Skip to main content

Novel microfluidic based sample preparation methods for rapid separation and detection of viable bacteria from blood for sepsis diagnostics

Time: Fri 2021-12-03 10.00

Location: Air Fire, Science for Life laboratory, Tomtebodavägen 23A, Solna

Language: English

Subject area: Biotechnology

Doctoral student: Sharath Narayana Iyengar , Nanobioteknologi, Proteinvetenskap, Science for Life Laboratory, SciLifeLab, Nanobiotechnology

Opponent: Prof. Jonas Tegenfeldt, Lund university

Supervisor: Prof. Aman Russom, Science for Life Laboratory, SciLifeLab, Nanobioteknologi

QC 2021-11-10


Sepsis is a serious medical condition characterized by a whole-body inflammatory response caused by bloodstream infection. The final stage of sepsis can lead to septic shock, multiple organ failure, and death. In early sepsis, the concentration of bacteria in the bloodstream is typically low, making diagnosis challenging. Rapid diagnosis of sepsis is crucial as there is an exponential increase in mortality for every hour delay in the appropriate antibiotics administration. Common culture-based methods fail in fast bacteria determination as it takes up to 24-72 hr. On the other hand, recent rapid nucleic acid-based diagnostic methods are prone to false positives from human DNA mainly due to a lack of efficient sample preparation methods.

 This Ph.D. work was aimed at the development of novel sample preparation methods for rapid and efficient separation and identification of bacteria from  blood for sepsis diagnostics. To address this, two different approaches were explored. In the first approach, a label-free, size-based, passive elasto-inertial microfluidics (visco-elastic flows) method was developed (Paper I and II). Initially, behavior of particles were studied in solution containing polyethylene oxide (PEO) using different spiral designs (Paper I). By using the knowledge from paper I, a spiral design was used to preposition the particles at the outer wall of the inlet using PEO as sheath and we showed that a particle of a certain size remains fully focused at the outer wall throughout the channel length. The optimized parameters were extended to demonstrate that when bacteria is spiked into diluted blood, blood remains fully focused at the outer wall throughout the channel length while smaller bacteria differentially migrate towards the inner wall for rapid separation. Using E.coli spiked into the diluted blood sample, bacteria separation is demonstrated at an efficiency of 82 to 90% depending on the blood dilution using a single spiral chip (Paper-II). The second approach (Paper III) involves a selective cell lysis method where lysis buffer composition is optimized to selectively lyse blood cells in 5 min while maintaining bacterial viability. The lysed blood cells were filtered through a filter paper to capture viable bacteria. The captured bacteria on the filter paper was detected using Prussian blue (PB)  colorimetric analysis. In PB color-based assay, viable bacteria metabolically reduce iron (III) complexes, initiating a photo-catalytic cascade toward PB formation on the filter paper visible to the naked eye. Using this approach it was possible to detect bacteria by the naked eye. This approach was also further optimized to perform antibiotic susceptibility testing to determine the minimum inhibitory concentration (MIC). 

Furthermore, as a step towards rapid genomic analysis, a novel method combining ITP-RCA (Isotachophoresis – Rolling Circle Amplification) was studied and optimized for real-time amplification (RCA), focusing and detection of bacterial DNA in a microfluidic channel (Paper IV). In this study we demonstrate rapid and increased sensitivity of bacterial DNA detection. This method has a huge potential to accelerate the time needed for DNA based analysis for infectious diseases.

 All in all, the ability of these sample preparation methods for rapid and effective separation and detection of key pathogens in blood will help in decreasing the time of sepsis diagnosis and aid towards efficient phenotypic or genotypic analysis.