Abstract: Engineering microbial hosts to express plant natural product (NP) pathways is a sustainable and scalable method of producing high-value commodity chemicals, including food additives and pharmaceuticals. While Saccharomyces cerevisiae is a suitable host for expression of plant alkaloid enzymes and production of target NPs, generating these products at sufficiently high titer for industrial application remains challenging due to enzymatic bottlenecks unique to the metabolic environment of the native host. Advances in synthetic biology have given researchers powerful tools to generate massive bio-enzymatic mutant or genome-editing libraries that can be screened for production-enhancing targets. However, optimization of secondary product output is hindered by a lack of technology to isolate strains with production-enhancing mutations at a rate comparable with strain library generation.
We applied the PicoShell microbial screening platform as a method to directly interrogate intracellular metabolite production at several orders of magnitude higher rates than current colony picking techniques – hundreds of thousands of clones per hour. PicoShells are hollow, polymeric microparticles with porous outer shells that function as picolitre-scale culture chambers for proliferating monoclonal microbial colonies. The shells are made from biologically inert, functionalizable polymers that encapsulate hundreds of single cells per second. We utilized a naturally fluorescent terpenoid as a reporter molecule for the desired production pathway in our yeast. The output of in vivo fluorescent probes or pathway-related fluorescent metabolites is often not detectable at single-cell levels. Single-cell expansion into hundred-to-thousand-cell colonies within PicoShells amplifies the fluorescent signal by orders of magnitude, enabling clones to be selected through fluorescence activated cell sorting (FACS). Large scale cytometry runs are of particular use when measuring protein expression of heterologous cells in response to an environmental stimuli or for screening large populations of cells for a particular phenotype.
In a single sort workflow, we performed multiplexed colony selection using growth and metabolite production phenotypes as selection probes. We were able to sort >250,000 colonies in a single FACS run. PicoShell amplification enables visualization of the metabolite at two-three orders of magnitude above the fluorescence of a single cell. Using only intracellular serpentine accumulation, we differentiated colonies with different genotypes and different production profiles. We then selected subpopulations of differently producing yeast colonies, released and recultured them, resulting in populations enriched for a desired set of phenotypic traits. On average, we see an order of magnitude increase in fluorescence between encapsulate colonies producing serpentine and control, “background” colonies. Through repeated selection of high-producing colonies, PicoShells can also be used for accelerated “continuous evolution” to enhance target metabolite yields for the purpose of sustainable microbial biomanufacturing.
.png)
.png)
.png)
