Acinetobacter therapeutics: Inhibitor of LPS synthesis

Much to like in a new study published in the journal mBio today. This study (mBio, 2012, 3[5]:e00312-12) by Lin et al. [1] (led by Brad Spellberg of UCLA) discovered that, for the gram-negative pathogenic bacteria Acinetobacter baumannii, notorious for its association with nosocomial ('hospital-associated') infection the world over and highly resistant to antibiotics, shutting down the production of an extracellular antigen released by the bacteria can effectively silence its effects on the host body and enhance its clearance, to the benefit of the host.

Two main reasons why I found this study interesting. First, it falls squarely in line with the hypothesis of Damage Response Framework (DRF) of microbial pathogenesis, put forth by Arturo Casadevall and Liise-anne Pirofski (who is also the editor for the mBio article under discussion), in their seminal Nature Reviews Microbiology paper in 2003 [2] [N.B. COI declaration: I have worked with Liise-anne.] The DRF represents a simple, yet elegant, attempt to synthesize the existing views of infectious disease pathogenesis that are either microbe-centered (i.e. the ability to cause disease - a.k.a. pathogenicity - is an attribute of the microbe, by virtue of its products and/or ability to replicate in the host) or host-centered (i.e. a disease represents a breakdown in the host defence mechanisms, which allows microbes either to cause disease immediately or to colonize niche areas in the host to activate and cause disease later). There are important lacunae in both views taken separately; for instance, a microbe-centric view doesn't explain occasional pathogenicity of ordinarily benign, commensal organisms such as Candida albicans, while a host-centric view doesn't explain disease caused by certain microbes in apparently immune-competent individuals. An example of the latter is nosocomial candidiasis associated with intravascular catheters, prostheses and implanted devices.

Dispensing with complicated functional definitions that abound for microbes, the DRF simply defines a 'pathogen' as a microbe capable of causing damage to the host, and 'virulence' as a relative measure of that capacity, underscoring the concept that "only in a susceptible host is a microorganism a pathogen, and virulence can be expressed." How does this help the understanding of disease and pathogens? According to the DRF, "...outcome of microbial pathogenesis is the result of a host–microorganism interaction, and that the relevant outcome of this interaction is host damage." Although the usefulness of the framework may be somewhat restricted until 'host damage' is quantified, but for the sake of understanding, a clinically manifest disease may be taken as evidence of damage. Most importantly, the DRF could reclassify common pathogens according to whether or not they could cause disease in presence or absence of a strong or weak host immune response. For instance, the fungal pathogen Aspergillus that causes disease in situations of both weak and strong immunity is in a different class (Class 4) than, say, the bacteria Staphylococcus epidermidis that causes disease only in the setting of weak immunity (Class 1), or the gastric pathogen, Helicobacter pylori, that causes damage if the immune responses to it are strong (Class 6) [2].

Consider Acinetobacter baumannii, a ubiquitous coccobacillus (shaped intermediate to spherical and rod-shaped) known to be the world's most prevalent bacterial species present in clinical settings, responsible for a variety of hospital-associated infectious syndromes - many of them life-threatening - such as nosocomial pneumonia (often affecting ICU patients on mechanical ventilation), nosocomial meningitis, sepsis, urinary tract infection, as well as traumatic skin, soft-tissue and bone infection [3]. A. baumannii has the dubious distinction of being one of the few bacterial pathogens that have become resistant to all available FDA-approved antibiotics ('pan drug resistance' or PDR), although majority of the hospital isolates are considered extensively drug resistant (XDR) (i.e., resistant to carbapenems and all other antibiotics except colistin or tigecycline) [1]. The unique ability of A. baumannii to survive in a hospital environment and develop resistance has been attributed to several factors: (i) biofilm formation and resistance to dessication on non-living surfaces; (ii) adhesion, colonization and invasion of human epithelial cells; (iii) considerable repertoire of antibiotic resistance mechanisms that can be promptly up-regulated as required; and (iv) acquisition of foreign genetic material via lateral gene transfer to promote its own survival under antibiotic and host selection pressures [3].

Scanning electron micrograph of a highly magnified cluster of Gram-negative, non-motile Acinetobacter baumannii; Mag-13331x;
Source: CDC's Public Health Image Library Image #6498; Photo Credit: Janice Carr, a CDC microbiologist and electron microscopist.

So how does A. baumannii cause disease? It appears that this bacteria behaves much like the DRF Class 5 or Class 6 pathogens. In the mammalian host, response to A. baumannii infection is in part hinged on the immune-reactive lipopolysaccharide (LPS) antigen that this bacterium expresses on its cell surface. Bacterial LPS (variants of which are elaborated by A. baumannii and various other bacteria) is a known, potent inducer of inflammatory responses, working via the adapter molecule CD14 and the pathogen signature-recognition molecule TLR4 (which I have written about earlier) present on the surface of various immune defence cells of the host. TLR4-dependent LPS signaling processes are involved in release of inflammatory mediators such as Tumor Necrosis Factor (TNF)-α and interleukin (IL)-8 from macrophages (phagocytic cells active in innate immune defence) [1]. Therefore, ordinarily, TLR4-signaling induced by A. baumannii LPS and the subsequent inflammatory response (which calls in reinforcements) are important in protection against the bacterial infection and help clear the microbe from the tissues (via mechanisms such as phagocytosis, in which immune cells internalize and kill the microbe). However, like DRF Class 5 or 6 microbes, A. baumannii has the ability to cause excessive inflammation in the tissues, leading to host damage; for example, in case of ventilator-associated nosocomial infections in immune-competent hosts, the microbe is known to cause severe acute and chronic inflammation of the lung and tissue damage consistent with pneumonia - as also corroborated by animal experiments. At least one recent clinical report has observed a severe inflammatory response, called the Macrophage Activation Syndrome (in which uncontrollably activated macrophages devour other blood cells!), subsequent to A. baumannii infection.

Clearly, a way to control the unmitigated inflammation caused by A. baumannii infection would be beneficial to the host. Conventional wisdom says that reducing the number of bacteria in the tissue by pharmacotherapy, i.e. administration of antibiotics, would also help. This is where A. baumannii throws a spanner into the works by turning XDR, or worse, PDR.

And this is where the study under discussion becomes interesting, by espousing evidence-driven, targeted strategies for therapeutics and drug discovery. The authors had access to XDR A. baumannii strains of two different mouse virulence phenotypes, highly virulent (strain HUMC1) and less virulent (strain ATCC 17989), and to two types of mice, genetically modified or mutated mice that lack TLR4, and their normal 'wild-type' counterparts (with a functional TLR4). Surprisingly, the highly virulent strain, that caused wild-type mice to go into septic shock (with attendant hypothermia; acidemia, i.e. lowering of blood pH; higher levels of various cytokines; apoptotic neutrophils in spleen and lung, consistent with Gram-negative LPS-induced sepsis) and was ultimately lethal, did nothing significant in TLR4-deficient mice, despite being in similar numbers in the tissues (spleen, lungs, kidney) of the two types of mice; there was also no evidence of the microbe invading deep into the tissue. The less virulent strain, similar in number in both mouse types, but overall lower (= better clearance), had no significant effect on these mice.

Therefore, regardless of virulence, A. baumannii didn't actually invade the tissues to cause host damage. So what differentiated between highly virulent (=more damage) and less virulent (=relatively less or no damage) strains? The authors compared a panel of clinical isolates of the bacteria. It didn't appear that LPS density on the microbe was related to strain virulence in vivo, but the authors discovered that the amount of LPS shed during growth of the strains was strongly correlated to the strain's potency to induce TLR4 activation. In other words, the most virulent strains shed the most LPS during growth in vitro (and presumably in vivo, too) and correspondingly caused maximum TLR4 activation.

I hope you've understood the significance of that observation, gentle reader. It seems to indicate that if there is a means to stop the LPS from being shed or produced, the virulence of the A. baumannii strain may perhaps be attenuated. And that's exactly what the authors found out: "Inhibition of LPS biosynthesis did not kill A. baumannii but enhanced opsonophagocytosis and decreased inflammation, resulting in protection of mice from lethal infection. Which means, that shutting down the production of LPS did not kill the bacteria in the traditional way (the way, say, an antibiotic works), but reduced the excessive inflammation in the host, giving a chance to host immune cells, such as neutrophils and macrophages, to destroy the bacteria.

So, how is LPS biosynthesis inhibited? A few years ago, a key enzyme - a zinc-dependent metalloamidase - in LPS biosynthesis was identified; the enzyme, produced by an essential, single copy gene, is referred to as 'LpxC', a.k.a. UDP-3-O-(R-3-hydroxymyristoyl)-N-acetylglucosamine deacetylase - charming name, isn't it? - and it catalyzes the committed step of synthesis of lipid A (the lipid component of LPS) in virtually all Gram-negative bacteria [4]. Since then, given the role of LPS in the pathogenicity of many Gram-negative organisms, LpxC has been designated a good target for developing therapeutics for these group of microbes, and several small molecule inhibitors of this enzyme have been designed and studied. The study by Lin et al. used an investigational LpxC inhibitor that is in advanced preclinical development, called LpxC-1.

LpxC: Image modified for educational purpose from Figure 2 of
Barb AW and Zhou P, Curr Pharm Biotechnol, 2008, 9(1): 9-15.

LpxC inhibitors have been around for a short while and have been studied by many investigators. But the beauty of this study by Lin et al. lies in (a) connecting an essential bacterial component, LPS, to virulence in a novel way, i.e., via quantity released; and (b) recognizing the therapeutic importance of LpxC inhibitors in reducing host immunopathogenesis ('over-reaction by the host body') as well as bacterial virulence - even in absence of any bacteria-killing ('bactericidal') activity - by simply lowering LPS quantity. This study also opens the door to other related and important investigations; as the authors conclude:

The molecular genetics and structure of LPS that result in greater shedding by the more-virulent strains merits investigation, since elucidating these factors should result in novel targets for therapeutic intervention.

---

Sources:

1. Lin Lin, Brandon Tan, Paul Pantapalangkoor, Tiffany Ho, Beverlie Baquir, Andrew Tomaras, Justin I. Montgomery, Usa Reilly, Elsa G. Barbacci, Kristine Hujer, Robert A. Bonomo, Lucia Fernandez, Robert E. W. Hancock, Mark D. Adams, Samuel W. French, Virgil S. Buslon, and Brad Spellberg (2012). Inhibition of LpxC Protects Mice from Resistant Acinetobacter baumannii by Modulating Inflammation and Enhancing Phagocytosis mBio, 3 (5) DOI: 10.1128/mBio.00312-12
2. Casadevall A, and Pirofski LA (2003). The damage-response framework of microbial pathogenesis. Nature reviews. Microbiology, 1 (1), 17-24 PMID: 15040176
3. Cerqueira GM, and Peleg AY (2011). Insights into Acinetobacter baumannii pathogenicity. IUBMB life, 63 (12), 1055-60 PMID: 21989983
4. Barb AW, and Zhou P (2008). Mechanism and inhibition of LpxC: an essential zinc-dependent deacetylase of bacterial lipid A synthesis. Current pharmaceutical biotechnology, 9 (1), 9-15 PMID: 18289052