Steve C. C. Shih and Christopher Moraes

Process for realizing new biological systems follows:
* design - selecting biological parts and pathways
* built - using genomic modification tools to generate a library of strain variants
* test - assessing the performance of constructed strains
* learn - evaluating results to see if the design was successful or needs more improvement

Design Phase
Developing initial specifications for a system to achieve a certain task

Two critical design decisions:
* the host has an impact on the functional performance of system, so deciding what type of cell you want to transplant DNA into
* deciding which biological pathway to engineer for you to meet a specific pathway
    * TinkerCell and gro are two software tools that help picking pathways

Build Phase

Assembling individual gene fragments (i.e. parts) into plasmids and genomes (i.e. bioparts)

Two methods for assembly:
* one-pot Golden Gate
* Gibson assembly
* in vivo in yeast

Usually favour in vivo since it doesn’t need expensive enzymes to join DNA together, and can have much bigger final products (up to 100 kb).

Yehezkel et al. built a digital microfluidic platform in a cell-free system.
* the more parts you added to the system, Gibson assembly did worse (more non-specific construction products) than their cell-free method
* Yehezkel’s methods were high-fidelity and were able to achieve an error rate of 1/450
* a bottleneck in the build process is in cloning —> the group went with in vitro cloning using single molecule PCR
    * to get single molecule PCR, diluted the assembly product iteratively until one target DNA molecule per droplet was obtained
* demonstrated:
    * reduction in costs by 50x
    * reduction in time through cell-free cloning
    * improvement in parallelization of assembly reaction with automation

Although using cell-free DNA assembly and in vitro cloning have helped with some parts of the build process, usually the engineering is done within cells.

Introducing DNA into the host (transformation) is really hard. Some traditional methods that don’t have automated/standardized/robust tools:
* electroporation
* heat-shock treatment

Some microfluidic systems have been applied to electroporation and heat shock, but don’t integrate between transformation, plating colonies and selecting for single clones.

On-chip cell culture is important for getting successful transformation b/c it gives time for cells to product antibiotic resistance proteins.

Test Phase
* need more large-scale analysis for engineered organisms
* current methods like microscopy and mass spec are too slow
* need real-time evaluation of implemented designs

Fiore et al. created a microfluidic control loop that automatically read an output and modulated input such that output reached preset level. Used 3 different control algorithms:
* to make cells emit a constant fluorescence level - set-point control
* time-varying: signal-tracking

Could potentially integrate “test" and “learn” phases into one platform.

* one-pot Golden Gate
* Gibson assembly
* in vivo yeast DNA synthesis
* single molecule PCR