Contribution of the Immune Response to Phage Therapy Oleg Krut and Isabelle Bekeredjian-Ding This information is current as J Immunol 2018; 200:3037-3044; ; of September 24, 2021. doi: 10.4049/jimmunol.1701745 http://www.jimmunol.org/content/200/9/3037 Downloaded from References This article cites 70 articles, 17 of which you can access for free at: http://www.jimmunol.org/content/200/9/3037.full#ref-list-1

Why The JI? Submit online. http://www.jimmunol.org/ • Rapid Reviews! 30 days* from submission to initial decision

• No Triage! Every submission reviewed by practicing scientists

• Fast Publication! 4 weeks from acceptance to publication

*average by guest on September 24, 2021 Subscription Information about subscribing to The Journal of Immunology is online at: http://jimmunol.org/subscription Permissions Submit copyright permission requests at: http://www.aai.org/About/Publications/JI/copyright.html Email Alerts Receive free email-alerts when new articles cite this article. Sign up at: http://jimmunol.org/alerts

The Journal of Immunology is published twice each month by The American Association of Immunologists, Inc., 1451 Rockville Pike, Suite 650, Rockville, MD 20852 Copyright © 2018 by The American Association of Immunologists, Inc. All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. Th eJournal of Brief Reviews Immunology

Contribution of the Immune Response to Phage Therapy Oleg Krut and Isabelle Bekeredjian-Ding Therapeutic phages are being employed for vaccination regard to the impact of the immune response on efficacy and and treatment of cancer and bacterial infections. Their safety are important obstacles to defining the target product natural immunogenicity triggers intertwined interac- profiles and obtaining regulatory approval. tions with innate and adaptive immune cells that Despite the use and effectiveness of antibiotics, physicians might influence therapy. Phage- and bactierial-derived are well aware of the role of the immune system in successful pathogen-associated molecular patterns released after treatment of bacterial infections. Notably, an important bacterial lysis have been proposed to stimulate local principle of treatment with bacteriostatic antibiotics is that innate immune responses, which could promote antitu- antibiotic control of bacterial proliferation and spread enables bacterial clearance by host immune cells. Furthermore,

mor immunity or bacterial clearance. Conversely, im- Downloaded from munogenicity of phages induces phage-specific humoral antibiotic-mediated damage of bacterial cells facilitates memory, which can hamper therapeutic success. This phagosomal lysis and release of pathogen-associated molecular review outlines the current knowledge on the different patterns (PAMPs), which enhance bacterial clearance by ac- types of immune responses elicited by phages and their tivating the innate immune system (3–6). In cases of high potential benefits and adverse side effects, when applied bacterial burden, excess release of high amounts of endotoxin (i.e., LPS and other PAMPs) following administration of the therapeutically. This review further summarizes the http://www.jimmunol.org/ knowledge gaps and defines the key immunological antibiotic promotes a systemic inflammatory reaction that can require medical intervention (7). questions that need to be addressed regarding the clin- Similarly to antibiotics, it has been suggested that activation ical application of antibacterial phage therapy. The of innate immunity by phage-mediated lysis of bacterial cells Journal of Immunology, 2018, 200: 3037–3044. could contribute to the efficacy of phage therapy. By contrast, adverse effects of phage therapy have been attributed to toxicity ncreased antibiotic resistance and subsequent therapeutic of LPS contaminations in phage preparations (8, 9). However, failure of antibiotics have fueled the discussion on po- unlike antibiotics, phages themselves are immunogenic mi- by guest on September 24, 2021 I tential alternatives to antibiotics for treatment of bacterial croorganisms that can stimulate an adaptive immune response infections. Many of these alternative approaches are based on and bear the potential to interfere with repetitive treatments. development of biologicals. These include 1) passive immu- Table I provides a summary and a comparison of the rele- nization with Ig preparations enriched for pathogen-specific vant characteristics of antibiotics and antibacterial phage Abs or mAbs directed at known pathogens, 2) induction of therapeutics. a pathogen-specific memory response by vaccination, 3) tar- Moreover, recent mathematical modeling of phage–bacteria geted modulation of the host immune response via immune interactions suggests that the phage alone is unable to exter- stimulatory treatments that facilitate host-mediated clearance minate the whole bacterial population and that cooperation of the extracellular and/or intracellular infecting microbe, and with the immune system is a prerequisite for successful phage 4) elimination of bacteria by lytic phages. With the exception therapy (10, 11). Considering the potential clinical applica- of bacteriophage therapy, these concepts are currently referred to tions for phage therapy, this would imply that either all or as host-directed therapies (1). Among biologicals, phage therapy certain subgroups of immunosuppressed patients (e.g., the is currently experiencing a renaissance because lytic phages patients with highest need of antibacterial therapy) would be targeting antibiotic-resistant bacterial pathogens might repre- excluded from phage therapy, resulting in, overall, only little sent an efficient therapy if development is driven to match the benefit in light of the desperate search for new antibacterial clinical needs and meet regulatory requirements (2). However, therapies targeting nosocomial infections. It was, therefore, an despite concerted efforts from academia, regulators, and bio- important purpose of this review to deliver an analysis of the tech companies, medicinal products for phage therapy are not available data in search of potential limitations for a broad available in countries with regulatory requirements demanding indication of phage therapy that would be grounded in effects a high level of clinical evidence. In this context, uncertainties in arising from phage–immune system interaction.

Division of Microbiology, Paul-Ehrlich-Institute, D-63225 Langen, Germany Abbreviations used in this article: PAMP, pathogen-associated molecular pattern; PRR, pattern recognition receptor. ORCID: 0000-0001-6646-5888 (I.B.-D.). Received for publication December 15, 2017. Accepted for publication February 12, Copyright Ó 2018 by The American Association of Immunologists, Inc. 0022-1767/18/$35.00 2018. Address correspondence and reprint requests to Prof. Isabelle Bekeredjian-Ding, Division of Microbiology, Paul-Ehrlich-Institute, Paul-Ehrlich-Strasse 51–59, D-63225 Langen, Germany. E-mail address: [email protected]

www.jimmunol.org/cgi/doi/10.4049/jimmunol.1701745 3038 BRIEF REVIEWS: THE IMMUNE RESPONSE TO PHAGES

Table I. Comparison of antibacterial therapies with antibiotics or lytic phages

Antibiotics Lytic Phages Mechanism of action Inhibition of cell wall synthesis, Infection and subsequent lysis of bacteria DNA replication, or protein synthesis Specificity (Usually) broad spectrum: Gram-negative Narrow spectrum: one or many individual or Gram-positive species or both strains within a bacterial species Vital microorganism No Yes (inactivation by heat or low pH) Innate immune stimulation No direct effect on innate immune cells Phages contain PAMPs such as DNA and RNA. Release of PAMPs upon loss of Release of PAMPs upon bacterial cell lysis bacterial cell wall integrity Ab induction No Yes (phages are complex biological organisms bearing immunogenic proteins) Half-time life Several hours up to 1 d Depending on host immunity and target species bioburden, hours to weeks Resistance development Natural resistance (target missing) Natural resistance (presence of nonsusceptible strains) Acquired resistance (accessory Acquired resistance (selection of genomic elements encoding nonsusceptible strains based on resistance mechanisms) CRISPR-Cas system, target modification, etc.) Resistance development Downloaded from upon exposure (mutations) Elimination of intracellular bacteria Possible with cell Questionable, possible by concomitant permeant antibiotics uptake of bacteria and phages CRISPR-Cas is a genome editing system found in bacteria and archaea that hinders integration of foreign DNA, characterized by clustered regularly interspaced short palindromic repeats (CRISPR) and endonuclease activity (Cas).

The overall aim of the present review is to provide an of phage application (e.g., oral, intranasal, and i.p.) tested in http://www.jimmunol.org/ overview of the published data and the conclusions to be drawn different rodent models resulted in the rapid appearance of in regard to the beneficial and the unfavorable immunological phages in the bloodstream and their accumulation in the effects of therapeutic phages, with an emphasis on antibacterial kidneys, spleen, liver, and thymus (reviewed in Ref. 15). phage therapy. Nevertheless, summarizing the data reveals Moreover, parental administration of phages in human subjects their major limitation: generalization of findings obtained with led to rapid distribution to the spleen and other organs, where a specific phage or in a specific model is often difficult because phages persisted for more than 4 d (reviewed in Ref. 16). Following the criteria defining when extrapolation would be appropriate oral administration to human subjects, phages were detected in and permissible are unknown. blood and urine samples (see Refs. 15 and 16 for review). How- by guest on September 24, 2021 ever, the ability of phages to penetrate mucosal barriers remains not Physiological interactions of phages with the human immune system fully understood. Several hypotheses have been generated: 1) Phages live where their prey live (e.g., in unsterile areas of the body). transcytosis of phage particles via epithelial cells; 2) circumvention They form part of the physiological mammalian microflora. In the of the epithelial barrier by a Trojan Horse mechanism, in which human intestine, dsDNA and ssDNA phages infect Firmicutes, phages hide inside bacteria; 3) direct sampling of the luminal Bacteroidetes, Proteobacteria, and Actinobacteria, whereas RNA content by intestinal dendritic cells; and 4) translocation of phages phages are thought to be ingested with food, and are present only through a damaged epithelial barrier (17, 18). The latter seems transiently (12, 13). Furthermore, phages appear to adhere to the most convincing because the published data suggest that the in- mucosal surfaces of diverse animals, thus reducing microbial testinal barrier efficiently restricts the translocation of the naturally colonization and pathology at these interfaces and providing a occurring phages into the bloodstream and tissues of healthy ani- non–host-derived layer of immunity (14). It has further been mals (19). Also, the descriptions of naturally occurring phagemia suggested that lytic phages regulate the composition of the are scarce (16). In contrast, application of high doses of phages or microbiome and support its diversity and resilience. Interaction of antibiotics may lead to massive lysis of susceptible bacterial host phages with host innate immune cells and epithelial cells, cytokine cells with consecutive release of endotoxin, which triggers inflam- profiles, and anti-phage Abs might, in turn, control the phageome mation and may thus promote leakiness of the intestinal barrier composition (12). This concept is new and exciting, but it re- (20). Additionally, the receptors responsible for translocation re- mains to be investigated in much more detail. Similarly, concerns main elusive (21), and the significance of phage adherence to that phage therapy could alter the composition of the microbiome mucin for transepithelial translocation needs to be elucidated (22). or that the preexisting immune response to natural phages could In regard to phage therapy, a better understanding of the processes interfere with phage therapy need to be addressed in relevant involved in translocation might allow the distinction of patients at models to clarify the in vivo relevancy of these hypotheses. risk for the formation of systemic adaptive responses that target To initiate the production of Abs and generate a long-lasting orally or topically administered phage preparations. memory response, phages have to be processed and presented to T cells by APCs. Therefore, it is important to understand Phage–phagocyte interactions how phages circumvent the epithelial barriers and reach the The main function of innate immune cells is the recognition normally sterile lymphoid organs. A series of studies were and elimination of foreign material and, when appropriate performed to demonstrate the translocation of phages through and necessary, the mounting of an adaptive immune re- the mucosal barriers in the living organism. Different routes sponse against it. Neutrophils and granulocytes are critical The Journal of Immunology 3039 components of the innate immunity and represent the frontline comparison of phages targeting Pseudomonas aeruginosa and defense against tissue invasion by bacteria and viruses. Since Escherichia coli, respectively (30–32). the 1960s, it has been acknowledged that leukocytes are not The adjuvant properties attributed to the phages are being only able to bind phages in a time-, concentration-, and exploited for phage-based approaches to vaccination and temperature-dependent manner but also internalize and therapy of cancer but may also influence other types of phage eventually eliminate the phages (reviewed in Ref. 23). Whether therapy (33, 34). TLR dependency has been observed in specific receptors are involved in phage endocytosis (the term different types of phage therapies: 1) recruitment of tumor- phagocytosis is used for particles .500 nm) is not completely associated macrophages and tumor regression was absent in resolved, but recognition of the lysine–glycine–aspartic acid MyD88-deficient mice treated with genetically engineered b motif in the p26 capsid protein of the T4 phage by 3 phages that target tumor cells in vivo (35, 36), 2) neutrophils integrins on phagocytes might be implicated in this process were essential for successful treatment of pneumonia with (16, 24). Fig. 1 visualizes the different cellular mechanisms anti–P. aeruginosa phages (11), 3) phage-mediated provision that promote uptake of phages into macrophages and subse- of Ag to dendritic cells and subsequent T cell activation were quent macrophage activation (Fig. 1). strongly impaired in the absence of MyD88 or TLR9 (37), Once activated, macrophages produce an array of micro- and 4) vaccine responses to peptide-displaying phages were bicide effectors and immunoregulatory cytokines that act in abrogated in the absence of MyD88 and altered in the ab- concert to eliminate the invasive agent and influence the course sence of TLR9 (28). These findings support the hypothesis of the developing immune response. Lytic phages induce that interactions between phage-derived ligands and host Downloaded from massive LPS release from Gram-negative bacteria; however, PRR contribute to efficacy of phage therapies. This raises the this does not exceed the amounts released under antibi- question of the actual clinical impact of PRR activation and otic treatment (25). Degradation of phages by poly- the concern that phage therapy could be less effective in morphonuclear cells and macrophages has been described for patients with immune deficiencies that affect their innate T2, wX174, l, and P22 phages (see Ref. 26 for review). These immune responsiveness. To tackle this issue, clinical study studies demonstrate rapid phagocyte-mediated clearance of protocols would need to include patients with different types http://www.jimmunol.org/ phages from organs, in particular the liver, where a rapid of immune deficiencies. decline of highly infective phage titers was observed. Murine Kupffer cells, the resident liver phagocytes, were subsequently Phage-specific Ab responses shown to be responsible for this process (27). Furthermore, in Initial evidence for phage-mediated induction of adaptive the spleen, high titers of phages were detectable irrespective of immunity traces back to studies describing background neu- the route of phage immunization. Here, splenic macrophages tralizing activity in sera and active immunization with well- were found to mediate clearance of phages, but they displayed described phages, such as fX174 and T4-like phages. f f an approximately four-times-slower kinetic than those ob- X174. The enterobacterial phage X174 contains a small by guest on September 24, 2021 served with Kupffer cells (27). circular (+) ssDNA genome. The viral particles display speci- Notably, the degradation process is also a prerequisite for Ag ficity for LPS on the outer membrane of Gram-negative presentation and initiation of adaptive immune responses. It can pathogens. Priming and boostering of the humoral immune thus be envisioned that improvement of phage therapy could response were initially demonstrated in guinea pigs and rabbits include selection of phages with natural resistance to phagosomal immunized with fX174 (38–40). These experiments provided degradation or rational design of recombinant phages to achieve evidence for the presence of phage-specific short-lived IgM this. This could avoid or delay induction of phage-specific (19S) and memory responses (IgG/7S) with a long t1/2 life. adaptive immune responses and possibly prolong persistence In mice, immunization with this phage elicited neutralizing of phages in immune-competent individuals. Ab responses in a dose-dependent manner (41). Of note, background levels of serum neutralizing activity were higher in Phage-mediated activation of pattern recognition receptors conventional breeding than in germ-free mice. In humans, A recent transcriptome analysis revealed that lysates from five fX174 was used as a model Ag for induction of T cell– different endotoxin-free phage preparations induced compa- dependent Ab responses by diagnostic immunization with rable cytokine signatures in human PBMCs (9). These cyto- fX174 in patients with suspected common variable immune kine profiles significantly differed from those induced by deficiency or hypogammaglobulinemia, HIV, or those who had potentially contaminating LPS, and addition of LPS costim- undergone a splenectomy (42–45). fX174-specific IgM and ulated but did not alter the phage-induced cytokine profiles. IgG responses were elicited in patients with intact B cell Nevertheless, phage-mediated immune stimulation can most responses and absent or impaired in patients with B cell likely be attributed to the activation of pattern recognition immunodeficiency or HIV. Furthermore, IgG responses to receptors (PRR) by PAMPs present in phages. Phages contain fX174 were absent in patients with defective Ig class genomic DNA and RNA (e.g., ligands for nucleic acid– switching, including splenectomized patients. sensing receptors, such as TLRs 7, 8, 9, and 13, RLR, and T4 and T4-like phages. T2 and T4 phages have a dsDNA ge- cytosolic DNA sensors) as well as compounds exerting TLR2 nome and target and lyse enterobacteria. Abs against the T2 and TLR4 activity (28). Additionally to PAMPs, some phages phage were demonstrated in lymph node cultures from rats and express proteins that mediate interaction with mammalian in rabbit sera (46, 47). Binding of anti-T2 IgM and IgG to the host cells and thereby promote immune responses (29). To- head and tail of the phage was demonstrated with electron gether with possible differences in endotoxin contamination microscopy (48). Neutralizing Abs were found to react levels, these virulence factors may account for the recently with the distal phage tail. Notably, a serum complement described differences in cytokine induction observed upon component was found to be necessary for neutralization of 3040 BRIEF REVIEWS: THE IMMUNE RESPONSE TO PHAGES Downloaded from http://www.jimmunol.org/ by guest on September 24, 2021

FIGURE 1. Phage–host immune cell interactions. The graph summarizes the multiple sites of interaction between phages (dark blue), bacteria (purple), and host macrophages (light blue). Upper panel, Phage-mediated lysis of bacterial cells. (A) Phages infect bacterial cells. (B) Phages replicate in the bacterial host. (C) Phages induce lysis of bacteria. (D) Phage-specific Abs (orange) block bacterial infection by phages. Lower panel, Phage-mediated activation of host macrophages in the presence (yellow endosomes) and absence (orange endosomes) of Abs (anti-phage Abs in orange; antibacterial Abs in red). (1) Phages expressing proteins that mediate host–phage interactions bind to macrophage surface receptors and activate macrophages. (2) Macrophages phagocytize extracellular bacterial cells and endocytose phages; bacterial and phage-derived PAMPs stimulate macrophage activity. (3) Macrophages phagocytize phage-infected bacterial cells; phage- derived PAMPs costimulate bacteria-induced macrophage activity. (4) After phage-induced bacterial lysis macrophages phagocytize bacterial debris and phages; bacterial and phage-derived PAMPs stimulate macrophage activity. (5) Macrophages phagocytize opsonized bacteria upon recognition of bacteria-specific Ig by Fc receptors (yellow); this enhances clearance of bacteria. (6) Phage-Ab complexes bind to Fc receptors (yellow) on macrophages; this triggers endocytosis and subsequent clearance of phages. The Journal of Immunology 3041

T2 but not FX174 in mice (49). Abs against T4-like phages Although clearance occurred within an hour in wild type were detected in 81% of tested human sera. IgG was directed mice, phage titers persisted in both SCID and B cell–deficient against several phage head proteins and displayed neutralizing mice. More recent in vivo studies demonstrated that anti- activity (29). In vivo experiments in mice revealed that phage Abs mediate clearance of phages from the gut and immunization induced high titers of anti-Hoc Abs with circulation: increasing levels of phage-specific IgA in feces neutralizing activity against T4; this correlated with a loss of correlated with gradual absence of orally ingested phages (62) T4-mediated protection against E. coli infection (29). Of note, and high levels of anti-phage IgG, with elimination of phages the Hoc capsid glycoprotein was previously described as a from the blood and tissues after i.p. administration (63). highly immunogenic protein containing Ig-like domains but Furthermore, kinetic studies of phage circulation in the as nonessential in regard to infectivity of the phage (50, 51). peripheral blood of healthy and X-linked agammaglobulin- It was recently shown that heat-inactivated phages lose emia patients provided evidence that elimination of phages in their immune stimulatory potential, whereas UV irradiation humans may similarly require opsonization by Abs. Whereas in preserves immunogenicity (e.g., phage-induced Ab responses) normal subjects injected with fX174, phages were completely elicited by live phage preparations (52). Interestingly enough, in eliminated within 4 d, in patients with very low Ab titers, 1961, Fishman described that a RNAse-sensitive substance in phages remained detectable in the circulation for up to 7 wk macrophage or T2 phage cocultures was capable of stimu- (64). lating Ab synthesis (46). Further studies suggested a stimu- latory role of phage-derived RNA in the induction of IFN Association of neutralizing Abs with therapeutic outcome Downloaded from (53). Because IFN-I triggers TLR7 expression in B cells, Neutralizing Abs constitute a risk for therapeutic failure in phage-derived RNA could directly contribute to B cell ac- phage therapy. Notably, early electron microscopy studies tivation and the synthesis of anti-phage Abs (54). Well in using the E. coli–specific ssDNA phage M13 associate Ab- line with these observations, in vivo experiments demon- mediated neutralization with mechanical hindrance of bacte- strate that Ab production induced by the ssDNA phage M13 rial cell penetration due to bound Ig (65). is TLR dependent (e.g., absent) in MyD88-deficient mice Two studies on patients receiving phage therapy are avail- http://www.jimmunol.org/ and decreased in TLR2-, TLR4-, and TLR7-deficient mice able. In a study from 1987, anti-phage Ab titers were studied in (28). Notably, in mice, M13 induced IgG2b/c and IgG3 re- 57 patients before and after phage therapy for different types of sponses but only weak IgG1 responses; the latter were in- infections (66). In 13 cases, anti-phage Abs were present be- creased in TLR9-deficient mice, thus confirming earlier fore initiation of therapy and increased upon therapy; in 17 observations that IgG1 is increased in the absence of cases, anti-phage Abs became detectable under therapy. The TLR9 (55). authors emphasize that only in two cases did the Ab titers Based on these reports, phages deliver all necessary com- before therapy exceed a 1:80 dilution of the serum. In these ponents for stimulation of naive B cells (56, 57): 1) im- two cases, the outcome of therapy was unsatisfactory. For all by guest on September 24, 2021 munogenic phage proteins enabling specific recognition via others, the presence and induction of anti-phage Abs did not the BCR, 2) ligands for activation of costimulatory TLRs in lead to therapy failure. However, because of the low number B cells (e.g., nucleic acids), such as RNA derived from (vi- of cases and in the absence of a rigorous statistical analysis, the able) phages (danger signal), and 3) APC-mediated activa- results remain descriptive. The data collected in a more recent tion of helper T cells. Albeit there are no studies available study on 20 patients with Staphylococcus aureus infections that describe the Ag-specific T cell response to phages, the reveal that there is significant induction of anti-phage IgM sequential induction of IgM and IgG upon immunization and, in particular, IgG titers caused by therapy (67). Neu- provides evidence that T cells specific for phage Ag develop tralizing activity in sera increases with the Ab titers. In ad- and enable class switch recombination (45). In theory, this dition, generation of IgG is strongly dependent on the bears the potential that adaptive immune responses could be individual phage. The data do not allow conclusions on the avoided by preselection of phages with low immunostimu- predictive value of neutralizing Ab titers in regard to patient latory potential. outcome. Taken together, these studies indicate that the presence of The role of anti-phage Abs in the clearance of phages phage-specific Abs can interfere with therapeutic efficacy. This In a murine model of bacteremia with antibiotic resistant is particularly relevant in chronic infections in which repeated E. coli and P. aeruginosa strains, an increase in anti-phage treatments with the same phages boost the humoral immune IgG levels (58, 59) was detectable on day 10 after admin- response. Conversely, the induction of phage-specific Ig seems istration of 109 PFU of the respective phages but had no irrelevant for treatment of acute infections because antibac- effect on survival because of the short disease course, which terial effects of phages become effective before Abs are formed. only lasted for a few days. IgG levels reached the maximum However, prescreening of patient sera for the absence of anti- ∼20-fold increase after 30 and 40 d, respectively. In an phage Abs could be advisable before initiation of therapy. The earlier report on acute bacteremia, the authors showed that delay may, however, not be compatible with acute treatment. A despite using similar doses (1010 PFU), anti-phage IgM and possible solution for pre-existent anti-phage Abs is presented in IgG only became detectable after three repetitive phage in- a recent report that proposes that packaging of phages into jections and levels (5-fold and 3800-fold over baseline, re- liposomes not only promotes cellular uptake into infected spectively). The levels remained unchanged despite further macrophages but avoids binding of anti-phage Abs and neu- phage injections (60). tralization (68). Nevertheless, this approach might not be An early study in immune-deficient mice highlighted a role feasible in infections with extracellular bacteria. Here, genetic for B cells in clearance of phages after i.v. injection (61). deletion of phage genes encoding immunogenic proteins 3042 BRIEF REVIEWS: THE IMMUNE RESPONSE TO PHAGES might represent an interesting alternative for sustenance or clearance of phages from the human body (62, 63). However, rescue of phage efficacy. This approach would, however, re- to date, it is not well understood whether this regulatory role quire an extensive characterization of phages with the scope of anti-phage Abs could also be important for prevention of of comprehensive identification of potentially immunogenic resistance development to phages and, additionally, whether phage proteins, which, considering the currently applied pre-existing immunity to natural phages could affect phage methods, is difficult to achieve. Ultimately, genetically engi- therapy. Furthermore, it is unclear which phage-specific fac- neered lytic phages would be applied to patients, which could tors have impact on this specific mechanism of clearance. raise additional regulatory concerns. Thus, more in-depth investigations are warranted to elucidate whether phage persistence due to lack of immune control (64) Conclusions might favor selection of phage-resistant bacterial strains in Taken together, the vast majority of data indicate that ther- vivo. apeutic application of high-titer bacteriophages exposes them Together, the compiled data indicate that there are gaps in to and stimulates the host immune system. Although some our understanding of the clinical relevancy of the phage- studies describe dose- and time-dependent immunological immune interaction. Nevertheless, immunogenicity of phages effects, few were conducted with phages used for antibacterial itself does not seem to represent a relevant safety risk for therapy. Moreover, the relevancy of the systemic anti-phage patients. However, many studies refer to phages selected for immune response has to be assessed in conjunction with diagnostic immunization or tumor therapy. The results ob- duration of treatment until recovery and route of adminis- tained in these studies cannot be extrapolated to lytic phages, Downloaded from tration. This has, so far, not been systematically addressed in and reports on the immune effects of clinically applied lytic patients or animal models. phages are very limited. Furthermore, validated assays for Three major areas of phage-immune interaction can be standardized measurement of in vitro and in vivo immune discerned: effects exerted by therapeutic candidate phages are outstanding First, immune recognition via PRR mainly contributes to despite the importance of these findings from a clinical and resolution of infection by recruiting phagocytes to the in- regulatory perspective. http://www.jimmunol.org/ fection site (11). Phage-mediated activation of innate im- At present, the clinical value of phage therapy remains mune cells is mainly based on recognition of phage-derived uncertain. This literature review reveals that, more recently, DNA and RNA by PRR. Specific PRR engagement and the reports are increasingly addressing the major gaps in our extent of immune activation, therefore, will differ depend- understanding of the immunological aspects of phage therapy, ing on the phage type, the phage dose, and nucleic acid mainly in in vivo infection models. However, despite a con- synthesis activity. Not surprisingly, successful treatment of tinuously increasing number of publications on phage therapy, infections with multidrug-resistant pathogens was primarily there are few studies using validated methods and rigorous dependent on phage specificity and susceptibility of the controls. The introduction of validated in vitro and in vivo by guest on September 24, 2021 infecting bacterial strains (59, 60, 69). However, although methods is indispensable, as it will permit the assessment of phage therapy may be fully effective in lymphopenic pa- comparability of immune effects of different phages and phage tients, it might be impaired in patients with myeloid cell combinations. Only these validated methods will allow valid deficiencies. conclusions on the key issues, such as 1) the value of immune- Second, immunogenicity of phages promotes the formation based parameters for selection of phages and identification of phage-neutralizing Abs that can hamper therapeutic success of responsive patient populations and 2) exchangeability of and increase with repeated administration (60). Although phages, an important prerequisite for individualized phage nucleic acid–sensing PRR drive the process of Ab generation therapy concepts. (46), the arising Abs are directed at immunogenic proteins present on the phages (29, 70). At present, it is not known whether selection of natural phages from different sources is Disclosures The authors have no financial conflicts of interest. biased toward more or toward less immunogenic phages. Protein expression is considered a characteristic of the indi- vidual phage; Ab induction is, therefore, predicted to be References highly variable, and immunogenicity should be taken into 1. Zumla, A., M. Rao, R. S. Wallis, S. H. E. Kaufmann, R. Rustomjee, P. Mwaba, account in the selection of phages for therapeutic purposes. C. Vilaplana, D. Yeboah-Manu, J. Chakaya, G. Ippolito, et al; Host-Directed Therapies Network consortium. 2016. Host-directed therapies for infectious dis- Additionally, and despite its challenges, genetic engineering of eases: current status, recent progress, and future prospects. Lancet Infect. Dis. 16: phages to abrogate the expression of immunogenic proteins e47–e63. 2. Expert round table on acceptance and re-implementation of bacteriophage therapy. might represent an interesting approach to avoid therapeutic 2016. Silk route to the acceptance and re-implementation of bacteriophage therapy. inefficacy due to the presence of phage-specific Ig in repeated Biotechnol. J. 11: 595–600. 3.Hilmi,D.,M.Parcina,D.Stollewerk,J.Ostrop,M.Josten,A.Meilaender, administration courses of well-characterized immunogenic U. Zaehringer, T. A. Wichelhaus, G. Bierbaum, K. Heeg, et al. 2014. Hetero- phages. For the time being, avoidance of PRR-mediated im- geneity of host TLR2 stimulation by Staphylocoocus aureus isolates. PLoS One 9: mune effects could lead to reduction of phage-induced neu- e96416. 4. Muller,€ S., A. J. Wolf, I. D. Iliev, B. L. Berg, D. M. Underhill, and G. Y. Liu. 2015. tralizing Ab formation. This could be achieved by dose Poorly cross-linked peptidoglycan in MRSA due to mecA induction activates the reductions based on the use of self-amplifying phages or inflammasome and exacerbates immunopathology. Cell Host Microbe 18: 604–612. 5. Wolf, A. J., G. Y. Liu, and D. M. Underhill. 2017. Inflammatory properties of refined dose-ranging studies. antibiotic-treated bacteria. J. Leukoc. Biol. 101: 127–134. Third, phage-specific IgG or IgA can limit the multiplication 6. Wolf, A. J., A. Arruda, C. N. Reyes, A. T. Kaplan, T. Shimada, K. Shimada, M. Arditi, G. Liu, and D. M. Underhill. 2011. Phagosomal degradation increases rate of phages. High Ab levels and Fc receptor–mediated TLR access to bacterial ligands and enhances macrophage sensitivity to bacteria. uptake of phage/Ab complexes by macrophages promote the J. Immunol. 187: 6002–6010. The Journal of Immunology 3043

7. Butler, T. 2017. The Jarisch-Herxheimer reaction after antibiotic treatment of 35. Eriksson, F., P. Tsagozis, K. Lundberg, R. Parsa, S. M. Mangsbo, M. A. A. Persson, spirochetal infections: a review of recent cases and our understanding of patho- R. A. Harris, and P. Pisa. 2009. Tumor-specific bacteriophages induce tumor de- genesis. Am. J. Trop. Med. Hyg. 96: 46–52. struction through activation of tumor-associated macrophages. J. Immunol. 182: 8. Szermer-Olearnik, B., and J. Boratynski. 2015. Removal of endotoxins from bac- 3105–3111. teriophage preparations by extraction with organic solvents. PLoS One 10: 36. Eriksson, F., W. D. Culp, R. Massey, L. Egevad, D. Garland, M. A. A. Persson, and e0122672. P. Pisa. 2007. Tumor specific phage particles promote tumor regression in a mouse 9. Van Belleghem, J. D., F. Clement, M. Merabishvili, R. Lavigne, and melanoma model. Cancer Immunol. Immunother. 56: 677–687. M. Vaneechoutte. 2017. Pro- and anti-inflammatory responses of peripheral blood 37. Sartorius, R., L. D’Apice, M. Trovato, F. Cuccaro, V. Costa, M. G. De Leo, mononuclear cells induced by Staphylococcus aureus and Pseudomonas aeruginosa V. M. Marzullo, C. Biondo, S. D’Auria, M. A. De Matteis, et al. 2015. Antigen phages. Sci. Rep. 7: 8004. delivery by filamentous bacteriophage fd displaying an anti-DEC-205 single-chain 10. Leung, C. Y. J., and J. S. Weitz. 2017. Modeling the synergistic elimination of variable fragment confers adjuvanticity by triggering a TLR9-mediated immune bacteria by phage and the innate immune system. J. Theor. Biol. 429: 241–252. response. EMBO Mol. Med. 7: 973–988. 11. Roach, D. R., C. Y. Leung, M. Henry, E. Morello, D. Singh, J. P. Di Santo, 38. Uhr, J. W., and M. S. Finkelstein. 1963. Antibody formation. IV. Formation of J. S. Weitz, and L. Debarbieux. 2017. Synergy between the host immune system and rapidly and slowly sedimenting antibodies and immunological memory to bacte- bacteriophage is essential for successful phage therapy against an acute respiratory riophage phi-X 174. J. Exp. Med. 117: 457–477. pathogen. Cell Host Microbe 22: 38–47.e4. 39. Uhr, J. W., M. S. Finkelstein, and J. B. Baumann. 1962. Antibody formation. III. 12. Mirzaei, M. K., and C. F. Maurice. 2017. Me´nage a` trois in the human gut: in- The primary and secondary antibody response to bacteriophage phi X 174 in guinea teractions between host, bacteria and phages. Nat. Rev. Microbiol. 15: 397–408. pigs. J. Exp. Med. 115: 655–670. 13. Zhang, T., M. Breitbart, W. H. Lee, J.-Q. Run, C. L. Wei, S. W. L. Soh, 40. Ha´jek, P. 1970. Neutralization of bacterial viruses by antibodies of young animals. M. L. Hibberd, E. T. Liu, F. Rohwer, and Y. Ruan. 2006. RNA viral community in 3. The development of the avidity of 19S and 7S neutralizing antibodies in the human feces: prevalence of plant pathogenic viruses. PLoS Biol. 4: e3. course of primary and secondary response in young rabbits immunized with PhiX 14. Barr, J. J., R. Auro, M. Furlan, K. L. Whiteson, M. L. Erb, J. Pogliano, A. Stotland, 174 bacteriophage. Folia Microbiol. (Praha) 15: 9–16. R. Wolkowicz, A. S. Cutting, K. S. Doran, et al. 2013. Bacteriophage adhering to 41. Stashak, P. W., P. J. Baker, and B. S. Roberson. 1970. The serum antibody response mucus provide a non-host-derived immunity. Proc. Natl. Acad. Sci. USA 110: to bacteriophage phi chi 174 in germ-free and conventionally reared mice. I. Assay 10771–10776. of neutralizing antibody by a 50 per cent neutralization method. Immunology 18: 15. Barr, J. J. 2017. A bacteriophages journey through the human body. Immunol. Rev. 295–305. Downloaded from 279: 106–122. 42. Rubinstein, A., Y. Mizrachi, L. Bernstein, J. Shliozberg, M. Golodner, G. Q. Liu, 16. Go´rski, A., E. Wazna, B.-W. Dabrowska, K. Dabrowska, K. Swita1a-Jelen, and and H. D. Ochs. 2000. Progressive specific immune attrition after primary, sec- R. Miedzybrodzki. 2006. Bacteriophage translocation. FEMS Immunol. Med. ondary and tertiary immunizations with bacteriophage phi X174 in asymptomatic Microbiol. 46: 313–319. HIV-1 infected patients. AIDS 14: F55–F62. 17. Duerkop, B. A., and L. V. Hooper. 2013. Resident viruses and their interactions 43. Smith, L. L., R. Buckley, and P. Lugar. 2014. Diagnostic immunization with with the immune system. Nat. Immunol. 14: 654–659. bacteriophage FX 174 in patients with common variable immunodeficiency/hy- 18. Nguyen, S., K. Baker, B. S. Padman, R. Patwa, R. A. Dunstan, T. A. Weston, pogammaglobulinemia. Front. Immunol. 5: 410.

K. Schlosser, B. Bailey, T. Lithgow, M. Lazarou, et al. 2017. Bacteriophage 44. Sullivan, J. L., H. D. Ochs, G. Schiffman, M. R. Hammerschlag, J. Miser, http://www.jimmunol.org/ transcytosis provides a mechanism to cross epithelial cell layers. [Published erratum E. Vichinsky, and R. J. Wedgwood. 1978. Immune response after splenectomy. appears in 2018 MBio 9: e02207-17.] MBio 8: e01874-17. Lancet 1: 178–181. 19. Geier, M. R., M. E. Trigg, and C. R. Merril. 1973. Fate of bacteriophage lambda in 45. Pyun, K. H., H. D. Ochs, R. J. Wedgwood, X. Q. Yang, S. R. Heller, and non-immune germ-free mice. Nature 246: 221–223. C. B. Reimer. 1989. Human antibody responses to bacteriophage phi X 174: se- 20. Weber-Dabrowska, B., M. Dabrowski, and S. Slopek. 1987. Studies on bacterio- quential induction of IgM and IgG subclass antibody. Clin. Immunol. Immunopa- phage penetration in patients subjected to phage therapy. Arch. Immunol. Ther. Exp. thol. 51: 252–263. (Warsz.) 35: 563–568. 46. Fishman, M. 1961. Antibody formation in vitro. J. Exp. Med. 114: 837–856. 21. Hamzeh-Mivehroud, M., A. Mahmoudpour, H. Rezazadeh, and S. Dastmalchi. 47. Toussaint, A. J., and L. H. Muschel. 1962. Studies on the bacteriophageneutralizing 2008. Non-specific translocation of peptide-displaying bacteriophage particles activity of serums. I. An assay procedure for normal antibody and complement. J. across the gastrointestinal barrier. Eur. J. Pharm. Biopharm. 70: 577–581. Immunol. 89: 27–34. 22. Barr, J. J., R. Auro, N. Sam-Soon, S. Kassegne, G. Peters, N. Bonilla, M. Hatay, 48. Ho¨glund, S. 1967. Electron microscopic investigations of the interaction between

S. Mourtada, B. Bailey, M. Youle, et al. 2015. Subdiffusive motion of bacteriophage the T2-phage and its IgG-and IgM-antibodies. Virology 32: 662–677. by guest on September 24, 2021 in mucosal surfaces increases the frequency of bacterial encounters. Proc. Natl. Acad. 49. Ha´jek, P., and L. Mandel. 1966. Antibody response of young animals to bacte- Sci. USA 112: 13675–13680. riophages of different immunological behaviour: phi X 174 and T2. Folia Microbiol. 23. Kurzepa, A., K. Dabrowska, G. Skaradzinski, and A. Go´rski. 2009. Bacteriophage (Praha) 11: 282–289. interactions with phagocytes and their potential significance in experimental ther- 50. Fokine, A., M. Z. Islam, Z. Zhang, V. D. Bowman, V. B. Rao, and apy. Clin. Exp. Med. 9: 93–100. M. G. Rossmann. 2011. Structure of the three N-terminal immunoglobulin do- 24. Go´rski, A., R. Mie˛dzybrodzki, J. Borysowski, K. Da˛browska, P. Wierzbicki, mains of the highly immunogenic outer capsid protein from a T4-like bacterio- M. Ohams, G. Korczak-Kowalska, N. Olszowska-Zaremba, M. Łusiak- phage. J. Virol. 85: 8141–8148. Szelachowska, M. K1ak, et al. 2012. Phage as a modulator of immune responses: 51. Robertson, K., Y. Furukawa, A. Underwood, L. Black, and J. L. Liu. 2012. Deletion practical implications for phage therapy. Adv. Virus Res. 83: 41–71. of the Hoc and Soc capsid proteins affects the surface and cellular uptake properties 25. Dufour, N., R. Delattre, J.-D. Ricard, and L. Debarbieux. 2017. The lysis of of bacteriophage T4 derived nanoparticles. Biochem. Biophys. Res. Commun. 418: pathogenic Escherichia coli by bacteriophages releases less endotoxin than by 537–540. b-lactams. Clin. Infect. Dis. 64: 1582–1588. 52. Samoylova, T. I., M. D. Norris, A. M. Samoylov, A. M. Cochran, K. G. Wolfe, 26. Kaur, T., N. Nafissi, O. Wasfi, K. Sheldon, S. Wettig, and R. Slavcev. 2012. V. A. Petrenko, and N. R. Cox. 2012. Infective and inactivated filamentous phage as Immunocompatibility of bacteriophages as nanomedicines. J. Nanotechnol. 2012: carriers for immunogenic peptides. J. Virol. Methods 183: 63–68. 1–13. 53. Ta¨borsky´, I., and V. Dolnı´k. 1977. Ability of human polymorhonuclear blood cells 27. Inchley, C. J. 1969. The activity of mouse Kupffer cells following intravenous in- to produce interferon after induction with phage double-stranded RNA. Acta Virol. jection of T4 bacteriophage. Clin. Exp. Immunol. 5: 173–187. 21: 499–502. 28. Hashiguchi, S., Y. Yamaguchi, O. Takeuchi, S. Akira, and K. Sugimura. 2010. 54. Bekeredjian-Ding, I. B., M. Wagner, V. Hornung, T. Giese, M. Schnurr, S. Endres, Immunological basis of M13 phage vaccine: regulation under MyD88 and TLR9 and G. Hartmann. 2005. Plasmacytoid dendritic cells control TLR7 sensitivity of signaling. Biochem. Biophys. Res. Commun. 402: 19–22. naive B cells via type I IFN. [Published erratum appears in 2005 J Immunol. 174: 29. Da˛browska, K., P. Miernikiewicz, A. Piotrowicz, K. Hodyra, B. Owczarek, 5884.] J. Immunol. 174: 4043–4050. D. Lecion, Z. Kazmierczak, A. Letarov, and A. Go´rski. 2014. Immunogenicity 55. Jegerlehner, A., P. Maurer, J. Bessa, H. J. Hinton, M. Kopf, and M. F. Bachmann. studies of proteins forming the T4 phage head surface. J. Virol. 88: 12551–12557. 2007. TLR9 signaling in B cells determines class switch recombination to IgG2a. J. 30. Shiley, J. R., K. K. Comfort, and J. B. Robinson. 2017. Immunogenicity and an- Immunol. 178: 2415–2420. timicrobial effectiveness of Pseudomonas aeruginosa specific bacteriophage in a hu- 56. Bekeredjian-Ding, I., and G. Jego. 2009. Toll-like receptors–sentries in the B-cell man lung in vitro model. Appl. Microbiol. Biotechnol. 101: 7977–7985. response. Immunology 128: 311–323. 31. Khan Mirzaei, M., Y. Haileselassie, M. Navis, C. Cooper, E. Sverremark-Ekstro¨m, 57. Ruprecht, C. R., and A. Lanzavecchia. 2006. Toll-like receptor stimulation as a and A. S. Nilsson. 2016. Morphologically distinct Escherichia coli bacteriophages third signal required for activation of human naive B cells. Eur. J. Immunol. 36: differ in their efficacy and ability to stimulate cytokine release in vitro. Front. 810–816. Microbiol. 7: 437. 58. Wang, J., B. Hu, M. Xu, Q. Yan, S. Liu, X. Zhu, Z. Sun, E. Reed, L. Ding, 32. Dufour, N., M. Henry, J.-D. Ricard, and L. Debarbieux. 2016. Commentary: J. Gong, et al. 2006. Use of bacteriophage in the treatment of experimental animal morphologically distinct Escherichia coli bacteriophages differ in their efficacy and bacteremia from imipenem-resistant Pseudomonas aeruginosa. Int. J. Mol. Med. 17: ability to stimulate cytokine release in vitro. Front. Microbiol. 7: 1029. 309–317. 33. Gambashidze, K., P. Khorava, T. Azaladze, K. Kalandarishvili, E. Jaiani, 59. Wang, J., B. Hu, M. Xu, Q. Yan, S. Liu, X. Zhu, Z. Sun, D. Tao, L. Ding, E. Reed, B. Lasareishvil, A. Azaladze, and M. Tediashvili. 2012. Antitumor and adjuvant et al. 2006. Therapeutic effectiveness of bacteriophages in the rescue of mice with effects of phagelysates of E.coli in mice with Ehrlich carcinoma. Exp. Oncol. 34: extended spectrum beta-lactamase-producing Escherichia coli bacteremia. Int. J. Mol. 107–111. Med. 17: 347–355. 34. Pajtasz-Piasecka, E., J. Rossowska, D. Duś, B. Weber-Dabrowska, A. Zab1ocka, and 60. Biswas, B., S. Adhya, P. Washart, B. Paul, A. N. Trostel, B. Powell, R. Carlton, and A. Go´rski. 2008. Bacteriophages support anti-tumor response initiated by DC-based C. R. Merril. 2002. Bacteriophage therapy rescues mice bacteremic from a clinical vaccine against murine transplantable colon carcinoma. Immunol. Lett. 116: 24–32. isolate of vancomycin-resistant Enterococcus faecium. Infect. Immun. 70: 204–210. 3044 BRIEF REVIEWS: THE IMMUNE RESPONSE TO PHAGES

61. Srivastava, A. S., T. Kaido, and E. Carrier. 2004. Immunological factors that affect 66. Kucharewicz-Krukowska, A., and S. Slopek. 1987. Immunogenic effect of bacte- the in vivo fate of T7 phage in the mouse. J. Virol. Methods 115: 99–104. riophage in patients subjected to phage therapy. Arch. Immunol. Ther. Exp. (Warsz.) 62. Majewska, J., W. Beta, D. Lecion, K. Hodyra-Stefaniak, A. K1opot, 35: 553–561. _ Z. Kazmierczak, P. Miernikiewicz, A. Piotrowicz, J. Ciekot, B. Owczarek, et al. 67. Zaczek, M., M. Łusiak-Szelachowska, E. Jonczyk-Matysiak, B. Weber-Da˛browska, 2015. Oral application of T4 phage induces weak antibody production in the gut R. Mie˛dzybrodzki, B. Owczarek, A. Kopciuch, W. Fortuna, P. Rogo´z,_ and and in the blood. Viruses 7: 4783–4799. A. Go´rski. 2016. Antibody production in response to Staphylococcal MS-1 phage 63. Hodyra-Stefaniak, K., P. Miernikiewicz, J. Drapa1a, M. Drab, E. Jonczyk-Matysiak, cocktail in patients undergoing phage therapy. Front. Microbiol. 7: 1681. D. Lecion, Z. Kazmierczak, W. Beta, J. Majewska, M. Harhala, et al. 2015. 68. Singla, S., K. Harjai, O. P. Katare, and S. Chhibber. 2016. Encapsulation of bac- Mammalian host-versus-phage immune response determines phage fate in vivo. Sci. teriophage in liposome accentuates its entry in to macrophage and shields it from Rep. 5: 14802. neutralizing antibodies. PLoS One 11: e0153777. 64. Ochs, H. D., S. D. Davis, and R. J. Wedgwood. 1971. Immunologic responses to 69. Vinodkumar, C. S., Y. F. Neelagund, and S. Kalsurmath. 2005. Bacteriophage in bacteriophage phi-X 174 in immunodeficiency diseases. J. Clin. Invest. 50: 2559– the treatment of experimental septicemic mice from a clinical isolate of multidrug 2568. resistant Klebsiella pneumoniae. J. Commun. Dis. 37: 18–29. 65. Beaudoin, J., and D. Pratt. 1974. Antiserum inactivation of electrophoretically 70. Dabrowska, K., K. Swita1a-Jelen, A. Opolski, and A. Go´rski. 2006. Possible asso- purified M13 diploid virions: model for the F-specific filamentous bacteriophages. J. ciation between phages, Hoc protein, and the immune system. Arch. Virol. 151: Virol. 13: 466–469. 209–215. Downloaded from http://www.jimmunol.org/ by guest on September 24, 2021