Novel tools for the study and diagnosis of bacterial infections : exploring the intersection of microbiology and microfabrication

The alarming increase in antibiotic resistance calls for new approaches to study and diagnose bacterial infections. In my thesis, I worked at the intersection of microbiology, microfabrication, and optical probes to develop novel diagnostic and screening assays for microbiology and elucidate bacterial pathogenesis in urinary tract infections (UTI).

Conventional antibiotic susceptibility testing (AST) has an 18 h turnaround time and low throughput. Papers I & II describe the development and validation of the nanowell AST (nwAST). This method was based on the nanowell slide (nwSlide), which is a miniaturized nanotiter well plate featuring 672 nanowells of 500 nl each. Owing to this large number of wells, we designed a quantitative AST that determines a minimum inhibitory concentration for up to 6 antibiotics. Validation with 70 UPEC clinical isolates showed a 97.9 % overall categorical agreement with agar disc diffusion and a turnaround time between 3 h 40 min - 8 h 10 min. Key to this short turnaround time was the implementation of the Tlag algorithm, which identified the exact time point when bacteria started to grow. This analysis delivered results up to 5 times faster compared to conventional AST. Overall, our high throughput nwAST surpassed FDA’s requirement of > 90 % categorical agreement.

The agar plate and 96-well plate offer a limited resolution for bacterial phenotypic screening. To improve screening resolution, in paper III we took advantage of the nwSlide’s versatility and combined it with fluorescence-activated cell sorting. We developed a rapid workflow to directly select and single-sort bacterial mutants in individual nanowells, based on a fluorescence-encoding transposon. Sorted mutants were phenotypically screened for growth, morphology, and metabolism already during the first incubation on the nanowell slide using spectrophotometry, algorithmic analysis and microscopy. Selected phenotypes were retrieved and screened with single-primer PCR and sequencing to identify the transposon insertion site. By leveraging the versatility of the nwSlide, we developed a high-resolution screening platform, with higher throughput and less reagent consumption compared to the agar plate and 96-well plate.

In paper IV, we developed a proximal tubule-on-a-chip (PToC) based on a microfluidic device. We delineated UPEC's adhesion to renal epithelial cells under shear stress with temporal and single-cell resolution. We demonstrated that only a minority of cells adhered and withstood shear stress for > 30 min. This binding was PapG-independent. Adherent bacteria divided rapidly and eventually formed microcolonies, which were mediated by FimH adhesin. Microcolonies expanded colonization beyond the cell surface, enhanced infection spread, and extended bacterial binding under shear stress from minutes to hours. Although the absence of PapG and FimH delayed infection, UPEC eventually colonized renal cells causing them to round up and slough off. These results showed that UPEC has a repertoire of redundant adhesion organelles that help bacteria to withstand shear stress in the urinary tract.

Detection of biofilm is absent in clinical diagnostics, despite its association with antibiotic tolerance. In paper V, we developed a rapid diagnostic assay for biofilm-associated UTI called optotracing. This assay is based on heptamer formyl thiophene acetic acid, a luminescent conjugated oligothiophene. This molecule produces a unique spectral signature upon binding to cellulose, which is an extracellular component in biofilms of UPEC. We first optimized optotracing’s performance in PBS and healthy urine spiked with UPEC biofilm or purified cellulose. Next, we developed a workflow to isolate and screen urine sediment for cellulose within 45 min. Optotracing of 182 urine samples from UTI patients and interpretation of results with principal component analysis and k-means clustering identified 27 urine samples as positive for cellulose. This result provided the first direct evidence of biofilm formation in UTI. With a short turnaround time and minimum equipment requirement, this diagnostic assay could guide clinicians when choosing antibiotics for biofilm-associated infections.

List of scientific papers

I. Weibull E, Antypas H, Kjäll P, Brauner A, Andersson-Svahn H, Richter-Dahlfors A. Bacterial nanoscale cultures for phenotypic multiplexed antibiotic susceptibility testing. Journal of Clinical Microbiology. 2014;52(9):3310-3317.
https://doi.org/10.1128/JCM.01161-14

II. Veses-Garcia M, Antypas H, Löffler S, Brauner A, Andersson-Svahn H, Richter-Dahlfors A. Rapid phenotypic antibiotic susceptibility testing of uropathogens using optical signal analysis on the nanowell slide. Frontiers in Microbiology. 2018;9:1530.
https://doi.org/10.3389/fmicb.2018.01530

III. Antypas H, Veses-Garcia M, Weibull E, Andersson-Svahn H, Richter-Dahlfors A. A universal platform for selection and high-resolution phenotypic screening of bacterial mutants using the nanowell slide. Lab on a Chip. 2018;18(12):1767-1777.
https://doi.org/10.1039/c8lc00190a

IV. Antypas H & Richter-Dahlfors A. Single-cell studies of uropathogenic Escherichia coli fitness determinants during colonization of epithelial cells under shear stress. [Manuscript]

V. Antypas H, Choong FX, Libberton B, Brauner A, Richter-Dahlfors A. Rapid diagnostic assay for detection of cellulose in urine as biomarker for biofilm-related urinary tract infections. [Submitted]