ABSTRACT
Clinical outcomes in invasive gram-negative bacterial infections are determined by the interplay of patient,
treatment, and bacterial variables. Significant progress has been made in understanding patient and treatment
factors; however, little is known about how bacterial genetic variation influences patient outcomes. This is an
important gap in understanding microbial pathogenesis and limits the ability to identify drug targets critical in
human pathogenesis. To address this problem, the long-term goal is to interrogate bacterial pathogenesis in
humans by defining the impact of bacterial genetic variation on the outcome of patients with gram-negative
bacteremia. This knowledge can be leveraged to identify novel treatments. The career development plan
reflects this long-term goal as the emphasis is to develop quantitative and lab expertise to independently
identify and explore associations between bacterial genetics and patient outcomes. The objective in this
proposal is to characterize two putative pathogenicity islands (PPI), variably present in E. coli, that are
promising for their roles in pathogenesis and potential as drug targets. The two PPI were identified by the
applicant and encode genes homologous to a type III secretion system (T3SS) structural apparatus (PPI-1)
and translocases/adhesin (PPI-2). PPI-1/PPI-2 are not present in typical laboratory strains of E. coli and so
have not been significantly studied. Presence of PPI-1/PPI-2 associated with increased mortality in patients
with E. coli bacteremia, and deletions in PPI-1/PPI-2 allowed for complement-mediated killing of E. coli in
serum and impaired E. coli-host cell interactions (decreased invasion). Further, PPI-1/PPI-2 functioned as a
prognostic biomarker that improved ability to identify patients at high risk of mortality. The central hypothesis is
that PPI-1 and PPI-2 function together as a T3SS that mediates complement-mediated serum resistance and
tissue invasiveness to influence patient outcomes, and presence of PPI-1/PPI-2 provides prognostic value in
patients with E. coli BSI. This will be tested through two specific aims (SA): 1) Determine how PPI-1 and PPI-2
affect virulence to influence patient outcome, and 2) Define value of PPI-1 and PPI-2 genotype as prognostic
biomarkers for E. coli bacteremia. SA1 will identify how PPI-1/PPI-2 mediates resistance to complement
activation, verify that PPI-1/PPI-2 is a functional T3SS, and determine the relative contributions of serum killing
versus T3SS function to virulence. SA2 will use an existing external validation cohort to show that incorporating
bacterial genetics with clinical variables improves prognostic models of clinical outcomes in E. coli bacteremia.
Characterization of PPI-1/PPI-2 will increase our understanding of E. coli pathogenesis and pave the way for
novel therapeutics. For example, a protein in PPI-2 is homologous to a drug target that the applicant and
others previously exploited in Pseudomonas aeruginosa. The skills and mentoring acquired during this award
will scaffold development into an independent clinician-scientist focused on the critical area of gram-negative
infections.