HTGAA 2026 · Homework Week 11 · April 14, 2026
Week 11

Bioproduction & Cloud Labs

A 1,536-well collaborative pixel art experiment, cell-free protein synthesis reagent design, and fluorescent protein optimization — connecting global science to the chemistry of living light.

Reshma Shetty Ronan Donovan Cell-Free Cloud Lab Fluorescent Proteins
Part A · Collective Artwork

The 1,536-Pixel Canvas

Global pixel art canvas at an early stage — approximately 84 pixels placed across a 1536-well plate grid. Visible clusters: green sfGFP forming a letter K shape in upper left, red mRFP1 forming partial heart shapes in lower regions, cyan and blue isolated pixels scattered across the canvas, yellow mKO2 pixels in the center cluster.
Global pixel art canvas · Step 84 · 84 pixels placed · sfGFP (green), mRFP1 (red), Electra2 (blue/purple), mKO2 (yellow), mTurquoise2 (cyan) · Early stage — no mScarlet_I visible yet
My contribution
What I added to the collaborative bioart project

I contributed at least one pixel — likely the isolated blue pixel visible to the left of the heart cluster in the lower-right quadrant of the plate, and possibly one cyan pixel placed in the same early session. I was among the first contributors to the canvas, at a stage (around step 84 out of the final 1,536) when the collective image had no defined shape yet — just scattered early commitments from students around the world trying to figure out what the thing would become. Not appearing in the top contributors list is accurate: one pixel placed early, when the coordinate system still felt abstract and there were no instructions on how to add more than one.

The final canvas distribution at step 84 was: mRFP1 (47 pixels, 56%), sfGFP (18, 21%), Electra2 (8, 10%), mKO2 (6, 7%), mTurquoise2 (5, 6%). The sfGFP cluster in the upper left was already forming the letter K — one of the earliest coordinated contributions. The heart shape in the lower right was still fragmentary.

What I liked
About the collaborative experiment

The most interesting thing about this project is that it works at two levels simultaneously and neither can be separated from the other. At the surface level it is a collective art project — people placing colored pixels to form an image. At the functional level it is a real cell-free protein synthesis experiment: each pixel is a well, each color is a fluorescent protein, and the image that emerges is the actual layout of a 1,536-well plate that will be run in the Cloud Lab. The art is the experiment. That inversion — where aesthetic choices have biochemical consequences and scientific parameters determine what the image looks like — is genuinely unusual and worth preserving.

The early-stage chaos also reflects something real about how collaborative science actually works: people commit early without knowing what the collective result will be, coordination happens through a shared interface rather than through top-down design, and the final image emerges from thousands of individual decisions that no one planned in full. That's not just art. It's how microbiomes work, and how research communities work.

What could be better
Improvements for next year
  • Multi-pixel contribution from the start. The interface only allowed one pixel per session initially, which made coordination difficult and led to isolated contributions that felt disconnected from the emerging image. Allowing 3–5 pixels per student from the beginning would enable more intentional design without losing the emergent quality.
  • Live preview of the plate layout. Seeing the well coordinates alongside the visual canvas would connect the art more explicitly to the science — students would understand that they're not just placing a colored dot but choosing a specific well in a real experimental plate.
  • Closing the loop faster. The fluorescence data from the Cloud Lab run should be overlaid on the original canvas as a follow-up — showing where high expressors landed, where proteins failed to fluoresce, where the biochemistry diverged from the artistic intent. That comparison is the most scientifically interesting result of the whole experiment and it should be visible to everyone who contributed.
Part B · Cell-Free Protein Synthesis

Master Mix Components

Question 1
Role of each component in the cell-free reaction
E. coli Lysate
BL21 (DE3) Star Lysate (includes T7 RNA Polymerase)
The lysate provides the complete biological machinery for transcription and translation: ribosomes, tRNAs, aminoacyl-tRNA synthetases, elongation factors, release factors, and T7 RNA Polymerase. BL21 (DE3) Star is chosen because it carries the DE3 prophage encoding T7 RNAP under lacUV5 control, and the Star mutation eliminates the mRNA-degrading RNase E, extending mRNA half-life and improving protein yield.
Salts / Buffer
Potassium Glutamate · HEPES-KOH pH 7.5 · Magnesium Glutamate · Potassium phosphate monobasic/dibasic
  • Potassium Glutamate — provides K⁺ ions for ribosome function and ionic strength, while glutamate acts as a crowding agent that mimics cytoplasmic conditions and stabilizes macromolecular complexes.
  • HEPES-KOH pH 7.5 — buffers the reaction against pH drift caused by metabolic byproduct accumulation (organic acids from energy metabolism), maintaining the optimal pH for ribosome activity and protein folding.
  • Magnesium Glutamate — provides Mg²⁺, essential for ribosome tertiary structure, RNA polymerase activity, and as cofactor for ATP hydrolysis during translation.
  • Potassium phosphate monobasic/dibasic — provides inorganic phosphate for energy metabolism and secondary buffering, and serves as substrate for phosphate-dependent energy regeneration reactions.
Energy / Nucleotide System
Ribose · Glucose · AMP · CMP · GMP · UMP · Guanine
  • Ribose and Glucose — carbon sources that feed into glycolysis and the pentose phosphate pathway to regenerate ATP and reduce cofactors (NADH, NADPH) continuously during the reaction, extending its productive lifetime beyond simple ATP-bolus approaches.
  • AMP, CMP, GMP, UMP — nucleoside monophosphates that serve as substrates for RNA synthesis during transcription. They are phosphorylated to NTPs by kinases present in the lysate using the energy from the glucose/ribose catabolism.
  • Guanine — a nucleobase (not nucleotide) that is salvaged by the lysate's purine salvage pathway (hypoxanthine-guanine phosphoribosyltransferase, HGPRT) to regenerate GMP, supplementing the GMP directly added and preventing guanosine depletion during long reactions.
Translation Mix
17 Amino Acid Mix · Tyrosine · Cysteine
Amino acids are the building blocks that ribosomes incorporate into the growing polypeptide chain during translation. The 20 canonical amino acids are supplied exogenously because the lysate preparation depletes endogenous pools. Tyrosine and Cysteine are added separately because they are poorly soluble at the concentrations required for efficient translation — Tyrosine precipitates at neutral pH and Cysteine oxidizes rapidly in air — so they require individual preparation and careful handling to maintain effective concentration in the reaction.
Additives
Nicotinamide
Nicotinamide (the amide form of niacin) is a precursor and salvage substrate for NAD⁺ biosynthesis. NAD⁺ is an essential electron carrier for glycolysis and oxidative energy metabolism in the lysate — without continuous NAD⁺ regeneration, the glucose-based energy system stalls and ATP production collapses, halting both transcription and translation. Nicotinamide also inhibits NAD⁺-consuming enzymes (sirtuins, PARPs) that would otherwise deplete the cofactor pool during the reaction.
Backfill
Nuclease-Free Water
Nuclease-free water adjusts the final reaction volume to the target (20 μL in this experiment) after all concentrated components have been added. Its nuclease-free certification is critical — trace RNase or DNase contamination would degrade the mRNA transcribed from the template DNA or the template itself, silencing expression before any protein can accumulate.
Question 2
Main differences between the 1-hour PEP-NTP master mix and the 20-hour NMP-Ribose-Glucose master mix

The two master mixes differ primarily in their energy regeneration strategy, which determines how long the reaction remains productive. The 1-hour PEP-NTP mix uses phosphoenolpyruvate (PEP) as a high-energy phosphate donor — pyruvate kinase regenerates ATP from PEP directly, providing a rapid burst of energy that drives fast, high-yield expression in the first hour. However, PEP is consumed irreversibly and cannot be regenerated from downstream metabolites, so the reaction exhausts its energy supply quickly. The 20-hour NMP-Ribose-Glucose mix replaces PEP with a metabolic energy loop: glucose is catabolized through glycolysis and the pentose phosphate pathway, continuously regenerating ATP and NADH, while ribose feeds NMP phosphorylation to maintain the NTP pool. This slower but self-sustaining energy system allows transcription and translation to continue for 20+ hours at lower instantaneous rates. The practical consequence for fluorescent protein experiments is significant: slow-maturing proteins (mRFP1, mKO2) that need hours to develop their chromophore benefit from the NMP-Ribose-Glucose mix, while the 1-hour PEP-NTP mix favors fast-folding reporters like sfGFP where speed of readout matters more than total accumulated protein.

Bonus — how transcription occurs without GMP in the NTP pool

GMP is not directly usable by RNA polymerase, which requires GTP. However, the lysate contains nucleoside monophosphate kinases (NMKs) that phosphorylate GMP → GDP using ATP, and nucleoside diphosphate kinases (NDKs) that convert GDP → GTP. Separately, the Guanine base added to the master mix is converted to GMP by HGPRT (hypoxanthine-guanine phosphoribosyltransferase) using PRPP, and this GMP then enters the same kinase cascade. The system therefore does not require pre-formed GTP: it builds GTP from the guanine base through the purine salvage pathway, using the ATP generated from glucose catabolism as the phosphate donor at each kinase step.

Part C · Cell-Free Master Mix Design

Six Fluorescent Proteins — Biophysical Properties

Question 1
Key biophysical or functional property of each fluorescent protein in cell-free systems
sfGFP — Superfolder GFP
Folding speed and robustness. sfGFP was engineered with six point mutations that dramatically accelerate barrel closure and chromophore formation, allowing it to fold correctly even when fused to aggregation-prone partners or expressed at sub-optimal temperatures. In cell-free systems this means it produces a fluorescent signal almost in real time after translation begins — making it the ideal reporter for monitoring expression kinetics rather than endpoint fluorescence. Its tolerance to the variable redox environment of the lysate further distinguishes it from wild-type GFP.
mRFP1
Slow chromophore maturation requiring oxidation. mRFP1 belongs to the DsRed family of red fluorescent proteins whose chromophore formation requires a three-step oxidation cascade — dehydration, cyclization, and an additional oxidation step absent in green FPs. In a 36-hour cell-free assay this produces a measurable lag between protein synthesis (which is complete within hours) and fluorescence readout. The reaction must maintain sufficient dissolved oxygen throughout the incubation for chromophore oxidation to proceed fully. In practice, mRFP1 wells may show lower apparent fluorescence at early timepoints than sfGFP wells with equivalent protein yield.
mKO2
pH sensitivity of the chromophore. Kusabira Orange-type proteins contain a phenolate anion in the chromophore whose protonation state determines fluorescence intensity — when protonated (acidic conditions), the chromophore shifts out of its absorbing state and fluorescence is quenched. Cell-free reactions accumulate organic acids (pyruvate, lactate, acetate) as metabolic byproducts during the 36-hour incubation, progressively lowering the pH of the lysate from its initial 7.5. Without active buffering, this acidification can quench mKO2 fluorescence premature before the full protein yield has been translated — making buffer capacity the single most important variable for optimizing mKO2 readout.
mTurquoise2
Exceptionally high quantum yield. mTurquoise2 has a quantum yield of 0.93 — the highest of any monomeric fluorescent protein — meaning it converts 93% of absorbed photons into emitted fluorescence. In cell-free systems where total protein yield per reaction is typically 10–100 μg/mL (lower than in vivo), this high emission efficiency allows reliable fluorescence detection even when template DNA concentration is limiting or when the reaction has partially slowed due to energy depletion. It requires relatively little protein to produce a detectable signal, making it robust in low-yield conditions.
mScarlet-I
Accelerated chromophore maturation. mScarlet-I was engineered specifically to address the maturation lag of earlier red fluorescent proteins — its chromophore forms and oxidizes substantially faster than mRFP1 or mCherry under the same conditions. This is critical in cell-free systems because the reaction's energy supply is finite: proteins synthesized late in a 36-hour reaction may not have time to mature before ATP is exhausted and the reaction stops. mScarlet-I maximizes the fraction of translated protein that converts to fluorescent form before the reaction terminates, directly increasing the fluorescence-per-molecule efficiency of the experiment.
Electra2
Oxygen dependence for chromophore formation. Electra2 is a very rapidly maturing fluorescent protein whose chromophore formation requires molecular oxygen as a substrate for the oxidation step. In 20 μL wells on a 1,536-well plate, oxygen diffusion from the plate surface into the reaction volume is limited by the high surface-area-to-volume ratio and by any seal or evaporation barrier on the plate. If oxygen becomes limiting during the 36-hour incubation — particularly in central wells of the plate where diffusion distances are longest — Electra2 will accumulate translated but non-fluorescent apoprotein despite high mRNA and ribosome activity. Oxygenation strategy (plate sealing, shaking, headspace) is therefore the dominant variable for this protein.
Question 2
Hypothesis — reagent adjustment to maximize mKO2 fluorescence over 36 hours
Hypothesis

Selected protein: mKO2 — chosen because its pH sensitivity makes it the most vulnerable to the progressive acidification that occurs during long cell-free incubations.

Reagent adjustment: Increase HEPES-KOH concentration from the standard 50 mM to 100 mM, and supplement the reaction with 10 mM 3-phosphoglycerate (3-PGA) as a secondary energy regeneration substrate alongside the ribose-glucose system.

Expected effect: Doubling the HEPES buffer capacity directly counteracts the organic acid accumulation that drives pH from 7.5 toward 6.5–6.8 during the 36-hour reaction, maintaining the mKO2 chromophore in its deprotonated, fluorescent state throughout the incubation. The 3-PGA supplement provides an additional phosphate donor for ATP regeneration that extends productive translation past the point where glucose-ribose metabolism typically slows, ensuring that mRNAs produced late in the reaction are still translated and that the chromophore has time to mature fully. The combined intervention addresses both temporal axes: maintaining the chemical environment for fluorescence, and extending the window during which new fluorescent protein accumulates.

Question 3
Custom reagent supplement design — 2 μL for mKO2 wells

No well assignment email was received by the April 24 deadline as a global student at the SynBio USFQ Node. The following custom reagent composition is designed for mKO2 wells, directly implementing the hypothesis from Q2:

ReagentConcentration in supplementFinal conc. in 20 μL rxnRationale
HEPES-KOH pH 7.51 M stock+50 mM (total 100 mM)Doubles buffer capacity to counteract organic acid accumulation over 36 h
3-Phosphoglycerate (3-PGA)500 mM stock10 mMSecondary ATP regeneration substrate extending productive translation window
Nuclease-free waterto 2 μL totalBackfill to reach 2 μL supplement volume

The 2 μL supplement volume is used entirely for the two active reagents — HEPES boost and 3-PGA — with water backfill. No additional amino acids or nucleotides are added because the standard 2X master mix already provides saturating concentrations; the limiting variables for mKO2 are pH stability and energy duration, not substrate availability.

Question 4
Fluorescence data analysis
Global Student Context — Pending

The fluorescence dataset from the Ginkgo Nebula Cloud Lab run had not been released to committed listener students by the time of this submission. Analysis will be completed and added here when the data becomes available through the node. The analytical approach would be: compare RFU (relative fluorescence units) at each timepoint across wells with different custom supplement compositions for the same fluorescent protein, identify whether the HEPES + 3-PGA intervention produces higher endpoint fluorescence or delays the fluorescence plateau relative to control wells, and calculate the fold-improvement over baseline for mKO2 specifically.

The 1,536-pixel canvas at step 84 shows 84 pixels placed across 5 proteins — mRFP1 (56%), sfGFP (21%), Electra2 (10%), mKO2 (7%), mTurquoise2 (6%) — with no mScarlet-I pixels yet placed, consistent with the early-stage canvas captured before the heart and DNA motifs were coordinated.

Part D · Optional Bonus

Build-A-Cloud-Lab

Cloud lab design using the Ginkgo Reconfigurable Automation Carts simulator at racs.rcdonovan.com — a minimal implementation of the Nebula RAC system that allows spatial arrangement of automation cart units.

Ginkgo Nebula RAC cloud lab design — V-shaped arrangement of approximately 40 automation cart units over a jellyfish-textured floor background. Two parallel lanes converge at a central processing hub, forming a symmetrical V pattern. The design mirrors the bilateral symmetry of a synthetic consortium where two processing arms feed into a shared output node.
Ginkgo Nebula RAC simulator · V-shaped layout · ~40 RAC units · Jellyfish floor texture · racs.rcdonovan.com · April 2026
Design description

The layout arranges approximately 40 Reconfigurable Automation Carts into a V-shape — two parallel processing lanes that converge at a central hub. The bilateral symmetry was not planned as a reference to anything, but in retrospect it mirrors the architecture of a two-input AND gate: two independent signal arms feeding into a single output point. The jellyfish floor texture was chosen as a nod to the bioluminescent GFP lineage — Aequorea victoria, the jellyfish from which the original GFP was isolated, being the indirect ancestor of every fluorescent protein used in this week's cell-free experiment.

The simulator did not allow individual instrument removal or addition of specialized modules (lunatic, plate readers) beyond the base RAC units in this implementation. The V topology does represent a valid automation layout: two parallel liquid handling lanes sharing a central storage and incubation hub reduces robot travel distance between steps and allows parallel processing of two plate types simultaneously — consistent with the 1,536-well plate format of the global pixel art experiment.