1. Historical overview
The bacteriophage, commonly referred to as ‘phage’, was first identified 100 years ago by a British Microbiologist Frederick Twort while studying how the vaccinia virus can be grown in artificial, cell free media (Duckworth, 1976). During this work, he noticed that upon inoculation of agar with unfiltered fluid from smallpox vaccinations, a micrococcus bacterium grew but were affected with something as they showed “watery looking areas” (Twort, 1915), which contained no bacteria and could be transmitted to other micrococcus colonies for many generations. Twort concluded that the cause was a filterable agent that could kill bacteria, multiply only in the presence of bacteria (Twort, 1915). However he thought it was secreted by the bacteria and did not pursue his finding (Summers, 1999).
Two years after this Felix d’Herelle rediscovered bacteriophages during an investigation into an outbreak of haemorrhagic dysentery in French troops in 1915 (Sulakvelidze, 2001). In this investigation, he observed clear areas called plaques on agar plates containing faecal sample filtrates mixed with Shigella strains. d’Herelle proposed that this anti-shiga substance was a virus that could parasitize bacteria (d’Herelle, 1917) and subsequently named it bacteriophage. Furthermore, he noted that the phage titers in faecal samples of the dysentery patients peaked as the patients started to recover, therefore proposing that their recovery was mediated by bacteriophage and that phages are “agents of natural immunity” (Summers, 2011).
2. Phage Biology
Phages are parasites of bacteria and are the most abundant life form on the planet and play important roles in the biology and balance of microbes in all environments (Sulakvelidze, 2011). Population numbers are estimated to be within the range of 104 to108 viruses per ml and 109 viruses per g in the soil (Weinbauer, 2004). To date, more than 6000 phages have been discovered, including bacterial and archael viruses (Wittebole, 2014) and they are classified according to morphology, genetic type, specific host, habitat and life cycle (Wittebole, 2014).
Bacteriophages consist of genetic material (such as double stranded DNA) contained within an icosahedral protein or lipoprotein capsid that is attached to a collar (Fig. 1). The majority of phages (more than 95%) are also tailed (Fig. 1) with tail fibers and tips that interact with bacterial surfaces and receptors (Deresinski, 2009).
Phages have different life cycles within the bacterial host: lytic, lysogenic, pseudo-lysongenic and chronic (Weinbauer, 2004). In the lytic life cycle, the phage infects and kills the host cell, generating new phage progeny that are released when the cell bursts. The lysogenic cycle occurs when temperate phages integrate their genetic information (prophage) into the host genome where it can remain silent and is replicated along with the host genome. This lysogeny can last for many generations until the lytic cycle is induced. However the phage can also alter the bacterium’s phenotype via the transfer and expression of genes not usually expressed and which may be of selective advantage to the host bacterium (Clokie, 2011). Such genes can include virulent genes and antibiotic resistant genes, for example, the gene encoding the CTXϕ toxin associated with cholera (Fortier, 2013) and shiga toxins in E. coli (Kutter, 2014). In pseudolysogeny, development of the bacteriophage within the host is stalled, with no multiplication of the phage genome or its replication occurring (Los, 2012). Chronic infection occurs when phage progeny are released constantly from the host cell but without host cell lysis (Weinbauer, 2004). Phage therapy mainly focuses on lytic phages of the Caudovirales order, as temperate phages are associated with the dangers of introducing virulent factors into the pathogenic bacteria.
The first stage of all phage infections involves attachment to specific receptors on bacterial cell surfaces (Fig.2). For example, Bacteriophage λ only interacts with the LamB receptor on E. coli cells (Wittebole, 2014). It is this specific phage-receptor interaction that causes the narrow host range of each phage. After adsorption, a pore is induced in the cell wall and the DNA is injected into the cell (Fig. 2). In lytic infections the expression of early genes occurs, redirecting the host machinery to the replication and synthesis of viral nucleic acids and proteins, allowing the assembly and packaging of new phage progeny. Late enzymes, such as holin and endolysin are then utilized causing the bacterial cell to burst, releasing the progeny.
3. Phage Therapy
Phage therapy has mainly involved the use of living phages for the targeted treatment of bacterial infections, with most studies and clinical trials focusing on this area of clinical use. However, phage therapy can also involve using phage products such as viral proteins and enzymes. Recently, research has focused on engineering or creating synthetic phages to express genes encoding specific enzymes, which, for example can be used to sensitise bacterial cells to antibiotics.
3.1 Conventional Phage Therapy
Conventional phage therapy utilises the lytic life cycle of phages to kill bacteria and self-replicate to release progeny to amplify and continue the killing cycle of other cells until the infection is cleared (Fig. 3). As phages are so specific the correct phage must first be isolated by enrichment with their bacterial strain and then the recovery and purification of plaques (Gill, 2011). Recently, assays have been developed for detection of phage for infections including M. tuberculosis and MRSA (Monk, 2010). For broader use, phage cocktails consisting of multiple phage types can be used for treatment. The use of lytic living phages to treat bacterial infections has been practised since 1919 when d’Herelle used phages to successfully treat dysentery in children (Sulakvelidze, 2001). In 1921 Richard Bruynoghe and Joseph Maisin used phages to treat a skin disease caused by Staphylococcus, in which infections were reported to regress within 24 to 48 h after phage injection in the lesions (Bruynogne, 1921). There were further successful studies carried out and d’Herelle continued to use phage to treat various types of infections, including surgical infections (d’Herelle, 1931) and Cholera (Summers, 1999).
3.2 Use of phage derived products
Phage derived products such as phage encoded enzymes can also be used in phage therapy. For example, endolysins, which are peptidoglycan hydrolases released by phage, are used to digest peptidoglycan in the cell wall resulting in osmotic imbalance and lysis, allowing progeny release. Currently, phase I and II clinical trials are on going for the topical and intravenous use of endolysins to treat S. aureus infections in humans. Additionally, an endolysin (Micreos) is now commercially available for the treatment of S. aureus skin infections (Salmond, 2015).
3.3 Engineered/Synthetic phage
Engineering phages often involves engineering phage to encode enzymes such as the previously mentioned endolysins. For example, endolysin domains can be fused to other domains to alter their host range and increase their activity (Salmond, 2015). Phage can also be engineered to express diffusible polysaccharide- depolymerizing enzymes, such as phage T7, which was engineered to express an exopolysaccharide (EPS)- degrading enzyme in an E. coli biofilm. Removing EPS has several effects on the bacterial host including the disruption of biofilms, reducing virulence and assisting the immune system for bacterial clearance (Salmond, 2015). Biofilms are created when bacteria adhere to surfaces to create slimy layers and the generation of an EPS matrix (Abedon, 2011, book). Biofilms contribute to bacterial pathogenicity, and are also highly resistant to antibiotics, which is problematic when treating infections caused by biofilm- associated bacteria. Therefore the use of these biofilm-degrading enzymes is a potential solution.
4. Why use phage therapy?
Antimicrobial resistance in bacteria has increased rapidly since its emergence and is now a significant and growing threat to healthcare worldwide. Currently, 700,000 people die as a result of resistant infections every year, with this figure expected to rise to 10 million per year if resistance is not slowed or alternatives are not found (AMR, 2016). Antimicrobial resistance is increasing, threatening the treatment of common, simple to treat infections such as gonorrhoeal infections caused by N. gonorrhoea, with resistance to last resort cephalosporins reported in 10 countries, prompting it to be listed as an urgent threat by the CDC (CDC, 2015). Additionally, many hospital-acquired infections are caused by highly resistant bacteria such as methicillin resistant Staphylococcus aureus (MRSA), which causes 11,000 deaths per year, and Clostridium difficile, which causes 14,000 deaths per year (CDC). A faction of these hospital acquired bacteria have been called the ESKAPE pathogens (Enterococcus faecium, Staphyloccocus aureus, Klebisella pneumonia, Acinetobacter baumanii, Pseudomonas aeruginosa and Enterobacter spp.) as they cause many of the infections in hospitals and can effectively escape the effects of antibacterial drugs (Rice, 2008).
With this spread of resistance, new antibiotics or alternatives are urgently needed. It is also likely that resistance would spread to new antibiotics eventually meaning that new antibiotics will need to be produced almost constantly to meet demand. An alternative therapy for the treatment of bacterial infections is phage therapy, with the National Institute of Allergy and Infectious Diseases (NIAID) listing it as one of their 7 strategic approaches to combat resistance, and the recent AMR review also listing it as an alternative, with the potential to come to market within the next 10 years (AMR, 2016).
- talk about sections
II. Bacteriophage resistance mechanisms
Although phage therapy is a potential alternative to antibiotics in the treatment of bacterial infections, phages and bacteria are known to have extensive co-evolutionary dynamics. Bacteria evolve mechanisms to overcome phage infection and killing, with the phage further evolving to re-infect the host, resulting in an evolutionary arms race between the host and phage (Stern, 2012). This has prompted the concern that long-term use of phages as antimicrobials, could again, like antibiotics result in the development of general resistance to phages (Örmälä, 2013). Some studies have looked at the effect phage have on host evolvability and resistance. For example, one study showed that phage-driven diversification of resistance conferred an evolutionary benefit to the host, enabling bacteria to discover fitter genotypes, and thus concluded that phage therapy may promote rapid bacterial adaptation. Another concern is of the use of phage cocktails, as although they would reduce the emergence of resistance in the short term, long-term use of phages could eventually select for more general, broad-host resistance (Örmälä, 2013) like that of antibiotics. However, a study on the co-evolution between P. syringae and lytic phages showed that bacteria grown in multiple phage environments (like that of a phage cocktail) were no more resistant to novel phages than bacteria grown in a single phage environment, thus there was no evidence that phage cocktails would select for a general resistance mechanism (Koskella, 2012). Furthermore, this study also suggested that the use of phage cocktails could select for less fit individuals that are more likely to be lost over time, thus removing resistance from the population.
Bacteria have evolved several antiphage defence mechanisms, which affect different parts of the phage infection process, including adsorption, DNA entry, transcription, translation and viral replication. However in many cases, phages have evolved methods to evade these resistance mechanisms. Understanding bacterial defence mechanisms and bacteriophage counter-measures will help to exploit phages as antimicrobials by further understanding the infection dynamics of phages and their bacterial hosts.
2. Preventing phage adsorption
2.1 Blocking phage receptors
To infect a host, phages bind to specific receptors on the bacterial cell surface. This adsorption step is often the target of resistance mechanisms in bacteria. One mechanism in which they do this is by altering the structure of cell surface receptors via mutations in receptor genes. For example, missense mutations in the coliphage lambda, lamB receptor gene affects phage adsorption, causing resistance (Werts, 1994). However, lambda phage can then gain point mutations in their J tail protein, which restores infection (Bickard, 2011). Another mechanism is by concealing phage receptors such as Staphylococcus aureus, which produces a virulence factor, immunoglobin G-binding protein A, which has been suggested to mask the phage receptor, leading to reduced phage adsorption.
Bacteria can also use phase variation (Fig. 3), which allows ON-OFF reversible switching of receptor expression, enabling bacteria to switch their receptor type quickly to prevent phage infection (Seed, 2015). The mechanisms of phase variation include genomic rearrangements, slipped-strand mispairing of repetitive sequences during DNA replication and epigenetic modifications such as DNA methylation (Bickard, 2011). Bordetella spp. use phase variation to alter their cell surface by switching between Bvg+ phase and Bvg- phase. In the Bvg+ phase, there is expression of colonization and virulence factors such as the adhesion pertactin (Seed, 2015), which is a phage receptor. As this receptor is only expressed in the Bvg+ phase, phage infection is much higher than in Bvg- cells (Labrie, 2010).
Diversity- generating retroelements (DGRs) are a mechanism used by phages to overcome adsorption resistance. For example Bordetella phages can use a template repeat (TR) retroelement present in the tail fiber protein mtd (major tropism determinant) gene responsible for receptor binding (Bickard, 2011). A reverse transcription process introduces nucleotide substitutions into the variable region, promoting genetic variability (Labrie, 2010).
2.2 Production of extracellular matrix
Bacteria can also produce an extracellular matrix or a capsule, which provides a physical barrier between phages and receptors. The K1 capsule of E. coli has been observed to interfere with phage attachment to its LPS receptor (Scholl, 2005). One example of exopolysaccharides are alginates, produced by Pseudomonas spp. and Azotobacter spp. Phage resistance has been shown to increase in alginate producing bacteria, however some phage can produce an alginate lyase, degrading it and facilitating phage dispersal in the matrix (Hanlon, 2001).
Hyaluronan, a glycosaminoglycan, is a virulence factor produced by streptococci that helps cells escape the immune system. Genes encoding hyaluronan-degrading enzymes (hyaluronidases) are found in prophages in the genome of bacterial strains and in streptococcal phages, suggesting that hyaluronan is a barrier involved in preventing phage adsorption with phage evolving to overcome this resistance (Labroe, 2010).
Phages have also managed to overcome receptor blockage by exopolysaccharides by evolving to recognise the polysaccharides themselves. For example, some phages have coevolved to specifically recognise the lipopolysaccharide O antigen and the polysaccharide K antigen produced by E. coli (Stirm, 1968).
2.3 Production of competitive inhibitors
Molecules produced by bacteria can bind to phage receptors, concealing them and preventing phage adsorption. FhuA, an iron transporter in E. coli and a receptor for coliphages is also used by the antimicrobial molecule microcin J25, produced under nutrient depletion. Microcin outcompetes the phage for FhuA binding, thus allowing the bacterium to avoid phage infection. Outer membrane vesicles (OMVs) also reduce phage levels, suggesting that releasing OMVs into the environment acts as a decoy for phage infection (Seed, 2015).
3. Preventing DNA entry
Following binding to a receptor, bacteria have other defences to block DNA injection into the cell. These defences are called superinfection exclusion (Sie) systems. The genes that encode these proteins are usually phage encoded and found in prophages, suggesting that they typically act to prevent infection from another related phage.
An example of a Sie system is that of coliphage T4, which has two Sie systems, encoded by imm and sp. Imm changes the conformation of the DNA injection site, thus preventing the transfer of phage DNA into the cytoplasm, whereas Sp inhibits T4 lysozyme activity, preventing peptidoglycan degradation and entry of DNA (Lu, 1994).
The SieA system in the prophage TP-J34 of Streptococcus thermophilus encodes a membrane-localised lipoprotein (LTP), which is believed to interact with the tape measure protein of other phages and block the injection of phage DNA into the cell (Bebeacua, 2013).
4. Restriction Modification Systems
If a phage successfully injects its DNA into the cell, many bacteria possess intracellular innate defences called restriction-modification (R-M) systems that destroy the invading DNA. R-M systems consist of 2 or more genes encoding a sequence specific restriction endonuclease (REase) that cleaves exogenous dsDNA into fragments, and a DNA methyltransferase (MTase) that modifies specific bases of the host genetic material to protect it from the restriction endonuclease (Bickard, 2011). R-M systems are very diverse and are classified into groups (type I – type IV).
If phage DNA enters the cell and is unmethylated, the REase will recognise and degrade it, or it will be methylated by the MTase to avoid degradation. These two enzymes have different processing rate, with the REase usually more active than the MTase, resulting in cleavage of phage DNA while the host DNA is protected by methylase activity (Labrie, 2010). However, if the phage DNA is methylated, new progeny will be insensitive to the restriction enzyme and remain so until it infects a bacterium that encodes a different methylase gene.
Phase variation, as well as changing receptors, can also alter R-M systems. For example, H. influenza uses phase variation to introduce mutations in the mod methylase gene, which produces an inactive methylase. It’s related cognate restriction enzyme res is only active in the Mod/Res heterodimer, therefore inactive Mod causes loss of this R-M system. Although this seems deleterious to the bacteria, it is thought that this switch lowers the selective pressure on phages to slow down their process of acquiring counter-mechanisms (Bayliss, 2006). In other species such as H. pylori, phase variation in mod produces different variants that confer different methylation patterns on DNA, resulting in different gene expression patterns in outer-membrane proteins and thus can cause resistance to phage.
Phages have evolved several anti-restriction strategies to overcome R-M systems. One such strategy is the absence of endonuclease recognition sites via accumulating point mutations. For example, Staphylococcus phage does not contain Sau 3A recognition sites in its genome (Tock, 2005). Another mechanism is through the acquisition of the methylase gene (Labrie, 2010). Phage T4 also contains the base hydroxymethylcytosine (HMC) instead of cytosine. This results in the phage no longer being vulnerable to R-M systems that recognise sequences containing cytosine (Labrie, 2010). However, some bacteria have further evolved to attack modified phage DNA using modification-dependent systems (MDSs), which recognise and degrade methylated or hydroxymethylated DNA. Examples of these systems include DpnI in Streptococcus pneumonia and McrA in E. coli (Labrie, 2010).
5. CRISPR- Cas system
Bacteria also have adaptive immunity against bacteriophages in the form of CRISPR-Cas (clustered, regularly interspaced short palindromic repeats–CRISPR-associated proteins), a sequence specific defense mechanism (Barrangou, 2007). CRISPR loci are composed of short repetitive sequences (21-48 bp long), separated by short spacer sequences from bacteriophages or plasmids and flanked by cas genes (Bickard, 2011). It is these spacers that store the immunological memory in bacteria, allowing the recognition and degradation against invading complementary sequences (Westra, 2014).
The CRISPR mechanism consists of 3 stages: adaptation, expression and maturation, and interference. In the adaptation stage, Cas proteins identify the target DNA and acquire a spacer sequence from this, which is integrated into the CRISPR array to form the memory. In expression and maturation the CRISPR array is transcribed into a precursor RNA transcript that is then processed into smaller CRISPR RNAs (crRNAs). Each of these crRNAs contains a spacer flanked by the repeat sequence, and are combined with Cas proteins to form the active Cas-crRNA complex. Lastly, in the interference stage, foreign targets such as phage DNA are recognised via complementary base pairing between foreign DNA and crRNA sequences, resulting in cleavage and degradation of the target (Amitai, 2016). Examples of CRISPR-Cas systems include the type I-F CRISPR-Cas system in Pseudomonas aeruginosa, which has been found to provide adaptive resistance to both native phages and engineered phage (Cady, 2012), and recently type I-B CRISPR-Cas system in Clostridium difficile (Boudry, 2015).
Phage have also co-evolved counter-defense mechanisms against the CRISPR-Cas system. S. thermophilus phages can evade resistance by gaining a point mutation or deletion in the targeted spacer sequence (Deveaux, 2008), yet, bacteria can gain new spacers that target the resistant mutant, exemplifying the co-evolution between phage and bacteria. Recently (Bondy-Denomy, 2015), phage produced proteins were identified that inhibit the CRISPR-Cas system by inhibiting different components of this system.
6. Abortive Infection
In contrast to the resistance mechanisms already described, abortive infection (Abi) systems result in the death of the infected bacterial cell to protect the surrounding population. Abi systems are encoded by mobile genetic elements such as prophages and target steps in phage multiplication such as replication, transcription and translation.
One characterised Abi system is the RexAB system in phage lambda E. coli strains. After phage infection, a phage protein-DNA complex is produced that activates RexA that then activates ion channel RexB. RexB reduces membrane potential; causing a decrease in cellular ATP levels, reducing cell multiplication thereby aborting phage infection (Molineux, 1999).
The E. coli Abi system Lit is activated by a Gol peptide present in the major capsid protein of phage T4. Activated Lit cleaves elongation factor-Tu, part of the translation machinery, hindering protein synthesis and causing bacterial cell death and phage abortion (Snyder, 1995).
7. Assembly Interference
Some gram-positive bacteria contain phage-inducible chromosomal islands (PICIs), which are able to interfere with reproduction of some phages. One family of PICIs are the Staphylococcus aureus pathogenicity islands (SaPIs), which carry virulence factors (Novick, 2010). Upon infection by phages, the SaPIs are induced, allowing them to affect phage assembly and DNA packaging. However, compared to other resistance mechanisms, the phage programme is allowed to progress to the production of new phage particles, but with SaPI DNA rather than phage DNA. There are several ways in which SaPIs interfere with phage reproduction. These include: remodelling the phage capsid proteins to exclude the larger phage genome, using phage packaging interference (Ppi) proteins, and interrupting late gene activation thereby hindering phage packaging and lysis (Ram, 2014). Similarly to Abi systems, the infected cell dies but allows the spread of SaPIs to surrounding cells (Seed, 2015)
Although some resistance mechanisms have been described as highlighted here, it is likely that many more will be uncovered, since most studies are performed on E. coli, though bacteria and phage are so diverse with a long co-evolutionary history. Phage resistance mechanisms could have serious implications in the use of phage therapy, especially since current use of phages to treat infections is very limited. However if adopted as a main therapy, extensive use could result in the rapid evolution of resistance mechanisms as with antimicrobial resistance in bacteria. It is likely that host-phage interactions will need to be studied in vivo, to comprehend the whole cell dynamics. Currently most resistance studies are performed in laboratory environments, which only look at single phage-host models and only analyse one mechanism at a time. In vivo studies will be able to investigate the effects of multiple resistance mechanisms on phage population and evolution, as well as the effects of phage counter mechanisms on bacteria.
If phages are to be used as a treatment for bacterial infections, it is important to understand both the evolutionary dynamics of host-phage interactions and the resistance mechanisms utilised by bacteria, as overcoming these defences could be vital to the use of phage therapy as a viable, long-term treatment of infections in humans.
...(download the rest of the essay above)