Ultrahigh-throughput Enzymatic Screening System Based On Fluorescence Activated Droplet Sorting

As an important method for protein engineering, directed evolution is widelyused to modify enzymes at molecular level. Lack of genuine high-throughput,high-versatility screening methods limits the success of directed evolution.Fluorescence Activated Droplet Sorting (FADS) is one of the most promisingscreening methods with high potential. The core to FADS is the preparation of microreactor, or micro droplets, which reaction happens inside. Due to the technicallimitation in droplet generation, the poor uniformity is always a bottleneck, whichaffects the accuracy and efficiency of FADS.To solve this problem, this thesis focused on exploring two droplet generationmethods, known as membrane-extrusion and microfluidic chips. And their feasibilityfor FADS was investigated, respectively.On one hand, based on the previous study in our laboratory, an “extrusion-FADS”system was deeply optimized and explored. After optimization for emulsificationtimes, the ideal generation process was established, including15.5-20.5for w/odroplets and25.5-30.5for w/o/w droplets, and the uniformity was significantlyimproved. Then this method was compared with homogenizing, which was the mostcommonly used for droplet generation, and the advantages were proved. The statusfor single cell encapsulation in droplets generated by mini-extruder was given asPoisson distribution. To grope for the suitable conditions of enzyme reactions withindroplets, several experiments including different substrate concentrations, differentenzyme concentrations, etc. were implemented. The results showed that this systemcould make a distinction between enzymes with different substrate concentrations, enzyme concentrations, and activities, thereby indicated the feasibility for mutantscreening. Finally, the model screening assay (using esterase AFEST as positiveindividual and pGF101blank as negative one) revealed that the esterase-displayingcells could be enriched out of a large excess of non-active ones (positive: negative=1:1000) in a single round of sorting (with the highest enrichment factor of330).On the other hand, a microfluidic-based experimental platform was initiallyestablished for FADS. Through careful design of the internal channel, and search forappropriate oil phase, highly-uniformed droplets (w/o and o/w) could be generatedusing PDMS chips. Different droplets size could be obtained as well by changing thewidth of orifice. The stability of droplets was tested at37℃and the integralityremained good overnight. A “trick” to prevent the E.coli cells from sedimentationduring encapsulation was also developed by using hydroxyethyl cellulose asthickening agent.In this thesis, facing the target of establishing a high-throughput andhigh-versatility enzymatic screening system, we set up two different droplet systemsby implement several explorations in methodology. Membrane extrusion was simple,flexible and easy to operate, but with relatively large deviation. Microfluidic chip wasaccurate, and being the direction of future development. However, with highcomplexity the system remains under discussion. Practical application of the newmethod, it will greatly shorten the time for directed evolution, significantly reduce thecost required to facilitate enzymatic properties modification.