Design-Build-Test-Learn (DBTL)

---

The development of the Riceguard project for iGEM 2025 followed the iterative Design-Build-Test-Learn (DBTL) cycle, allowing refinement of the cell-free arsenic biosensor in response to new insights and constraints. The following sections outline the key pivots across seven cycles, structured according to the four DBTL steps, with a focus on adjustments to protocols, incubation times, plasmid concentration ratios, and related parameters.

Cycle 1: Transition from GMO-Based Biosensor to Cell-Free System

• Design: We initially conceptualized the biosensor using engineered E. coli Nissle 1917, a probiotic strain, for arsenic detection in settings such as rice fields or community areas.
• Build: We shifted construction to a cell-free system employing purified cellular components, with initial assemblies exploring lyophilization for enhanced stability.
• Test: We modified evaluation protocols to examine cell-free reactions, incorporating adjustments to incubation times and plasmid ratios for optimal performance.
• Learn: Regulatory constraints and safety considerations from iGEM guidelines and university consultations indicated the impracticality of GMO deployment, leading us to focus on non-living systems for improved safety and applicability.

Cycle 2: Shift from Multi-Cassette Plasmid to Dual-Plasmid System

• Design: Our early plans incorporated a multi-cassette plasmid integrating various genetic elements, such as reporters, repressors, and reductases, for concurrent expression.
• Build: We focused assembly on constructing the multi-cassette plasmid to enable simultaneous production of required proteins and reporter elements.
• Test: We conducted no laboratory testing; instead, discussions with a university professor highlighted concerns regarding our team's capacity to harvest proteins effectively or control their quantities when expressed on the same cassette as the reporter gene, which would necessitate evaluating numerous design variants—a process we deemed excessively time-consuming due to the need to outsource synthesis to Twist Bioscience and await 2–3 weeks for delivery to Taiwan.
• Learn: These insights prompted us to pivot to a dual-plasmid system, comprising a sense plasmid and a reporter plasmid, allowing variation in their concentrations during addition to cell-free reactions to serve as a tunable control mechanism for regulating expression levels.

Cycle 3: Redirection from Rice Farmers to Household Users

• Design: We originally intended the biosensor for rice farmers as a diagnostic for post-harvest rice, paddy water, or soil, to guide irrigation practices like flood-and-drain methods. However, after speaking with a parent who had herself tested for excessive Arsenic in a medical test, we were convinced to refocus the objective to include both parents and rice farmers

• Build: We integrated user-friendly elements into development suitable for household application, with preparations adapted for everyday conditions.
• Test: We revised protocols to simulate household scenarios, including rice extract processing and modified incubation times.
• Learn: Insights from health reports, such as the Healthy Babies Bright Futures study, underscored the priority of addressing dietary arsenic exposure for families, resulting in our reorientation toward accessible home use.

Cycle 4: Focus from Dual Analyte to Arsenic-Only Detection

• Design: Our initial concepts encompassed detection of both arsenic and cadmium to cover multiple heavy metals prevalent in rice.
• Build: We emphasized construction of arsenic-specific genetic constructs to ensure targeted functionality.
• Test: We refined evaluations to measure sensitivity and specificity, with adjustments to plasmid ratios and reaction conditions.
• Learn: Data highlighting arsenic as the primary contaminant, combined with project timelines, supported our concentrated effort on arsenic alone for deeper characterization and registry contribution.

Cycle 5: Validation Testing to Select the Best Variant Pair

• Design: We considered multiple sense and reporter plasmid pairs to determine the most effective combination for arsenic sensing; specifically, we designed a few plasmids using Benchling, utilizing three sense plasmids (A, B, and E) and reporter plasmids (NoProm and OC2), where the sense plasmid produces ArsC and ArsR, and the reporter plasmid generates fluorescence via an RNA aptamer that binds to a fluorescent dye, with different combinations tested for clearest fluorescence output.
• Build: Our assemblies involved various plasmid combinations, with preparations adjusted for repeated evaluations; we prepared a master mix consisting of buffer (81.25 μL), lysate (39 μL), RNA polymerase (3.25 μL), RNase inhibitor (3.25 μL), and nuclease-free water (35.75 μL), followed by incubation of sense plasmids A, B, and E at 37°C for one hour to produce ArsC and ArsR in separate 1.5 mL test tubes, after which we added reporter plasmids NoProm and OC2 and left them at 4°C overnight.
• Test: Our protocols incorporated concentration gradients and multiple trials to verify efficacy across pairs; we inserted mixtures into 12 wells of a 96-well plate (12.5 μL master mix per well), with 1.1 μL of DFHBI-1T fluorescent dye added, and tested fluorescence in a plate reader across configurations: 2 wells with sense A and 0 ppb arsenic, 2 with sense A and 800 ppb arsenic, 2 with sense B and 0 ppb arsenic, 2 with sense B and 800 ppb arsenic, 2 with sense E and 0 ppb arsenic, and 2 with sense E and 800 ppb arsenic (sense E lacking a promoter served as the negative control for background noise).

• Learn: Performance variations, including differences in sensitivity and leakiness, led us to select the optimal pair (WIST_SENSE_MedArsR_StrArsC_001_A and WIST_REPORT_OC2) based on reliable activation at 50 ppb; after one hour of kinetic testing, fluorescence was detected in 0 ppb arsenic wells, indicating insufficient time for sense plasmids to produce ArsR repressors.

Cycle 6: Refining Incubation Temperatures and Times

• Design: We set broad incubation parameters, including temperatures from 25°C to 37°C and durations extending several hours.
• Build: We maintained system preparations with a focus on stability under varying conditions.
• Test: We conducted kinetic monitoring through iterative adjustments to temperature and duration to identify optimal reaction windows.
• Learn: Our observations of fluorescence degradation or incomplete reactions under certain conditions resulted in standardization at 37°C for 2–4 hours to enhance efficiency and reproducibility.

Cycle 7: Adjusting Plasmid Concentrations to a 1:10 Ratio

• Design: We initially set plasmid concentrations equally or varied them flexibly, without a predefined ratio; we planned adjustments to the sense-to-reporter plasmid ratio (1:5 and 1:10) without modifying plasmids, focusing on the sense A and reporter OC2 combination for clearest expression.
• Build: We modified preparations to establish a balanced ratio, emphasizing reporter dominance; we prepared a 250 μL master mix with the same ingredients and proportions as prior experiments, including an extra 25 μL to account for evaporation losses.
• Test: We performed titrations of ratios, with fluorescence assessed across arsenic levels to evaluate outcomes; we divided 18 reactions into five groups, with half receiving 1.1 μL of 800 ppb or 0 ppb arsenic solutions and all receiving 1.1 μL of DFHBI-1T dye before plate reader insertion: Group A (4 wells, 1:5 ratio), Group B (4 wells, 1:10 ratio), Group C (4 wells, 1:5 ratio without reporter plasmids as negative control for unactivated fluorescence), Group D (4 wells, 1:10 ratio without sense plasmids as positive control for maximum fluorescence), and Group E (2 wells, Nebexpress GFP Control Plasmid to verify kit functionality).
• Learn: Inconsistent expression from unbalanced concentrations indicated our need for a 1:10 sense-to-reporter ratio to optimize dynamic range and minimize background noise; after two hours of testing at one-minute intervals, noticeable but fluctuating fluorescence without clear trends suggested that 800 ppb arsenic concentration was excessively high and potentially impacting results.

Final Pivot: Simultaneous Addition and Kinetic Reading

• Design: We originally planned components for sequential addition, followed by separate incubation and measurement.
• Build: We integrated all reagents—lysate, T7 polymerase, plasmids, DFHBI-1T, and rice extract—into a single master mix for unified processing.
• Test: We implemented real-time kinetic analysis over 90 minutes in an ELISA plate at 37°C to observe transcription dynamics and response plateaus.
• Learn: Variability from sequential methods necessitated a streamlined simultaneous protocol, confirming the biosensor's effectiveness with a 5–100 ppb dynamic range suitable for practical contamination assessment.