Logo: to the web site of Uppsala University

Bacterial evolution is closely intertwined with our lives. As their hosts, we shape how bacteria evolve by imposing numerous selective pressures during the time bacteria spend in our bodies. As a result, they adapt in various ways to colonize us or infect us better. In this thesis, I present studies aimed to expand the knowledge on the pathoadaptive changes in Klebsiella pneumoniae, which is a bacterial pathogen of critical importance worldwide. 

In Paper I, we present a new 3D-printed device for growing and studying surface-attached bacterial biofilms. The special aim was to increase the ease of use and versatility, and we have used this biofilm device to screen for biofilm capacity, perform experimental evolution and fundamental biofilm analysis in subsequent studies.

In Paper II, we study within-host evolution by analyzing 110 isolates originating from the same multidrug-resistant K. pneumoniae clone that caused an outbreak at Uppsala University Hospital between 2005 and 2010. We whole-genome sequenced these isolates and phenotypically characterized them to show that the clone has undergone extensive changes in individual patients, leading to increased biofilm formation capacity, attenuation of systemic virulence, and altered colonization potential.

In Paper III, we exploit an experimental evolution approach to decipher evolutionary trajectories towards increased biofilm formation. We show how fast this trait can be acquired in different K. pneumoniae strains by a strong convergent evolution, mostly targeting genes involved in capsule, fimbriae, and c-di-GMP-related regulatory pathways. Importantly, this genetic parallelism extends beyond in vitro observations as we find an extensive overlap with clinical outbreak isolates that carry signatures from within-host evolution.

The experimental evolution experiments revealed interesting genetic changes not only in the known structures or pathways but also in completely novel factors. In Paper IV, we explore a previously uncharacterized T6SS effector that is involved in biofilm formation in K. pneumoniae and strongly enhances this phenotype upon acquiring a single and specific point mutation. We demonstrate that the toxin acts as a DNase and that this mutation results in changes at multiple levels, including protein stability, toxicity, and transcriptional profiles, which collectively lead to the formation of biofilms.

Medical device-related biofilms are a major cause of hospital-acquired infections, especially chronic infections. Numerous diverse models to study surface-associated biofilms have been developed; however, their usability varies. Often, a simple method is desired without sacrificing throughput and biological relevance. Here, we present an in-house developed 3D-printed device (FlexiPeg) for biofilm growth, conceptually similar to the Calgary Biofilm device but aimed at increasing ease of use and versatility. Our device is modular with the lid and pegs as separate units, enabling flexible assembly with up- or down-scaling depending on the aims of the study. It also allows easy handling of individual pegs, especially when disruption of biofilm populations is needed for downstream analysis. The pegs can be printed in, or coated with, different materials to create surfaces relevant to the study of interest. We experimentally validated the use of the device by exploring the biofilms formed by clinical strains of Escherichia coli and Klebsiella pneumoniae, commonly associated with device-related infections. The biofilms were characterized by viable cell counts, biomass staining, and scanning electron microscopy (SEM) imaging. We evaluated the effects of different additive manufacturing technologies, 3D printing resins, and coatings with, for example, silicone, to mimic a medical device surface. The biofilms formed on our custom-made pegs could be clearly distinguished based on species or strain across all performed assays, and they corresponded well with observations made in other models and clinical settings, for example, on urinary catheters. Overall, our biofilm device is a robust, easy-to-use, and relevant assay, suitable for a wide range of applications in surface-associated biofilm studies, including materials testing, screening for biofilm formation capacity, and antibiotic susceptibility testing.

Biofilm infections caused by Pseudomonas aeruginosa are frequently treated with ciprofloxacin (CIP); however, resistance rapidly develops. One of the primary resistance mechanisms is the overexpression of the MexCD-OprJ pump due to a mutation in nfxB, encoding the transcriptional repressor of this pump. The aim of this study was to investigate the effect of subinhibitory concentrations of CIP on the occurrence of nfxB mutants in the wild-type PAO1 flow cell biofilm model. For this purpose, we constructed fluorescent reporter strains (PAO1 background) with an mCherry tag for constitutive red fluorescence and chromosomal transcriptional fusion between the P-mexCD promoter and gfp leading to green fluorescence upon mutation of nfxB. We observed a rapid development of nfxB mutants by live confocal laser scanning microscopy (CLSM) imaging of the flow cell biofilm (reaching 80 to 90% of the whole population) when treated with 1/10 minimal biofilm inhibitory concentration of CIP for 24 h and 96 h. Based on the observed developmental stages, we propose that nfxB mutants emerged de novo in the biofilm during CIP treatment from filamentous cells, which might have arisen due to the stress responses induced by CIP. Identical nfxB mutations were found in fluorescent colonies from the same flow cell biofilm, especially in 24-h biofilms, suggesting selection and clonal expansion of the mutants during biofilm growth. Our findings point at the significant role of high-enough antibiotic dosages or appropriate combination therapy to avoid the emergence of resistant mutants in biofilms.