Collective Artwork

The 1,536 Pixel Canvas

Contribution

1. Drawing a bacteriophage from the concentric circle: the attempt was hacked by “competitor 01” who was able to submit more than one pixel per minute.
2. Drawing some sort of deconstructed time-lapse diagram of the transcription/translation process including DNA/RNA strands, elongating chain of amino acids and a giant polymerase: this attempt was slowed down by “competitor 02” who was trying to finish a duck pattern that I did not identify at first.
3. Drawing a giant pair of scissors in reference to restriction enzymes, but also lesbian visibility :) The pattern design used the already present ducks, concentric circles and parallel lines. The upper scissor hole was adapted into a “spitting bacteriophage” and the geometrical design of “competitor 03” into a DNA helix passing through the lower scissor hole. The design also respected patterns already present such as LOVE, O and heart.

Impressions

A fun, satisfying, and addictive project that also reveals some very interesting layers.

My experience felt more competitive than collaborative, with a general tendency to push through one’s own idea vs. letting things develop collectively into a creative direction (e.g. exquisite corpse concept).

Direct communication with my “competitors” (via DMs on the forum) helped orientate the creative flux but despite the frustration, I found the dynamic of the “territorial tensions” more interesting.

The experience made me curious about how the collected data might inform topics such as patterns (spacial colonization, spontaneously emerging pixel creativity…), competitive dynamics, and flocking behaviors. It also made me wonder whether - and how - collective intelligence can manifest in a society which is more driven by productivity than observation.

https://forum.htgaa.org/t/global-pixel-artwork-cooperation-guidelines/559/16?u=2026a-flo-razoux

Cell-Free Reagents

1. Composition and role of the different reagents

BL21 (DE3) Star Lysate is a ready-to-use cell extract from E. coli that contains all the machinery needed to carry out transcription and translation processes, such as ribosomes, tRNAs, and enzymes etc. The star refers to a mutation that reduces the activity of the RNase E and thus, protects the system from mRNA degradation. This lysate also contains a special enzyme, the T7 RNA polymerase. This enzyme recognizes the T7 promoter and thus enables it to focus the transcription on target genes in a more efficient way than E. coli’s own polymerase.

Salts and buffers

Potassium Glutamate (KGlu) provides potassium ions (K⁺) that help ribosomes and enzymes work properly. It helps stabilize proteins and RNA and keeps the system balanced. The use of glutamate-based instead of Cl-based salts is gentler and less harmful to enzymes.

HEPES-KOH (pH 7.5) act as a buffer agent, meaning it keeps the pH stable (typically around 7.5). This prevents the acidification of the system and is crucial for the enzymes to keep functioning well while the reaction is happening.

Magnesium Glutamate (MgGlu2) supplies magnesium ions (Mg²⁺), which are essential for many reactions in the system. Notably, it helps ribosomes and enzymes to work properly and stay stable. It is also needed for energy-related reactions like using ATP.

Potassium Phosphate (Monobasic, KH2PO4 & Dibasic, K2HPO4) helps maintain a stable pH (it can be considered as a secondary buffer system). It provides inorganic phosphate (Pi), which is needed to regenerate ATP and thus, supports the ongoing production of energy during the reaction.

Energy / Nucleotide System

Ribose (C5H10O5) is the main energy source. It is involved in the creation of ATP and it also provides building blocks (C ring) to rebuild nucleotides. Glucose also generate ATP but it mainly supports long-lasting energy supply.

ATP, CTP, GTP and UTP are the basic building blocks for making RNA. Their high-energy level powers the transcription process but AMP, CMP, GMP and UMP can also be used to reduce costs and increase the stability of the system. Remark: In the NMP-ribose setup though, GMP is not added directly but through the conversion of guanine (a base molecule).

Translation Mix (Amino Acids) The 17 amino acids provide most of the building blocks needed to build the proteins. The tyrosine and the cysteine are added separately because they need a special preparation (the tyrosine doesn’t dissolve well at normal pH and the cysteine oxidizes easily).

Additives

Nicotinamide (NAM) serves as a precursor for the biosynthesis of nicotinamide adenine dinucleotide (NAD+) and its reduced form (NADH). NAD+/NADH is essential for the energy regeneration (ATP) of the system.

Backfill

Nuclease Free Water is ultra-purified water that is used to fill up the reaction to the right volume and doesn’t contain enzymes that can digest DNA nor RNA.

2. PEP-NTP vs NMP-Ribose-Glucose Master Mixes

The PEP/NTP mix (1-hour incubation) gives the system “ready-to-use” energy and building blocks (high-energy NTPs and PEP with additional boosters), so it can produce proteins very fast but only for a short time (1 hour).

In contrast, the NMP-Ribose mix (20-hour incubation) provides low-energy precursors that the system can slowly metabolize and regenerate, enabling longer-lasting (20 hours), more cost-efficient protein production.

3. Documentation

Main references: https://pmc.ncbi.nlm.nih.gov/articles/PMC6481089/ HW Week 09

AI support Gemini: as a starting base, prompt: “role of reagent X in cell-free system” ChatGPT: proofreading/summary

Master Mix Design

1. Properties of the fluorescent proteins

In cell-free systems, each fluorescent protein differs in how quickly it folds and matures, how bright its signal is, and how sensitive it is to conditions like oxygen or pH.

sfGFP matures very quickly (~13.6 minutes) and folds reliably even under difficult conditions or when fused to other proteins, giving a fast and strong fluorescence readout that closely tracks protein production. However, it still requires oxygen for chromophore formation, which can limit fluorescence in low-oxygen environments.

mRFP1 has a slow maturation time (~60 minutes) and low yield, so its fluorescence appears late and remains relatively dim compared to newer variants. As a result, in short cell-free reactions much of the protein may be present but not yet fluorescent, leading to delayed and weaker signal readout.

mKO2 is sensitive to both oxygen and pH, so its fluorescence decreases in low-oxygen or acidic conditions commonly found in cell-free reactions. It also matures relatively slowly (~108 minutes), leading to delayed signal appearance after protein production.

mTurquoise2 has an exceptionally high quantum yield (~0.93), making it very bright and enabling strong fluorescence even at low expression levels. Its low pKa (~3.1) also makes it highly stable to pH changes, so it maintains a consistent signal in cell-free reactions.

mScarlet is very bright (high quantum yield and extinction coefficient), giving a strong fluorescence signal even at low expression levels. However, its slow maturation (~174 minutes) delays signal appearance, making it less suitable for short experiments.

Electra2 achieves strong fluorescence due to its high brightness (~61, quantum yield ~0.76), making it easier to detect than many blue fluorescent proteins. Unlike most fluorescent proteins, it relies on binding an external chromophore (bilirubin), allowing rapid, oxygen-independent fluorescence if the ligand is supplied.

2. Optimizing Electra2 Fluorescence (36-hour incubation)

Intuitively, I would have suggested to adapt the master mix composition for mKO2 (maintaining a stable pH because the protein is sensitive to acidification) or for mRFP1 (ensuring an efficient metabolic activity and energy regeneration as this protein presents low yield properties). However, when asking Gemini and Claude “which from the following proteins (sfGFP, mRFP1, mKO2, mTurquoise2, mScarlet_I, Electra2) is the one that requires the most critical adjustment of the cell-free mix composition, both designated Electra2. This protein is designed for high brightness and stability, but this superior performance comes at the cost of higher demand on cellular machinery, making its synthesis more sensitive to energy depletion.

Suggested adjustments for the composition mix to sustain the expression and fluorescence of Electra2 for up to 36 hours:

1. Decrease Potassium Phosphate: As phosphate accumulates during long reactions, it inhibits riboflavin kinase, the enzyme needed to produce FMN, which is essential for Electra2 fluorescence. Starting with high phosphate levels worsens this inhibition and reduces FMN production over time. Lowering the initial phosphate concentration helps maintain FMN synthesis and supports stronger Electra2 fluorescence.

2. Increase Magnesium Glutamate: Increasing magnesium glutamate helps maintain sufficient free Mg²⁺ over long reactions, where Mg²⁺ is progressively depleted by binding to nucleotides and their breakdown products. This is important because Mg²⁺ is required both for ribosome function and for stabilizing FMN binding and riboflavin kinase activity needed for Electra2 fluorescence. Starting with more Mg²⁺ prevents loss of enzymatic activity and supports sustained protein synthesis and signal over time.

3. Increase Ribose: Increasing ribose helps sustain FMN biosynthesis because it provides key carbon precursors needed to build the flavin cofactor required for Electra2 fluorescence. Over long reactions, continued protein production creates ongoing demand for FMN, which can deplete available precursors. Supplying more ribose prevents this bottleneck and supports consistent fluorescence over time.

A schematic showing the conversion of riboflavin to flavin mononucleotide (FMN) and flavin adenine dinucleotide (FAD). Image source: https://www.researchgate.net/figure/A-schematic-showing-the-conversion-of-riboflavin-to-flavin-mononucleotide-FMN-and_fig1_361819429

3. Cell-Free Master Mix: Reagents concentrations

Optimization suggested for a Master Mix specific to Electra2:

• Increase MgGlu to 11 mM, ribose to 120 mM, glucose to 12 mM, nicotinamide to 5 mM
• Add GMP (200 µM, if no Guanine)
• Reduce potassium phosphate to 6 mM total
• All other components: unchanged

These changes collectively enhance FMN biosynthesis, stabilize FMN binding, maintain redox balance, and sustain energy metabolism over 36 hours, while minimizing phosphate inhibition and Mg²⁺ depletion, thereby maximizing Electra2 fluorescence.

Intended design for the 8 wells (line Q2-O15 to Q2-O22):

When filling in the composition of the wells in HTGAA: 1536, I turned into an issue: reagent concentrations could be increased but not decreased.

Next step: rethink the compositions with concentration increase only.

4. Documentation

Claude Prompts

“What reagent (or reagents) an expert in molecular biology would suggest to adjust in a cell-free mix composed of Potassium Glutamate, HEPES-KOH pH 7.5, Magnesium Glutamate, Potassium phosphate, Ribose, Glucose, AMP, CMP, GMP, UMP, 17 Amino Acid Mix, Tyrosine, Cysteine and Nicotinamide in order to improve maximize the fluorescence of Electra2 over a 36-hour incubation, and why?”

“Here is the composition of the master-mix used for 20-hours incubation: Potassium Glutamate : 312 mM ; HEPES-KOH pH 7.5 : 45 mM ; Magnesium Glutamate : 7 mM ; Potassium phosphate monobasic : 5.6 mM ; Potassium phosphate dibasic : 5.6 mM ; Ribose : 77.4 mM ; Glucose : 6.9 mM ; AMP : 600 microM ; CMP : 400 microM ; GMP : 0 microM ; UMP : 400 microM ; Guanine : 200 microM ; 17 Amino Acid Mix : 4.10 mM ; Tyrosine pH 12 : 4.10 mM ; Cysteine : 4 mM and Nicotinamide : 3.10 mM. Please adapt the concentrations of each reagent of the master-mix for a 36h-incubation, only if it’s needed and knowing that Guanine: 0 microM in that mix. Please justify any change by providing either publication references or solid argumentation. Version 01: master mix is optimized for all the 6 proteins: sfGFP, mRFP1, mKO2, mTurquoise2, mScarlet_I, Electra2. Version 02: master mix is optimized for Electra2.”

Claude Concentration Values

1. Magnesium Glutamate: 7 mM → 11 mM

Mg²⁺ depletion over 36h justifies increase; 11 mM is safe for Electra2 translation.

Chelation load from NTPs: AMP (600 µM) + CMP (400 µM) + UMP (400 µM) + GMP (200 µM) = 1.6 mM nucleotides at start. Each binds ~1 Mg²⁺. That is 1.6 mM Mg²⁺ immediately chelated → effective free Mg²⁺ at t=0 is already ~9.4 mM at 11 mM total. Phosphate chelation: 6.0 mM total Pi (post-reduction) forms Mg-phosphate. At pH 7.5 and 37°C, Mg-phosphate Ksp is not exceeded at these concentrations Progressive NTP hydrolysis over 36h generates additional AMP/ADP/Pi - further depleting free Mg²⁺. Upper safety limit: Caschera & Noireaux (2015) showed inhibition begins above 12–14 mM in extended CFPS. Ribosome sensitivity: at 11 mM total (effective free ~8–9 mM initially), ribosome fidelity is maintained - Mg²⁺ optimum for E. coli ribosomes is 7–12 mM free (Johansson et al., 2011, Cell)

2. KH₂PO₄: 5.6 → 3.0 mM | K₂HPO₄: 5.6 → 3.0 mM (total Pi: 11.2 → 6.0 mM)

Reducing Pi relieves RFK inhibition; pH buffering remains adequate; energy metabolism unimpaired.

RFK Ki for Pi: reported 2–5 mM (Kambourakis & Rozzell 2004; Bauer et al. 2003). Starting at 6.0 mM Pi places initial [Pi] just above Ki - marginal but better than 11.2 mM However: Pi will accumulate from NTP hydrolysis over 36h regardless. Each NTP hydrolysis cycle releases one Pi. With ~1.6 mM NTP pool being recycled, and assuming ~10–20 turnover cycles over 36h, Pi accumulation could add 5–15 mM on top of starting 6.0 mM. This is a concern but not correctable without adding a Pi scavenger (not in the mix). The reduction to 6.0 mM at least delays RFK inhibition and reduces the inhibition severity at early timepoints - which is when most Electra2 translation and FMN loading occurs. pH buffering check: KH₂PO₄/K₂HPO₄ ratio maintained at 1:1 → pH contribution ~7.2 (phosphate pKa₂ = 7.2). HEPES-KOH at 45 mM (pKa 7.48) provides dominant buffering. Phosphate buffering is dispensable. Potassium contribution: reducing from 2×5.6 mM to 2×3.0 mM reduces K⁺ by ~5.2 mM total. Against 312 mM background KGlu, this is negligible (<2% change in ionic strength)

3. Ribose: 77.4 mM → 120 mM

120 mM is within safe osmotic range and provides flavin precursor support.

Sun et al. (2013, Biotechnol. J.): optimal ribose range 75–120 mM. 120 mM is the stated upper bound - not beyond it Osmolarity check: additional 42.6 mM ribose adds ~43 mOsm. Total solution osmolarity estimate: KGlu 312 mM → ~624 mOsm (dissociates to K⁺ + Glu⁻) Mg Glu 11 mM → ~33 mOsm Phosphates ~12 mOsm Ribose 120 mM → ~120 mOsm Glucose 12 mM → ~12 mOsm NTPs + AA mix ~15 mOsm Total ≈ ~816 mOsm

This is hypertonic relative to standard E. coli cytoplasm (~300 mOsm) but cell-free systems are not membrane-bound and tolerate high osmolarity well. CFPS systems routinely operate at 600–900 mOsm (Pardee et al. 2016). Ribose as flavin precursor: ribose-5-phosphate → DHBP via RibA/RibB in the flavin pathway. The increased substrate availability directly supports FMN biosynthesis flux.

4. Glucose: 6.9 mM → 12 mM

12 mM extends energy metabolism; stays below Maillard reaction threshold.

Glucose depletion rate in CFPS: ~0.3–0.5 mM/h at 37°C (Caschera & Noireaux 2015). At 6.9 mM, glucose depleted by hour 14–23 At 12 mM: depletion expected by hour 24–40 Maillard reaction (glucose + amino acids at 37°C): significant above ~15–20 mM glucose over long incubations. 12 mM is safely below this threshold Osmolarity contribution: +5.1 mM → +5.1 mOsm NADPH regeneration via G6PDH: directly supports oxidized FMN maintenance for Electra2

6. GMP: 0 µM → 200 µM (if no Guanine)

GMP correctly replaces Guanine as GTP precursor; amount is sufficient.

Guanine salvage: Guanine + PRPP → GMP (via HGPRT). Guanine = 0 µM (change mandatory) GMP → GDP (via GMK + ATP) → GTP (via NDK + ATP): 2 ATP consumed per GTP generated. ATP is regenerated from AMP via the ribose/glucose energy system Solubility of GMP at pH 7.5: ~10 mM - 200 µM is well within solubility Amount: 200 µM GMP matches the original 200 µM Guanine pool GTP role in FMN biosynthesis: GTP cyclohydrolase II consumes GTP as first committed step. 200 µM GMP → GTP pool supports both translation AND flavin synthesis

7. Nicotinamide: 3.10 mM → 5.0 mM

Claim being checked: Higher NAD⁺ precursor maintains oxidized FMN for Electra2 fluorescence over 36h.

NAD⁺ half-life at 37°C, pH 7.5: ~18–24h (Bernofsky & Swan 1973) — at 3.10 mM, NAD⁺ pool significantly degraded by hour 18–24, well before 36h endpoint Nicotinamide → NAD⁺ via NMN intermediate (requires NAMPT + NMNAT). These enzymes are present in E. coli cell extracts 5.0 mM nicotinamide: provides sufficient NAD⁺ precursor to maintain pool throughout 36h FMN oxidation state: NAD⁺-rich environment maintains FMN in oxidized (fluorescent) form via flavin reductase equilibrium. Mechanistically sound for Electra2

Upper limit concern: nicotinamide at high concentrations (>10 mM) inhibits PARP and sirtuins — not relevant in cell-free context. No toxicity concern Caveat The phosphate accumulation problem (Pi building up from NTP hydrolysis over 36h regardless of starting concentration) cannot be solved within the constraints of this mix without adding a phosphate scavenger such as creatine phosphate or PEP — neither of which is in the listed components. This is an acknowledged irreducible limitation of Version 02. The reduction of starting Pi to 6.0 mM mitigates but does not eliminate progressive RFK inhibition. This should be flagged if experimental results show declining Electra2 fluorescence signal after an initial peak in the 12–18h window.