Generating Isogenic Deletions (Knockouts) in Francisella tularensis, a Highly-infectious and Fastidious Gram-negative Bacterium.

Generating bacterial gene deletion mutants, also known as knockouts (KOs), is a powerful tool to investigate individual gene functions. However, fastidious bacteria such as Francisella tularensis (F. tularensis) often are difficult to genetically manipulate. Indeed, many different approaches have been tested to generate F. tularensis mutants. First, Tn5-based EZ::TN transposons have been successfully used to generate transposon libraries in F. tularensis (Qin and Mann, 2006; Weiss et al., 2007). However, creating a comprehensive transposon library with saturating mutations can be laborious, screening for gene disruption requires high-throughput assays where known phenotypes can be measured, and transposons may not completely inactivate the gene of interest or may alter downstream gene expression. Second, group II introns (also referred to as Targetron) have been used to inactivate F. tularensis genes of interest (Rodriguez et al., 2008; Rodriguez et al., 2009). Targetron functions by forming a complex between plasmid-encoded RNA and chromosomal DNA, followed by group II intron insertion into the gene of interest. The main advantage of Targetron is that it does not require an antibiotic resistance marker. However, as noted for transposons, targetron gene insertions may not eliminate all gene functions or may affect downstream gene expression. Third, homologous recombination can be used to completely replace the chromosomal target gene with a selectable marker, such as an antibiotic resistance marker. This classical genetic technique has been used in many F. tularensis studies (Ramakrishnan et al., 2008; Ren et al., 2014; Mohapatra et al., 2008; Robertson et al., 2013). To accomplish this, a suicide plasmid is engineered to include a selectable marker flanked by regions upstream and downstream of the gene of interest. This KO plasmid can be delivered into host bacteria by many methods, including electroporation, chemical transformation, or conjugation. Here, we describe an optimized procedure to generate KO plasmid constructs, use E. coli to conjugatively transfer the plasmid into F. tularensis, select for F. tularensis KOs using a series of kanamycin-, hygromycin-, and sucrose-resistance steps, and confirm that the gene of interest has been deleted (general overview of the knockout protocol diagramed in Figure 1). This optimized procedure is relatively simple, rapid, and, more importantly, includes a series of both positive and negative selection steps to increase the chances of deleting a target gene from F. tularensis.