LOADING
The core measurements in our project included bacterial density via OD600, GFP fluorescence, and colony-forming unit (CFU) counting. To ensure the reproducibility of our work and the general applicability of our conclusions, we established standardized protocols for these assays and calibrated our instruments using reference standards, serving as future reference for other iGEM teams and research groups.
Optical density at 600 nm (OD600) is a common method for quantifying cell density in cultures. In this project, we primarily use a microplate reader for OD600 measurements. To ensure a linear relationship between OD600 and cell count, and to guarantee reproducible data, we calibrated the OD600 readings using standard suspensions of silica microspheres.
Silica microspheres with a diameter of 1 μm are suitable for this purpose as their size and absorbance characteristics are similar to those of bacterial cells, and standard suspensions can be conveniently prepared. According to the Beer-Lambert law, the absorbance of a solution is linearly correlated with the concentration of the solute at lower concentrations. Specifically for cells and microparticles, a linear relationship between particle number and light absorption generally holds within the OD600 range of 0.2 to 0.8.
The wavelength of 600 nm was selected as interference from pigments, metabolites, or biomolecules like nucleic acids and proteins which absorb strongly in the UV range is minimal. Shorter wavelengths are more susceptible to increased light scattering caused by cell clumping or sedimentation, leading to greater measurement deviations. The 600 nm wavelength mitigates these scattering effects, providing a more reliable reflection of cell density changes. We used 1 μm silica microspheres as a calibration standard to establish a universal "reference currency" for converting OD600 readings across different laboratory instruments.
Figure 1. OD600 calibration via microplate readers based on the principle of Lambert-Beer Law. (Note: A is absorbance, ε is the molar attenuation coefficient, c is concentration, and l is the path length.)
•Silica powder (1 μm), Shanghai Hangshu New Material Co., Ltd.
•Black-walled, clear-bottom 96-well plates
•ddH2O
•PBS Buffer (1×)
•Thermo Scientific Varioskan LUX microplate reader
The detailed experimental procedure was performed according to the protocol referenced in the following document, which provided the basis for deriving our OD600 calibration formula.
A suspension was prepared by mixing 100 μL of silica microspheres with 900 μL of ddH2O, followed by serial dilution. The absorbance at 600 nm was measured for each dilution in quadruplicate using the microplate reader. Using a unified procedure for both bacterial OD600 and silica particle measurements, and with 1× PBS buffer as a blank control, we obtained a calibration curve of OD600 versus standard particle concentration. Linear regression fitting yielded an R2 value greater than 0.99, indicating an excellent fit on both linear and logarithmic scales. Within the mid-to-high OD600 range of 0.12 to 0.7, a quantitative relationship between OD600 value and silica microsphere count was established. Notably, an OD600 of 1.00 corresponded to 3.81×109 particles/mL. Deviations outside the optimal linear range may be due to non-uniform distribution of silica particles in the solution, violating the assumptions of the Beer-Lambert law for dilute solutions.
Figure 2. Calibration of OD600 using silica microsphere standard suspension. (A) Linear regression fit in the standard coordinate system. (B) Linear regression fit performed on logarithmically transformed data.
When the measured OD600 value falls within the range of 0.2 to 0.8, the counts of E. coli (typically 1–3 μm in length and 0.5 μm in diameter) can be approximately equated to that of similarly sized silica microspheres (1 μm). This also allows the silica microsphere count to be used to calibrate OD600 readings across different instruments. The bacterial concentration can be estimated using the following formula:
Potential sources of error in the measurement include: inadequate cleaning of the 96-well plates, non-uniform particle size, uneven distribution of microspheres in suspension, pipetting errors during serial dilution, and inconsistent sample loading. For OD600 measurements of E. coli cultures, additional interfering factors can include background luminescence from metabolic impurities in the medium and incomplete resuspension of cells after washing or centrifugation. To minimize these interferences, we washed the samples three times by centrifugation and resuspension in 1× PBS, included blank controls, and conducted replicate measurements. These practices help to corroborate results across parallel experiments, reducing random and systematic errors, thereby enhancing data accuracy and reliability.
Calibrating OD600 values ensures that the measured cell density closely reflects the true biological properties, preventing instrument-specific variations from introducing errors that could lead to misinterpretation of cell count and growth status, which would ultimately affect characterization and optimization outcomes. We propose the use of 1 μm silica microspheres as a standardized "currency" for data conversion and calibration. This approach enhances the universality of experimental results and conclusions, which is particularly crucial for our project involving extensive fluorescence/OD600 measurements and circuit optimization. We hope to promote the wider adoption of OD600 calibration within the iGEM competition.
GFP is the reporter protein we use to characterize gene circuits and the integration of unnatural amino acids. Its fluorescence intensity directly affects the reliability of our experimental conclusions and the optimization of subsequent experimental conditions. In gene circuit construction, we typically quantify the expression level of target genes to evaluate the effectiveness of our genetic circuits.
In wild-type GFP, the amino acids at positions 65, 66, and 67 (serine, tyrosine, and glycine) on the α-helix form a chromophore through cyclization, dehydrogenation, and other reactions. The gene circuit modules we use primarily employ GFP(S65T) as the reporter gene, which has maximum excitation and emission wavelengths of 490 nm and 510 nm, respectively. Due to the close proximity of these two wavelengths, our microplate reader uses 485 nm (excitation) and 520 nm (emission) for fluorescence measurements.
FITC is a highly efficient and stable fluorochrome with a maximum absorption wavelength of 490-495 nm and a maximum emission wavelength of 520-530 nm, emitting bright yellow-green fluorescence, which is similar to the GFP we used. We aim to use FITC’s fluorescence signal as an intermediate to standardize GFP expression levels, thereby calibrating and unifying the expression signals of our engineered bacteria across different instruments. Since the fluorescence efficiency of FITC is closely related to the pH of the solution, we subsequently adjusted the solution pH to about 10 using Na2CO3-NaHCO3 buffer before measuring fluorescence at 495 nm excitation and 520 nm emission.
Figure 3. Chemical Structure of FITC.
•FITC (97% purity, Macklin)
•Black-walled, clear-bottom 96-well plates
•PBS Buffer
•NaHCO3 (≥99.5% purity)
•Sodium hydroxide pellets
•Thermo Scientific Varioskan LUX microplate reader
The detailed experimental procedures were performed according to the protocol referenced in the following document, and we established a calibration conversion formula for fluorescence intensity measurement of our experimental data.
We prepared serial 2-fold dilutions of a 15.98 μM FITC stock solution to generate gradient standard solutions. The pH of these solutions was adjusted with Na2CO3-NaHCO3 buffer, and their fluorescence intensities were measured using a microplate reader at 495 nm excitation and 520 nm emission. After plotting the curve of fluorescence intensity against FITC concentration and performing linear regression fitting, the results showed that the R2 value was greater than 0.99, indicating a strong linear relationship on both linear and logarithmic scales at lower concentrations. For high-concentration fluorescent solutions (≥4 μM), the fluorescence intensity gradually deviated from the linear fit, which is consistent with the violation of the dilute solution assumption in the Lambert-Beer law.
Figure 4. Fluorescence calibration using FITC standard solution. (A) Linear regression fit in the standard coordinate system. (B) Linear regression fit performed on logarithmically transformed data.
Within the fluorescence intensity range of 0-300 relevant to our experiments (indicating relatively low GFP fluorescence in our system), the signal showed high linearity with FITC concentration, enabling the calculation of the correlation between FITC and GFP fluorescence yields. From this, we derived the following conversion formula:
The main interfering factors in this fluorescence measurement experiment include: the cleaning efficiency of 96-well plates, the accuracy of FITC dissolution and serial dilution, the uniformity of solution pH adjustment, and uneven sample loading in some experiments. In actual measurements of E. coli fluorescence expression, we typically use 1×PBS buffer for washing without adjusting the pH of our sample solution. Additionally, there are inherent differences between FITC and GFP fluorescence, and intracellular substances may interfere with GFP fluorescence detection. To more accurately quantify GFP production and further optimize the expression system, it is advisable to measure fluorescence after cell lysis or use a flow cytometer to avoid errors caused by cell walls. In our actual experiments, we minimized errors by repeated washing with 1×PBS via centrifugation, including blank controls, and performing measurements in replicates to ensure consistency.
By calibrating both fluorescence values and OD600 readings as described previously, we can calculate a more accurate Fluorescence/OD600 ratio (unit fluorescence) to assess the signal strength of our GFP reporter protein. This helps to provide a reproducible data reference for other teams. In actual experiments, we also set up blank controls and negative controls to evaluate induction levels and background fluorescence. In the future, we will further improve the fluorescence characterization system and apply these calibration formulas to standardize experimental data, thereby enhancing the reliability and universality of conclusions drawn for genetic circuit construction and characterization.
In the characterization of hydrogel-based bacterial delivery systems, we employed the serial dilution plating method to quantify viable bacteria in liquid samples. This approach is cost-effective and provides greater accuracy than turbidimetry, though it requires a longer incubation period for results. The core principle involves serially diluting the sample, spreading it evenly on a solid medium, and allowing viable cells to grow into visible colonies. The colony count then reflects the live bacterial concentration in the original sample.
The plating method directly measures colony-forming units (CFUs), defined as the number of visible colonies formed by one or more viable microorganisms on a standardized solid medium after incubation. CFU counting excludes non-viable and non-proliferating cells, offering a more accurate estimate of reproductively active bacteria. By applying appropriate dilution factors, the colony count on at least one plate can be adjusted to fall within the valid range of 30–300 colonies, enabling back-calculation of the original live bacterial concentration.
It should be noted that CFU counts generally underestimate the true viable concentration, as a single colony may originate from multiple cells or clusters. Where feasible, techniques such as flow cytometry or electron microscopy can be used for more precise quantification.
Figure 5. Schematic diagram of serial dilution plating.
Samples were diluted to an estimated concentration between 300 and 3000 CFU/mL to avoid overcrowding or insufficient colony formation. Since the initial bacterial concentration was unknown, multiple dilution gradients were tested. A 100 μL aliquot of each diluted sample was spread on a 10 cm plate and incubated at 37 °C for 24 hours, followed by colony counting.
We analyzed time-course data from hydrogel release experiments involving 1% SA + 1% CaCl2 and 1% SA + 0.5% CaCl2 hydrogels. The colony counts generally followed expected trends when compared to theoretical standard curves. However, plates with low colony numbers were susceptible to stochastic deviations, while heavily overgrown plates were uncountable. These limitations of the plating method introduced some experimental variability and increased the workload for valid data selection. The use of larger Petri dishes could expand the accurate counting range and improve result reliability where resources allow.
The original bacterial density can be calculated using the following formula:
where is the colony count on a countable plate, is the corresponding dilution factor, and is the number of plates counted.
Figure 6. Relationship between colony count and dilution degree from countable plates in the hydrogel controlled-release experiment. (A) 1% SA + 1% CaCl2 hydrogel. (B) 1% SA + 0.5% CaCl2 hydrogel.
While constraints prevented us from correlating CFU with absolute viable counts using advanced instruments such as flow cytometry or electron microscopy, we demonstrated that our dilution and plating protocols could yield reproducible data. We hope this methodology offers a practical reference for future iGEM teams conducting similar bacterial quantification assays.
[1]Jacob Beal, Traci Haddock-Angelli, Markus Gershater, Vishal Sanchania, Russell Buckley-Taylor, Geoff Baldwin, Natalie Farny, Richard Tennant, Paul Rutten 2020. Calibration Protocol - Plate Reader Abs600 (OD) Calibration with Microsphere Particles. protocol.io. https://dx.doi.org/10.17504/protocols.io.bgy6jxze
[2]Heim R, Cubitt AB, Tsien RY. Improved green fluorescence. Nature. 1995;373(6516):663-664. doi:10.1038/373663b0.
[3]Paul Rutten, Cheryl Telmer, Geoff Baldwin, Dennis Mishler, Traci Haddock-Angelli, Jacob Beal, Ari Dwijayanti, Marko Storch, Natalie Farny, Alejandro Vignoni, Richard Tennant 2020. iGEM Calibration Protocol - Red Fluorescent Proteins in Plate Readers. protocol.io. https://dx.doi.org/10.17504/protocols.io.bcdjis4n
[4]Chaganti, L. K., Venkatakrishnan, N., & Bose, K. (2018). An efficient method for FITC labelling of proteins using tandem affinity purification. Bioscience reports, 38(6), BSR20181764. https://doi.org/10.1042/BSR20181764
[5] Ben-David A, Davidson CE. Estimation method for serial dilution experiments. J Microbiol Methods. 2014;107:214-221. doi:10.1016/j.mimet.2014.08.023