Lysins : the arrival of pathogen-directed anti-infectives

Lysins represent a novel class of anti-infectives derived from bacteriophage. Lysins are bacterial cell-wall hydrolytic enzymes that selectively and rapidly kill (¢3 log c.f.u. in 30 min) specific Gram-positive bacteria providing a targeted therapeutic approach with minimal impact on unrelated commensal flora. The potential for bacterial resistance to lysins is considered low due to targeting of highly conserved peptidoglycan components. Through cutting-edge genetic engineering, lysins can be assembled into large libraries of anti-infective agents tailored to any bacterium of interest including drug-resistant Gram-positive pathogens such as meticillinand vancomycin-resistant Staphylococcus aureus, vancomycin-resistant Enterococcus faecalis and Enterococcus faecium, and penicillin-resistant Streptococcus pneumoniae. Lysins can eliminate bacteria systemically and topically from mucosal surfaces and biofilms, as evidenced by experimental models of sepsis, endocarditis, pneumonia, meningitis, and nasopharyngeal, skin and vaginal decolonization. Furthermore, lysins can act synergistically with antibiotics and, in the process, resensitize bacteria to non-susceptible antibiotics. Clinical trials are being prepared to assess the safety and pharmacokinetic properties of lysins in humans.


Introduction
The introduction of a new class of antimicrobial agents known as bacteriophage-encoded lysins into clinical use will be an important advance in the field of infectious disease therapy.Lysins, are bacterial cell-wall-hydrolytic enzymes that selectively and rapidly kill specific Grampositive bacteria on contact (Fischetti, 2008).This review will focus on bacteriophage endolysins (lysins) used by phage to exit bacterial cells and not on lytic structural proteins used by phage during the infection process.Generally, lysins are genus-or species-specific, i.e. a Staphylococcus aureus phage lysin may have activity only against Staphylococcus aureus, providing a targeted therapeutic approach with minimal impact on unrelated commensal flora.Less frequently, lysins may have activity on several Gram-positive genera or even on both Grampositive and Gram-negative bacteria (Morita et al., 2001;Lai et al., 2011).A large body of work has emerged over the past decade describing the development, characterization and application of lysins to control pathogens systemically and topically on mucosal surfaces.
Lysins have been studied in several animal models of sepsis, endocarditis, pharyngitis, pneumonia, meningitis and mucosal and skin decolonization (Loeffler et al., 2001;Schuch et al., 2002;Loeffler et al., 2003;Entenza et al., 2005;Grandgirard et al., 2008;Witzenrath et al., 2009;Daniel et al., 2010;Pastagia et al., 2011).As antimicrobial agents, several features distinguish lysins from small molecule antibiotics including: (1) rapid bactericidal activity against both stationary-and exponential-phase bacteria within minutes of contact with highly conserved peptidoglycan substrates, (2) efficacy in treating biofilmbased infections such as endocarditis, (3) synergistic lethal activity in the presence of cell-wall-active antibiotics, and (4) the ability to 'resensitize' pathogens to formerly ineffective antibiotics (Daniel et al., 2010).As molecules, lysins are proteins and, as such, are immunogenic.However, in vitro and in vivo studies thus far have shown that antibodies are non-neutralizing (Nelson et al., 2001;Schuch et al., 2002;Jado et al., 2003;Loeffler et al., 2003;Rashel et al., 2007;Daniel et al., 2010;Pastagia et al., 2011).Human studies will be initiated to examine the safety and pharmacokinetic profile of lysins.Here, we will review the current status of lysin therapy, considering mechanism of action, potential for resistance, spectrum of activity, in vitro and in vivo models, synergy with antibiotics, plausible clinical uses and potential risks.the normal human microbiome.It is estimated that there are 10 31 phages in the biosphere, representing over 10 6 distinct species (Bru ¨ssow & Hendrix, 2002).Predating the discovery of penicillin in 1928 by Fleming, scientists have attempted to harness the lytic activity of phage to treat bacterial infections for nearly a century.The life cycle of a bacteriophage centres upon interaction with the bacterial host as a means for replication.Typical DNA tailed phages have icosahedral heads (encasing the nucleic acid genome) and tube-like tails, the tips of which bind to specific receptor molecules on the surface of target bacteria (Fig. 1).The viral DNA is then injected through the tail into the host cell, directing production of progeny phages in a life cycle that may take less than 30 min.These newly produced phages burst from the host bacterial cell (killing it) and infect other host cells for the production of more phage (Thurber, 2009).

Mechanism of action of lysins
Phage exit infected bacterial host cells by expressing two proteins -holin and lysin.At the end of the phage replicative cycle within the bacterial cell, the holin creates a pore in the cytoplasmic membrane allowing the lysin to pass through and interact with the peptidoglycan matrix.The lysin enzymically cleaves specific peptidoglycan bonds, disrupting the integrity of the structure.The disrupted peptidoglycan matrix can no longer withstand the internal osmotic pressure, resulting in extrusion and rupture of the cytoplasmic membrane through hypotonic lysis (Figs 1 and 2), and release of progeny phage particles (Wang et al., 2000).
Whole phage therapy was used commonly in the Soviet Union and Eastern Europe prior to and after the widespread introduction of antibiotics.A major drawback of phage therapy is rapid development of resistance to phage attachment -one of the defensive tools of bacteria (Labrie et al., 2010).Recently, researchers have started to focus only on the bacteriolytic proteins encoded by phage (i.e. the lysins).Using purified, recombinant lysins, researchers have demonstrated the 'lysis from without' of target bacterial organisms (Fischetti et al., 2006).In this manner, lysins can be used outside of the context of the bacteriophage as a potent antimicrobial agent.As a 'stand alone' agent applied exogenously to bacteria, lysins can cause bacterial lysis on contact (Fig. 2).
The ability of exogenously applied lysin to kill bacteria is typically limited to organisms without an outer membrane or surface lipids and waxes (i.e.Gram-positive bacteria).In recent years, considerable progress has been made in search of lysins with Gram-negative bacterial activity.Researchers are working on identifying naturally occurring Gramnegative lysins that work from the outside, bioengineering lysins to have Gram-negative activity by fusing bacterial cell surface binding proteins to lysin catalytic domains, and combining lysins with outer membrane-permeabilizing agents (Briers et al., 2011;Lai et al., 2011;Lukacik et al., 2012).
Lysins are generally greater than 25 kDa in size (antibiotics are typically 0.3-1.6 kDa) and display a modular organization of at least two distinct domains (Fischetti et al., 2006) (Fig. 3a).Lysins must perform two basic functions: substrate recognition and enzymic hydrolysis.Generally, the N-terminal domain enzymically cleaves specific peptidoglycan bonds.Included among the lysin catalytic domains are N-acetyl-b-Dglucosaminidases and N-acetylmuramidases (glycosidases) which hydrolyse glycosidic bonds in the glycan strand, endopeptidases which cleave the cross-bridge, and N-acetylmuramoyl-L-alanine amidases which cleave the amide bond connecting the glycan moiety and the stem peptide (Fig. 3b) (Diaz et al., 1990).Albeit less frequent, domain architectures other than one Nterminal catalytic and one C-terminal binding domain have been described in the literature (Briers et al., 2007;Nelson et al., 2012).The C-terminal cell-wall targeting domain specifies binding to peptidoglycan ligands or secondary cell-wall polymers like teichoic acids and neutral polysaccharides (Schuch et al., 2002) that are often restricted to particular bacterial species or even strains, imposing a level of specificity to the binding (and thus, activity) of lysins (demonstrated in Fig. 3c).This concept of determining binding ligands and specificity of cell-wall binding domains has been reported extensively in the literature for Listeria spp.(Loessner et al., 2002;Schmelcher et al., 2010;Eugster et al., 2011;Eugster & Loessner, 2012).

Lysin bioengineering
The catalytic and cell-wall targeting domains are functionally distinct and are connected by a short linker domain (Fig. 3).Domains can be swapped to create chimeric lysins with altered catalytic activities and binding specificities.The ability to shuffle lysin domains brings the potential to create designer lysins with optimized functional properties related to solubility, thermostability, binding specificity and catalytic efficiency (Fig. 3d).For example, Donovan and co-workers (Mao et al., 2013)  Recombineering different cell-wall binding domains targeting Listeria bacteria provides extended recognition and binding properties (Schmelcher et al., 2011).Based on these findings, one could envision a large toolbox of lysin domains to be mixed and matched in every combination to ultimately optimize potent antibacterial agents directed against any Gram-positive pathogen of interest.

Potential for resistance
No specific resistance mechanisms to lysins have been described.Exposure of staphylococci, pneumococci and Bacillus cereus to low (i.e.subminimum inhibitory) concentrations of lysin does not lead to the recovery of resistant mutants.In experiments in which colonies were picked at the periphery of a lysin drop on a lawn of staphylococci and grown, repeated for 40 cycles, there was no isolation of resistant bacteria (Loeffler et al., 2001;Schuch et al., 2002) (Fig. 4).Exposing MRSA cells in vitro to increasing concentrations of a staphylococcus-specific lysin or mupirocin resulted in the appearance of low-level mupirocin resistance within 8 days as opposed to no resistance to the lysin (Pastagia et al., 2011).The expression of thick polysaccharide capsules by streptococci or Bacillus anthracis or the formation of dense biofilms by staphylococci or streptococci does not block lysin activity (Loeffler et al., 2001;Nelson et al., 2001;Schuch et al., 2002;Cheng et al., 2005;Kokai-Kun et al., 2009;Domenech et al., 2011).Even after mutagenesis of a lysin-sensitive B. cereus strain with a powerful DNA alkylating agent (and a concomitant 1000-10 000-fold increase in streptomycin and novobiocin resistance), no lysin-resistant derivatives were found (Schuch et al., 2002).
One theory for the apparent lack of resistance postulates that for phage survival lysins evolved over hundreds of millions of years to bind and cleave targets in host peptidoglycan that are essential for bacterial viability (Fischetti et al., 2006).Thus, lysin-resistant mutants have yet to be isolated because they are not viable.This is supported by findings that the lysin's cell-wall binding ligands in Streptococcus pneumoniae (specifically, choline), Streptococcus pyogenes (polyrhamnose) and Bacillus anthracis (neutral polysaccharide) are required for viability (Yother et al., 1998;Schuch et al., 2002;Fischetti et al., 2006).

Spectrum of activity
In general, lysins have bactericidal activity against the bacterial hosts of the phages from which they were isolated.For example, lysins produced from streptococcal phage kill streptococci, and lysins produced by staphylococcal phage kill staphylococci (Loeffler et al., 2001;Daniel et al., 2010).In contrast to antibiotics, lysins may be identified to kill only the disease organism with little to no effect on unrelated commensal flora.Lysins have excellent potency in vitro against low to heavy inoculum of antibiotic susceptible and resistant Gram-positive bacteria, including B. anthracis and B. cereus (Schuch et al., 2002), Clostridium difficile (Mayer et al., 2008), Clostridium perfringens (Schmitz et al., 2011), Enterococcus faecalis and Enterococcus faecium (Yoong et al., 2004), Listeria monocytogenes (Loessner et al., 2002), Staphylococcus aureus and Staphylococcus epidermidis (Daniel et al., 2010), Streptococcus agalactiae (Cheng et al., 2005), Streptococcus pneumoniae (Loeffler et al., 2001) and Streptococcus pyogenes (Nelson et al., 2001).
An example of a highly specific lysin is the chimeric lysin, ClyS, which targets only staphylococci, including those resistant to meticillin (MRSA) and vancomycin (VRSA) (Daniel et al., 2010).Although unconfirmed by experimental evidence, it has been suggested that ClyS is an endopeptidase that cleaves the polyglycine cross-bridge specific to staphylococci (Daniel et al., 2010).ClyS contains the catalytic CHAP domain of the Twort endolysin which is similar in architecture to LysK and the phi11 lysin.LysK and phi11 lysin both feature identical cleavage sites of their CHAP domains between the stem peptide and the pentaglycine bridge found in staphylococci (Navarre et al., 1999;Becker et al., 2008).In some cases, however, phage enzymes may be identified with broader lytic activity.For example, the enterococcal phage lysin PlyV12 has been reported to kill not only enterococci but also a number of other Gram-positive pathogens such as Streptococcus agalactiae and Streptococcus pyogenes, making it one of the broader-acting lysins identified (Yoong et al., 2004).The binding site for such broad-acting lysins is under investigation.Loessner et al. (2002) have shown that lysins bind to their cell-wall substrate with affinity constants K a 53-6610 8 , similar to affinity-matured antibodies, which suggests that lysins are 'one-use' enzymes that do not disengage from the cell wall after binding and cleavage.This high affinity for bacterial cell-wall substrates allows lysins to identify their target bacteria quickly.Lysins are effective against dividing (logarithmic) and non-dividing (stationary) bacterial cells and thus can be used systemically and to decolonize mucosal surfaces.Multiple studies confirm that binding, cleavage and lysis occur rapidly after contact (Loeffler et al., 2001;Schuch et al., 2002;Daniel et al., 2010).For example, nanogram quantities of PlyC reduce the viability of a culture of 10 7 Streptococcus pyogenes by .6 logs only minutes after its addition (Nelson et al., 2001).Similar findings have also been reported using lysins against B. anthracis (Schuch et al., 2002), E. faecalis and E. faecium (Yoong et al., 2004), Staphylococcus aureus (Daniel et al., 2010) and Streptococcus agalactiae (Cheng et al., 2005).No known biological compounds other than chemical antiseptics such as bleach are able to kill bacteria this quickly.

In vitro and in vivo activity
The bactericidal activity of phage lysins has been characterized using a variety of in vitro assays.Crude assays to screen for lysin activity are based on the simple clearance of bacterial lawns on agar Petri dishes (Schuch et al., 2002(Schuch et al., , 2009) ) (Figs 1 and 4).Other techniques for examining lysin activity involve: (1) turbidometric methods to follow decreases in the optical density of lysintreated cultures (Nelson et al., 2001), (2) luminescencebased methods to measure ATP released from lysin-treated bacteria for quantification of bacteriolysis (Orito et al., 2004) and (3) variations of time-kill methods, whereby known quantities of bacteria and lysin are combined in liquid growth media, and bacterial viability is followed over time (Yang et al., 2012) (Fig. 5).
In vitro lytic assays have demonstrated that lysins rapidly kill bacteria.In Fig. 5, in vitro viability assays for MRSA isolates reveal a 3 log reduction of bacteria in approximately 30 min using the staphylococcal-specific lysin ClyS.
In the literature, it takes approximately 6 h for daptomycin and .24h for vancomycin to demonstrate similar killing ability in vitro (Marconescu et al., 2012).Although lysins have minimal inhibitory concentrations (MIC) that are relatively high (i.e.30-40 mg ml 21 for ClyS) (Daniel et al., 2010;Pastagia et al., 2011), on a molar basis, the activity of lysins is more potent than that of conventional antibiotics.More importantly, the killing rate of lysins is significantly faster than that of conventional antibiotics with susceptible MICs that are lower (MIC 90 ¡1 mg ml 21 ).In an in vitro biofilm model, the staphylococcal-specific lysin ClyS, at 16MIC is able to eradicate an MRSA biofilm within 24 h whereas antibiotics at concentrations of 10006MIC show minimal clearance over 24 h.Sass & Bierbaum (2007) have also reported the ability of lysins to disrupt staphylococcal biofilms in vitro.Streptococcus pyogenes, which is known to colonize human tissues and form biofilms, is readily destroyed by the lytic activity of PlyC (Shen et al., 2013).Collectively, these data suggest that MICs are not entirely predictive of activity for this novel class of antimicrobial agents.
Using the aforementioned methods, in vitro activity has been examined against pathogenic and multidrug-resistant Gram-positive bacteria.Many of these lysins were further studied using in vivo models of infection.Table 1 depicts lysin activity against drug-resistant Gram-positive bacteria.
Lysins have been evaluated in animal models of pneumonia, endocarditis and sepsis with a focus on efficacy and host immune response (Loeffler et al., 2001(Loeffler et al., , 2003;;Entenza et al., 2005;McCullers et al., 2007;Grandgirard et al., 2008;Witzenrath et al., 2009) (Fig. 6).McCullers et al. (2007) used Cpl-1 to reduce intranasal Streptococcus pneumoniae and ultimately prevent the development of acute otitis media following infection with a concomitant pathogen (influenza virus).Taking these results together, lysins are effective at eliminating bacteria from mucosal surfaces and from deeper tissues, including bacteria in biofilms.
Lysins as decolonizing agents may potentially be used for prophylaxis prior to surgical procedures or for household contacts.The findings shown in Fig. 6, taken along with the data from Table 1, present a large body of work that substantiates the role of lysins to treat systemic human infections, highlighting their ability to reach sites of infection quickly and effectively.

Synergy and antibiotic resensitization
Two different methods are standardly used to determine lysin synergy in vitro: (1) time-kill assays (Yang et al., 2012) and (2) chequerboard broth microdilution analysis (isobolograms) (Daniel et al., 2010;Garcı ´a et al., 2010).Using each of these methods, a clear synergistic effect has been observed between lysins and multiple distinct antibiotics.When used in combination with lysins, within 8 h, antibiotics with limited susceptibility had increased activity in the presence of subinhibitory concentrations of lysins (Djurkovic et al., 2005) (Table 2).It has been established that the combination of two lysins with different cleavage specificities has a synergistic effect and is highly lethal against Streptococcus pneumoniae (Loeffler & Fischetti, 2003).It has been postulated that the increased access of these lysins to their respective cleavage sites or the enhanced destructive effect of a two-dimensional digestion in the three-dimensional peptidoglycan is responsible for the observed synergy (Djurkovic et al., 2005).
The mechanism of synergistic reactions between lysins and antibiotics remains unclear.In vitro studies have shown enhanced daptomycin bactericidal activity and increased membrane daptomycin binding in the presence of antistaphylococcal beta-lactam antibiotics, and the combination is reported to be clinically beneficial in patients with refractory MRSA bacteraemia (Dhand et al., 2011).Lysins may enhance antibiotic activity in similar ways, through cell-wall interactions that facilitate antibiotic uptake.The right combination of lysin and antibiotic can, therefore, not only treat antibiotic-resistant bacteria but also resensitize bacteria to antibiotics against which there is resistance.The use of lysin/antibiotic combinations may provide the dual benefit of reducing the emergence of enzyme-or antibiotic-resistant mutants and enabling the use of antibiotics for shorter durations at possibly lower doses (thus reducing dose-related toxicity).

Clinical use of lysins
The initial systemic use of lysin therapy would likely be limited to circumstances in which the pathogen is known, or as adjunct/combination therapy with antibiotics.The availability of rapid diagnostic tests for infectious diseases (Ince & McNally, 2009) will coincide nicely with the introduction of lysins as a new class of therapeutics.Initial indications for lysins will likely be against life-threatening infections with multi-drug-resistant Gram-positive bacteria such as MRSA, especially since treatment failures are considerably common with these infections when antibiotics are used alone (Fowler et al., 2006).
The use of lysins may also be clinically applicable as salvage therapy for patients who experience recurrent or relapse infection despite appropriate use of antibiotics.Data illustrating the ability of lysins such as ClyS to resensitize oxacillin against MRSA and the improvement of vancomycin killing ability against vancomycin intermediateresistant Staphylococcus aureus (VISA) infections (Daniel et al., 2010) should allow for the repositioning of antibiotics that are or will become obsolete.
Disinfection of implanted medical devices (i.e.implantable cardiac devices, cardiac valves, orthopaedic prostheses and indwelling catheters) represents another practical application of lysins.Lysins that target clinically relevant bacteria can remove lingering pathogens in the lumen of catheters and implanted medical devices before suturing, thus preventing and reducing bacterial colonization of these surfaces.This could potentially avoid consequent invasive infections necessitating catheter or device removal.

Using the binding domain to identify novel targets for antibiotic development
The biosynthetic pathways for the assembly of cell-wall substrates may be targets for antibiotic development.As a proof-of-concept experiment, the pathway for the cell-wall receptor of the anthrax lysin, PlyG, was identified; the pathway for its synthesis (a 13 gene cassette) was located and one of the enzymes in that pathway (a UDP-Nacetylglucosamine 2-pimerase) was crystallized (Velloso et al., 2008).A small molecule inhibitor for that enzyme was engineered with an 8 mM MIC and used to successfully protect mice from fatal B. anthracis sepsis.When B. anthracis was tested for its rate of resistance to the compound, the frequency was found to be ,10 211 (Schuch et al., 2013).These studies strongly suggest that using the binding domain of lysins may be a more rational approach to identifying essential biosynthetic targets in bacteria.

Safety
Bacteriophage have a symbiotic relationship with bacteria, and as such have evolved concomitantly with bacteria for about a billion years.Available data indicate that most DNA viruses in the gastrointestinal tract are bacteriophage (Reyes et al., 2010), forming an integral part of the human microbiome.This indicates that humans are in constant contact with phage and lysins without discernible consequence.Lysins target structures unique and highly conserved to bacterial cells and as such should not affect mammalian cells (Fenton et al., 2010b).For example, histological analysis of skin and mucosal tissue of mice administered lysin for 7 consecutive days did not reveal any abnormalities (Fischetti, 2003).However, as proteins, lysins have the potential of generating an antibody response, and the release of various proinflammatory cell-wall-and membrane-associated components as a consequence of the lytic action on treated bacteria may result in host immune system effects as discussed below.

Immunogenicity
A potential concern with lysin therapy is the generation of neutralizing antibodies during a treatment course that could impede subsequent applications of the lysin.Unlike antibiotics, which are generally non-immunogenic small molecules, lysins are proteins that generate an antibody response when delivered mucosally or systemically.These antibodies could result in the neutralization of lysin activity.To examine this, rabbit hyperimmune serum was raised against lysins specific for B. anthracis, Staphylococcus aureus, Streptococcus pneumoniae or Streptococcus pyogenes.In each case, antibodies produced against the corresponding lysin did not block lysin activity in vitro (Nelson et al., 2001;Schuch et al., 2002;Loeffler et al., 2003;Rashel et al., 2007;Daniel et al., 2010;Pastagia et al., 2011) (Fig. 6).Similarly, in vivo, Jado et al. (2003) demonstrated that antibodies that were generated in mice during treatment with either of two therapeutic pneumococcal lysins, Cpl-1 and Pal, did not impair the animals' ability to recover after repeat infection and repeat administration of lysin treatment, nor were signs of anaphylaxis or other adverse effects observed.These results may be partially explained by the fact that the binding affinity of a lysin for its cell-wall substrate is higher than that of IgG antibodies for the lysin.Collectively, these data suggest that antibodies generated to lysins are nonneutralizing and lysins can be used repeatedly to treat bacterial infections without loss of efficacy or adverse effect.

Cytokine response
Because lysins kill Gram-positive bacteria on contact, release of bacterial debris including immunostimulatory peptidoglycan and toxins could potentially cause an increase in cytokine production and host inflammatory response.To examine this, mice were transnasally infected with Streptococcus pneumoniae and therapeutically treated with Cpl-1 (50 mg kg 21 ) or amoxicillin (20 mg kg 21 ) twice daily for 3 days by intraperitoneal injection (Witzenrath et al., 2009).The inflammatory response for Cpl-1 and amoxicillin-treated mice was similar (Fig. 6).
However, in an endocarditis model, when rats received Cpl-1 (250 mg kg 21 ) by continuous infusion for 6 h as opposed to vancomycin (1 g every 12 h for 48 h), there was a marked increase in the serum cytokines IL-1b, IL-6, IFN-c and TNF-a 6 h after infection (Entenza et al., 2005) (Fig. 6).The discrepancy of cytokine production between the pneumonia and endocarditis models is unclear.It may be related to animal species, method of dosing, or the dose itself.Further data are needed regarding dosing parameters and the potential for a cytokine response using different animal species and infection models.7 2 8 4 9 6 1 2 0 1 4 4 1 6 8 1 9 2 2 1 6 2 4 0 (murine)

Infection model
Efficacy of lysin Immune effects of lysin Fig. 6.In vivo models demonstrating efficacy and immune effects of lysin therapy.In a pneumococcal pneumonia model, infected mice treated with Cpl-1 lysin (50 mg kg "1 ) twice daily for 3 days resulted in sustained 10 day survival (Cpl-1-treated, 100 % survival, versus buffer-treated, 0 % survival) and a serum cytokine profile similar to amoxicillin [data adapted from Witzenrath et al. (2009)].In a rat endocarditis model, continuous infusion of high dose Cpl-1 lysin (250 mg kg "1 ) was able to decrease pneumococcal levels by nearly 4 log c.f.u.ml "1 in the blood and approximately 6 log c.f.u.ml "1 in cardiac vegetations as compared with buffer control [data adapted from Entenza et al.

Conclusions
As clinical studies commence, our understanding of the optimal use of lysins will evolve.In vitro and in vivo studies have demonstrated that lysins are bactericidal on contact, provide rapid killing of stationary and logarithmically growing bacteria, are a potent eradicator of biofilm infections, kill antibiotic-resistant bacteria, and exhibit synergy with other antibiotics.Lysins generally have a narrow spectrum of activity and as such should not affect unrelated commensal flora, thereby potentially avoiding infections such as C. difficile colitis.These positive attributes of lysins may allow for optimizing current treatment of serious life-threatening infections by reducing hospital stay, duration of concurrent antibiotic use and morbidity and mortality associated with these infections.To date, risk of resistance to lysins is very low due to targeting of critical peptidoglycan components.Through cutting-edge technology, lysins are amenable to genetic engineering which may be used to generate large libraries of anti-infective agents directed against any pathogen of interest.Lysins also have the potential to resensitize resistant and partially resistant antibiotics to their active form.
With any new class of drugs, much remains unknown in terms of effects in humans.As such, lysins face particular challenges with the FDA to substantiate safety.Being a large protein (.25 kDa), lysins are similar to biological agents used to treat autoimmune diseases and malignancies.Lysins, however, target specific bacteria, rather than components of the endogenous human immune system.Adverse risks of lysins may be quite different from those of traditional biologics (e.g.infection and malignancy), but similar in other ways (e.g.rash and infusion site reaction).
It will be important to determine the proper doses and dosing schedules of lysins to administer to humans that are safe and effective.
Clinical trials are being prepared to assess the safety and pharmacokinetic properties of this novel class of antiinfectives in humans (Gravitz, 2012).Given the pressing need for new agents to effectively combat multi-drugresistant bacteria and to overcome the problem of infection relapse, lysins are an exciting addition to current therapeutic options for the management of Gram-positive infections.

Fig. 1 .
Fig. 1.Lysin activity in the bacteriophage life cycle.(a) Bacteriophage infect bacteria, replicate, and express lysins to exit the host bacterium.(b) Lysins can be expressed recombinantly and purified by column chromatography.When applied exogenously to bacterial lawns, a zone of clearing can be seen arising from the potent bactericidal effect of purified lysin.

Fig. 2 .Fig. 3 .
Fig. 2. Lysin activity against MRSA.Exogenous application of the Staphylococcus aureus lysin, ClyS, causes peptidoglycan disruption and hypotonic lysis within 1 min.The cytoplasmic membrane is shown extruding through regions of the cell wall weakened by ClyS.
and Fernandes et al. (2012) constructed chimeric lysins to solve the solubility problems associated with many anti-staphylococcal lysins.Daniel et al. (2010) fused a catalytic domain of the Staphylococcus aureus phage Twort lysin to the cell-wall binding domain of the Staphylococcus aureus phage phiNM3 lysin, creating the highly soluble hybrid, ClyS, which was shown to have potent staphylococcus-specific activity in a murine sepsis model of meticillin-resistant Staphylococcus aureus (MRSA) infection.ClyS was shown to be superior to mupirocin at eradicating staphylococcal skin infections in mice (Pastagia et al., 2011).Rodrı ´guez-Rubio et al. (2012) were able to improve the staphylococci lytic activity of HydH5, via fusion with lysostaphin.

Fig. 4 .
Fig. 4. In vitro assay for lysin resistance.An aliquot of lysin is placed on a fresh bacterial lawn and incubated overnight, resulting in a zone of clearing the next day.Bacterial colonies at the periphery of the clearing zone (indicated by arrows) are subcultured and exposed to subinhibitory concentrations of lysin, are grown as a bacterial lawn on agar plates, and a drop of lysin is placed on the lawn.This procedure was repeated 40 times with PlyG with no detectable resistance.

Fig. 5 .
Fig. 5. Rapid killing ability of lysins.Time-kill experiments using the staphylococcal-specific lysin ClyS (MIC 90 32 mg ml "1 ) against MRSA reveal a 3 log c.f.u.ml "1 reduction of bacteria within 30 min [data adapted from Daniel et al. (2010)].Biofilm assays with Staphylococcus aureus demonstrate clearance within 24 h at MIC of ClyS and minimal clearance with antibiotics at 1000¾MIC.Unlike conventional antibiotics, MIC may not be the most appropriate assay for measuring activity of lysins.
Fig.6.In vivo models demonstrating efficacy and immune effects of lysin therapy.In a pneumococcal pneumonia model, infected mice treated with Cpl-1 lysin (50 mg kg "1 ) twice daily for 3 days resulted in sustained 10 day survival (Cpl-1-treated, 100 % survival, versus buffer-treated, 0 % survival) and a serum cytokine profile similar to amoxicillin [data adapted fromWitzenrath et al. (2009)].In a rat endocarditis model, continuous infusion of high dose Cpl-1 lysin (250 mg kg "1 ) was able to decrease pneumococcal levels by nearly 4 log c.f.u.ml "1 in the blood and approximately 6 log c.f.u.ml "1 in cardiac vegetations as compared with buffer control [data adapted fromEntenza et al. (2005)].High dose continuous infusion of Cpl-1 resulted in increased levels of cytokines, but lower intermittent dosing did not.In an MRSA sepsis model in mice, one dose of a subinhibitory concentration of ClyS lysin combined with oxacillin was able to increase survival significantly better than either agent alone (ClyS alone 25 % survival, oxacillin alone 30 %, and ClyS+oxacillin 75 %).ClyS maintained in vitro activity (i.e.no change in optical density in preimmune vs immune serum) when exposed to hyperimmune rabbit serum (final ELISA titre of 10 000) [data adapted fromDaniel et al. (2010)].

Table 1 .
Characteristics of lysins effective against antibiotic-resistant Gram-positive bacteria

Table 2 .
In vitro synergy between antibiotics and subinhibitory concentrations of lysins within 8 h Staphylococcus aureus; VISA, vancomycin intermediate-resistant Staphylococcus aureus.