Week 3 HW: Lab Automation

Lab Automation

Design links:

http://opentrons-art.rcdonovan.com/?id=14ylm4e9u480k3w

http://opentrons-art.rcdonovan.com/?id=jsx1tq762tv3s6q

Post-Lab Questions

1. Find and describe a published paper that utilizes the Opentrons or an automation tool to achieve novel biological applications.

Malcı, K., Meng, F., Galez, H., Franja Da Silva, A., Caro-Astorga, J., Batt, G. and Ellis, T., 2026. Slowpoke: An Automated Golden Gate Cloning Workflow for Opentrons OT-2 and Flex. ACS Synthetic Biology.

This publication outlines a golden gate cloning workflow automation process with Opentrons that would greatly improve my current labs workflow. I think the automation of such work would increase our output by 1-2 orders of magnitude. I think combining such a tool with more robust golden gate kits for assembling more complex genetic devices would greatly speed up synthetic biology work more broadly.

The paper presents Slowpoke, a low cost, modular automation workflow for Golden Gate cloning on Opentrons OT 2 and Flex, aimed at making routine synthetic biology cloning faster and more reproducible while remaining accessible through open source protocols and a simple graphical interface. It automates key steps of the cloning pipeline, including assembly setup, transformation, plating, and colony PCR, so that users mainly handle plate movements and colony picking rather than complex protocol programming.

Using the MoClo Yeast Toolkit and the SubtiToolKit on both OT 2 and Flex, the authors show that Slowpoke achieves high assembly efficiency in basic transcription unit builds and maintains strong performance in more complex, multi part combinatorial assemblies. By releasing all code, templates, and documentation for community use and adaptation, the work positions Slowpoke as a generalizable platform for robust, everyday Golden Gate cloning in synthetic biology labs.

Slowpoke implements an end to end Golden Gate automation pipeline on Opentrons OT 2 and Flex that covers assembly setup, transformation, plating, and colony PCR, with users mainly intervening for plate transfers and colony picking. A free graphical interface at https://slowpoke.streamlit.app generates robot protocols from simple file uploads, making protocol design largely point and click and enabling non specialists to automate cloning workflows. Using the MoClo Yeast Toolkit on OT 2, the workflow achieved 17 of 17 correct colonies for basic transcription unit assemblies, while on Flex 11 of 12 colonies were correct, demonstrating high efficiency on both platforms. With the SubtiToolKit on OT 2, 8 of 13 colonies were correct for basic assemblies, showing that the same pipeline can be ported across Golden Gate toolkits with good success rates. In a higher throughput test on Flex using YTK compatible parts, 55 out of 57 multi part combinations yielded correct constructs, indicating strong reliability for combinatorial six part assemblies. The authors release all code, templates, and documentation in the Slowpoke GitHub repository, providing an open, modular foundation that laboratories can adapt to their own Golden Gate toolkits and construct designs.

2. Write a description about what you intend to do with automation tools for your final project.

Project Ideas

SOD Sunscreen Evolution Screen In this project, I’m evolving CuZn SOD variants as intracellular “bioactive sunscreens” that protect cells from reactive oxygen species generated during UV exposure. I start from a baseline SOD1 (human or consensus) fused to a trafficking module such as a cell‑penetrating peptide or organelle‑targeting signal, and then build a focused mutant library around catalytic and stability‑relevant residues. I use E. coli as the initial expression and screening chassis, where I can score variants by enzymatic activity and survival under oxidative stress, and later port top candidates into a mammalian context for proof‑of‑concept protection assays.

My experimental workflow begins with designing the SOD library using error‑prone PCR plus targeted mutations near the active site and key stability positions. I clone these variants into a standardized expression vector and express them in 96‑ or 384‑well format. I then quantify SOD activity in lysates using a plate‑based assay that measures inhibition of a superoxide‑driven colorimetric reaction, and I complement that with growth assays under superoxide‑generating compounds like paraquat. By ranking variants across both enzymatic and functional stress assays, I can identify sequence features that enhance activity and robustness in oxidative environments that resemble UV‑induced stress in skin cells.

This project is designed from the ground up for lab automation. I use automated liquid handling to set up PCR and assembly reactions for the library, perform transformations, and distribute transformants into multiwell plates. Induction, lysis, and reagent addition for the SOD activity assay are all scripted and executed in parallel, and a plate reader continuously records absorbance or fluorescence to generate kinetic activity curves. I also integrate an automated incubator/shaker for consistent growth and induction conditions and use simple analysis scripts to automatically flag top‑performing wells for re‑arraying and sequencing. In effect, I’m building a small, closed‑loop directed‑evolution platform for ROS‑protective enzymes.

CO₂ Co-Culture Toggle Project In this project, I’m engineering a cyanobacteria–E. coli co-culture where the phototroph fixes CO₂ and secretes organic carbon that E. coli converts into a storable product such as PHB. I’m designing a bistable genetic toggle in E. coli that switches between a “carbon uptake and storage” state and a “carbon burn” state, controlled by small-molecule inducers or light. The cyanobacterium supplies fixed carbon (for example sucrose or lactate), while E. coli’s state determines whether that carbon is hoarded as product or consumed for growth, letting me explicitly test how dynamic carbon partitioning shapes community stability and overall CO₂ fixation.

Experimentally, I first validate the toggle in E. coli alone, using mutually repressing transcription factors that each drive a fluorescent reporter plus either a “storage” or “burn” enzyme module. I then work on a cyanobacterial partner engineered to secrete a defined carbon source at tunable rates. Finally, I assemble defined co-cultures in microplates or small bioreactors and measure CO₂ uptake, product formation, and community composition across different toggle states. My goal is to build a model platform for programmable carbon flux in microbial consortia that can later be generalized to other substrates and host pairs.

For this project, I’m leaning heavily on lab automation. I use automated DNA assembly to build toggle variants and transporter constructs in 96‑well format, followed by automated colony picking and growth in deep‑well plates. A liquid handler sets up a matrix of co‑culture conditions by varying initial species ratios, inducer concentrations, and light regimes. Plate readers then track OD and fluorescence over time for both partners, and automated sampling feeds into simple assays for secreted metabolites and storage products. The result is a high‑density dataset of co‑culture and induction conditions that lets me map how the toggle controls carbon flux in a fully programmable, automation‑friendly way.

Gut Microbiome for Mood and Health Regulation In this project, I’m engineering a “mood‑ and craving‑aware” probiotic chassis—such as E. coli Nissle or a GRAS Lactobacillus—to modulate multiple neuroactive pathways linked to anxiety, mood, and addiction. One module produces GABA and/or 5‑HTP to support anxiolysis and prosocial mood; a second module modulates dopamine‑related metabolism to smooth extreme reward spikes associated with craving; and a third module expresses a short oxytocin‑mimetic peptide to reinforce bonding and compassionate affect. Each module is controlled by gut‑relevant sensors (for example bile salts, lactate, or inflammatory markers), so the system is most active under physiological states associated with stress, post‑exercise recovery, or post‑use withdrawal.

At the bench, I start by building and characterizing each pathway independently. I place GABA and 5‑HTP biosynthesis pathways under inducible or sensor‑driven promoters and quantify their outputs with metabolite assays. In parallel, I design circuits that either consume dopamine precursors or express mild inhibitors of dopamine biosynthesis, and I validate those using in vitro assays or bioreporters. For the oxytocin‑mimetic module, I fuse a short peptide to an appropriate signal peptide to ensure secretion and verify production and export. Once I’m confident in the individual modules, I combine them in one or more strains using orthogonal sensors and simple logic to activate different outputs under distinct input profiles that approximate real gut states.

Automation is central to how I explore this design space. I use a liquid handler to assemble promoter–RBS–gene combinations for each module and to transform and array the resulting clones. I then set up microtiter plate assays with controlled input conditions such as gradients of pH, lactate, bile salts, and carbon sources. Automated sampling into metabolite assays or bioreporter strains gives me dose–response curves across hundreds of construct–condition pairs. With basic scripting, I automatically identify designs that show high production under “desired” states (for example high lactate and bile salts) and minimal leakiness at baseline. As an extension, I can connect the system to a simple gut‑on‑a‑chip or microfluidic device to automate media switching and simulate feeding, exercise, or withdrawal cycles, turning this into a small but realistic testbed for programmable neuroactive probiotics.