Introduction

Reliable and repeatable measurement is a key component to all engineering disciplines. The same holds true for synthetic biology, which has also been called engineering biology. However, the ability to repeat measurements in different labs has been difficult. The Measurement Committee, through the InterLab study, has been developing a robust measurement procedure for green fluorescent protein (GFP) over the last several years. We chose GFP as the measurement marker for this study since it's one of the most used markers in synthetic biology and, as a result, most laboratories are equipped to measure this protein.

The aim to improve the measurement tools available to both the iGEM community and the synthetic biology community as a whole. One of the big challenges in synthetic biology measurement has been that fluorescence data usually cannot be compared because it has been reported in different units or because different groups process data in different ways. Many have tried to work around this using “relative expression” comparisons; however, being unable to directly compare measurements makes it harder to debug engineered biological constructs, harder to effectively share constructs between labs, and harder even to just interpret your experimental controls.

The InterLab protocol aims to address these issues by providing researchers with a detailed protocol and data analysis form that yields absolute units for measurements of GFP in a plate reader.

Goal for the Fifth Interlab

The goal of the iGEM InterLab Study is to identify and correct the sources of systematic variability in synthetic biology measurements, so that eventually, measurements that are taken in different labs will be no more variable than measurements taken within the same lab. Until we reach this point, synthetic biology will not be able to achieve its full potential as an engineering discipline, as labs will not be able to reliably build upon others’ work.

In the previous interlab studies, it was shown that by measuring GFP expression in absolute fluorescence units calibrated against a known concentration of fluorescent molecule can greatly reduce the variability in measurements between labs. However, when taking bulk measurements of a population of cells (such as with a plate reader), there is still a large source of variability in these measurements: the number of cells in the sample.

Because the fluorescence value measured by a plate reader is an aggregate measurement of an entire population of cells, we need to divide the total fluorescence by the number of cells in order to determine the mean expression level of GFP per cell. Usually this is done by measuring the absorbance of light at 600nm, from which the “optical density (OD)” of the sample is computed as an approximation of the number of cells. OD measurements are subject to high variability between labs, however, and it is unclear how good of an approximation an OD measurement actually is. If a more direct method is used to determine the cell count in each sample, then potentially another source of variability can be removed from the measurements.

This year, teams participating in the interlab study helped iGEM to answer the following question: Can we reduce lab-to-lab variability in fluorescence measurements by normalizing to absolute cell count or colony-forming units (CFUs) instead of OD?

In order to compute the cell count in the different teams samples, two orthogonal approaches were be used:

1. Converting between absorbance of cells to absorbance of a known concentration of beads.

Absorbance measurements use the way that a sample of cells in liquid scatter light in order to approximate the concentration of cells in the sample. In this year’s Measurement Kit, teams were provided with a sample containing silica beads that are roughly the same size and shape as a typical E. coli cell, so that it should scatter light in a similar way. Because the concentration of the beads is known, each lab’s absorbance measurements can be converted into a universal, standard “equivalent concentration of beads” measurement.

2. Counting colony-forming units (CFUs) from the sample.

A simple way to determine the number of cells in a sample of liquid media is to pour some out on a plate and see how many colonies grow on the plate. Since each colony begins as a single cell (for cells that do not stick together), we can determine how many live cells were in the volume of media that we plated out and obtain a cell concentration for our sample as a whole. Each team will have to determine the number of CFUs in positive and negative control samples in order to compute a conversion factor from absorbance to CFU.

By using these two approaches, Interlab Measurement Study will be able to determine how much they agree with each other, and whether using one (or both) can help to reduce lab-to-lab variability in measurements. If it can, then together we will have brought synthetic biology one step closer to becoming a true, reliable engineering discipline.

Calibration Reference

Fig. I The growth curve of BL21 induced by IPTG from 0 to 4 hours. The concentration of BL21 reached stationary phase at 4 hours.

Fig. II The growth curve of BL21 from 0 to 8 hr. The concentration of BL21 reached stationary phase at 4 hours and then declined slightly.

Fig. III Cell lysates from E. coli BL21 with Synthetic MSMEG5998 from 0 to 8 hours and 0 to 4 hours were analyzed by Western blot. The amount of Synthetic MSMEG5998 increased consistently with time.

Discussion

1. According to the data shown above, the growth curve of E.coli BL21 with Synthetic MSMEG_5998 reached the ceiling when the O.D. value was approximately at 2 while the amount of Synthetic MSMEG_5998 were still increasing.

2. Though the amount of Synthetic MSMEG_5998 increased consistently with time, we could not jump to conclusions that it was proper to incubate E.coli as long as possible. Another consideration was the time it would take. Just as our expected, it growed fast at the first 2.5 hours. That’s why we also chose 2.5hr after induced by IPTG when we extracted Synthetic MSMEG_5998 from total cell lysate in other experiments.

3. Based on previous experience, if the E.coli was incubated over 4 hours, the protein that it expressed may be degraded or mis-folded, leading to malfunction. As a result, it was also an important issue for this modeling. However, because of the lack of F420, we did not have the chance to check the enzyme activity of each time spot. It was still unknown whether the titer of the Synthetic MSMEG_5998 would change or not and awaited further research.