Methicillin-resistant Staphylococcus aureus
(MRSA) is a Gram-positive pathogen that can cause skin abscess, bloodstream infections, and pneumonia [1
]. Infections associated with MRSA are among the leading hospital-acquired infections [3
] They are associated with high mortality and increased hospital stays that result in a higher cost burden [3
Vancomycin is a glycopeptide able to inhibit cell wall synthesis by binding to the ends of D-Ala-D-Ala moieties of un-crosslinked Lipid II molecules [4
]. Vancomycin is an antibiotic effective at treating Gram-positive multidrug-resistant pathogens, including MRSA [4
]. However, strains such as vancomycin-intermediate Staphylococcus aureus
(VISA, minimal inhibitory concentration (MIC) = 4–8 µg/mL), vancomycin-resistant Staphylococcus aureus
(VRSA, MIC ≥ 16 µg/mL), as well as vancomycin-resistant enterococci (VRE) have emerged [5
During normal cell wall synthesis, penicillin-binding proteins (PBPs) are able to attach to terminal D-Ala-D-Ala moieties of un-crosslinked Lipid II and link them. Vancomycin is able to bind them and thus block PBPs’ attachment and crosslinking. This eventually leads to osmatic stress and bursting of the cell wall making vancomycin a potent bactericidal antibiotic [4
Initially, it had been thought that resistance to vancomycin would be minimal given that it does not target enzymatic cell processes [4
]. However, we now understand that vancomycin resistance is achieved by a group of genes encoding various enzymes and regulatory proteins that alter the original structure of Gram-positive bacterial walls [4
]. These groups of genes are usually referred to as resistance cassettes. For vancomycin, the originally discovered cassette was named the “VanA”-type cassette which is composed of vanHAX
cluster encoding enzymes and vanR
genes that work as a two-component regulation system [12
]. Numerous other resistance cassettes to vancomycin have been discovered and described, each of which includes VanA homologs [4
]. These resistance cassettes encode genes that facilitate the conversion of D-Ala to D-Lac. In addition, other cassettes exist that help replace D-Ala with D-Ser [13
Regardless of the change in the amino acid, the basic mechanism of resistance stays the same. By altering the original composition of Gram-positive Lipid II amino acid D-Ala-D-Ala vancomycin is no longer able to attach to the end of these glycopeptides and thus is unable to inhibit cell wall synthesis, creating bacteria mildly susceptible or resistant to vancomycin [4
]. Given this threat, vancomycin resistance in Gram-positive bacteria poses a great risk to health care systems worldwide. As a last resort, antibiotics such as linezolid and daptomycin are clinically in use [14
]. However, resistance to both drugs has become more prevalent throughout the decades [14
]. Therefore, the development of new antibiotics to combat these drug-resistant bacteria is necessary and in dire need.
We screened ~82,000 small molecules to identify anti-infective agents that block Caenorhabditis elegans
from a MRSA infection [19
]. We identified several bioactive compounds, of which biological activities have been previously determined [20
]. For example, the selective retinoic acid receptor γ (RARγ) agonist CD437 and CD1530 [19
], the selective peroxisome proliferator-activated receptor γ (PPARγ)-agonist nTZDpa [21
], the anti-parasite drug bithionol [22
], and insulin-like growth factor receptor inhibitor PQ401 [23
]. Each show promising antimicrobial potency against multidrug-resistant Gram-positive pathogens and their persister cells. Considering that many hit compounds are excluded for further investigation due to their in vivo inactivity and toxicity, bioactive compound hits in particular have a high potential to become lead compounds because their in vivo efficacy and in vivo toxicity have been previously proven in several animal models [21
]. Therefore, we further investigated other bioactive compound hits.
In this study, we explore another bioactive compound hit N
-[3,5-Bis(trifluoromethyl)phenyl]-5-chloro-2-hydroxybenzamide (IMD0354), previously described as an inhibitor of nuclear factor-kappa B (NF-κB) that works by directly blocking Iκκβ phosphorylation [24
]. IMD0354 is known to have multiple biological activities. For instance, it has shown anti-cancer properties by directly inhibiting cell invasion and viability as well as acting as an adjuvant with other chemotherapy drugs without showing detectable toxicity [25
]. In addition, it has also shown anti-inflammatory properties by blocking NF-κB and subsequent cytokine production [24
]. Recently, it has been shown that IMD0354 potentiates colistin antimicrobial activity against colistin-resistant Acinetobacter baumannii
]. However, the activity of this compound against Gram-positive bacteria and the activity of IMD0354 alone is not known to have antimicrobial potency. Here, we report, for the first time, that IMD0354 is notably potent against Gram-positive multidrug-resistant bacteria VRSA and VRE. We report that IMD0354 inhibits initial VRSA cell attachment and biofilm formation and is able to induce rapid membrane permeabilization at high concentrations of ≥ 4 µg/mL. Furthermore, we demonstrate that IMD0354′s antimicrobial activity is superior to its anti-cancer activity.
Here, we report the novel finding that the kinase inhibitor IMD0354 is able to prolong the life of C. elegans
during a lethal MRSA infection [19
]. We show that IMD0354 inhibits MRSA growth at MIC levels as low as 0.06 µg/mL and can inhibit growth of other multidrug-resistant Gram-positive bacteria including VRSA, VISA, and VRE. At high concentrations, IMD0354 permeabilizes Gram-positive bacterial membranes at concentrations ≥ 4 µg/mL (Figure 6
In the previous report by Barker et al., IMD0354 was found to have no antimicrobial activity on its own against Gram-negative bacteria, which is consistent with our antimicrobial susceptibility test on Gram-negative ESKAPE pathogens (Table 1
). Barker et al. demonstrated that the enhanced potency of colistin by IMD0354 results from its ability to reverse the colistin-resistance modification of lipid A of colistin-resistant bacteria [29
]. Nonetheless, at this time, we cannot find any parallels between the Gram-negative mode of action of IMD0354 given that Gram-positive bacteria do not produce lipid A [47
]. This may explain why we did not find any synergistic effect even with cell wall- or cell membrane-targeting antimicrobial agents, such as vancomycin or daptomycin, against Gram-positive VRS1.
In addition, IMD0354 inhibited the initial cell attachment for biofilm formation in a dose-dependent manner and completely inhibited biofilm formation at sub-MIC levels and above (Figure 7
). It is possible that this phenotype is partially dependent on the antimicrobial activity of IMD0354. However, we find this unlikely given our experimental design and supporting data. For example, based on our killing kinetics assay, we show that IMD0354 has no antibacterial effect at 1 h post drug incubation. In addition, initial cell attachment assays are run using 102
more bacteria than our killing kinetics studies. Therefore, we conclude that the reduction in initial cell attachment is independent of the antimicrobial activity by IMD0354. Biofilms are a significant threat given their high resistance to antibiotic therapy and common target of medical devices [41
]. Biofilm formation and development consists of five stages, the first being cell attachment [42
]. Attachment is controlled via various cell wall-anchoring proteins, including microbial surface components recognizing adhesive matrix molecules (MSCRAMMs) [42
]. Our working hypothesis is that IMD0354 targets these associated genes or proteins and thus results in a decrease in biofilm attachment.
Furthermore, we also found that IMD0354 is a more potent antimicrobial than a cell toxicity agent. Previous studies have shown that IMD0354 exhibits anti-cancer activity by inhibiting cell invasion, viability, as well as acting as an adjuvant with other chemotherapy drugs [25
]. In particular, our studies show that the average LC50
of IMD0354 toward two cancerous cell lines is 1.04 µg/mL, approximately 17× greater than its MIC of 0.06 µg/mL. IMD0354 is known to selectively suppress the proliferation of cancerous cells over normal cells [48
]. For instance, unlike on neoplastic mast cells, it did not affect the proliferation of normal human mast cells at 1 µM (0.4 µg/mL) [48
]. Furthermore, various groups have used IMD0354 in murine in vivo models from 1 mg/kg up to 30 mg/kg over several weeks and have reported no detectable toxicity [24
]. These previous in vitro and in vivo results demonstrate that IMD0354 is relatively non-toxic to normal cells. Consistently, we observed the cytotoxic effect of IMD0354 on cancerous cells at 1 µg/mL (Figure 3
), while it did not cause cytotoxicity to the model animal C. elegans
at ~7 µg/mL (Figure 1
b). It is worth noting the MIC of IMD0354 against the VRSA strain VRS1 is 0.06 µg/mL (Table 2
), which is about one order of magnitude lower than its effective concentration on cancer cell lines. Given these findings, we propose that IMD0354 has greater promise in terms of being repurposed as an antimicrobial rather than an anti-cancer agent.
The antibacterial activity of other anti-cancer drugs, such as mitomycin C and cisplatin, has been validated and these drugs have been proposed as antimicrobial candidates against multidrug-resistant bacteria [49
]. For instance, mitomycin C demonstrates an LC50
of 27 µM (9.03 µg/mL) against HepG2 cancer cells and a MIC ranging between 0.2–15 µg/mL to multiple Gram-negative and Gram-positive bacteria [49
]. Alternatively, cisplatin shows an IC50
of 2 µg/mL against HepG2 cancer cells and a MIC > 50 µg/mL for both Gram-positive and Gram-negative bacteria [50
]. Nonetheless, low-dose administration of cisplatin to septic mice improves their bacterial clearance [52
]. From these studies, we find that certain anti-cancer agents might have antimicrobial activity at concentrations similar to or greater than their anti-cancer activity. In contrast, IMD0354 has an LC50
of 1.1 µg/mL to HepG2 cells and a MIC of 0.06 µg/mL. These data demonstrate that IMD0354 has a greater antimicrobial to anti-cancer activity ratio than both mitomycin C and cisplatin.
Interestingly, the antimicrobial and anti-cancer mechanism of action (MOA) of mitomycin C and cisplatin appear to be similar as they both cross-link mammalian and bacterial cell DNA, thus leading to cell death [49
]. On the other hand, the anti-cancer MOA of IMD0354 has been shown to be both NF-κB dependent and independent [24
]. Given that bacteria have no NF-κB it is reasonable to assume that the antibacterial MOA of IMD0354 is different from its anti-cancer activity. Importantly, this information allows us to speculate that IMD0354 could be a promising lead compound that can be structurally optimized to abate or nullify anti-cancer activity while retaining its antimicrobial properties.
In addition to anti-cancer activity, IMD0354 has other bioactivities. Onai et al. and Sugita et al. used IMD0354 to directly inhibit NF-κB and subsequently target inflammation [24
]. From these in vitro studies, we glean that IMD0354 can significantly inhibit cytokine production at 1 µM (0.4 µg/mL) [24
], nearly six times more than IMD0354′s MIC (0.06 µg/mL). Furthermore, in vivo data from Onai et al. showed a significant reduction in inflammation in a rat myocardial ischemia/reperfusion injury model after treatment with IMD0354. In these studies rats, were treated with either 1 mg/kg, 5 mg/kg, or 10 mg/kg of IMD0354 over 4 weeks. After treatment, it was found that only 5 mg/kg and 10 mg/kg had significant differences in reducing infarction size. We therefore suggest that there is a low concentration window in which even IMD0354 can be administered as an antibiotic with low cross-activation of other bioactivities.
Overall, the continued evolution of antibiotic resistance to last-resort therapeutics such as vancomycin persist as a primary threat. The C. elegans–MRSA high-throughput screening (HTS) system has become an invaluable tool in drug discovery research as this model is unique in its ability to find antimicrobial kinase inhibitors that would otherwise be neglected due to their toxicity, such as NF-κB inhibitors. Moving forward, we find it important to distinguish the structural relationship between NF-κB inhibition and antimicrobial effect. Further research into analogs would be beneficial in advancing our understanding of the mechanism of action of kinase inhibitors, which can illuminate new antimicrobial targets against multidrug-resistant bacteria. In additional, experiments testing the efficacy of IMD0354 as an antimicrobial in a murine in vivo model would be insightful. Given IMD0354′s low MIC, we speculate that there may be a tritiated dose that does not induce significant NF-κB inhibition but is still able to inhibit bacterial growth. However, at this time, these experiments fall outside the focus of our study.