In-Vitro Activity of Dimercaptosuccinic Acid in Combination with Carbapenems Against Carbapenem-Resistant Pseudomonas aeruginosa

Introduction

Pseudomonas aeruginosa is one of the so-called “ESKAPE” pathogens and carbapenem-resistant P. aeruginosa (CRPA) is labeled as “Priority 1; Critical” on the WHO’s global priority list of antibiotic-resistant bacteria to guide research, discovery, and development of new antibiotics. ¹ Multidrug-resistant P. aeruginosa isolates are a major cause of nosocomial infections, particularly in immunocompromised patients, leaving very few therapeutic options. Carbapenem resistance in CRPA can be largely attributed to the production of carbapenemases; however, other mechanisms such as permeability defects (e.g., porin OprD loss/modifications) and the over-expression of efflux pumps can also be contributing factors.^2,3 Diverse types of carbapenemases have been identified in P. aeruginosa, but predominantly metallo-ß-lactamases (MBLs). VIM-, NDM-, and IMP-type enzymes are the most common MBL-types in CRPA, with VIM-2 being the most reported variant globally. ⁴ Other non-MBL carbapenemases have been reported in P. aeruginosa such as KPC and GES variants. ⁵ Infections caused by MBL-producing P. aeruginosa (MBL-PA) are particularly challenging since MBLs typically confer resistance to most clinically available ß-lactams, the only exceptions being aztreonam, and to a certain extent, cefiderocol.^2–6 In addition, MBL-PA also often harbor acquired or modified chromosomal genes encoding resistance to other antimicrobial classes such as the aminoglycosides and the fluoroquinolones, rendering most current antipseudomonal drugs ineffective.

Recently, several novel ß-lactam/ß-lactamase inhibitor (BL/BLI) combinations have been developed for the treatment of MBL-PA. These include aztreonam-avibactam and cefepime-taniborbactam, that are currently in phase three clinical trials, and meropenem-xeruborbactam. ⁷ While these new potential treatment options look promising, it is always necessary to use these sparingly to limit and/or slow the inevitable development of resistance, which has been observed with almost all new antimicrobials previously. One of the most recent examples is the dissemination of cefepime-taniborbactam resistant NDM-producing Gram negative bacteria, a BL/BLI that has not yet been approved. ⁸ Subsequently, there is a constant ongoing search for new antipseudomonal drugs, particularly those for the treatment of MBL-PA. In this context, we investigated the potential of dimercaptosuccinic acid (DMSA; generic name succimer), a thiol-containing compound that belongs to the thiol group of molecules suggested as inhibitors of MBLs. ⁹ DMSA is a heavy metal chelator, which has been approved for over 30 years for the treatment of lead and mercury intoxication. MBLs contain zinc ions at their active site, which are a key determinant for their β-lactamase activity. Chelating agents, such as ethylenediaminetetraacetic acid and dipicolinic acid, have long been used in the clinical laboratory settings to identify MBL-production in bacteria, usually through disk potentiation tests. However, these compounds are not safe for use in human use. In contrast, DMSA is safe in adults and children at relatively high doses (30 mg/kg/day).^10,11 Recently, we found that DMSA restores the activity of MBL-producers using isogenic strains of Escherichia coli expressing the MBL NDM-1, IMP-1, and VIM-2. ¹²

The aim of the study was to conduct an in-vitro analysis of the effect of DMSA in combination with carbapenems against carbapenemase-producing P. aeruginosa, both with and without defects in the outer membrane porin OprD. This was pursued due to the noted differences in the permeability coefficients of carbapenems between P. aeruginosa and E. coli, and the specificity of the OprD porin in allowing the penetration of carbapenems in P. aeruginosa only. ¹³

Materials and Methods

Results and Discussion

MICs of DMSA alone were tested for all isolates, both clinical and isogenic, against concentrations of 1.5, 3, and 4.5 mM. 3 mM was determined to be the highest concentration at which DMSA did not significantly inhibit bacterial growth when used alone. Interestingly, most strains, both isogenic and clinical isolates, did not grow with a DMSA concentration of 4.5 mM without any addition of antibiotics, indicating the potential antimicrobial activity of this compound itself at high concentration. Therefore, further testing was performed using DMSA at 3 mM in order to test the inhibitor potential of DMSA of MBL activity.

MIC testing of the recombinant strain PA14 expressing different carbapenemases (GES-5, KPC-2, IMP-1, NDM-1, VIM-2, AIM-1, and SPM-1) showed that the addition of 3 mM DMSA to both IMI and MER against MBL-producers resulted in 4–32-fold decreases of MICs of IMI and 8–32 decrease of MICs values MER (Table 1). This resulted in FICIs ranging from 0.59–0.78 (IMI) and 0.59–0.66 (MER) for MBL-producers (an additive effect) and 1.03–1.53 (both IMI and MER) for non-MBL-producers (an indifferent effect). Interestingly, the most significant decreases were observed for the clones producing the MBLs AIM-1, and SPM-1. However, this effect was largely negated for the PA14ΔoprD strains against IMI/DMSA with fold decreases for the IMP-1, NDM-1, and VIM-2-producing strains becoming insignificant (1–2-fold), resulting in the FICIs increasing significantly. This was not surprising since OprD is the primary pathway for IMI entering the P. aeruginosa cell. ¹⁶ However, MER/DMSA remained effective against PA14ΔoprD (≥4-fold reduction) against all tested MBLs underlining the minor role of OprD in the entry of MER. As expected, there was no change of MIC values of carbapenems for strains producing the non-MBL carbapenemases, KPC-2, and GES-5.

Within the 59 clinical isolates tested, all exhibited non-susceptibility to IMI according to the MICs values (59/59; 100%), and most were non-susceptible to MER (50/59; 85%) (Supplementary Table S1). Within the MBL-producers (n = 42), the addition of 3 mM DMSA resulted in significant decreases in MICs to IMI and MER in 24/42 (57%) and 27/42 (64%) of isolates respectively. Within the isolates that did not exhibit significant MIC decreases, most were OprD deficient (17/18 for IMI and 13/15 for MER). Interestingly, most of the strains that were not affected by DMSA were members of ST773 (11/18 and 12/18 isolates respectively), a strain lineage that does not have a functional OprD, explaining the lack of efficacy. Within the 17 non-MBL-producing isolates tested, none exhibited significant decreases in IMI MICs in the presence of DMSA although three isolates showed a 4-fold MIC decrease to MER. These results suggest, as indicated by the isogenic clone testing results, that the effect of DMSA is largely MBL-specific.

Time-kill assays were performed using strain PA14/VIM-2 to assess the effect of DMSA and MER under the following conditions; DMSA alone, MER alone, DMSA in combination with 4 mg/L MER (0.5 × MIC), and DMSA in combination with 8 mg/L MER (1 × MIC) (Fig. 1). A significant growth deficit could be observed using MER/DMSA at both half of and 1x the MIC value when compared with MER alone, indicating the inhibitory efficacy of this drug.

OPEN IN VIEWER

FIG. 1.

Kill-curve performed on strain PA14/pUCPVIM-2 in the presence no antimicrobials, 3 mM DMSA alone, 4 and 8 mg/L MER alone, and 3 mM DMSA/with MER at 4 and 8 mg/L. Data are expressed as the mean and the standard deviation.

The addition of DMSA at 3 mM to both IMI and MER was shown to be the most effective concentration, resulting in a significant decrease in MIC to both IMI and MER in MBL-producing strains, without any OprD modification. The concentration of DMSA at 3 mM, which exhibited a significant inhibitory effect on MBL-PA, was similar to the concentration (6 mM) previously identified for the inhibition of MBL producers in E. coli. This study shows the potential of DMSA or DMSA-like chelator compounds to be developed for use in treating infections caused by MBL-producing bacteria. Such a repurposing strategy may speed up the development of novel antibiotics for multidrug-resistant P. aeruginosa for which solutions are so far scarce. However, the impact of the OprD modification observed suggests that further studies combining DMSA with other type antimicrobials that are not affected by OprD-related permeability defects, are necessary to elucidate the full potential of these compounds. Previously, we found that the beneficial antibacterial effect of adding DMSA could be achieved in vivo with plasma concentrations 1,000 times lower than those used in vitro, using a peritonitis model of infection with MBL-producing in E. coli. ¹² Subsequently further studies in an animal infection model with P. aeruginosa will be necessary.