Indiana University-Purdue University Indianapolis (IUPUI) / Haemophilus ducreyi resists the cytotoxic effects of human antimicrobial peptides (APs), including α-defensins, β-defensins, and the cathelicidin LL-37. Resistance to LL-37, mediated by the sensitive to antimicrobial peptide (Sap) transporter, is required for H. ducreyi virulence in humans. Cationic APs are attracted to the negatively charged bacterial cell surface. In other gram-negative bacteria, modification of lipopolysaccharide or lipooligosaccharide (LOS) by the addition of positively charged moieties, such as phosphoethanolamine (PEA), confers AP resistance by means of electrostatic repulsion. H. ducreyi LOS has PEA modifications at two sites, and we identified three genes (lptA, ptdA, and ptdB) in H. ducreyi with homology to a family of bacterial PEA transferases. We generated non-polar, unmarked mutants with deletions in one, two, or all three putative PEA transferase genes. Mutants with deletions in two PEA transferase genes were significantly more susceptible to β-defensins, and the triple mutant was significantly more susceptible to both α- and β-defensins, but not LL-37; complementation of all three genes restored parental levels of AP resistance. Deletion of all three PEA transferase genes also resulted in a significant increase in the negativity of the mutant cell surface, suggesting these three genes contribute to the addition of positively charged moieties on the cell surface. Mass spectrometric analysis revealed that LptA was required for PEA modification of lipid A; PtdtA and PtdB did not affect PEA modification of LOS. In human inoculation experiments, the triple mutant was as virulent as its parent strain. While this is the first identified mechanism of resistance to α-defensins in H. ducreyi, our in vivo data suggest that resistance to cathelicidin may be more important than defensin resistance to H. ducreyi pathogenesis.

Most gram-negative organisms require lipopolysaccharide and its membrane anchor, lipid A, for growth and survival. Also known as endotoxin, lipid A is synthesized via a nine-step enzymatic process, culminating in a conserved hexa-acylated, bis-phosphorylated disaccharide of glucosamine. This framework is often altered by condition- or species-specific lipid A modifications, which change the biochemical properties of the molecule in response to and to defend against environmental stress signals. Here, we expound on two stories in different gram-negative organisms, both involving novel or unanticipated biochemistry that impacts lipid A structure. First, the missing acyltransferase in the Epsilonproteobacterium Helicobacter pylori lipid A biosynthesis pathway is identified. This enzyme transfers a secondary acyl chain to the 3'-linked primary acyl chain of lipid A like E. coli LpxM, but shares almost no sequence similarity with the E. coli acyltransferase. It is reannotated as LpxJ and demonstrated to possess an unprecedented ability to act before the 2'-secondary acyltransferase, LpxL, as well as the 3-deoxy-D-manno-octulosonic acid transferase, KdtA. LpxJ is one member of a large class of acyltransferases found in a diverse range of organisms that lack an E. coli LpxM homolog, suggesting that LpxJ participates in lipid A biosynthesis in place of an LpxM homolog. The second story focuses on regulation of modifications to endotoxin structure that occur after the conserved biosynthesis pathway. E. coli pmrD is shown to be required for PmrAB-dependent lipid A modifications in conditions that exclusively activate PhoPQ; this result proves that PmrD connects PhoPQ and PmrAB despite previous reports that it is an inactive connector in this organism. Further, RNA sequencing and polymyxin B survival assays solidify the role of E. coli pmrD in influencing expression of pmrA and its target genes and promoting survival during exposure to cationic antimicrobial peptides. Notably, the presence of an unknown factor or system capable of activating pmrD to promote lipid A modification in the absence of the PhoPQ system is also revealed. In all, the findings presented here expand our understanding of alternative approaches to lipid A biosynthesis and the complex systems that regulate modifications of this dynamic molecule.

The growing problem of antibiotic resistance and the lack of promising prospective antibiotics have forced us to search for new classes of antibiotics. Among the candidates to develop into future antibacterials are antimicrobial peptides (AMPs). These potent, broad spectrum compounds are important components of innate immunity of organism from all kingdoms of life. One such family of mini-proteins from plants is called cyclotides, whose members are defines by cyclic backbone and a cystine knot (CCK), which confers to them extreme stability in the face of biological, chemical and physical insults.     Some cyclotides possess Gram-negative specific antibacterial activity; the purpose of this thesis was to characterize how these molecules kill bacteria, and how bacteria would respond to treatment with cyclotides. For this purpose, Salmonella enterica and Escherichia coli mutants resistant to the cyclotides cycloviolacin O2 and cycloviolacin O19, respectively, were selected. These mutants were characterized by whole genome sequencing, genetic reconstitution, fitness measurements, and cross-resistance studies. These studies identified a number of genetic pathways for resistance development to cyclotides. These mutants displayed variable fitness profiles in laboratory growth media and in mice competition experiments, with some mutants possessing a fitness advantage in mice. Cross-resistance studies resulted in the identification of several cases of cross-resistance and collateral sensitivity between cyclotides and other AMPs/antibiotics.      Antimicrobial effects of cyclotides were assayed in different conditions and in bacterial organisms with different surface characteristics. In addition, immunolocalization experiments were performed to explore the biological distribution of cyclotides in plants and to determine the mechanism of action of cyclotides in bacteria, respectively. Antibodies raised against cyO2 were used for this purpose. Immunohistochemical techniques applied to plant cells, tissues and organs provided the information that cyclotides were distributed in all plant organs, and were found in tissues vulnerable to pathogen attack, and that cyclotides were stored in the vacuoles of plant cells. Immunogold staining of cyclotide treated cells of S. typhimurium, showed effects of cyclotide treatment on the cell envelope components as well as cytoplasm. A higher number of cyclotide molecules was associated with the cell envelope, but a considerable fraction of them penetrated into the cytoplasm.