Apotransferrin in Combination with Ciprofloxacin Slows Bacterial Replication, Prevents Resistance Amplification, and Increases Antimicrobial Regimen Effect

There has been renewed interest in combining traditional small-molecule antimicrobial agents with nontraditional therapies to potentiate antimicrobial effects. Apotransferrin, which decreases iron availability to microbes, is one such approach.

ABSTRACT

There has been renewed interest in combining traditional small-molecule antimicrobial agents with nontraditional therapies to potentiate antimicrobial effects. Apotransferrin, which decreases iron availability to microbes, is one such approach. We conducted a 48-h one-compartment in vitro infection model to explore the impact of apotransferrin on the bactericidal activity of ciprofloxacin. The challenge panel included four Klebsiella pneumoniae isolates with ciprofloxacin MIC values ranging from 0.08 to 32 mg/liter. Each challenge isolate was subjected to an ineffective ciprofloxacin monotherapy exposure (free-drug area under the concentration-time curve over 24 h divided by the MIC [AUC/MIC ratio] ranging from 0.19 to 96.6) with and without apotransferrin. As expected, the no-treatment and apotransferrin control arms showed unaltered prototypical logarithmic bacterial growth. We identified relationships between exposure and change in bacterial density for ciprofloxacin alone (R² = 0.64) and ciprofloxacin in combination with apotransferrin (R² = 0.84). Addition of apotransferrin to ciprofloxacin enabled a remarkable reduction in bacterial density across a wide range of ciprofloxacin exposures. For instance, at a ciprofloxacin AUC/MIC ratio of 20, ciprofloxacin monotherapy resulted in nearly 2 log₁₀ CFU increase in bacterial density, while the combination of apotransferrin and ciprofloxacin resulted in 2 log₁₀ CFU reduction in bacterial density. Furthermore, addition of apotransferrin significantly reduced the emergence of ciprofloxacin-resistant subpopulations compared to monotherapy. These data demonstrate that decreasing the rate of bacterial replication with apotransferrin in combination with antimicrobial therapy represents an opportunity to increase the magnitude of the bactericidal effect and to suppress the growth rate of drug-resistant subpopulations.

INTRODUCTION

Antimicrobial resistance endangers the practice of modern medicine (1, 2). New strategies are critically needed that can prolong the effective life spans of antibiotics, especially those with lower intrinsic barriers to emergence of resistance, like the fluoroquinolones. Combination therapy represents one path to increasing antimicrobial regimen potency.

Recently, there has been renewed interest in combining traditional small-molecule antimicrobial agents with nontraditional therapies to potentiate their antimicrobial effects (3, 4). Nontraditional therapies broadly include phages, detergents (e.g., colistin-related agents), antibodies, bacterial-growth-rate modulators (e.g., catecholamine-like agents and ATP synthase inhibitors), and others. These approaches are of particular interest, as they hold the promise of sidestepping traditional antimicrobial resistance mechanisms.

With regard to bacterial-growth-rate modulation, recent studies have suggested that speeding up bacterial replication in the context of effective antimicrobial exposure increases the rate and extent of bactericidal activity and shortens therapy duration (5, 6). In the context of ineffective antimicrobial exposure, such regimens increased bactericidal activity early in the course of therapy but ultimately failed due to antibiotic-resistant bacterial subpopulations (6). An alternative to speeding up bacterial replication to increase antimicrobial regimen activity is slowing replication with the same end in mind, that is, slowing bacterial replication so that the drug-susceptible subpopulation is eliminated by the antibiotic’s action while growth of the drug-resistant bacterial subpopulation is suppressed.

Here, we explore iron depletion via apotransferrin as a means to slow bacterial replication, its impact on the rate and extent of ciprofloxacin’s bactericidal effects, and the resultant changes in emergence of antibiotic resistance. Beyond investigating how a reduced bacterial replication rate is coupled with the extent of ciprofloxacin bactericidal effects, this study also demonstrates how ciprofloxacin-apotransferrin combination therapy decreased the total population density and thereby suppressed the escape of antibiotic-resistant mutants.

RESULTS

In vitro susceptibility and growth rate studies.

Challenge isolate ciprofloxacin and apotransferrin MIC values determined in RPMI 1640 broth medium are presented in Table 1. The ciprofloxacin MIC values ranged from 0.008 to 32 mg/liter, while those for apotransferrin ranged from 8 to 128 mg/liter.

TABLE 1.

Susceptibility results for the four clinical isolates utilized in the PK-PD in vitro infection model studies

K. pneumoniae isolate	MIC (mg/liter)
	Ciprofloxacin	Apotransferrin
KP4	0.008	128
KP3	0.03	8
KP4.1	0.25	128
KPC-KP1	32	32

Drug assay and pharmacokinetic studies.

Assessments of the assay performance demonstrated interassay percent coefficients of variation (CV) for the quality control samples at concentrations of 0.0100, 0.160, and 1.00 μg/ml of 7.23%, 8.79%, and 8.09%, respectively, for ciprofloxacin. The interassay percent CV for the quality control samples at concentrations of 25.0, 400, and 2,500 μg/ml were 11.4%, 13.7%, and 18.7%, respectively, for apotransferrin. The correlation coefficient values ranged from 0.9924 to 0.9993 and from 0.9945 to 0.9974 for ciprofloxacin and apotransferrin, respectively.

The targeted and observed ciprofloxacin and apotransferrin concentrations are presented in Fig. 1A and B, respectively. The observed drug concentrations were concordant with those expected for both agents. These data indicate that the targeted drug concentration for each study agent was well simulated within the one-compartment in vitro infection model.

FIG 1.

Targeted versus observed ciprofloxacin (A) and apotransferrin (B) concentrations simulated within the one-compartment in vitro infection model.

One-compartment infection model studies.

Figure 2 shows the relationship between ciprofloxacin expsoure and changes in bacterial density with and without apotransferrin against four Klebsiella pneumoniae challenge isolates. Note that the bacterial growth in the no-treatment control arms at 48 h ranged from 0.6 to 2 log₁₀ units across the four challenge isolates, while that in the apotransferrin monotherapy control arms was approximately 2 log₁₀ CFU. As expected, there was a relationship between ciprofloxacin exposure and change in bacterial density (R² = 0.64).

FIG 2.

Relationship between ciprofloxacin exposure and change in bacterial density with and without apotransferrin against four K. pneumoniae challenge isolates.

Across the apotransferrin dose range studied (0.09 to 1.5 mg/liter continuous infusion) and the four challenge organisms, there was no within-isolate exposure-response relationship in the context of a given ciprofloxacin free-drug area under the concentration-time curve over 24 h divided by the MIC (AUC/MIC ratio) (Fig. 2), as evidenced by inconsistent signals of increased effect, as the apotransferrin concentration increased. However, it is critical to note that the bactericidal effects of ciprofloxacin-apotransferrin combination therapy were greatest for the most ciprofloxacin-susceptible isolate (KP4; 0.008 mg/liter) and least for the most ciprofloxacin-resistant isolate (KPC-KP1; 32 mg/liter). Across the four challenge isolates, there was a relationship between ciprofloxacin exposure and change in bacterial density in the context of transferrin exposure (R² = 0.84).

Figure 3 shows the time course of growth for each challenge isolate (total population and ciprofloxacin-resistant subpopulation) exposed to no-treatment, ciprofloxacin, and apotransferrin monotherapy control arms and ciprofloxacin-apotransferrin combination therapy regimens. There are four important observations that can be garnered from the figure. First, while ciprofloxacin monotherapy resulted in an initial 4- to 6-log₁₀ CFU/ml reduction in the bacterial burden at 4 h into therapy, the regimen ultimately failed over time due to the selection for a drug-resistant subpopulation. Second, relative to the no-treatment control arm, apotransferrin monotherapy attenuated bacterial growth (total population and ciprofloxacin-resistant subpopulation) by approximately 1 to 2 log₁₀ CFU/ml across challenge isolates. Third, ciprofloxacin-apotransferrin combination therapy resulted in less cell killing early in therapy relative to the ciprofloxacin monotherapy arm, but the combination regimens did not fail over the 48-h experiment duration. Fourth, and critically, often a significant proportion of the total bacterial population was ciprofloxacin resistant following ciprofloxacin-apotransferrin combination therapy, but that population did not grow out to the same degree as with ciprofloxacin monotherapy.

FIG 3.

Dose range study results for the four K. pneumoniae isolates evaluated in the one-compartment in vitro infection model displayed over time. The solid symbols represent total bacterial burden, and the open symbols represent the resistant subpopulation observed on the drug-supplemented agar plates. Ciprofloxacin doses of 23.4, 46.9, 93.8, and 750 mg every 12 h (q12h) were utilized for KP3, KP4, KP4.1, and KP1, respectively.

Figure 4 shows the change in log₁₀ CFU per milliliter of the ciprofloxacin-resistant subpopulation at 48 h. Note that ciprofloxacin monotherapy resulted in 1- to 6.5-log₁₀ CFU/ml growth of the drug-resistant subpopulation over 48 h. Moreover, note that apotransferrin monotherapy and ciprofloxacin-apotransferrin combination therapy generally resulted in suppression of growth of the drug-resistant subpopulation over 48 h.

FIG 4.

Dose range study results for the four K. pneumoniae isolates evaluated in the one-compartment in vitro infection model displayed as changes in the drug-resistant population from control at 48 h. Each symbol represents one of the different treatment regimens evaluated within the system.

DISCUSSION

Our objectives for these studies were 2-fold. The first was to demonstrate that a diminishing bacterial replication rate with a growth modulator was coupled with the extent of an antibiotic’s bactericidal effects. The second was to demonstrate that combination therapy with an antibiotic plus a bacterial growth modulator decreases the total bacterial population density while suppressing growth of the antibiotic-resistant subpopulation.

We selected the one-compartment in vitro infection model rather than the hollow-fiber in vitro infection model because apotransferrin’s mass was too large to pass through dialysis membranes. The antibiotic we selected for these studies was ciprofloxacin. Ciprofloxacin was selected due to its long history in the therapy of patients with serious infections by Gram-negative bacteria, because quinolone resistance has significantly eroded its clinical utility and because ciprofloxacin has a relatively low barrier to resistance emergence. Apotransferrin was selected as the bacterial growth modulator.

Apotransferrin was selected due to its physiological role in iron transport and distribution in the body and its ability to limit the outgrowth of rifampin-resistant Staphylococcus aureus on exposure to rifampin (7).

In contrast to small-molecule iron chelators, which sequester iron away from bone marrow and deliver it to the kidneys, transferrin delivers iron to myeloid cells for sequestration and storage. Transferrin therefore avoids the bone marrow suppression and nephrotoxicity caused by small-molecule chelators (8, 9). Indeed, in several clinical trials, administration of transferrin to patients with excess iron levels, including neutropenic and stem cell transplant patients, was safe while effectively reducing unbound iron in the blood and inhibiting microbial growth in the blood ex vivo (10–13).

Apotransferrin may have two antibacterial mechanisms of action. First, it has been demonstrated that apotransferrin suppresses antibiotic free-radical DNA damage via the Fenton reaction, a process that has been shown to promote antibiotic-resistant mutants (14–16). Second, apotransferrin may reduce resistance emergence by inhibiting microbial growth through iron starvation. Thus, through free-iron sequestration, apotransferrin may limit the production of reactive oxygen species and/or simply starve bacteria, either of which may result in decreased antibiotic resistance emergence.

There were four major conclusions of the studies described here. First, apotransferrin monotherapy does not reliably reduce bacterial growth (Fig. 3). In the in vitro infection model utilized in these studies, it was only when apotransferrin was administered in combination with an antibiotic that appreciable bacterial killing was observed. Second, we demonstrated that lowering the bacterial replication rate with apotransferrin decreased ciprofloxacin’s initial bactericidal effect. Specifically, relative to ciprofloxacin monotherapy, ciprofloxacin-apotransferrin combination therapy resulted in a shallower nadir of cell killing during the first few hours of the 48-h study (Fig. 3). Third, we demonstrated that lowering the bacterial replication rate with apotransferrin increased ciprofloxacin’s bactericidal effect over a 48-h study. That is, the ciprofloxacin exposure-response relationship when administered in combination with apotransferrin is shifted left relative to that for ciprofloxacin monotherapy (Fig. 2). Fourth, we demonstrated that ciprofloxacin-apotransferrin combination therapy markedly suppressed growth of the antibiotic-resistant subpopulation (Fig. 4).

Together, these data suggest that it is possible that the ciprofloxacin-apotransferrin combination is much more effective than ciprofloxacin alone against isolates with elevated MIC values. Moreover, these data may also support the notion of sequential therapy, that is, administering ciprofloxacin first to maximize early cell killing and adding apotransferrin thereafter to suppress the growth of an antibiotic-resistant subpopulation. Clearly, both of these strategies will need to be explored further using preclinical in vivo and in vitro infection models.

In conclusion, we demonstrated that apotransferrin modulated ciprofloxacin’s bactericidal effects against quinolone-susceptible and -resistant K. pneumoniae isolates. Moreover, we demonstrated that apotransferrin in combination with an antibiotic decreases the total bacterial population density while suppressing growth of the antibiotic-resistant subpopulation. These data provide proof of concept for new adjunctive iron-sequestering therapeutic treatment options. Such adjunctive-to-antibiotic therapies hold the promise of reducing the emergence of on-therapy antibiotic resistance and improving patient outcomes.

MATERIALS AND METHODS

Bacteria and study drug.

The challenge isolates utilized in these studies were supplied by collaborating scientists from the University of Southern California. They included three clinical isolates of K. pneumoniae (KPC-KP1 [a ciprofloxacin-resistant isolate] and KP3 and KP4 [both highly ciprofloxacin susceptible]). Isolate KP4.1 was derived from KP4 in a previously conducted in vitro study and was selected based upon its ciprofloxacin MIC value being between that of KPC-KP1 and those of KP3 and KP4. The study drugs (ciprofloxacin and apotransferrin) were obtained from Sigma-Aldrich (St. Louis, MO).

Media and in vitro susceptibility studies.

Susceptibility studies were performed according to Clinical and Laboratory Standards Institute (CLSI) guidelines (17) and interpretive criteria recommendations of the U.S. Committee on Antimicrobial Susceptibility Testing (USCAST) (18) using a microdilution method in RPMI 1640 medium supplemented with 2 g/liter sodium bicarbonate from Gibco Laboratories (Gaithersburg, MD). Apotransferrin (iron-depleted transferrin) was obtained from Sigma-Aldrich (catalog no. T1147). The stock solutions of apotransferrin and ciprofloxacin utilized in the susceptibility testing were prepared in OmniTrace Ultra water from Millipore Sigma (Burlington, MA) to minimize binding of free-iron ions.

Pharmacokinetics-pharmacodynamics (PK-PD) in vitro infection model and sample processing.

The one-compartment in vitro infection model used in this study has been described previously (19). The one-compartment in vitro infection model consisted of a central infection compartment, which contained RPMI 1640 growth medium maintained at pH 8.01, the challenge isolate, and magnetized stir bars to ensure homogeneity of the drug, medium, and organism. The central infection compartment sat on a magnetic stir plate within a temperature-controlled incubator set to 35°C. Computer-controlled peristaltic pumps infused drug-free growth medium into the central infection compartment while at the same time removing medium through an exit port into a waste container. The challenge isolate was inoculated directly into the central infection compartment, and the test compound was infused by a computer-controlled syringe pump. Peristaltic diffusion rates were set so that the desired concentration-time profile of ciprofloxacin mimicked human concentration-time profiles. Apotransferrin profiles were administered as continuous infusions. Sample aliquots for assessing bacterial density and drug concentration were aseptically collected from the central infection compartment at predetermined time points (see below).

Each challenge isolate was prepared from a culture grown overnight on plates of Trypticase soy agar supplemented with 5% lysed sheep blood (blood agar) obtained from BD Laboratories. Isolates were collected from overnight cultures, suspended in Mueller-Hinton broth medium, and grown to mid-logarithmic growth phase in an Erlenmeyer flask immersed in a shaking water bath set to 35°C and 125 rotations per minute. The bacterial concentrations of subcultures were estimated by measuring the optical density at 600 nm and making comparisons to previously confirmed growth curves for each challenge isolate. The isolates were inoculated into the system at a target of 1.0 × 10⁶ CFU/ml and allowed to grow uninhibited within the system for 2 h before initiating treatment, at which point the inoculum had entered logarithmic phase, achieving a bacterial burden of ∼1.0 × 10⁷ CFU/ml.

In the antibiotic-containing treatment arms, bacteria were exposed to changing ciprofloxacin concentrations that simulated free-drug concentrations, assuming 30% binding to human serum proteins and a 4.0-h half-life (20). Ciprofloxacin doses of 23.4, 46.9, 93.8, and 750 mg administered every 12 h for KP3, KP4, KP4.1, and KP1, respectively, were simulated alone and in combination with apotransferrin at concentrations of 0.09, 0.19, 0.38, 0.75, or 1.5 mg/ml. All the regimens were compared to the no-treatment and apotransferrin monotherapy control arms. Sample aliquots of 1 ml were collected for CFU quantification at 0, 2, 4, 8, 12, 24, 30, and 48 h. The samples were washed (centrifuged twice, aspirated, and resuspended with sterile saline) to prevent drug carryover and subsequently cultured onto blood agar. A portion of the washed 0-, 24-, and 48-h time point samples were also plated on Mueller-Hinton II (MHII) agar supplemented with ciprofloxacin at a concentration equivalent to 3× MIC to evaluate the change in density of the drug-resistant subpopulation within the system. If growth was observed on the drug-supplemented MHII agar plates, a subset of isolates was stored for future microbiological studies. To measure drug concentrations, a second sample aliquot of 1 ml was collected at 1, 3, 5, 7, 11, 23, 25, 27, 29, and 48 h postinoculation and immediately frozen at −80°C until it was assayed.

Drug assay.

Calibration standards and quality controls were prepared in the study matrix and processed concurrently with collected samples. All the samples were assayed by liquid chromatography and tandem mass spectrometry (LC–MS-MS) on a Sciex 5500 mass spectrometer, performed separately for apotransferrin and ciprofloxacin analyses. Prior to analysis, apotransferrin was digested using a Waters Protein Works express digestion kit. For ciprofloxacin, samples were diluted with water prior to analysis. Analytical curves were linear from 25.0 to 2,500 μg/ml for apotransferrin and from 0.0100 to 1.00 μg/ml for ciprofloxacin. The lower limit of quantitation was 25.0 μg/ml for apotransferrin and 0.0100 μg/ml for ciprofloxacin.