Date of Award


Document Type


Degree Name

Doctor of Philosophy (PhD)


Pharmaceutical Sciences


Microbiology, Immunology, and Biochemistry

Research Advisor

Jason W. Rosch, PhD


Elizabeth A. Fitzpatrick, PhD Richard E. Lee, PhD Elaine I. Tuomanen, MD Joshua Wolf, PhD, MBBS


antibiotic resistance, antibiotic tolerance, heteroresistance, Streptococcus pneumoniae


Streptococcus pneumoniae is a prominent human pathogen that causes both invasive and non-invasive diseases, such as otitis media, pneumonia, meningitis, and bacteremia. Although it is frequently an asymptomatic colonizer of the human nasopharynx, S. pneumoniae is a major cause of morbidity and mortality in the immune compromised population, young children, and the elderly. Up until the 1970s, S. pneumoniae was susceptible to almost all antibiotics. Since then, this pathogen has gained resistance to a variety of antibiotic treatments, including beta-lactams, macrolides, and fluoroquinolones.

In the first chapter, we focused on fluoroquinolone resistance in S. pneumoniae. Fluoroquinolones are one of the most frequently prescribed antibiotics, yet fluoroquinolone resistance in S. pneumoniae is still rare compared to other antibiotics resistance, such as beta-lactams. In this study, we investigated the mechanism(s) underlying this intriguing case by assessing the efficiency and fitness costs of horizontal transfer of fluoroquinolone resistance determinants. We hypothesized that the fitness tradeoffs incurred by resistance determinants would define the likelihood of such resistance to emerge in a clinical setting. Clinically relevant fluoroquinolone resistance requires both on-target mutations in topoisomerase IV parC and DNA gyrase gyrA. The wild-type S. pneumoniae TIGR4 was not readily transformed with single mutations in gyrA or parC; however, it was readily transformed with double on-target mutations in gyrA and parC. Compared to the wild type, the single on-target mutants were attenuated, whereas the double on-target mutant was virulent. This suggests that clinically relevant, high-level fluoroquinolone resistance requires the combination of several on-target mutations, which could be acquired via horizontal transfer. The combination of the extremely low probability of acquiring two or more mutations simultaneously from different target genes and the deleterious fitness tradeoffs imposed by individual on-target mutations in gyrA or parC likely result in the infrequent prevalence of fluoroquinolone resistance in S. pneumoniae. Through in vitro serial passaging, we identified a novel mutation (N291D) in the efflux pump patA that facilitated the acquisition of the on-target mutations in parC and gyrA via horizontal transfer with minimal fitness tradeoffs. We also modeled the evolution of fluoroquinolone resistance in a murine host and identified mutation(s) that arose and fixated during in vivo passaging. Interestingly, the experimentally-evolved isolates from the in vivo passaging study did not encode on-target mutations for fluoroquinolone resistance and instead displayed tolerance, which potentially facilitated the subsequent acquisition of fluoroquinolone resistance.

In the next chapter, we investigated how fitness tradeoffs and horizontal transfer play a role in the emergence and spread of another mainstay of treatment of pneumococcal infection, beta-lactams- specifically, penicillin, which inhibit wall synthesis. We found that recombination with related viridans species via horizontal transfer may be preferable to de novo on-target mutations in penicillin-binding proteins in S. pneumoniae to acquire resistance more rapidly without initially losing in vivo fitness. Initial recombinants retained virulence in vivo and could readily acquire higher resistance via subsequent transformation. The final recombinants displayed tolerance to penicillin, having reduced kill kinetics compared to the wild type. This suggests that S. pneumoniae might have minimized fitness tradeoffs by developing tolerance via horizonal transfer with related viridans group streptococci, which would serve as a stepping stone for subsequent development of resistance.

In the next study, we explored an underlying mechanism of antibiotic tolerance in S. pneumoniae. In our model of the evolution of antibiotic resistance, rny that encodes ribonuclease Y (RNAse Y) was a mutational hotspot across multiple antibiotics. The rny knockout mutant was fully virulent, indicating that deletion of this gene imposed minimal to no fitness tradeoffs. Disruptions in RNA degradation resulted in tolerance to several classes of antibiotics and reduced antibiotic treatment efficacy in vivo.

In the final chapter, we investigated whether other phenomena that allow bacteria to withstand antibiotic killing, such as heteroresistance, can affect antibiotic treatment outcomes clinically. We found that vancomycin heteroresistance is associated with treatment failure and poor outcomes in coagulase-negative staphylococci (CoNS) from pediatric leukemia patients.

Taken together, this dissertation provides insights into strategies of S. pneumoniae for striking a balance between maximizing resistance potential while minimizing fitness tradeoffs, thereby potentially contributing to the development of more-effective antibiotics for treatment of pneumococcal disease. It also provided insights into the association between heteroresistance in CoNS and clinical outcomes.

Declaration of Authorship

Declaration of Authorship is included in the supplemental files.




DaoDOA.pdf (326 kB)
Declaration of Authorship