Screen design for Borrelia burgdorferi lipoprotein secretion pathway components using periplasmic protein capture and suppressor mutations

Abstract

Borrelia burgdorferi infection, aka Lyme disease, is one of the most prevalent vector borne illnesses in the world. With an estimated 300,000 infections per year in the United States alone, this pathogen places a large burden on the health care system, as well as causing debilitating symptoms in the individuals who are not adequately treated for the disease in a prompt manner. While the normal life cycle is based upon transmission by ticks from small mammals to large mammals, mainly deer, human infection results in a dead end for Borrelia. The maintenance of this life cycle is a key feature of Borrelia, and understanding its ability to live in two contrasting environments gives researchers a framework for understanding the disease. Genes for bacterial lipoproteins (lipid modified proteins) make up a large portion of Borrelia species’ genomes and play important role in the infection/transmission cycle. While lipoprotein modification and localization is understood in several bacterial species, such as Escherichia coli and Bacillus subtilis, a few key components of the lipoprotein localization pathway are missing in Borrelia. Furthermore, the signals involved in proper lipoprotein sorting are less clear than in other species. Based on earlier studies in the laboratory, there are two proposed pathway components that are missing from a minimally functioning lipoprotein sorting machinery in Borrelia: a lipoprotein chaperone, which stabilizes the newly translocated lipoprotein in the outer membrane, and a lipoprotein “flippase”, which allows lipoproteins to pass through the outer membrane to the surface. To facilitate identification of these two hypothetical pathway components in functional screens, we embarked on two approaches. First, a mutational analysis of the complement binding protein CspA was performed with the aim to generate a subsurface mutant of CspA. CspA is a complement factor H binding protein that renders the bacteria serum-resistant. The resulting subsurface mutant would provide a basis for a functional suppressor screen, selecting for pathway mutants that are capable of transporting the mutant CspA to the surface. Additionally, the analysis would reveal the amino acids involved in localization of this lipoprotein, adding to the existing dataset with other Borrelia lipoproteins. In a parallel secondary approach, two separate binding proteins (Calmodulin and LysM) were setup to bind their respective targets (Calmodulin Binding Peptide and peptidoglycan) within the periplasm of B. burgdorferi, in an effort to stall the localization process. Based on our earlier work with Borrelia surface lipoproteins, a series of deletions within the intrinsically disordered N-terminal CspA tether were generated. Deletions of both the first and second halves of the tether peptide in CspA failed to yield a subsurface phenotype. Starting with a second-half tether deletion, escalating stepwise 4-amino acid deletions in an N-terminal direction were made. Still, no subsurface CspA mutant was recovered, with the first 4 CspA tether amino acids still supporting the lipoprotein’s surface exposure. Additionally tether swapping with a subsurface localizing lipoprotein tether yielded no subsurface CspA. Second, two protein trapping methods were attempted. These were designed to stall the interaction between pathway components and lipoproteins inside of the periplasm. The introduction into the bacterial periplasm of the mammalian binding protein Calmodulin and surface sorting lipoprotein fused to a tag derived from a Calmodulin Binding Peptide yielded no transformants in single or multiple plasmid systems. Plasmids encoding for each of the individual components could be readily transformed separately. Alternatively, introduction of a fusion protein that links a surface red fluorescent protein reporter lipoprotein to a native peptidoglycan binding peptide tag yielded significant filamentous growth as well as overall growth defects in E. coli, and no B. burgdorferi transformants were recovered.