Search for: All records

Abstract Droplet microfluidics enable high-throughput screening, sequencing, and formulation of biological and chemical systems at the microscale. Such devices are generally fabricated in a soft polymer such as polydimethylsiloxane (PDMS). However, developing design masks for PDMS devices can be a slow and expensive process, requiring an internal cleanroom facility or using an external vendor. Here, we present the first complete droplet-based component library using low-cost rapid prototyping and electrode integration. This fabrication method for droplet microfluidic devices costs less than $12 per device and a full design-build-test cycle can be completed within a day. Discrete microfluidic components for droplet generation, re-injection, picoinjection, anchoring, fluorescence sensing, and sorting were built and characterized. These devices are biocompatible, low-cost, and high-throughput. To show its ability to perform multistep workflows, these components were used to assemble droplet “pixel arrays, where droplets were generated, sensed, sorted, and anchored onto a grid to produce images. 

Abstract Droplet microfluidics enables kHz screening of picoliter samples at a fraction of the cost of other high-throughput approaches. However, generating stable droplets with desired characteristics typically requires labor-intensive empirical optimization of device designs and flow conditions that limit adoption to specialist labs. Here, we compile a comprehensive droplet dataset and use it to train machine learning models capable of accurately predicting device geometries and flow conditions required to generate stable aqueous-in-oil and oil-in-aqueous single and double emulsions from 15 to 250 μm at rates up to 12000 Hz for different fluids commonly used in life sciences. Blind predictions by our models for as-yet-unseen fluids, geometries, and device materials yield accurate results, establishing their generalizability. Finally, we generate an easy-to-use design automation tool that yield droplets within 3 μm (<8%) of the desired diameter, facilitating tailored droplet-based platforms and accelerating their utility in life sciences. 

Abstract Microbial production of cannabinoids promises to provide a consistent, cheaper, and more sustainable supply of these important therapeutic molecules. However, scaling production to compete with traditional plant-based sources is challenging. Our ability to make strain variants greatly exceeds our capacity to screen and identify high producers, creating a bottleneck in metabolic engineering efforts. Here, we present a yeast-based biosensor for detecting microbially produced Δ⁹-tetrahydrocannabinol (THC) to increase throughput and lower the cost of screening. We port five human cannabinoid G protein-coupled receptors (GPCRs) into yeast, showing the cannabinoid type 2 receptor, CB2R, can couple to the yeast pheromone response pathway and report on the concentration of a variety of cannabinoids over a wide dynamic and operational range. We demonstrate that our cannabinoid biosensor can detect THC from microbial cell culture and use this as a tool for measuring relative production yields from a library of Δ⁹-tetrahydrocannabinol acid synthase (THCAS) mutants. 

Abstract Despite the availability of Cas9 variants with varied protospacer-adjacent motif (PAM) compatibilities, some genomic loci—especially those with pyrimidine-rich PAM sequences—remain inaccessible by high-activity Cas9 proteins. Moreover, broadening PAM sequence compatibility through engineering can increase off-target activity. With directed evolution, we generated four Cas9 variants that together enable targeting of most pyrimidine-rich PAM sequences in the human genome. Using phage-assisted noncontinuous evolution and eVOLVER-supported phage-assisted continuous evolution, we evolved Nme2Cas9, a compact Cas9 variant, into variants that recognize single-nucleotide pyrimidine-PAM sequences. We developed a general selection strategy that requires functional editing with fully specified target protospacers and PAMs. We applied this selection to evolve high-activity variants eNme2-T.1, eNme2-T.2, eNme2-C and eNme2-C.NR. Variants eNme2-T.1 and eNme2-T.2 offer access to N₄TN PAM sequences with comparable editing efficiencies as existing variants, while eNme2-C and eNme2-C.NR offer less restrictive PAM requirements, comparable or higher activity in a variety of human cell types and lower off-target activity at N₄CN PAM sequences. 

This work presents two new quality metrics for droplet generation, versatility and stability. 

Microfluidics has developed into a mature field with applications across science and engineering, having particular commercial success in molecular diagnostics, next-generation sequencing, and bench-top analysis. Despite its ubiquity, the complexity of designing and controlling custom microfluidic devices present major barriers to adoption, requiring intuitive knowledge gained from years of experience. If these barriers were overcome, microfluidics could miniaturize biological and chemical research for non-experts through fully-automated platform development and operation. The intuition of microfluidic experts can be captured through machine learning, where complex statistical models are trained for pattern recognition and subsequently used for event prediction. Integration of machine learning with microfluidics could significantly expand its adoption and impact. Here, we present the current state of machine learning for the design and control of microfluidic devices, its possible applications, and current limitations.