Abstract
Multidrug-resistant (MDR) Acinetobacter baumannii is a difficult pathogen due to its propensity to develop resistance to antibiotics. Antisense nucleic acid analogs have been proposed as a potential alternative anti-infective approach. We developed a peptide nucleic acid (PNA) compound that targets the essential Acinetobacter gene carA. The PNA oligomer was conjugated to the cell-penetrating peptide (RXR)4XB. In vitro testing of the PNA conjugate against four clinical strains of MDR-A. baumannii in minimal medium demonstrated that all four strains were inhibited at a concentration of 1.25 μM. In vivo testing of the PNA conjugate was done using a Galleria mellonella model of sepsis caused by one of the clinical strains. Preliminary testing of a variety of inocula demonstrated that an inoculum of 1 × 106 cfu was lethal to the majority of caterpillars by day 3, but not within 24 hours. The PNA compound was administered 30 minutes after an inoculum of 1 × 106 cfu at doses estimated to produce concentrations of ∼5 and 20 μM. The PNA compound had no effect at the lower dose. However, the higher dose reduced mortality from 5/28 (18%) to 0/28 (0%) at day 1 (p = 0.051) and from 19/28 (68%) to 9/28 (32%) at day 6 (p = 0.015). Antisense therapy is a novel approach to dealing with difficult MDR pathogens that could circumvent the problem of progressive resistance to available antibiotics. Further studies need to be done with additional strains and more complex in vivo model systems.
Introduction
Gram-negative bacterial resistance has become one of the greatest health care threats in hospitals worldwide. 1 Multidrug-resistance has grown among Enterobacteriaceae, Acinetobacter baumannii, and Pseudomonas aeruginosa (due to a variety of carbapenemases and other mechanisms). The recent emergence of plasmid-mediated polymixin resistance due to MCR-1 has further limited treatment options. 2 Acinetobacter species in particular are important contributors to the problem of gram-negative resistance, with A. baumannii serving as a common cause of nosocomial infections. Multidrug-resistant (MDR)-A. baumannii has been labeled a serious threat resulting in over 500 deaths per year in the United States and has been associated with substantial added health care costs.3,4 These growing concerns, along with the paucity of new antimicrobial agents active against MDR-A. baumannii, have created an urgent need for new approaches to anti-infective therapy.
One promising new strategy is the development of antisense nucleic acid analogs, which form complexes with complementary sequences of DNA or RNA, preventing the successful transcription or translation of genes that are essential to microbial growth. 5 Peptide nucleic acids (PNAs) are nucleic acid oligomers that have a N-(2-aminoethyl)-glycine backbone in place of the deoxyribose and ribose backbones of DNA and RNA, respectively. PNAs, and other antisense oligomers, have demonstrated antibacterial activity against antibiotic-resistant bacteria6,7 as well as inhibition of antibiotic resistance genes.8,9 Attacking pathogens by inhibiting expression of essential genes may circumvent many of the difficulties associated with resistance to antibiotics.
The gene target for antisense therapy should ideally be an essential gene for the pathogen but not the human host. We selected the carA gene that is involved in pyrimidine metabolism and has previously been demonstrated to be an essential gene for A. baumannii. 10 In that report, carA-deficient mutants demonstrated impaired growth in a variety of experimental systems, making carA an attractive target for an antisense approach. For in vivo testing, we used Galleria mellonella (caterpillars of the greater wax moth) that have previously been demonstrated to be a useful model of sepsis caused by A. baumannii. 11
Materials and Methods
Bacterial strains
Isolates of A. baumannii were chosen from a collection of clinical isolates from citywide surveillance studies conducted in Brooklyn, NY between 2001 and 2013. Four MDR isolates were chosen because they represented the four most common strains in the region and were resistant to most antibiotics (Table 1). Isolates were previously tested for the presence of the carbapenemases blaOXA23, blaOXA24, blaOXA58, blaIMP, blaVIM, blaNDM, and blaKPC. 12 Isolates previously underwent genetic fingerprinting by the rep-PCR method with the ERIC-2 primer using Ready-To-Go RAPD Analysis Beads (GE Health care Biosciences, Piscataway, NJ). Isolates tested by the rep-PCR method were considered related if there was a 0–1 band difference.
Multidrug-Resistant Acinetobacter baumannii Strains and In Vitro Activity of Peptide Nucleic Acid Conjugates
Rep-PCR method.
CA-MHB - cation-adjusted Mueller–Hinton broth.
Alpha - PNA targeting an alpha helix region (amino acids 147–149 of carA, Beta strand - PNA targeting a beta region (amino acids 192–194) of carA, Coil - PNA targeting a coil region (amino acids 57–60) of carA.
PNA, peptide nucleic acid; NA, not available.
PNA synthesis
The carA gene was chosen as a target from a group of essential A. baumannii genes identified in a prior study. 10 This gene, involved in pyrimidine metabolism, was essential for growth in several experimental systems and was the only gene whose deficiency resulted in impaired growth even in rich laboratory medium. 10 Ten-base oligomers were designed targeting sequences in the predicted 13 alpha helix and beta strand regions of the protein; in addition, an oligomer was designed targeting a region outside these conformations (coil region) of the carA gene (Table 1). The oligomers were conjugated to the cell-penetrating peptide (RXR)4XB, (“X” and “B” stand for 6-aminohexanoic acid and β-alanine, respectively).6,8 The (RXR)4XB-oligomer conjugates were synthesized by PNA Bio, Inc., (Newbury Park, CA).
Minimum inhibitory concentration testing
Susceptibility testing was initially performed in CA-MHB for two of the strains using standard CLSI methodology (Table 1). Because the PNA conjugate targeting the coil region had the best activity in broth, additional studies were done using this PNA conjugate. Susceptibility testing was repeated in minimal medium for Acinetobacter. 14 Glucose (20 mM) was used as the carbon source. All testings were performed with an inoculum of ∼5 × 105 cfu/mL in microwell plates. The PNA conjugates were dissolved in PBS and tested in serial two-fold dilutions from 40 μM to 1.25 μM. The Minimum inhibitory concentration (MIC) was the lowest concentration without visible growth after overnight incubation at 37°C. Minimum bactericidal concentrations (MBCs) were performed on aliquots from each of the clear wells. Serial 10-fold dilutions of the aliquots were plated onto Mueller-Hinton agar plates and incubated overnight. The lowest concentration resulting in a ≥3 log10cfu/mL decrease in the starting inoculum was considered the MBC. MIC and MBC testings for strain KB17 were performed twice.
Expression of the carA gene
Expression of the carA gene in strain KB17 was analyzed by real-time RT-PCR with and without exposure to the PNA conjugate targeting the coil region. KB17 was grown to log phase in Mueller–Hinton broth, diluted to ∼5 × 105 cfu/mL, and then grown overnight with and without 0.5 × MIC of the PNA conjugate (10 μM in Mueller–Hinton broth). Expression of the gene was normalized to that of a ribosomal housekeeping gene, as previously described. 15 The primers and probe for expression of the carA gene were (5′ to 3′): carA forward CAGCTGCTTGTGATTACGCT; carA reverse CACCGTGGTGACCATGATTC; and carA probe FAM-TGGCACTCGCTTCTGGTGCAA-TAMRA. Primer and probe concentrations were adjusted to achieve amplification efficiencies of ∼100%. A total of 5 ng of RNA was used, and samples were run in triplicate. Expression of carA in the PNA-exposed isolate was calibrated against the simultaneous growth control without the PNA conjugate.
G. mellonella sepsis model
Strain KB17 was chosen for the in vivo studies because it represents one of the two dominant strains in the region, possesses OXA23, and was the most resistant to conventional antibiotics (Table 1). G. mellonella caterpillars (Vanderhorst, Inc., Saint Marys, OH) were stored in the dark and used within 7 days of shipment. Caterpillars weighing between 250–330 mg were used for the experiments. All injections were performed with 10 μL aliquots into the left or right last proleg using a Hamilton syringe. 11 After the injections, caterpillars were stored in large Petri dishes in the dark at room temperature, and the number of survivors was assessed daily for 6 days.
Preliminary experiments were performed to determine the optimal inoculum for subsequent experiments. In all experiments, groups of 10–16 caterpillars were given two injections separated by 30 minutes. All injections were diluted in PBS. Three groups of caterpillars were inoculated with ∼107cfu, 106cfu, or 105cfu of A. baumannii KB17. An inoculum that resulted in the majority surviving 24 hours, but the majority dying by day 6 was considered optimal. In the treatment experiments, one group received two injections of PBS, one group received A. baumannii KB17 followed by PBS, and the third group received A. baumannii KB17 followed by the PNA conjugate targeting the coil region. The PNA conjugate was tested at a lower dose (10 μL of 150 μM PNA) and a higher dose (10 μL of 600 μM PNA). These doses were estimated to produce concentrations of 5 and 20 μM, respectively, in the caterpillars. Statistical differences in mortality between groups were compared by Fisher's exact test.
Results
The PNA conjugate targeting the coil region demonstrated excellent in vitro activity, with an MIC and MBC of ≤1.25 μM for all four A. baumannii strains (Table 1). The other PNA conjugates targeting the alpha helix and beta strand regions were inactive against strain KB17 (Table 1). Following exposure to the PNA conjugate targeting the coil region, expression of carA in strain KB17 was 23% of that of a simultaneous growth control, confirming inhibition of this gene. In the G. mellonella model, mortality rates following inoculation with A. baumannii KB17 were dose-dependent. Mortality rates at days 1 and 6 after an inoculum of ∼105cfu were 0% and 0%, respectively, after an inoculum of ∼106cfu were 20% and 60%, respectively, and after an inoculum of ∼107cfu were 80% and 100%, respectively. Therefore, an inoculum of ∼106cfu was selected for further studies.
In initial experiments, the lower dose of the PNA conjugate was used, which was estimated to produce a concentration of 5 μM in the caterpillars. At this dose, no effect on mortality was demonstrated (Table 2). Subsequently, the higher dose of the PNA conjugate was used, which was estimated to produce a concentration of 20 μM in the caterpillars. At this dose, a significant reduction in mortality was evident by day 1 and at the end of the study (Table 2). Using the higher dose, mortality in the PNA treatment group was similar to the uninfected PBS group throughout the study (day 6 mortality 32% vs. 18%, p = 0.36).
Activity of the Peptide Nucleic Acid Conjugate Targeting the carA Coil Region in the Galleria mellonella Model of Sepsis Due to Acinetobacter baumannii KB17
p = NS, PNA group versus A. baumannii alone.
p = 0.051, PNA group versus A. baumannii alone.
p = 0.029, PNA group versus A. baumannii alone.
p = 0.003, PNA group versus A. baumannii alone.
p = 0.015, PNA group versus A. baumannii alone.
Discussion
Infections caused by antibiotic-resistant bacteria are a growing public health problem worldwide. In the United States, the CDC estimates that 23,000 people die annually from antibiotic-resistant infections, including 500 deaths from MDR-Acinetobacter infections. 3 In addition to morbidity and mortality, antibiotic-resistant infections added in excess of $2 billion treatment costs in 2014 alone in the United States. 16 A. baumannii has been a particularly difficult pathogen due to its ability to acquire multiple mechanisms of resistance and its ability to persist on environmental surfaces for extended time periods. Several antibiotic agents with activity against KPC-producing Enterobacteriaceae have recently become available. However, newer agents with enhanced activity against MDR-A. baumannii are not available. Moreover, the progressive emergence of resistance suggests that antibiotic development alone will not completely solve the problem of antibiotic resistance.
Accordingly, there has been a great deal of interest in alternative approaches to battling MDR-bacterial pathogens. One such approach has been the use of antisense nucleic acid analogs to block specific bacterial genes necessary for growth or that confer antibiotic resistance. Prior studies have primarily utilized either peptide-conjugated phosphorodiamidate morpholino oligomers (PPMOs) or PNAs. Using PPMOs, growth of MDR-A. baumannii was inhibited in vitro and in vivo with a conjugate targeting acpP, 17 and carbapenem susceptibility was restored in vitro with a compound targeting the carbapenemase gene NDM-1. 9 In vitro activity was reported against one isolate of MDR-A. baumannii using a PNA compound targeting rpoD. 6 We chose to develop a PNA conjugate targeting the carA gene because this gene was shown to be essential in a variety of tested media, 10 suggesting the possibility of enhanced in vivo efficacy. To our knowledge, this is the first study to assess the activity of a PNA conjugate in an in vivo model of MDR-A. baumannii infection and targeting carA in particular. The greater in vitro activity of one of the PNA conjugates compared to the others suggests that the activity is related to the antisense effect rather than a nonspecific effect of the cell-penetrating peptide.
There are several limitations to this study. Although a significant difference was seen between the PNA and control groups in the sepsis model, the overall group numbers are relatively small. In addition, there is obviously a limit to the conclusions that can be drawn about potential human use from the invertebrate G. mellonella model. Another troubling issue is the relatively high dose of the PNA conjugate required to demonstrate activity in the sepsis model, given the very low MICs and MBCs of the compound. The estimates of the concentrations of PNA conjugate produced by the doses administered were based on crude estimates of the volume of the caterpillars and, more importantly, an assumption of an even distribution of the bacterial inoculum and the PNA conjugate within the caterpillars. In fact, the distribution may be quite uneven and the actual concentration of the PNA conjugate in vivo may be significantly lower than the estimated 5 and 20 μM. Finally, synthesis of the PNA conjugates is relatively expensive at this time and might present an issue for future use on a larger scale.
A peptide-PNA conjugate targeting the essential bacterial gene carA demonstrated promising activity against MDR-A. baumannii both in vitro and in vivo. These findings lend support to the use of antisense compounds as a strategy to overcome the problem of antibiotic resistance in MDR bacterial pathogens. Additional studies are warranted to determine the potential clinical utility of this approach.
Footnotes
Acknowledgment
No specific grant was received from any funding agency in the public or private sector.
Disclosure Statement
No competing financial interests exist.