High-Throughput Microfluidic Electroporation for Library Generation

Protein engineering has important applications that range from developing sensors in bacteria to producing antibodies for cancer therapeutics. Re-engineering protein function usually requires many amino acid changes, which results in libraries containing 10¹²-10²⁰ mutants. However, current genetic transformation tools limit bioengineers to making much smaller mutant libraries. This is relevant in applications where continuous transformation is desired such as in the creation of a library of mutants for drug discovery, metabolic engineering, and deep mutational scanning [1, 2]. In this presentation we will show recent efforts using a continuous flow microfluidic electroporation system to accelerate discovery in protein engineering.

Electroporation is a genetic engineering tool that uses pulsed electric fields for intracellular delivery of molecules such as drugs, proteins, DNA, or RNA and is traditionally performed in a single-batch cuvette system. A subset of the authors have developed a novel microfluidic electroporation platform for improving bacterial electrotransformation efficiency and throughput by up to 3 orders of magnitude [3]. The microfluidic system uses a non-uniform constriction to expose bacteria to a hydrodynamically controlled electric field that induces electroporation. The short residence time (~ms) within the constriction and fast flow rates employed (~mL/min) enables electrotransformation of sufficient bacteria to approach the required numbers needed for protein engineering.

In this study our efforts revolve around re-engineering bacterial two-component sensors, modulating the specificity of bacteria’s largest class of natural bio-sensors. We have designed a 7kb DNA construct that codes for Spectinomycin resistance and has a theoretical library size of 1x10¹⁶ mutants. E. coli cells in exponential growth phase (OD₆₀₀ = 0.5) were harvested, washed 3x in 4°C 10% glycerol, and concentrated 200x prior to electrotransformation. The 1 mL samples were transformed in a diverging microfluidic channels at 1.0 mL/min, 1.5 mL/min, and 2.0 mL/min with processing times of 60 s, 45 s, or 30 s, respectively. Preliminary quantification resulted in 5x10⁸ (1.0 mL/min), 2.5x10⁹ (1.5 mL/min), and 3.75x10⁹ (2.0 mL/min) total transformants in comparison to the 2x10⁷ total transformants from a 50 μL cuvette transformation. These results suggest that using a single microfluidic channel results in up to a 10x increase in transformation efficiency with an approximately 10-100x increase in throughput compared to cuvette based electroporation.

We envision this microfluidic platform as a tool to enable synthetic biologists to access the largest natural biosensors set in engineered systems. This study focuses on E. coli but could be easily applied to other microorganisms and adapted to parallel processing in order to get closer to the theoretical library size. Results of this study will broaden the complexity of engineered libraries that can be developed for protein engineering.

1. Smanski, M.J., et al., Functional optimization of gene clusters by combinatorial design and assembly. Nat Biotech, 2014. 32(12): p. 1241-1249.

2. Fowler, D.M. and S. Fields, Deep mutational scanning: a new style of protein science. Nat Meth, 2014. 11(8): p. 801-807.

3. Garcia, P.A., et al., High efficiency hydrodynamic bacterial electrotransformation. Lab Chip, 2017. 17(3): p. 490-500.