Phage therapy: eco-physiological pharmacology

Abstract

Bacterial virus use as antibacterial agents, in the guise of what is commonly known as phage therapy, is an inherently physiological, ecological, and also pharmacological process. Physiologically we can consider metabolic properties of phage infections of bacteria and variation in those properties as a function of preexisting bacterial states. In addition, there are patient responses to pathogenesis, patient responses to phage infections of pathogens, and also patient responses to phage virions alone. Ecologically, we can consider phage propagation, densities, distribution (within bodies), impact on body-associated microbiota (as ecological communities), and modification of the functioning of body "ecosystems" more generally. These ecological and physiological components in many ways represent different perspectives on otherwise equivalent phenomena. Comparable to drugs, one also can view phages during phage therapy in pharmacological terms. The relatively unique status of phages within the context of phage therapy as essentially replicating antimicrobials can therefore result in a confluence of perspectives, many of which can be useful towards gaining a better mechanistic appreciation of phage therapy, as I consider here. Pharmacology more generally may be viewed as a discipline that lies at an interface between organism-associated phenomena, as considered by physiology, and environmental interactions as considered by ecology.

Figure 1

Connections between physiology, ecology, and pharmacology. At best the distinction between an organism's physiology and its overall ecology can be ambiguous, though with body-pharmaceutical interactions representing one aspect of their interface. Such chemicals specifically can be viewed as abiotic components of an organism's environment, ones that have made contact with an organism's tissue, as can also toxins. Shown are body interactions not just with self (as body) but with nonself as well (as environment), with emphasis in physiology on the impact of these interactions on the functioning of self. Ecology also is the study of interactions between self and nonself, but with the consequences of such interactions considered with emphasis on both self, as body, and nonself, as environment. Pharmacology too is the study of interactions between self and nonself (particularly “environment-to-self”), though as with physiology there is an emphasis on impact on self (again, as body).

Figure 2

The realm of pharmacology includes an organism's gene products and their functioning, the overall physiology of an organism, the interactions that occur between a host organism and their associated microbiota, and even aspects of an organism's ecology. The realm of pharmacology. Pharmaceuticals interact with individual body molecules, including gene products as well as products of enzyme-mediated catalysis. The goal with pharmacology, in turn, is a modification of the physiology and particularly the pathophysiology of treated organisms. Organisms themselves generally consist of more than just the products of their own genomes but also the products of their associated microbial symbionts [221]. An important component of pharmacology therefore is the interaction of pharmaceuticals with this microflora. Not shown is the impact of drugs that serve either as mutagens or as nucleic acid damaging agents, which can affect genotype as well.

Figure 3

Life cycle of an obligately lytic bacteriophage. As this is a cycle, the “beginning” is arbitrary. A successful infection nonetheless progresses through adsorption, infection, release (here via lysis), and a period of extracellular “search” for new bacteria to infect. Deviations from this life cycle can include inactivation during the extracellular stage, a failure to successfully adsorb, and various forms of phage inactivation that can occur during infection, including as explicitly mediated by bacterial cells [17, 164]. Though lytic phages are released via lysis, other phages exist, most notably filamentous phages such as phage M13, that instead are released from infected bacteria chronically. Generally such nonlytic phages are not used for phage therapy. Another variation on the phage life cycle is lysogenic cycles, which are nonvirion productive extensions of the infection stage. Only temperate, particularly not obligately lytic phages display lysogenic cycles, and temperate phages typically also are not among the first choice for phage therapy purposes [222]. Shown too, in the middle, is reference to pharmacological aspects of phage infections. Particularly these are distribution throughout body tissues that can occur while in the free phage state (a.k.a., phage penetration to target bacteria) along with amplification of phage numbers in situ as can occur as a consequence of phage infection of bacteria, which is a component of what pharmacokinetically is known as metabolism.

Figure 4

Basics of phage therapy pharmacology. Absorption and distribution can have the effect of increasing antibacterial concentrations within the vicinity of target bacteria, though also they have a diluting effect on dosages. Phage infection too can increase phage numbers within the vicinity of target bacteria, which I have indicated as being an aspect of metabolism and which more generally is a description of the chemical modification of a drug. Together these pharmacokinetic mechanisms contribute to some peak phage density that may or may not be sufficient to substantially decrease densities of target bacteria [93, 117]. Particularly, peak densities must exceed some minimum effective concentration to effect net reductions in bacterial densities and these densities can be achieved through a combination of supplying sufficient phage numbers per individual dose, supplying multiple doses, and/or allowing for phages to replicate in situ. Ideally phage densities will not be so high that toxicity results. Exactly what phage densities are necessary to achieve toxicities is not well appreciated, except that impurities in phage formulations can contribute to such toxicities (as too can potentially the humoral immune system given nonnaive patients). As a consequence of this uncertainty, what constitutes a preferred upper limit of phage densities is not known in the same way that minimum toxic concentrations can be appreciated for specific small-molecule drugs, except that this upper limit may be high relative to minimum effective phage densities. Lastly, various mechanisms exist whereby phage densities may decrease over time, which include what pharmacologically are described as metabolism and excretion, though as noted dosage dilution plays a role as well. A modified version of this figure is found in Abedon [93] as well.

Figure 5

Molecular aspects of biology (left) give rise to organismal characteristics (middle) which in turn can give rise to ecological consequences (right). Ecological consequences include impact on environments as well as environment impact back onto organisms (not shown). Within phage therapy as a pharmacological process, these ecological consequences—with a patient's body serving as environment—can be viewed as being equivalent to considerations of pharmacodynamics (drug impact on body) and pharmacokinetics (body impact on drug), respectively. Physiology in turn is a description of how an organism's molecular aspects as well interactions with environments combine to give rise to organism functioning. Here physiological aspects are indicated, in the middle of the figure, particularly in terms of phage organismal properties. Phage physiology, within a phage therapy context, thus can be viewed as a highly complex elaboration on how chemical form, that is, of phages, gives rise to ecological properties (in terms principally of bacterial eradication), just as a pharmaceutical's chemistry gives rise to its pharmacological characteristics. Despite the complexity of a phage's chemical form as well as the process of translation of that form into so-called pharmacologically emergent properties, such properties as side effects can be less likely than the case with less-complex, small-molecule drugs. This often low phage propensity towards toxicity presumably is a consequence of phages consisting primarily of DNA (or RNA) and proteins that have been molded by evolution to be highly specific in their impact towards modification of bacterial metabolism and structure rather than that of eukaryotic organisms such as ourselves [66]. Note that this figure is modified from one found in Abedon [109].

Figure 6

The interface between different organisms, and their physiologies, basically is ecological. The quantity of these interactions as well as their impact on physiologies increases with the number of organism types involved. With phage therapy, this includes three distinct species. (1) the patient, host, subject, or body that is experiencing a bacterial infection; (2) the infecting bacterial pathogen; and (3) the bacterial virus, a.k.a., phage or bacteriophage that is being used to treat the bacterial infection. Phage therapy pharmacology lies at the interface of these three components and thus inherently straddles both ecological and physiological considerations (with that confluence indicated by the star but not limited to the star). Since physiologies change with varying infection conditions as well as treatment approaches, including in terms of physiological adaptation to these changing conditions, phage therapy pharmacology can be viewed as being inherently eco-physiological.

Figure 7

Phenomena impacting virions include their movement, their survival, and their adsorption. Shown also are various aspects of environments as relating to virion properties and functions (green text with subcategories in blue). The terms “free virion” and “free phage” are being used interchangeably in the figure.

Figure 8

Impacts of pharmacokinetic phenomena during phage therapy. These processes occur in approximate temporal order as indicated by blue or inner arrows. Metabolism, via amplification of virion numbers, is also shown as contributing to increases in phage numbers that then may be distributed to other locations or compartments within the body (green or outer arrow). Amplification can also give rise to increased phage titers in blood, though this is not indicated. Also not indicated is activation of phage bactericidal activity, which can be viewed as an aspect of metabolism and one that precedes amplification (though which leads to amplification only given successful lytic, productive infection). Note that absorption and distribution too have the effect of increasing phage concentrations in specific compartments, the blood and nonblood tissues, respectively, as these are a means by which access to these compartments is achieved. These increases in concentration, however, are not relative to the initial phage dose but instead are relative to concentrations within compartments as observed prior to dosing.

Figure 9

Summary of secondary pharmacodynamic concerns that are peculiar to phage- or protein-based antibacterial therapy. Issues are listed in brown at the corners of the figure, with more specific considerations listed in blue. Means by which these issues can be mitigated are presented as green blocks of text as found between concerns and the indication of  “Side effects” shown in the center. Thus, for example, the issue of bacterial lysis products in phage formulations may be mitigated through a combination of informed bacterial choice in some combination with sufficient post-lysis purification of resulting phages.

Figure 10

Time course of the pharmacokinetics of active phage therapy. Numerous pharmacokinetic processes–either through dilution, inactivation, or inefficiencies in penetration–have the effect of reducing phage densities in situ such as below minimum effective phage densities (MEPDs). These losses may be minimized by reducing the length of the chain of processes separating phage application from phage contact with target bacteria or instead can be addressed by supplying more phages, such as to counteract inevitable losses. Metabolism as a pharmacokinetic process, in the form of phage replication and therefore in situ amplification in density (auto dosing), can reverse these losses and allow an achievement of MEPDs, at least local to target bacteria. The process illustrated in the figure is an elaboration on the concept that otherwise has been described as active treatment, that is, supplying insufficient phage numbers through traditional dosing to achieve MEPDs, and thus relying on active phage replication instead to achieve these densities.

Figure 11

Comparing proximate outcomes during different categories of phage therapy progress. In passive treatment, only cell killing must occur as a proximate outcome, though cell lysis as well as in situ amplification of phage numbers may occur as well. By definition, though, they do not have to happen for passive treatment to successfully clear a bacterial infection (hence use of dashed, grayed arrows towards the bottom of this column). Phage active penetration into bacterial biofilms appears to be dependent on some form of phage enzymatic activity other than those required to physiologically or genetically kill bacteria. Here this is indicated as bactericidal infection occurring in combination with bacterial lysis, with the latter contributing to further phage penetration into biofilms and/or improved phage-infection physiology. Such phage activity potentially may also improve antibiotic [223] or disinfectant [224] penetration into biofilms or at least their effectiveness against biofilms. Phage in situ amplification, though potentially helpful towards further phage penetration, nonetheless in this case is not necessarily absolutely required (dashed, gray arrow). Payne and Jansen [52] describe an intermediate state between active and passive treatment that they term “mixed passive/active” (here, for clarity, “passive-active”). This treatment approach involves a combination of dosing with large numbers of phages and subsequent phage population growth. It is the opinion of this author that this latter approach, possibly in combination with multiple dosing, likely either should or does represent a default approach to effecting phage therapy treatments. This represents supplying relatively large phage numbers to bacterial infections—in single or multiple doses—though nonetheless supplying phage numbers that are less than completely overwhelming (i.e., less than completely inundative) with the assumption that phage in situ population growth will enhance those numbers local to either planktonic bacteria or instead bacterial biofilms or microcolonies. See Abedon [225, 226] for consideration of the latter. The reduced but not eliminated requirement for lysis and amplification in the case of “passive-active” is indicated using solid but gray arrows rather than dashed gray arrows. Lastly, active treatment by definition is dependent on both lysis and in situ phage amplification (black, solid arrows).

Figure 12

First-approximation comparison of bacteriophages and antibiotics in terms of their activity spectra in combination with various concentration and dosing issues. In short, phages tend to display narrower activity spectra but that activity can be less dependent on concentration issues, particularly given passive treatment, than the activity spectra displayed by small-molecule antibacterial agents.

Figure 13

Ecological as well as physiological perspective on bacterial resistance to phages as seen following phage adsorption. Specifically, there exist gradations in bacterial interference with phage productivity ranging from (i) no interference (“normal lytic infection”) to (ii) partial blocks on phage productivity and/or extension of the phage infection cycle that can slow down phage population growth (“reduced infection vigor” [17] as seen with “reduced burst size” or “extended latent period”) to (iii) bacterial self-sacrifice for the sake of phage elimination (“abortive infection”; see also [225]) to (iv) bacteria simply inactivating infecting phages but without loss of bacterial viability (“restriction, exclusion, or immunity”; hosts in any case are shown as green ovals). These various mechanisms are reviewed by Hyman and Abedon [17] and also Labrie et al. [164]. Not shown, bacteria can also block phage infection by resisting phage attachment following phage encounter, though generally this does not result in phage metabolism in either a physiological or pharmacological sense. For comparison of mechanisms of bacterial resistance to phages to the immunity displayed especially by animals against pathogens, see Abedon [227]. Note also that analogies exist between mechanisms of bacterial resistance to phages and mechanisms of bacterial resistance to antibiotics. These include as mediated by compound destruction or avoidance of interaction through changes in target structures, though notably absent is a phage-resistance equivalent to “efflux off the antibiotic from the cell” (p. 1451) [160].

Figure 14

Context of pharmacology as an ecological as well as a physiological phenomenon, with physiology in turn a manifestation of underlying genetics. Pharmacology literally is the exposure of an organism's physiology to an environmental component, that is, a drug, and the study of organism-with-environment interactions literally defines ecology. Similar though arguably less complex “bubbles” can be placed around both bacterial pathogens and their viruses (e.g., Figure 6). Pharmacology thus is inherently eco-physiological while both the pharmaceutical treatment of distinct living entities, such as pathogens, and the use of drugs that themselves are living, particularly bacteriophages, introduces additional aspects of interface between ecology, physiology, and pharmacology.