Views and CoMMentaries Views and CoMMentaries Bacteriophage 3:1, e24219; January/February/March 2013; © 2013 Landes Bioscience Phage therapy Should bacterial resistance to phages be a concern, even in the long run? Anni-Maria Örmälä1 and Matti Jalasvuori2,3,* 1Department of Biosciences; University of Helsinki; Helsinki, Finland; 2Center of Excellence in Biological Interactions; Department of Biological and Environmental Science; University of Jyväskylä; Jyväskylä, Finland; 3Division of Ecology, Evolution and Genetics; Research School of Biology; Australian National University; Canberra, Australia Bacteriophage therapy, the use of viruses such as conjugative plasmids,2 and a single that infect bacteria as antimicrobials, has element can confer resistance to many dif- been championed as a promising alterna- ferent antibiotics.3 Furthermore, plasmids tive to conventional antibiotics. Although spread and persist within populations in the laboratory bacterial resistance even in absence of antibiotics, suggesting against phages arises rapidly, resistance so that decrease in the use of antibiotics is far has been an only minor problem for the not likely to help the situation.4 Although effectiveness of phage therapy. Resistance there may be means to get rid of the resis- to antibiotics, however, has become a tance or prevent its spread,3 more gener- major issue after decades of extensive use. ally effective alternatives to conventional Should we expect similar problems after antibiotics would be useful for combatting long-term use of phages as antimicrobials? those bacterial pathogens that present as Like antibiotics, phages are often noted to antibiotic resistant. be drivers of bacterial evolution. Should we The ability of bacteriophages to expect phage-treated pathogens to develop clear and kill bacteria often inspires the a general resistance to phages over time, a idea that these bacterial viruses may be resistance against which only, for example, included among these possible alternatives hypothetically co-evolved phages might to conventional antibiotic agents, that is, be infective? Here we argue that the global as selectively toxic antibacterials. This ini- infection patterns of phages suggest that tial excitement, however, can fade when this is not necessarily a concern as envi- it transpires that bacteria also can rapidly ronmental phages often can infect bacteria develop resistance to phages, even within with which those phages lack any recent relatively small bacterial populations (of, co-evolutionary history. e.g., about 108 cells).5 Bacteria can become Bacterial pathogens have been treated resistant by several mechanisms, includ- effectively with antibiotics for several ing for example alteration of receptors to Keywords: evolution, ecology, decades. However, the extensive use of which phages attach or the development phage-therapy, antibiotic resistance, antibiotics, both properly and improperly, of adaptive immunity via interfering phage resistance has generated strong selection pressure for CRISPR sequences.6 Although mecha- Submitted: 06/29/12 bacteria to become resistant.1 Antibiotic nisms of bacterial resistance to phages dif- resistant strains are becoming increas- fer substantially from those of chemical Revised: 03/06/13 ingly more common and thus infections antibiotics, the overall result from the per- Accepted: 03/06/13 acquired especially within hospital set- spective of therapy is the same: bacterial Citation: Örmälä A.-M. , Jalasvuori M. ; Phage tings often are untreatable with conven- proliferation can no longer be controlled therapy: Should bacterial resistance to phages be tional therapies. Bacteria typically need by therapeutic agents. Noting the rap- a concern, even in the long run? Bacteriophage only few genes to become resistant to com- idly emerging resistance, the advocates of 2013; 3:e24219; http://dx.doi.org/10.4161/bact.24219 monly used antibiotics. These resistance- phage therapy usually bring up the benefits genes often move horizontally between of phage cocktails, in which many differ- *Correspondence to: Matti Jalasvuori; Email: [email protected] bacteria within selfish genetic elements, ent types of phages infect the same species www.landesbioscience.com Bacteriophage e24219-1 or strains, and thus make the emergence of molecular evolution of the coevolving trait of the bacterium. Similarly, viruses of a resistant bacterial cell substantially species even in laboratory conditions.11,18,19 were less infective to current bacteria than less likely.7,8 Yet, there remains a lingering Although not all viruses seem to be able future or past hosts, suggesting that virus concern that extensive use of phages will to adapt to infect resistant hosts16,20 this predation was not able to induce changes select in one way or another for bacterial scenario nonetheless gives the impres- that would allow only the coevolved strains that are resistant to phage-cock- sion that at least some phages can remain viruses to remain infective to their hosts. tails. As Krylov et al. has recently argued, infective only by constantly evolving in This, in turn, suggests that viral selec- it might be more reasonable to use only a terms of their host range.16 If pre-defined tion is not forcing genetic divergence that single lytic phage, one very specific to the phage cocktails were used globally and would also prevent the replication of other pathogen causing an infection, in order to continuously in a manner similar to that viruses in any particular host. Phage resis- prevent the emergence of bacteria that are of antibiotics, resulting in bacterial evolu- tance in many cases can be costly for the resistant to the valuable broad-host range tion but not antimicrobial evolution, then bacterium22,23 and it thus appears to be cocktails.9 Perhaps equivalently to chemi- it is possible that phages could with time disadvantageous for the bacteria to remain cal antibiotics, after decades of use of become ‘over-evolved’ by their hosts. resistant against bacteriophages that are phages numerous bacterial pathogens may Evolution of bacterial resistance to no longer present in their local environ- be resistant to various phage-cocktails, phages, may occur more slowly given ment.21 Moreover, phage-resistant bacteria resulting in major difficulties in treat- use of phage cocktails rather than indi- often lack important surface features that ing, e.g., hospital acquired (nosocomial) vidual phages, but nevertheless only delay are responsible for bacterial virulence.24,25 infections.10 the evolution of bacterial superbugs that Resistant bacteria, even if they are not Phages are often stated to drive bac- are highly resistant, generally, to phages. killed by phages, therefore can become terial evolution,11 with phages forcing Alternatively, to the extent that broad mostly harmless. However, sometimes adaptations within bacterial populations, bacterial resistance to phages can evolve, there has been no observable costs associ- especially phage resistance. Generally cocktails by supplying multiple phages ated with resistance,20,26,27 demonstrating phages have a narrow host range and this simultaneously could select precisely for that resistant phenotypes, at least under is so to such an extent that the specificity such hypothetically broadly phage-resis- certain conditions, may persist. of a set of different phages has been rou- tant bacteria. In any case, this concern, Even if continuous use of phages tinely used to distinguish closely related no matter how hypothetical, we feel is forced a bacterial population to become bacterial strains from one another.12 a reasonable interpretation of the more permanently resistant to specific phage- Given the narrow host range, the con- general concern that phage therapy may cocktails, biogeography studies of phage tinuous selection for resistance and the suffer the same fate of antibiotic therapy: infection patterns suggest that new infec- often mentioned evolutionary arms race the evolution of superbugs for which it tious phages will nevertheless be avail- between bacteria and phages,13,14 might is difficult to identify new antibacterials able. It has been shown that regardless we suspect difficulties in future attempts against which resistance is not already of the absence of recent contact, phages to isolate phages that are effective against present. Alternatively, it could be argued have remained infective to bacterial host potentially multi-phage-resistant bacte- that perhaps only co-evolved phages that cells on the other side of the world. This rial strains and may this compromise the have responded evolutionarily in course of is despite the fact that most bacteria are long-term viability of phage therapy? It ongoing bacterial evolution to phage resis- thought to be under constant evolutionary is specifically this latter question that we tance may remain infective against these arms race with the phages in their pres- address in this commentary. hospital specific strains, that is, phages ent environment.11 For example, meta- Phage impact on bacterial evolution responding mutation by mutation rather analysis of host-phage interactions of 38 can occur in the course of antagonistic than simultaneously to large numbers separate studies performed by Flores et co-evolutionary cycles.15,16 In these cycles, of anti-phage resistance alleles that have al. concluded that phages can often infect viruses evolve to re-infect hosts that have accumulated in target bacteria. many different host strains from different already become resistant to earlier types Despite of the above mentioned obser- origins.28 They pointed out that infection of the same virus.13,14 It has been shown vations on bacterial evolution, several patterns were nested, indicating that the moreover that the frequency of bacterial studies are suggestive that evolution of ‘hard to infect’ hosts were infected by gen- reversion back to phage-susceptible phe- phage-resistant superbugs is not going to eralist phages and not by selective special- notypes is relatively low.17 Surface compo- occur. First, the cycles of viral adapta- ists. Correspondingly, Wolf et al. showed nents of bacterial cells in particular evolve tion to evolved hosts and vice versa have in 2003 that they were able to isolate bacte- rapidly, which likely is due to the fact been experimentally shown in laboratory riophages from distant geographical loca- that they are utilized by phages for host conditions and natural habitats. Gomez tions for a single freshwater bacterial host attachment (though also, for pathogens, and Buckling (2011) in particular demon- (Sphingomonas sp) isolated in 1985.29 One could be a consequence of immune sys- strated that bacteria in soil were more resis- of us recently isolated phages from Finland tem evasion). Co-evolutionary arms races tant to their contemporary phages than to for bacterial strains that were originally between bacteria and their lytic viruses past or future phages.21 Host resistance to collected from Britain in the sixties. These also have been shown to accelerate the rate phage infection was thus only a temporary and other examples have demonstrated e24219-2 Bacteriophage Volume 3 issue 1 that a single bacteriophage can remain References 17. Capparelli R, Nocerino N, Iannaccone M, Ercolini D, Parlato M, Chiara M, et al. Bacteriophage therapy infective to bacteria around the world 1. Andersson DI, Hughes D. Evolution of antibiotic of Salmonella enterica: a fresh appraisal of bacte- resistance at non-lethal drug concentrations. Drug riophage therapy. J Infect Dis 2010; 201:52-61; and that a single bacterium can serve as a Resist Updat 2012; 15:162-72; PMID:22516308; PMID:19929381; http://dx.doi.org/10.1086/648478 host for bacteriophages derived from mul- http://dx.doi.org/10.1016/j.drup.2012.03.005 18. Kashiwagi A, Yomo T. Ongoing phenotypic and tiple distant geographical locations. While 2. Bennett PM. Plasmid encoded antibiotic resistance: genomic changes in experimental coevolution of acquisition and transfer of antibiotic resistance genes RNA bacteriophage Qβ and Escherichia coli. PLoS continuous artificial selection in hospital in bacteria. Br J Pharmacol 2008; 153 Suppl 1:S347- Genet 2011; 7:e1002188; PMID:21829387; http:// settings against an over-presented set of 57; PMID:18193080 dx.doi.org/10.1371/journal.pgen.1002188 phages may still have unpredictable out- 3. Jalasvuori M, Friman VP, Nieminen A, Bamford 19. Paterson S, Vogwill T, Buckling A, Benmayor JK, Buckling A. Bacteriophage selection against a R, Spiers AJ, Thomson NR, et al. Antagonistic comes, evolution in natural communities plasmid-encoded sex apparatus leads to the loss of coevolution accelerates molecular evolution. Nature do not seem to be driving bacteria locally antibiotic-resistance plasmids. Biol Lett 2011; 7:902- 2010; 464:275-8; PMID:20182425; http://dx.doi. 5; PMID:21632619; http://dx.doi.org/10.1098/ org/10.1038/nature08798 to diversify in such manner that would rsbl.2011.0384 20. Lenski RE, Levin BR. Constraints on the generally render all other phages without 4. Dahlberg C, Chao L. Amelioration of the cost of Coevolution of Bacteria and Virulent Phage: A recent co-evolutionary history ineffective. conjugative plasmid carriage in Eschericha coli K12. Model, Some Experiments, and Predictions for Genetics 2003; 165:1641-9; PMID:14704155. Natural Communities. Am Nat 1985; 125:585-602; The same is true for archaeal extremo- 5. Lenski RE. Two-step resistance by Escherichia coli http://dx.doi.org/10.1086/284364 philes dwelling in relatively isolated habi- B to bacteriophage T2. Genetics 1984; 107:1-7; 21. Gómez P, Buckling A. Bacteria-phage antagonis- tats and their viruses, indicating that the PMID:6373494 tic coevolution in soil. Science 2011; 332:106-9; 6. Hyman P, Abedon ST. Bacteriophage host range PMID:21454789; http://dx.doi.org/10.1126/sci- ‘global infectivity’ is not an evolutionary and bacterial resistance. Adv Appl Microbiol ence.1198767 trend that is restricted to bacteria alone.30 2010; 70:217-48; PMID:20359459; http://dx.doi. 22. Meyer JR, Agrawal AA, Quick RT, Dobias DT, org/10.1016/S0065-2164(10)70007-1 Schneider D, Lenski RE. Parallel changes in host Altogether this implies that despite 7. Gu J, Liu X, Li Y, Han W, Lei L, Yang Y, et al. A resistance to viral infection during 45,000 genera- of extensive use of therapeutic phages, method for generation phage cocktail with great tions of relaxed selection. Evolution 2010; 64:3024- a situation where new infectious phages therapeutic potential. PLoS One 2012; 7:e31698; 34; PMID:20550574 PMID:22396736; http://dx.doi.org/10.1371/jour- 23. Inal JM. Phage therapy: a reappraisal of bacterio- for pathogenic bacteria can no longer be nal.pone.0031698 phages as antibiotics. Arch Immunol Ther Exp found is unlikely. In a historical perspec- 8. Chan BK, Abedon ST. Phage therapy pharmacology (Warsz) 2003; 51:237-44; PMID:12956433 tive, phage therapy has already failed once phage cocktails. Adv Appl Microbiol 2012; 78:1-23; 24. Capparelli R, Nocerino N, Lanzetta R, Silipo A, PMID:22305091; http://dx.doi.org/10.1016/B978- Amoresano A, Giangrande C, et al. Bacteriophage- (as reviewed by Sulakvelidze et al.31). We 0-12-394805-2.00001-4 resistant Staphylococcus aureus mutant confers broad need to be careful with our second try, but 9. Krylov V, Shaburova O, Krylov S, Pleteneva E. A immunity against staphylococcal infection in mice. genetic approach to the development of new therapeu- PLoS One 2010; 5:e11720; PMID:20661301; http:// unnecessary pessimism is unwarranted. tic phages to fight pseudomonas aeruginosa in wound dx.doi.org/10.1371/journal.pone.0011720 Thus, whereas new ‘conventional’ anti- infections. Viruses 2013; 5:15-53; PMID:23420928; 25. Filippov AA, Sergueev KV, He Y, Huang XZ, Gnade biotics have become increasingly more http://dx.doi.org/10.3390/v5010015 BT, Mueller AJ, et al. Bacteriophage-resistant mutants 10. Bush K, Courvalin P, Dantas G, Davies J, Eisenstein in Yersinia pestis: identification of phage receptors difficult to develop or discover, current B, Huovinen P, et al. Tackling antibiotic resistance. and attenuation for mice. PLoS One 2011; 6:e25486; evolutionary as well as ecological under- Nat Rev Microbiol 2011; 9:894-6; PMID:22048738; PMID:21980477; http://dx.doi.org/10.1371/journal. http://dx.doi.org/10.1038/nrmicro2693 pone.0025486 standing of phage biology instead is sug- 11. Clokie MR, Millard AD, Letarov AV, Heaphy 26. Wei Y, Ocampo P, Levin BR. An experimental gestive of optimism in this regard, with S. Phages in nature. Bacteriophage 2011; 1:31- study of the population and evolutionary dynamics reasonable expectation that new, effective 45; PMID:21687533; http://dx.doi.org/10.4161/ of Vibrio cholerae O1 and the bacteriophage JSF4. bact.1.1.14942 Proc Biol Sci 2010; 277:3247-54; PMID:20538647; environmental phages will be continually 12. Sechter I, Mestre F, Hansen DS. Twenty-three http://dx.doi.org/10.1098/rspb.2010.0651 available indefinitely against most bacte- years of Klebsiella phage typing: a review of 27. Lythgoe KA, Chao L. Mechanisms of coexistence of rial targets. phage typing of 12 clusters of nosocomial infec- a bacteria and a bacteriophage in a spatially homoge- tions, and a comparison of phage typing with K neous environment. Ecol Lett 2003; 6:326-34; http:// serotyping. Clin Microbiol Infect 2000; 6:233-8; dx.doi.org/10.1046/j.1461-0248.2003.00433.x Disclosure of Potential Conflicts of Interest PMID:11168118; http://dx.doi.org/10.1046/j.1469- 28. Flores CO, Meyer JR, Valverde S, Farr L, Weitz No potential conflicts of interest were 0691.2000.00070.x JS. Statistical structure of host-phage interactions. 13. Buckling A, Rainey PB. Antagonistic coevolution Proc Natl Acad Sci U S A 2011; 108:E288-97; disclosed. between a bacterium and a bacteriophage. Proc PMID:21709225; http://dx.doi.org/10.1073/ Biol Sci 2002; 269:931-6; PMID:12028776; http:// pnas.1101595108 Acknowledgments dx.doi.org/10.1098/rspb.2001.1945 29. Wolf A, Wiese J, Jost G, Witzel KP. Wide geo- 14. Marston MF, Pierciey FJ Jr., Shepard A, Gearin G, Qi graphic distribution of bacteriophages that lyse the We are grateful to one of the reviewers J, Yandava C, et al. Rapid diversification of coevolv- same indigenous freshwater isolate (Sphingomonas sp. for his or her valuable help in editing and ing marine Synechococcus and a virus. Proc Natl strain B18). Appl Environ Microbiol 2003; 69:2395- Acad Sci U S A 2012; 109:4544-9; PMID:22388749; 8; PMID:12676728; http://dx.doi.org/10.1128/ improving this commentary. http://dx.doi.org/10.1073/pnas.1120310109 AEM.69.4.2395-2398.2003 15. Pal C, Maciá MD, Oliver A, Schachar I, Buckling 30. Atanasova NS, Roine E, Oren A, Bamford DH, A. Coevolution with viruses drives the evolution of Oksanen HM. Global network of specific virus-host bacterial mutation rates. Nature 2007; 450:1079- interactions in hypersaline environments. Environ 81; PMID:18059461; http://dx.doi.org/10.1038/ Microbiol 2012; 14:426-40; PMID:22003883; nature06350 http://dx.doi.org/10.1111/j.1462-2920.2011.02603.x 16. Dennehy JJ. What Can Phages Tell Us about 31. Sulakvelidze A, Alavidze Z, Morris JG Jr. Host-Pathogen Coevolution? Int J Evol Biol 2012; Bacteriophage therapy. Antimicrob Agents 2012:396165; PMID:23213618; http://dx.doi. Chemother 2001; 45:649-59; PMID:11181338; org/10.1155/2012/396165 http://dx.doi.org/10.1128/AAC.45.3.649-659.2001 www.landesbioscience.com Bacteriophage e24219-3