Team:Hong Kong HKU/Background

HKU EVOLVE - Description

Beyond Unicellular: Recombinase circuits to develop monoculture phenotypic heterogeneity with precise ratiometric control

Imagine you can engineer bacteria monoculture, which can work in a societal way. You start from a single colony, inoculating, and it automatically allocates complicated tasks to sub-populations. The population uses division of labor, which increases overall efficiency. You have a high yield product without complicated experimental procedures at the end of the day.

Our project aims to build a cell consortium from mono cultured cells.  
Let it be prokaryotes like bacteria or eukaryotic cells like yeast or mammalian cells. We strive to develop a novel circuit with minimal optimization and introduce stable ratiometric 
control of the consortium phenotypes. To put it in a sentence, we allow cells to reversibly 
differentiate into different phenotypes, in the meantime, maintaining the phenotypic ratio. 

Why do we need the tool? Compartmentalization and separation of complicated pathways were proven to improve yield[1-7]. The trend of separating biotransformation pathways was also demonstrated in nature circuits[8-9] like nitrification[10]. If you engineer a very 
metabolic stringent pathway in one cell system, it is bound to suffer low yield and slow growth. According to modeling, it is favorable to separate high metabolic burden pathways but not necessary for low metabolic burden pathways[11].

However, segregation of pathways in nature settings may not be applicable in industrial 
environments due to a lack of robustness, reproducibility, and intrinsic noise in 
communication modules. Efforts have been made to eliminate these problems with 
intercellular modules [12]. But we wish to test out another approach: To achieve sub-
population ratiometric control via introducing single-cell randomness to achieve a stable probability distribution. Our system works in various settings that do not require specific population density (contrary to quorum sensing (QS) systems) and are resistant to noise and metabolic differences in different phenotypes. 

The project utilizes the lambda red system to introduce single copied genes, pairing up with novel recombinase gene circuits with orthogonal sites to achieve phenotypic diversity. Inspired by the brainbow research[13], we modified their cre-lox systems to 
dividing cells with only inversion circuits. We tested the use of cre mutants (R32V, R32M) to counteract the downside of the wild type, which causes rapid loss of function of circuits. Interviewing with original paper author Dr. Nikolai Eroshenko, supervised by 
Dr. George Church, suggested that the mutant cre might help us achieve our goal, with longer sustainability via incorporating kill switches. 

Pairing up with crRNA arrays with CMR or cas13 systems, we could engineer and display transient phenotypes used in genome knock-out models. Since many metabolic engineering requires genome deletion models[15], the same thing can be achieved via transient knock-out in CRISPR based transcription level regulation for simple multiple gene knockdown, making our system more 
diverse in various applications. 

Multiplexing and Advantages
The system can multiplex to achieve multiple ratio control, with a straightforward design for 2-5+ phenotypes with ratiometric 
regulation. It is also compatible with other gene circuits with high tolerance variation in metabolic burden differences and noise demonstrated with dry lab. Maintaining recombinase circuitry itself is itself low burden [16] and the crRNA array, which is RNA based regulation. With our novel crRNA based biased modules, we could even introduce asymmetry to flipping to originally symmetrical cre-lox inversion. It shows the circuit's design space and multiplexing ability, allowing it to apply to several medical, industrial, and environmental applications.

Pairing with software and database development, users could input easily characterizable parameters like growth rate and 
desired ratio. The software would output an optimal recombinase circuitry catered to the user's gene circuits. The incorporation of multiple circuits into one single strain, also makes industrial 
registration much more manageable[17].