Proof-of-Concept Standardized Approach Using a Single-Disk Method Analogous to Antibiotic Disk Diffusion Assays for Routine Phage Susceptibility Testing in Diagnostic Laboratories

ABSTRACT

Application of bacteriophages is increasingly being implemented in clinical therapies. Prior susceptibility testing should be regarded as mandatory, but standards are lacking. The objective of this research was to develop a highly standardized methodology to facilitate phage susceptibility testing (PST) in clinical microbiology routine laboratories. Therefore, EUCAST methods established for single disk-based antibiotic susceptibility testing (AST) were adapted. In a first step, basic parameters were evaluated using well-studied Escherichia phage T4–Escherichia coli combinations. In addition, test results were compared to those from conventional spot test and efficiency of plating (EOP) approaches. In a second step, the applicability of the methodology and the most promising test parameters were demonstrated for five other frequently isolated clinical bacterial species and their corresponding phages. At present, the method predominantly leads to qualitative rather than quantitative results. This disk-based approach provides a standardized, easy-to-handle, reproducible and reliable PST protocol by relying on well-established routine procedures in diagnostic laboratories.

IMPORTANCE Application of bacteriophages in clinical therapies is attractive due to increasing rates of isolation of multidrug-resistant bacteria worldwide. As the phage effect is highly specific, prior susceptibility testing of target bacteria is mandatory. Of note, established standards are lacking. In this research, we adapted the single-disk method for antibiotic susceptibility testing to phage susceptibility testing (PST) in order to provide a standardized, easy-to-handle, reproducible, and reliable PST protocol for application in diagnostic routine laboratories.

INTRODUCTION

Since the first description of bacterium-specific viruses (bacteriophages) by Félix d’Hérelle in 1917 (1), their antibacterial therapeutic potential was quickly recognized (2–4). Human therapeutical applications followed for skin infections (5, 6), purulent wound infections, postoperative infectious complications (7), and biofilm-related infections (8–10).

With antibiotics gaining a role in infection treatment in the 1940s, especially in the Western world, the position of phage therapy subsequently declined. Parallel to the increasing availability of safe and effective antimicrobial drugs, the urge to develop phage therapy decreased (11).

Recently, both augmented knowledge of chronic infections and their association with biofilms containing bacteria with substantially reduced susceptibility to antibiotics and the increasing emergence of multidrug-resistant bacteria worldwide led to new therapeutic challenges. Combined with a dwindling development and production of new antibacterial drugs (12, 13), this encourages scientists to reconsider the therapeutic use of phages (14–16).

The classification of phage preparations as medical devices within the EU (17) entails many requirements, such as manufacturing according to good manufacturing practice (GMP) criteria, preclinical and clinical studies and centralized marketing authorization (18). However, these requirements hamper sustainable, tailored phage therapy approaches. Phage therapy medicinal products (PTMPs) produced on an industrial scale are unlikely to be able to respond in a timely manner to challenges such as the inevitable emergence of phage-resistant bacterial isolates (18–20).

Due to the lack of commercially available PTMPs, the present regulations create a particular challenge for clinicians for the use of phage therapy in critically ill patients. Especially in biofilm-associated chronic infection, antibiotics have limited efficacies. Thus, combinations with other antibacterial compounds or principles might circumvent these therapeutical challenges.

To date, individual therapeutic phage applications have been performed in accordance with Article 37 (Unproven Interventions in Clinical Practice) of the Declaration of Helsinki (21, 22). Nevertheless, randomized clinical trials are in preparation or already in progress to evaluate the efficacy and safety of phage therapy employing customized phage preparations (18, 23).

Prior to application, susceptibility testing is mandatory to demonstrate that the patient’s bacterial isolate is within the host spectrum of the administered phages. However, no method delivering results that are both normalized and comparable between laboratories has been established so far. To enable the attending doctor to refer to a proof of effectiveness of the administered phages, phage susceptibility testing (PST) is often based on the “spot test” method (24, 25). This assay is frequently adapted, with different parameters for agar medium, spot size, incubation time, and concentration of bacterial and phage inocula used (24–27), making the comparison of results extremely difficult. In contrast, for antibiotics, in vitro antimicrobial susceptibility testing (AST) by a single-disk test is completely standardized, for example, by the European Committee on Antimicrobial Susceptibility Testing (EUCAST) or by the Clinical and Laboratory Standards Institute (CLSI).

The objective of this study was to develop and evaluate a standard operating procedure for PST, enabling routine diagnostic laboratories to perform a rapid phage testing of individual phage suspensions or of PTMPs (e.g., commercially available phage cocktails) with the same levels of both straightforward processing and quality as current antimicrobial susceptibility testing. Therefore, we focused on a highly standardized protocol, also enabling interlaboratory comparison, and with the potential for laboratory accreditation. For this purpose, the AST method for single disk tests as formulated by EUCAST was adapted to PST. Basic parameters were investigated for Escherichia coli–Escherichia phage T4 combinations, and the most promising parameters were then evaluated with other bacterium-phage combinations.

RESULTS

Disk-based test with 20 μL phage suspension shows round lysis zones.

Comparing the modified spot test with the disk-based test, a irregularly shaped lysis zone is noticeable in the spot test when drops are applied directly to the agar plate. In contrast, disk placement followed by phage application to disks results in a round lysis zone. Lysis is macroscopically visible in the modified spot test irrespective of the volume used. When disks are used, a lysis zone can be detected after application of at least 10 μL of phage suspension. The diameter of the lysis zone increases in a volume-dependent manner with good readable results when 20 μL suspension is used.

Of note, for both settings, the drying time is considerably prolonged when 50 μL of phage suspension is applied. Macroscopic aspects of the modified spot test and disk-based test are shown in Fig. 1.

FIG 1.

Comparison of the modified spot test and disk-based test using different phage volumes. The E. coli inoculum was 0.2 McFarland with spatula agar inoculation; the phage concentration was 10⁹ PFU/mL. (A) Modified spot test. (B) Disk-based test. Representative images from 3 independent experiments are shown.

Turbidity standard of 0.2 McFarland and approximately 10⁶ PFU/mL are recommended for bacterial and phage suspension testing, respectively.

In the analysis for the optimal bacterial inoculum to be used, E. coli suspensions with McFarland turbidity standard densities of 0.1, 0.2, and 0.5 were tested. Twenty-microliter portions of phage suspension with concentrations of 10⁴ PFU/mL, 10⁶ PFU/mL, 10⁸ PFU/mL, and 10¹⁰ PFU/mL were applied. No lysis zone was detectable when a phage concentration of 10⁴ PFU/mL was used, irrespective of the applied bacterial amount. Using different turbidity standards of bacterial inocula led to comparable macroscopic characteristics, with occasional microcolonies being visible within lysis zones (Fig. 2).

FIG 2.

Comparison of different turbidity standards of applied bacterium inoculum and phage concentration. A disk-based test with spatula inoculation was used. The phage volume applied was 20 μL, with phage concentrations of 10¹⁰, 10⁸, 10⁶, and 10⁴ PFU/mL (clockwise from top right). Turbidity standards of 0.1 (A), 0.2 (B), and 0.5 (C) McFarland were tested. (D) Enlargement of a segment with clearly readable lysis zone (i.e. parameters: turbidity standard, 0.2 McFarland; phage concentration, ≥10⁸ PFU/mL). Representative images from 3 independent experiments are shown.

Plate inoculation by spatula results in more evenly dense bacterial lawns.

In a comparison of two common plate inoculation methods, a homogeneous bacterial lawn can be seen when a Drigalski spatula is used (Fig. 3A and C). In contrast, inoculation by swab results in a finely spotted bacterial lawn (Fig. 3B and D), potentially making it difficult to discern lysis zones.

FIG 3.

Comparison of agar inoculation by spatula and swab. The turbidity standard was 0.2 McFarland. The phage volumes applied were 5, 10, 20, and 50 μL (clockwise from top right), and the phage concentration was 10⁹ PFU/mL. (A) Disk-based test with spatula inoculation. (B) Disk-based test with swab inoculation. (C) Modified spot test with spatula inoculation. (D) Modified spot test with swab inoculation. Representative images from 3 independent experiments are shown.

Comparison of PST to liquid-based phage assay.

Liquid phage testing assay was performed by incubating Escherichia phage T4 and the corresponding host strain or the host strain without the phage (control) (Fig. 4). Starting from the time point of inoculation, the data show the course of optical density over a time course of 8 h for a 10⁸ CFU/mL E. coli culture exposed to Escherichia phage T4. In the combination of bacteria and phages, according to density curve, growth came to a halt after the first burst of phage replication. In contrast, the optical density of the culture medium without phage inoculation increased until stationary phase was reached (Fig. 4).

FIG 4.

Growth kinetics of E. coli–Escherichia phage T4. The E. coli host strain (10⁸ CFU/mL) was incubated with Escherichia phage T4 (10⁸ PFU/mL) (triangles) and without phage (squares [control]). Error bars represent standard deviations (n = 3 experiments).

Effects of drying and storage on phage-impregnated disks.

When phage-preimpregnated disks were stored for various times, lysis zones were not discernible after a storage time of 24 h (Fig. 5), even when 20 μL SM buffer was added for improved phage elution from the disks.

FIG 5.

Different storage conditions for phage-impregnated disks. (A) PST of the E. coli host strain and Escherichia phage T4. (B) PST of dried phage-impregnated disks after 24 h storage without (top) and with (bottom) addition of 20 μL SM buffer.

Combined testing of lytic phages and antibiotics.

Potential additive or synergistic effects of phages and antibiotics were tested by adding 20 μL of phage suspension to commercial antibiotic disks. For any combined testing of phages and antibiotics, no differences in inhibition zone diameters were seen (Table 1; Fig. 6).

TABLE 1.

Combined testing of lytic phage (Escherichia phage T4) and antibiotic activities using commercial antibiotic disks^a

Antibiotic	Inhibition zone diam (mm)b
	With phage	Without phage
Ampicillin (10 μg)	24.5 ± 0.5	24.7 ± 0.5
Ciprofloxacin (5 μg)	38.9 ± 0.7	39.3 ± 0.5
Gentamicin (10 μg)	24.3 ± 0.4	23.5 ± 0.5

^aDiscs were used with 20 μL phage suspension or SM buffer (no phage). Agar was inoculated with E. coli (100 μL, 0.2 McFarland).

^bEach value is the mean of at least three measurements, with the standard deviation.

FIG 6.

Combined compound testing of antibiotics and phages. Mueller-Hinton II agar was inoculated with 100 μL E. coli host strain suspension (McFarland 0.2), and 20 μL phage suspension (+ phage [top]) or SM buffer (− phage [bottom]) was added. Representative images from at least 3 independent experiments are shown.

Negative controls show no lysis zone.

Quality control for host specificity and/or potential growth inhibition was performed to exclude disk- and buffer-induced effects and to prove that the detectable lysis zones were specifically phage induced. Neither disk nor buffer effects were observed (Fig. 7).

FIG 7.

Quality controls. (A) Host specificity control. (B) Intrinsic growth inhibition control.

Lysis zone diameter is volume dependent.

Comparing diffusion of SM buffer into the agar, as visualized by China ink, with lysis zone diameters of phage suspensions resulted in similar or identical diameters. Of note, the application of 20 μL to the disks led to evenly round lysis zones, whereas addition of a volume of 50 μL resulted in distorted zones (Fig. 8).

FIG 8.

Visualization of agar diffusion. Different volumes (20 μL and 50 μL) of ink (left) and phage suspension (right) were applied. Mueller-Hinton II agar was inoculated with 100 μL E. coli host strain (McFarland 0.2) by spatula.

Suggested parameters for PST are transferable to other bacterium-phage combinations.

The final PST characteristics (Fig. 9) were evaluated for five frequently isolated bacteria with corresponding reference phages (Table 2) and compared to the spot test (Fig. 10). Phage-specific lysis zones varied in diameter but could clearly be read for all bacterium-phage combinations tested. The disk-based approach showed results similar to those of the spot test. PST reproducibility of lysis zone diameter in three test series for all bacterium-phage combinations is shown in Table 3 (for images comparing the PST and spot test, see Fig. S1 and S2 in the supplemental material).

FIG 9.

PST method and final PST characteristics.

TABLE 2.

Phages and bacteria used in this study

Name	Straina
Phages
Escherichia phage T4	DSM 4505
Staphylococcus phage MRLN	DSM 26857
Staphylococcus phage vB_SepM_Alex	DSM 108061
Pseudomonas phage vB_PaeM-PT-MAS01	DSM 109904
Proteus phage vB_PniP-Homer	DSM 107146
Enterococcus phage vB_EfaS_Strempel1	DSM 110103
Staphylococcus phage vB_SepP_Nepomuk	DSM 106137
Staphylococcus phage vB_SepP_Spree	DSM 108057
Staphylococcus phage vB_SepP_UKE3	DSM 108058
Bacteria
Escherichia coli	DSM 613/ATCC 11303
Staphylococcus aureus	DSM 104437
Staphylococcus epidermidis	DSM 3269
Pseudomonas aeruginosa	DSM 19880
Proteus mirabilis	DSM 30116
Enterococcus faecalis	DSM 32036
Serratia marcescens	ATCC 13880
Staphylococcus epidermidis	DSM 20044/ATCC 14990
Staphylococcus epidermidis	DSM 18857

^aDSM, German Collection of Microorganisms; ATCC, American Type Culture Collection.

FIG 10.

Comparison of PST and Spot testing of various bacterium-phage combinations. (Column 1) Disk-based phage susceptibility testing. (Column 2) Spot testing. (a) E. coli plus Escherichia phage T4. (b) S. epidermidis plus Staphylococcus phage vB_SepM_Alex. (c) E. faecalis plus Enterococcus phage vB_EfaS_Strempel1. (d) P. aeruginosa plus Pseudomonas phage vB_PaeM-PT-MAS01. (e) S. aureus plus Staphylococcus phage MRLN. (f) P. mirabilis plus Proteus phage vB_PniP-Homer.

TABLE 3.

PST reproducibility of lysis zone diameters with five phages^a

Phage	Lysis zone diam (mm)
	Test series 1	Test series 2	Test series 3	Mean ± standard deviation
Escherichia phage T4	9	9	10	9.3 ± 0.5
Staphylococcus phage MRLN	8	8	9	8.3 ± 0.5
Staphylococcus phage vB_SepM_Alex	10	10	10	10 ± 0
Pseudomonas phage vB_PaeM-PT-MAS01	10	10	9	9.7 ± 0.5
Proteus phage vB_PniP-Homer	27	24	23	24.7 ± 1.7
Enterococcus phage vB_EfaS_Strempel1	11	11	10	10.7 ± 0.5

^aPictures of compared PST and spot test (in triplicate) are shown in Fig. S1 and S2 in the supplemental material.

Of note, based on observations from all PST experiments, the regularity of the lysis zone shapes is optimum when the plates rest unmoved for 5 min after the phage suspensions have been applied, precluding uncontrolled spillover of the liquid to the immediate vicinity of the disks.

EOP-independent lysis zone diameters.

Assessing EOP is generally regarded as a suitable test for certifying the applicability of a specific phage for therapeutic purposes. Therefore, the comparability of PST and EOP test results was examined.

PST and EOP were simultaneously performed with four Staphylococcus phages and their reference Staphylococcus epidermidis strains. Although the bacterium-phage combinations displayed individual EOP values, lysis zone diameters differed much less if at all. However, for every bacterium-phage combination with EOP values above 0.001, the generally accepted baseline level for efficiency, a lysis zone could be read by PST. Results are summarized in Fig. 11.

FIG 11.

(Top) Table comparing PST and EOP for different Staphylococcus phages on their bacterial hosts. Relative efficiency was classified as high (0.1 ≤ EOP < 1), low (0.001 ≤ EOP < 0.1), or “no efficiency” (EOP < 0.001). EOP measurement was performed in triplicate. The plating on the reference host strain of isolation (EOP = 1) is in gray. +, lysis zone; −, no lysis zone. (Bottom) PST of Staphylococcus phages with different EOP. Phages applied include vB_SepP_Nepomuk, vB_SepP_UKE3, vB_SepP_Spree, and vB_SepM_Alex (clockwise from top right). Bacterial strains used were (A) DSM 20044, (B) DSM 18857, and (C) DSM 3269.

DISCUSSION

Due to the increasing emergence of biofilm-associated infections as well as of multidrug-resistant pathogens worldwide, there is a need for a therapeutic alternative to antibiotics. Thus, phage therapy could gain a new standing in infectious disease management (15, 16).

Phages are highly specific and usually infect only a subset of strains within a single bacterial species (28). While narrow host spectra of phages are already well studied, the scientific focus on naturally occurring as well as genetically engineered phages with a broad phenotypic host bacterial range is increasing (29). Such phages could be used for therapeutic applications in polymicrobial infections.

Therapeutic success of a calculated antibiotic therapy can be largely predicted based on mechanisms of action and extensive surveillance data. In contrast, the specificity for a bacterial target strain cannot currently be predicted by knowledge of the phage alone. For example, a comparative study investigating the host ranges of staphylococcal phages of the genus Kayvirus (also including the Staphylococcus phage MRLN, used in our study) demonstrated significant differences on variant methicillin-resistant Staphylococcus aureus (MRSA) strains (30). Also, polyvalent phages with the ability to infect related but distinct hosts exist, but they differ in their host ranges due to their specificity (31). Therefore, pretherapeutic susceptibility testing prior to phage application is crucial.

In routine microbiological diagnostics, antibiotic susceptibility testing according to internationally recognized standards, such as those published by CLSI and EUCAST, has become firmly established. However, international standards for phage testing are still lacking. Therefore, establishing a phage testing protocol based on AST methodology according to standard operating procedures of the above-mentioned institutions could facilitate and standardize this process.

In contrast to the rules for interpretation of antibiotic susceptibility testing, in the present experimental setting, there is no correlation between inhibition zone diameter and phage quantity and/or productivity in the PST. Lysis zone diameters predominantly depend on the volume of phage suspension applied to the disk rather than the amount of phages within this application. This means that once the optimum parameters for testing different phage preparations have been defined, the test is read as a qualitative assay, thereby facilitating routine diagnostic laboratory evaluation.

Using standard Mueller-Hinton (MH) II agar plates with a bacterial inoculation concentration of McFarland 0.2 in combination with disks inoculated with 20 μL of ≥10⁶ PFU/mL phage suspension resulted in clearly readable lysis zones for all bacterium-phage combinations tested. This phage concentration is in line with several case reports that describe the in vivo therapeutic application of phage suspensions with concentrations ranging between 10⁷ and 10⁹ PFU/mL (32–34).

Determining EOP is generally regarded as a suitable test format for therapeutic phage applications (35–37). Phages classified as efficient by EOP testing showed visible lysis zones in the PST assay, and phages classified as inefficient according to EOP results lacked inhibition zones in the PST. However, different EOP values did not correlate with lysis zone diameters in PST testing. Thus, the PST results should be read and interpreted in a qualitative rather than quantitative manner.

In future, therapeutic outcome should be correlated with the presence of detectable lysis zones by clinical therapy studies. However, prior to the initialization of such studies, the choice of phage testing system for that purpose should be agreed upon. Thus, establishing a PST method focusing on routine lab practicability and readability of lysis zones is a first step in this direction.

One limitation of the PST parameters described here is that there is a formal possibility that this may not be applicable to all phages and their host strains. Based on 70,000 isolates annually examined from patient materials at a tertiary-care hospital, we chose the 5 most frequent bacterial pathogens for PST evaluation in addition to E. coli. It may be necessary to modify this PST for fastidious or anaerobic microorganisms. It can also not be excluded that the sheer size of different phages or their physiochemical interactions with the media could affect their diffusion capacity and consequently the reading of the lysis zone. We are also aware that the sensitivity of the test could be lowered by effects such as abortive phage infections and persistent bacterial survival, as observed in different experimental settings (22, 38). However, such effects would result in withholding a potentially effective agent and not in the application of an ineffective therapy.

In addition, a large number of repeated tests will be necessary to raise the status from the current proof-of-concept study to statistically valid work, similar to the process of developing the Kirby-Bauer method (39) and to developing EUCAST clinical breakpoints and setting guidelines after the initial introduction of the single-disk diffusion method (40). Such work will also define the specificity of individual phage preparations.

Ideally, future routine laboratory phage tests will include sufficient genomic characterization and appropriate risk analyses, for instance to preclude transduction of antibiotic resistance or virulence genes. For risk assessment, both the criteria and the legal framework have yet to be defined. Phenomena such as horizontal transfer of antibiotic resistance genes and pathogenicity factors through transduction should therefore be excluded in the context of subsequent therapy. Thus far, the proposed concept cannot provide information on therapeutic safety beyond the information regarding bactericidal efficacy.

The proposed PST has similarities to the spot test. Previous studies describe potential advantages of the EOP method in comparison to the spot test (35, 41), but complex and time-consuming manual steps (e.g., pouring soft agar and carrying out individual dilution steps for phages and bacteria) complicate its implementation in routine diagnostics. The lack of quality certificates for the final product hampers the implementation of this method in accredited routine diagnostic laboratories. Special characteristics of individual phage strains such as lysis without events may not be detected by the present test format. However, the test is principally designed to identify phage preparations with the highest probability of exerting therapeutic effects. This could be expected for phages with clear-cut lytic activity. In contrast, for phage strains or variants without visible lytic activity, demonstration of therapeutic effects in a clinical setting is still pending. This parallels the testing of antibiotic activity by an agar diffusion assay: the assay misses some—clinically potentially negligible—in vitro effects of antibiotics identified only in specific setups in exchange for the ease of performance and the broad applicability of the test results.

Bacterial biofilm formation is a clinical challenge especially in chronic and foreign material-associated infections, resulting in dramatically reduced susceptibility to antibiotics, impaired healing and recurrent inflammation (42–44). Conventional routine susceptibility testing like AST by EUCAST cannot reflect the true resistance profile of biofilm-associated bacteria and consequently cannot provide recommendations for antimicrobial treatment (45). However, determining biofilm-specific antibiotic susceptibility could definitely improve therapeutic outcomes. Different screening test systems for bacterial biofilm-specific antibiotic resistance have been proposed as standard methods (46–49). This study does not provide an adapted PST format for that purpose but attempts another popular approach to address this challenge, i.e., the combination of two antibacterial strategies.

For this purpose, PST was combined with established biofilm-active and less active antibiotics (50, 51) to identify potential additive or synergistic effects. Such effects could not be visualized for any of the combinations. However, the chosen conditions may not have been optimal concerning, e.g., the disk material, diffusion parameters, or more generally the time points of adding the phages to the tests. Once the PST is broadly accepted as a reliable test format, renewed attention will be paid to the optimal test conditions.

The proposed PST method is based on the application of liquid preparations to a disk immediately after placing the disk on agar medium. For routine diagnostic purposes, preimpregnated and dry-stored disks comparable to antibiotic-impregnated disks would be easier to handle. With such disks, the production process, including a specific quality control, could be outsourced from the laboratory to a commercial producer, relieving the laboratory of handling of phage suspensions and guaranteeing maximum intertest comparability. In order to pave a way for the routine applicability of PST in a similar fashion, the effects of drying and storage were tested for phage-impregnated disks.

Unfortunately, unlike the stable lytic capacity of phages suspended in SM buffer at 4°C, the phage-impregnated disks lost their activity after only 24 h of storage at 4°C. This observation was independent of the amount of fluid added to the disks after they were applied to the agar surface. A potential explanation could be the quality of the disk material, which might irreversibly fix the phages once the suspension dries. If that is so, it would be prudent to search for better-suited material for the disk assay. Alternatively, some phages may lose their stability when drying on the disks, or the method might not work with phages that are intrinsically unstable and are even difficult to keep in solution.

Our proposed method provides regularly shaped lysis zones, easily detectable and with a measurable phage-specific diameter. The PST protocol steps are easy to handle and lead to reproducible results; thus, by relying on well-established routine procedures, a rapid implementation in diagnostic laboratories appears to be within reach. Quality standards could be defined by, e.g., establishing a minimum diameter to exclude medicinal phage therapy products with intrinsic inefficacy.

Conclusions.

In this study, we were able to establish basic parameters for performing a PST in a highly standardized and quality-controlled manner using the well-studied Escherichia T4 phage. When various bacterium-phage combinations were used, the methodological instructions were also highly applicable and showed reliable and reproducible results. Focusing on existing resources and established methods comparable to CLSI and EUCAST antimicrobial susceptibility testing, with similar manual steps and times to results, this PST can easily be instituted in routine diagnostic laboratories with the potential for laboratory accreditation. In clinical trials, outcomes of phage therapies should be correlated with laboratory-determined lysis zones, as has been established for antibiotic susceptibility testing for several decades. However, before initiating such laborious and time-consuming studies, which phage testing system will be used for that purpose should be agreed on. With the present PST method and its advantages with respect to routine lab practicability and readability of lysis zones, we are confident of introducing a well-suited approach for that purpose.

MATERIALS AND METHODS

The materials used for the method described here are listed in Table 4.

TABLE 4.

Media, chemicals, and devices used in this study

Type	Name	Description (supplier)
Chemicals	NaCl	Sodium chloride (Thermo Fisher Scientific, Waltham, MA, USA)
	MgSO4	Magnesium sulfate, dried (Thermo Fisher Scientific, Waltham, MA, USA)
	Tris-HCl	Tris hydrochloride, 1 M solution, Ultrapure (Thermo Fisher Scientific, Waltham, MA, USA)
Agar	Columbia agar	Columbia agar (BD Columbia agar with 5% sheep blood; Becton Dickinson GmbH, Heidelberg, Germany)
	MH II agar	Mueller-Hinton II agar (BD Mueller-Hinton II agar, Becton Dickinson GmbH, Heidelberg, Germany)
	Bottom agar	2.5% Miller's LB broth base powder (Invitrogen, Thermo Fisher Scientific, Waltham, MA, USA), 1% agar bacteriological no. 1 (wt/vol); 100 (Oxoid, Thermo Fisher Scientific, Waltham, MA, USA), 5 mM MgSO4
	Soft agar	2.5% Miller's LB broth base powder (Invitrogen, Thermo Fisher Scientific, Waltham, MA, USA), 0.4% agar bacteriological no. 1 (wt/vol); 100 (Oxoid, Thermo Fisher Scientific, Waltham, MA, USA)
Liquids	Saline solution	0.9% NaCl (wt/vol)
	SM buffer	100 mM NaCl, 8 mM MgSO4, 50 mM Tris-HCl (pH 7.5)
	LB medium	2.5% Miller's LB broth base powder (Invitrogen, Thermo Fisher Scientific, Waltham, MA, USA) (wt/vol)
Devices	Swab	PS stick with small viscose tip (nerbe plus, Winsen, Germany)
	Drigalski spatula	Hammacher Drigalski spatula, single ended (Hammacher, Solingen, Germany)
	Disk	Whatman antibiotic assay disks, 6 mm (GE Healthcare, Chicago, IL, USA)

Bacteriophage enrichment.

The Escherichia phage T4 was provided by the DSMZ (German Collection of Microorganisms and Cell Cultures GmbH, Braunschweig, Germany) in a liquid suspension. For enrichment, 100 μL of the initial suspension was diluted with 900 μL SM buffer. A suspension of an overnight culture of E. coli and saline (0.9%) was made and adjusted to a photometric optical density at 600 nm (OD₆₀₀) of 0.2. One hundred microliters of bacterial suspension and 100 μL of phage dilution were transferred into 4 mL warm liquefied soft agar. The mixture was poured on bottom agar using the double agar overlay method (35). The plates were incubated at 37°C for 18 h under an ambient atmosphere. When lysis occurred, 4 mL of phage buffer was added to the plate and repeatedly spread on the complete surface by employing a rotating table for 5 h. The supernatant was centrifuged (4,000 rpm, 5 min, 15°C) and sterile filtered (0.45 μm). The enriched phage suspension was stored at 6°C.

Spot test.

Following an overnight incubation in LB medium at 37°C, 200 μL of the reference bacteria suspension was added to the soft agar, mixed, and immediately poured on the bottom agar plates. After the bacterium-containing top agar solidified, 10 μL of phage suspension was randomly spotted onto the surface of the plates and allowed to dry. The inoculated plates were incubated overnight (18 h) at 37°C under an ambient atmosphere, followed by inspection for lysis zones (28).

Preparation of phage dilutions.

Utilizing 900 μL SM buffer for serial dilution, phage suspensions were prepared in a concentration range of 10¹⁰ to 10⁴ PFU/mL. Enumeration of bacteriophages was performed by using the double agar overlay plaque assay (35).

Determination of bacterial concentrations.

Overnight colonies of E. coli strain DSM 613 grown on Columbia agar were suspended in 0.9% saline and photometrically adjusted to various McFarland units in a range between 0.1 and 0.5 (also see below).

Development of the PST.

The PST was developed by using the following steps.

Step 1: evaluation of basic parameters.

First, the following basic parameters were evaluated using the combination Escherichia phage T4 (DSM 4505)–E. coli (DSM 613).

(i) Host inoculum. E. coli inocula were set to densities of 0.1, 0.2, and 0.5 McFarland units, approximately corresponding to 2 × 10⁶, 8 × 10⁶, and 1.2 × 10⁸ CFU/mL, respectively. The suspension was used within 15 min of preparation.

(ii) Agar inoculation method. Agar inoculation was performed by two different approaches. (i) One hundred microliters of the prepared E. coli inoculum was added dropwise onto dry MH II agar plate at room temperature. The inoculum was spread with a Drigalski spatula. (ii) A dry sterile cotton swab was dipped into the bacterial suspension. Excess fluid was removed by pressing and turning the swab against the inside of the tube, followed by streaking the bacterium-inoculated swab on the agar plate.

(iii) Phage inoculation volume. Volumes of 5 μL, 10 μL, 20 μL, and 50 μL phage suspension with a concentration of 10⁹ PFU/mL were utilized for phage susceptibility testing.

(iv) Phage concentration. Phage dilutions containing 10⁴ PFU/mL, 10⁶ PFU/mL, 10⁸ PFU/mL, and 10¹⁰ PFU/mL were tested.

(v) Phage application methodology. To perform the modified spot test, the above-mentioned volumes and concentrations of phage suspensions were added dropwise to E. coli-inoculated agar plates.

For implementation of the disk-based test, disks were placed on the agar surface using forceps. The number of disks on a plate was limited to 4. The above-mentioned volumes and concentrations of phage suspensions were added dropwise to the disks. After phage application to agar or disks, the lids were placed on the plates, and the plates were incubated at 37°C for 18 to 24 h under an ambient atmosphere. When 50 μL of phage suspension was used, plates were allowed to dry for up to 30 min in a laminar flow hood prior to overnight incubation.

(vi) Growth kinetics. Overnight cultures of E. coli were set to a photometric OD₆₀₀ of 0.2, corresponding to approximately 10⁸ CFU/mL. For each assay, 160 μL of LB medium was mixed with 20 μL of bacterial suspension and 20 μL of phage suspension (T4 phage, 10⁸ PFU/mL) or 20 μL of SM buffer in transparent 96-well plates (Micro test plate; Sarstedt, Germany). The plates were incubated at 37°C in a SpectraMax M3 Flash plate reader (Molecular Devices, USA). The OD₆₀₀ was recorded at regular intervals of 5 min for 8 h. Samples, controls, and blanks were always assayed in triplicate.

(vii) Stability of disks impregnated with phages. Twenty microliters of phage suspension was spotted onto a disk. The impregnated disks were dried for 120 min in an incubator at 37°C and thereafter stored at 4°C in a petri dish. PST was performed after storage times of 24 and 48 h. Disks were placed on the agar with and without the addition of 20 μL SM buffer.

(vii) Combined testing of antibiotic and phage activities. Agar inoculation with E. coli was done following the PST steps proposed below. For combined disk testing of antibiotics and phages, ready-to-use disks of ampicillin (10 μg), ciprofloxacin (5 μg), and gentamicin (10 μg) (BD BBL Sensi-Disk antimicrobial susceptibility test disks; Becton, Dickinson, USA) were placed onto the inoculated agar. Then, 20 μL of Escherichia phage T4 suspension was immediately spotted onto the disks. Controls were performed by spotting 20 μL SM buffer onto the disks.

Step 2: transfer of the PST parameters to other phages and bacterial strains.

In the second step, the most suitable parameters determined in step 1 were evaluated by using five other phages and bacterial strains (Table 2).

Step 3: quality controls.

(i) Host specificity. Quality control for host specificity involved inoculation of the agar plates with an Enterobacterales species not susceptible to T4 phages, i.e., Serratia marcescens (turbidity standard, 0.2 McFarland) using a Drigalski spatula. Disks were then placed on the agar, and the Escherichia phage T4 (5 μL, 10 μL, 20 μL, and 50 μL; 10⁹ PFU/mL) was added.

(ii) Intrinsic growth inhibition. For quality control regarding potential intrinsic growth inhibition, disks were placed on agar inoculated with E. coli (DSM 613), and phage-free SM buffer (5 μL, 10 μL, 20 μL, and 50 μL) was added.

(iii) Volume-dependent agar diffusion. To visualize the diffusion range of the applied volume, China ink in SM buffer (10% [vol/vol]; 20 μL and 50 μL) was added to the disks placed on the agar inoculated with E. coli (DSM 613).

(iv) Number of performed experiments. All experiments were conducted in triplicate in parallel (technical replicates) and on three different days (biological replicates).

Step 4: comparison of EOP and PST.

EOP testing has been established as a method for testing the lytic capacity of phages, potentially delivering more refined results than the spot test (35). EOP values depend on the existence of phage reference strains for a given bacterial species.

Four Staphylococcus phages and matching Staphylococcus epidermidis host strains (including S. epidermidis plus Staphylococcus phage vB_SepM_Alex, used in step 2 for PST transfer) were used for the comparison of PST and EOP. Thus, all bacterial strains to be tested were grown overnight (37°C), and 100 μL of each culture was used in double-layer plaque assays together with 100 μL of each step of a serially diluted phage suspension. EOP assay replicates for a particular phage were performed in parallel with all bacterial host strains. The plates were incubated overnight at 37°C, and the PFU were counted for each combination. Each phage/bacterium combination was tested three times for EOP. Additionally, PST was performed using the final characteristics summarized in Fig. 9.

Ethics.

The study was performed without using human or animal subjects or tissues.