Modifications to lipid A in Acinetobacter baumannii: Potential Biomarkers for Rapid Diagnosis of Colistin Resistance

Abstract

One of the major challenges in the management of A. baumannii infections is the rapid dissemination of multidrug resistant (MDR) strains and diminishing antibiotic choices available to treat this troublesome pathogen. In this context, physicians are turning to the potentially nephrotoxic cationic antimicrobial peptide, Colistin (polymyxin E) to treat patients infected with A. baumannii MDR strains. The emergence of A. baumannii Colistin-resistant clinical isolates has raised concerns among physicians as this antibiotic is used as a salvage therapy. More alarming, is the identification of Colistin-heteroresistant subpopulations in susceptible A. baumannii clinical isolates that may lead to the development of complete resistance to Colistin. Studies assessing reliability of current susceptibility test methods have suggested poor concordance and revealed a further problem: unreliable detection of heteroresistant subpopulations. Thus, rapid and reliable susceptibility testing methods are critical to deliver effective antimicrobial therapies and minimize the risk of failed treatments resulting in adverse clinical outcomes. Mass-spectrometry (MS) assays have proven to be an important tool to identify bacterial species. The MS-based identification method (Bruker Biotyper) is based on proteomic profiling using MALDI-TOF (Matrix Assisted Laser Desorption-Ionization Time-of-Flight), but the inability to distinguish antibiotic resistance profiles represents a significant limitation. Lipid profiles, especially modifications to lipid A, the membrane anchor of lipopolysaccharide can be directly linked to antimicrobial peptide resistance patterns. Using this MS approach to analyze modifications to lipid A has enabled us to determine unique diagnostic biomarkers to rapid identify Colistin-sensitivity or -resistance profiles in laboratory-adapted A. baumannii, including characterization of heterosensitivity. This work was further extended to analyze lipid A from patient samples to validate the clinical utility of this biomarker. The detection of a novel, modified lipid A with ethanolamine at the 4'position and a novel hexosamine addition. Subsequent LC/MS analysis identified the positively charged hexosamine as galactosamine attached at the 1 position of lipid A. Modification of lipid A with these positively charged constituents was concordant with antibiotic susceptibility profiles. Together, these findings clearly demonstrate the feasibility of this type of analysis as a diagnostic test for rapid determination of antimicrobial susceptibility profiles in clinical settings.