknown due to lack of experimental evidence.
Clearly, phage particles are visible to the immune system, and the phagocytic cells of the immune system are responsible for removal of foreign bodies. In mice, fluorescently labelled-T4 phage is taken up by macrophages in vivo within 30 min of exposure, and is carried to distant immune tissues such as the spleen and lymph nodes [
]. Despite the evidence that phage is taken up by macrophages, kinetic studies of T7 phage neutralisation in mouse models suggest that the majority of phage clearance in serum is B cell dependent, possibly due to production of antibodies [
Reassuringly, research does not support any significant upregulation of inflammatory molecules after phage treatment. T4 virions and phage proteins do not stimulate production of inflammatory proteins in vivo in mice or ex vivo in human blood, dendritic cell differentiation, or affect the production of reactive oxygen species in polymorphonuclear leukocytes [
]. Ex vivo human monocytes can produce IL-6 and IL-12 cytokines in response to phage lysates, but this effect is not seen in cells already stimulated with bacterial endotoxin, indicating that in the context of phage treatment of bacterial infection, no additional monocyte responses to the phage would be observed [
]. Classical markers of inflammation including serum C-reactive protein (CRP) and white blood cell count in human patients receiving phage therapy show significant decreases following treatment administration [
]. This suggests that phage therapy has an anti-inflammatory effect. Even lysis-deficient (i.e. unable to kill host) phages reduce inflammation and improve survival from E. coli in a mouse model of sepsis [
], through unknown mechanisms, though it suggests interference between bacterial inflammatory molecules and their immune receptors. Phages active against Burkholderia cenocepacia, another relevant CF pathogen, have been administered via inhalation to mice with no increase in TNF-alpha or MIP2 compared to untreated mice, and moreover, the levels of these proteins were reduced in infected mice compared to untreated mice with infection [
]. Although ingestion of phage is not recommended for CF therapy, mice fed T7 phage do not show increased expression of major inflammatory cytokines or histopathological changes in the gut that would raise concerns about this method of therapy administration [
Despite evidence suggesting that phage therapy is unlikely to cause significant inflammation, CF is associated with airways inflammation, which may be present prior to pathogen exposure, and compounded by infections [
]. As such, potential therapeutic agents must be investigated in CF-relevant models, given the altered epithelial surface liquid and influx of immune cells in infected airways. Potential sources of airway immune response interference in phage therapy are shown in Fig. 1.
Fig. 1Diagram depicting potential innate and adaptive immune interference with nebulised phage preparations in the lung in CF.
Evidence for clinical use of phages in CF patients already exists in the literature, with most studies published in non-English journals (reviewed in [
]). While the data from these studies is convincing, many lack appropriate controls and methodological detail expected in modern studies, making drawing definitive conclusions difficult. Nevertheless, relevant reports in cystic fibrosis include two separate case reports of young (5 and 7 year old) girls with CF who became chronically infected with P. aeruginosa and S. aureus and were successfully treated with a combination of nebulized phages and antibiotic, showing significant clinical improvement and reductions of bacterial numbers in the sputa [
]. In both cases, previous antibiotic therapy had not eradicated infection. Nebulised phage given only once per 4–6 weeks over nine treatments gave large reductions of colony forming units in one patient, which is a feasible treatment schedule for CF patients [
]. Elsewhere in the literature, there are reports of phage used to treat non-CF patients with airway infection with relevant CF pathogens with between 77 and 100% reported efficacy of treatment in all cases, as summarised recently by Abedon [
]. Inhaled phage therapy is consistent with the delivery of other nebulised drugs to the lungs in CF such as hypertonic saline and antibiotics, and therefore, it is appropriate to consider the effects of the phages on CF airways and the effects of nebulization on preparations.
Besides these reports, little data are available from humans relevant to CF. In vitro studies suggest that phage can penetrate expectorated sputa to reduce the quantity of viable P. aeruginosa present [
]. Although these clinical reports and sputum studies are very promising, the data are preliminary, and in the clinical setting, have to be interpreted for what they are; individual studies that have not been replicated, with little understanding of the host or human response to the treatment. Previously, bacterial biofilm research has involved growth of microbes on non-biological surfaces, which exclude the possibility of investigating innate immune responses. Extended studies would be required to demonstrate safety and efficacy of phage in the airways, in more CF-relevant biological models. In this sense, research in CF therapy stalled previously due to limited access to primary airway cells and reliance on available immortal cells lines, of questionable biological relevance. However, more recently, we and others have established techniques to culture CF-derived primary airway epithelial cells in models to investigate such questions [
Antimicrobial efficacy against Pseudomonas aeruginosa biofilm formation in a three-dimensional lung epithelial model and the influence of fetal bovine serum.
]. Despite this, the two techniques have not yet been incorporated to our knowledge. The personalised medicine approach of using CF-derived primary airway cells in combination with a biofilm of the infecting pathogen could be effective for predicting therapeutic efficacy of phage therapy in individual patients, but may be less appealing to pharmaceutical companies, that may perceive limited patent opportunities [
]. Therefore, the development of phage therapy may rely on investigator-driven evidence of efficacy and safety in controlled trials. The CF community's desire for phage therapy to be implemented is unstudied as far as we could determine. However, delays in development of a therapy that has potential to stem the tide of antibiotic-resistant and epidemic bacterial infections raises complex ethical dilemmas, which will require discussion amongst the medical community in consultation with CF community reference groups [
Although systemic application of drugs may be perceived to be more reliable in general, direct delivery of phage to the lungs may significantly reduce the clearance of phage from the circulation and development of anti-phage antibodies. The use of nebulisers to produce an aerosolised suspension of phage has been reported in several studies of bacteria with roles in CF pathogenesis. Sahota and colleagues reported superior retention of phage viability in aerosols produced by a jet nebuliser compared with a vibrating mesh nebuliser [
], and elsewhere a similar study of phage active against antibiotic-resistant Burkholderia cepacia, has reported similar viability of phage in either type of device [
]. Each of these demonstrated aerosol particles in the respirable size range.
Several studies have reported the efficacy of phage treatment in mouse models of lung infection with relevant CF pathogens. Endotoxin-depleted aerosolised preparation of B. cepacia phages delivered to infected mice gave decreases in bacterial numbers within two days of treatment [
]. However, the treatment efficacy was dependant on the infective dose of phage. In another mouse study, a curative intranasal dose of phage allowed 95% survival of P. aeruginosa infections in the treated group compared with no survival in the control group [
]. Of note, prophylactic treatment of mice with phage 24 h prior to challenge with P. aeruginosa gave 100% protection against mortality. This could be due to phage adhering to mucus in the lung. Adherence of phage to mucosal surfaces has been proposed as a symbiotic non-host derived antibacterial defence [
] showed that phage treatment of mice acutely infected with P. aeruginosa cleared infection, and that inflammatory markers in bronchoalveolar lavage fluid were reduced in the phage-treated mice compared with controls. Overall, the use of aerosol delivery to the lung appears feasible in CF, but the preparation and device used should be carefully investigated to ensure specific phages retain activity and can reach the airways in appropriate models or clinical studies of humans, who have different physiology and immune responses to mice.
Based on available evidence, the most efficient way to deliver phage to the lungs is likely to be through using aerosolised preparations, thereby reducing potential for immune clearance before therapeutic effects can occur. Fig. 2 shows a schematic diagram summarising how inhaled phage therapy might work to reduce the burden of pathogens in the microenvironment of infected lungs in an individual with CF.
Fig. 2Diagram depicting delivery of nebulised phage preparations to sites of bacterial infection in the lung in CF. Panel 1: Phage preparation is nebulised and inhaled by the person with cystic fibrosis. Panel 2: Inhaled aerosolised phage droplets (green) travel to the airways to the site of infection (yellow). Panel 3: Phage virions adhere to mucus layer on the airway epithelium at site of infection. Panel 4: Phages attach to Pseudomonas aeruginosa cells (purple) due to host range specificity and do not attach to other microbial cells. Panel 5: Phages replicate within the Pseudomonas aeruginosa cells and release more progeny, thus killing the bacterium (chequered purple cell). Panel 6: Pseudomonas aeruginosa numbers in the mucus are significantly depleted and balance is restored to microbiota.
Although there is a history of clinical use and evidence from animal studies in the literature, much evidence supporting phage therapy comes from non-CF specific disease models. It is critical for scientists to understand the effects of phages on the human CF airway, since inhaled therapies are most likely in these subjects. Therefore, studies that investigate the dosage and timing of delivery with other CF medicines, efficacy of treatment and immune modulating properties of phages in the context of the specific conditions encountered in the infected CF lung are most useful to ensure the reinvigorated therapy could be of maximum benefit to patients. The current status of research in CF phage therapy and suggestions for important future areas of research are summarised in Table 1. As antibiotic resistance becomes an insurmountable obstacle to health care, our over-reliance on these prescribed medications becomes clear with limited alternatives available. Encouragingly, phage therapy appears a viable alternative for success in treating persistent bacterial infections such as those experienced by people with CF. Phage therapy is undergoing a transformation from an unconventional alternative therapy toward a more conventional treatment in the wider medical and scientific community. The outcomes of contemporary randomized double-blind placebo controlled studies could allow registration of phages as new antibacterial agents. However, perhaps the most important application will be in chronic conditions like CF, where the burden of disease and potential to improve quality of life long-term is greatest.
Table 1Summary of knowledge in the utilisation of phages to treat CF airways infections.
What is known?
•
New classes of antimicrobials are urgently required for human therapies
•
A range of phages have potential medicinal use due to broad in vitro antibacterial activity against CF pathogens
•
Phages can be produced to GMP standards to minimise patient risk for clinical trials
•
Phage appear to have limited immunological effects in humans, though repeated use does provoke generation of specific antibody responses when applied intravenously
•
Phages have been applied to treat humans successfully in a number of diseases according to preliminary evidence from case reports and long-term use in clinics in parts of the former USSR
Potential areas of focus for research
•
Investigation of phage effectiveness in appropriate CF-relevant models of infection at epithelial surface
○
How will the CF airway respond to phage treatment in the context of specific disease-associated conditions such as excessive mucous and neutrophilia?
○
How will phage effectiveness be altered when given in combination with other therapies in the airways?
○
How will the CF lung microenvironment affect bacterial susceptibility to infection with phage?
•
The most appropriate and patient-acceptable treatment regime to give maximum benefit and compliance from any effective therapy developed?
•
How quickly will the immune system clear phage directly inhaled into lungs?
This work was supported by a Telethon Kids Institute Blue Sky Project Grant. We gratefully acknowledge the assistance of Caroline Wise at Telethon Kids Institute for her assistance with composing the images used in this manuscript.
References
Heimeshoff M.
Hollmeyer H.
Schreyogg J.
Tiemann O.
Staab D.
Cost of illness of cystic fibrosis in Germany: results from a large cystic fibrosis centre.
Pre-adapting parasitic phages to a pathogen leads to increased pathogen clearance and lowered resistance evolution with Pseudomonas aeruginosa cystic fibrosis bacterial isolates.
Antimicrobial efficacy against Pseudomonas aeruginosa biofilm formation in a three-dimensional lung epithelial model and the influence of fetal bovine serum.
30814-7&id=gr1.jpg){kind=link}
30814-7&id=gr2.jpg){kind=link}