Semester of Graduation
Master of Science in Chemical Engineering (MSChE)
Cain Department of Chemical Engineering
Molecularly-targeted therapeutics and personalized medicine have dramatically increased the median survival rate of patients suffering from cancer. However, cellular heterogeneity and the personalized nature of cancer have resulted in the limited success of single drug treatments which has led to the use of multiple therapeutic combinations. This has required the development of new analytical methods capable of multiplexed high-throughput screening (HTS) technologies necessary to identify is single or multi-agent therapies are effective in ex vivo samples like liquid biopsies. Droplet microfluidic devices have garnered significant interest to facilitate high-throughput, single cell analysis of heterogeneous populations. However, these devices are still limited in their ability to assess multiple input conditions such as combinations of multiple drugs or different doses of the same drug. Moreover, HTS approaches need to be coupled with automated image analysis metrics capable of rapidly processing raw data and quantifying it in an efficient manner.
The goal of this work is to address these two areas of need by developing a new method to track different inputs in a droplet microfluidic trapping array coupled with automated image analysis of single cell behavior. The first part of this study highlights the use of rare-earth (RE)-doped luminescent nanoparticles (NP) as novel method to track input conditions in droplets in a microfluidic device. The second part of the work deals with the development of an algorithm called FluoroCellTrack to efficiently analyze single cell data from high-throughput experiments in the droplet microfluidic trapping array. The β-hexagonal NaYF4 nanoparticles used for droplet tracking were doped with a rare-earth emitter with unique spectral properties that do not overlap with established fluorophores like GFP and Rhodamine. In this study, we employed europium as the dopants which has a luminescence emission spectrum in the red region upon UV excitation. We demonstrated that the RE-doped nanoparticles are biologically inert and spectrally independent with common fluorophores and fluorescent stains. This work provided a foundation for future applications using the combination of NPs and microfluidics for multiplexed droplet tracking to quantify tumor heterogeneity and assess the effectiveness of combinatorial therapies. To perform HTS of single cells, a Python algorithm (FluoroCellTrack) was developed to: (i) automatically distinguish droplets from cells, (ii) count cells in each droplet, (iii) quantify cell viability, and (iv) identify input conditions using the RE-doped nanoparticles. The performance of FluoroCellTrack was compared to manual image analysis with a difference in intracellular quantification of ~2% coupled with a decrease in analysis time ofquantification, droplet barcoding and biomarker detection.
Vaithiyanathan, Manibarathi, "High-throughput Droplet Barcoding and Automated Image Analysis in Microfluidic Droplet Trapping Array" (2017). LSU Master's Theses. 4363.
Melvin, Adam T.
Available for download on Friday, November 16, 2018