Discovery of Antibacterial Activity Targeting Fermenting Staphylococcus aureus
By: Carol A. Rouzer, VICB Communications
Published: May 7, 2013
The search for activators of heme biosynthesis reveals a compound that suppresses glycolysis and is toxic to fermenting bacteria.
Staphylococcus aureus poses one of the most challenging current threats to world-wide public health (Figure 1). Highly metabolically adaptable, S. aureus is capable of infecting almost any organ in the body, and the emergence of methicillin-resistant strains (MRSA), which are resistant to nearly all antibiotics, markedly amplifies the danger. The urgent need for new approaches to treat S. aureus infections led Vanderbilt Institute of Chemical Biology (VICB) members Eric Skaar, Gary Sulikowski, and Richard Caprioli to investigate the antibacterial action of novel compounds that activate S. aureus heme biosynthesis [L. A. Mike, et al. (2013) Proc. Natl. Acad. Sci. U.S.A., published online April 29, doi:10.1073.pnas.1303674110].
Figure 1. (Top) Scanning electron micrograph of S. aureus bacteria (yellow) and human white blood cells (blue). Image reproduced from Wikimedia Commons. Source: NIAID, public domain. (Bottom) Cutaneous abscess caused by methicillin-resistant S. aureus. Image reproduced from Wikimedia Commons. Source: Bruno Coignard and Jeff Hageman, CDC, public domain.
Key to S. aureus’s adaptability to multiple host environments is its ability to shift from respiration to fermentation for energy generation. Respiration, which yields the highest amount of energy (ATP) per molecule of fuel (e.g. glucose) through oxidative phosphorylation, is the preferred pathway when environmental conditions allow. Among the requirements for respiration are heme, a critical cofactor for many respiratory enzymes, menaquinone, an electron carrier, and oxygen or other suitable terminal electron acceptor. If any of these components is missing, the bacteria switch to fermentation via the glycolytic pathway, which produces ATP by substrate level phosphorylation. A switch to fermentation allows the bacteria to thrive under the anaerobic conditions found in many host tissues and in abscesses, which are the hallmark of S. aureus infection.
As noted above, heme is required for respiration, and S. aureus has evolved a transport system that acquires heme from the host as well as the enzymatic machinery necessary to synthesize its own heme. However, since excessive heme is toxic, S. aureus has also evolved a heme efflux pump HrtAB. Expression of HrtAB increases in the presence of high concentrations of heme through the regulatory activity of the HssRS two component system, comprising HssS, a heme sensor-kinase, and the HssR-response regulator. The mechanism by which HssS senses heme is a long-standing interest in the Skaar lab, leading the investigators to search for small molecule activators of HssS to probe its function.
To detect HssS activation, the Skaar group constructed a plasmid (phrt.lux) bearing a luciferase gene fused to the hrtAB promoter. Exposure of bacteria transfected with this plasmid to HssS activators leads to increased luciferase expression, easily detected by the resulting blue luminescence (Figure 2). The plasmid was used in a high-throughput screen (HTS) of the 160,000 compound VICB library in a search for activators of HssS. A secondary screen using a plasmid bearing the xylE gene (coding for catechol oxygenase) fused to the hrtAB promoter, provided an enzymatic readout to exclude compounds that caused nonspecific luminescence in the first screen. Compounds active in the XylE screen were tested for their ability to induce tolerance to heme toxicity in S. aureus through the induction of HrtAB efflux pump expression. The results yielded compound VU0038882 (‘882) as the most potent inducer of HrtAB expression (Figure 3).
Figure 2. The high-throughput screen for activators of HssRS used S. aureus transfected with phrt.lux, a plasmid carrying the gene for luciferase driven by the hrtAB promoter. Wild-type (WT) bacteria transfected with this plasmid expressed luciferase, leading to a bright blue luminescence when exposed to heme or other activators of HssRS. No luminescence was observed in the absence of heme, or in WT bacteria transfected with control plasmid (pXen-1) or in bacteria genetically deficient in HssR transfected with phrt.lux (ΔhssR phrt.lux). Image reproduced with permission from L. A. Mike, et al. (2013) Proc. Natl. Acad. Sci. U.S.A., published online April 29, doi:10.1073.pnas.1303674110, copyright 2013, L.A. Mike, et al.
Figure 3. Structures of ‘882 and ‘373.
The Skaar group hypothesized that ‘882 increased the expression of HrtAB through a direct interaction with the HssS-sensor kinase. They tested this hypothesis by exploring the effects of mutating key HssS residues on ‘882 activity. To their surprise, they found that each of the three mutations tested had the same effect on induction of HrtAB expression regardless of whether the inducing molecule was ‘882 or heme. This suggested that the two structurally different molecules were interacting with HssS in exactly the same way, a highly implausible conclusion. As an alternative, the investigators hypothesized that ‘882 might induce HrtAB expression by stimulating bacterial heme biosynthesis. In this scenario, the synthesized heme would then be the direct activator of HssS. Consistent with this hypothesis was their finding that ‘882 was not active in the XylE reporter assay if the bacteria used for the assay were genetically deficient for heme biosynthesis. Furthermore, exposure of wild-type S. aureus to ‘882 resulted in an increase in intracellular heme levels, which could be readily observed by the reddish-brown color visible in cell pellets (Figure 4).
Figure 4. Activation of HssRS by ‘882 requires heme synthesis. (A) WT S. aureus transfected with a plasmid linking xlyE expression to the hrtAB promoter (phrt.xylE) express XylE enzymatic activity in response to increasing concentrations of ‘882. No increase in XylE activity is seen in WT bacteria expressing a control plasmid (plgt.xylE) or in bacteria genetically deficient in heme synthesis (ΔhemB) expressing either plasmid. (B) Exposure of WT S. aureus to ‘882 causes a concentration-dependent increase in intracellular heme, which can be observed visibly as a reddish-brown color in cell pellets. Image reproduced with permission from L. A. Mike, et al. (2013) Proc. Natl. Acad. Sci. U.S.A., published online April 29, doi:10.1073.pnas.1303674110, copyright 2013, L.A. Mike, et al.
The finding that ‘882 stimulates heme synthesis led the Skaar lab to further investigate the compound’s effects on S. aureus metabolism. They discovered that ‘882 suppresses the glycolytic pathway in the bacteria and that inhibition of glycolysis, either by 2-deoxyglucose or by genetic mutation of 6-phosphofructokinase, antagonized ‘882’s effects on heme biosynthesis. These results suggested a tight association between heme biosynthesis and central metabolism.
The ability of ‘882 to suppress glycolysis suggested that it might be toxic to fermenting bacteria, which use this pathway for energy generation. The Skaar lab found that this was indeed the case, as ‘882 suppressed the growth of wild-type S. aureus incubated under anaerobic conditions, while having little effect on bacteria cultured aerobically. Bacteria genetically deficient in heme or menaquinone biosynthesis, which rely heavily on fermentation, were particularly sensitive to ‘882 toxicity, even when cultured in the presence of oxygen. These results demonstrated that ‘882’s antibacterial effects were due primarily to inhibition of central metabolism rather than to induction of heme biosynthesis.
Exposure of S. aureus to aminoglycoside antibiotics, such as gentamicin, leads to the development of small colony variants (SCVs), a slowly growing antibiotic-resistant form of the bacteria. SCVs can form in vivo, particularly in patients suffering from chronic S. aureus infections that have been extensively treated with antibiotics. SCVs are obligate fermenters, suggesting that they should be susceptible to ‘882 toxicity, and this proved to be correct. In fact, ‘882 markedly augmented the antibacterial efficacy of gentamicin, and prevented the development of gentamicin resistance in cultured S. aureus.
One of the first lines of defense against bacterial infection is the innate immune response. Early in this response, neutrophils are drawn to the infection site where they secrete large amounts of reactive oxygen and nitrogen species, which are directly toxic to bacteria. The toxicity of these agents depends, at least in part, on their ability to inactivate respiratory enzymes. Thus, bacteria may survive the onslaught by switching to fermentation for their energy needs. The Skaar lab investigators found that exposure to ‘882 augments the toxicity of nitric oxide and increased bacterial killing by neutrophils. These results are consistent with ‘882’s toxicity to fermenting bacteria.
As noted above, abscess formation is a characteristic of S. aureus infection, and the bacteria’s ability to survive in the anaerobic environment found in abscesses is attributed to its use of fermentation for energy production. The Skaar group hypothesized that ‘882 should be particularly useful in treating S. aureus-derived abscesses. However, the furan ring of ‘882 rendered it pharmacokinetically unsuitable for in vivo experiments. To solve this problem, the Sulikowski lab synthesized ‘882 analogs in search of a molecule that would retain its antibacterial activity in vivo. The result of their efforts was VU0420373 (‘373, Figure 3). Although not quite as potent as ‘882 in the in vitro assays, ‘373 exhibited good metabolic stability making it suitable for in vivo testing.
Treatment of mice systemically infected with S. aureus with ‘373 resulted in a one-log reduction in bacterial count in the liver and a significant decrease in the number of liver abscesses (Figure 5). Imaging mass spectrometry of liver slices from ‘373-treated mice performed in the Caprioli lab confirmed the dose-dependent presence of unaltered ‘373 in the tissue, which was further verified by liquid chromatography-mass spectrometric assay of liver homogenates.
Figure 5. In vivo efficacy of ‘373. Treatment of mice systemically infected with S. aureus using ‘373 caused a significant 1 log reduction in the number of bacteria in the liver (A) and decrease in liver abscesses (B). Liver slices from mice treated with varying amounts of ‘373 (C) were imaged by MALDI-MS (D), revealing the dose-dependent presence of the compound. Image reproduced with permission from L. A. Mike, et al. (2013) Proc. Natl. Acad. Sci. U.S.A., published online April 29, doi:10.1073.pnas.1303674110, copyright 2013, L.A. Mike, et al.
Together the results demonstrate that ‘882 and its analog ‘373 display promising antibacterial activity directed towards fermenting bacteria. Targeting central metabolic pathways is a novel approach to antimicrobial agent discovery, and these preliminary findings suggest that agents such as ‘882 may offer new ways to fight the growing problem of antibacterial resistance. Exactly how ‘882 and ‘373 work to suppress glycolysis remains a mystery; however, the Skaar lab is aggressively pursuing some interesting clues and expects to have an answer soon.