PHASE I : Cloning Experiments   

In the initial stages of our cloning experiments, we encountered an unexpected issue: entire pools of colony PCRs were either empty or produced no amplification products (very rare). We hypothesized that the assembly might have been hindered by contamination from unwanted fragments, potentially interfering with the reaction. To address this, we implemented stricter PCR conditions for the fragments production and proceeded with gel purification to ensure only the correct fragments were present.

Example of colony PCR results from the PcGMc cloning.

However, despite these precautions, the next round of results revealed that many of the clones contained inserts of incorrect sizes, with some even showing significant deletions. This led us to suspect that something more complex was at play that our constructs were toxic even though we didn’t see such things during our research.
Examples of colony PCR results for the XARcaf, XARctl, and XAG clonings (from left to right).

After nearly two months of troubleshooting, and having tested almost every conceivable variable, we tested two cloning methods (SLIC and Gibson HiFi assembly). Unfortunately, these attempts yielded similar results so we decided to try one last approach as the end of the internship drew near. We revisited our initial clones, where colony PCR had seemingly failed, and performed PCR colony screening using Q5 high-fidelity polymerase (red circles on the first image). This decision turned out to be a turning point. The results were markedly improved, with clear and correctly sized bands appearing more consistently, suggesting that the fidelity of the Q5 polymerase was crucial in resolving the earlier issues.Following these improvements, we successfully obtained 4 out of the 7 initially planned constructs:

Colony PCR results of XAG using Q5 polymerase.


Colony PCR results of XARcaf using Q5 polymerase.

Colony PCR results of XAGctrl using Q5 polymerase.

Colony PCR results of PcGMc using Q5 polymerase.

  PHASE II : Characterization of the Inducible Promoter and the VIRBAC Receptor   

Characterization of the XylA promoter
Analysis of Pxyl promoter induction by xylose revealed significant differences between conditions with and without glucose. The figure shows that in the absence of glucose, increasing the concentration of xylose leads to an increase in fluorescence. The KM parameter, estimated at 33.2 ± 18.5 mM, indicates the concentration of xylose at which half of the maximum fluorescence is observed.

Characterization of the caffeine receptor

We conducted a western blot to evaluate the functionality of the PxylA-RBS-caffeine receptor construct, specifically focusing on caffeine production. To determine the optimal xylose concentration for promoter induction and gene expression, we prepared a gradient of xylose concentrations (1%, 2%, 3%, 4%, 5%, and 10%), along with an uninduced control. However, no protein was detected via western blot, which raised concerns about potential issues with expression or translation.
Given the deletions observed in previous cloning attempts, we hypothesized that the receptor protein might have toxic effects on the cells, potentially due to promoter leakage. To investigate this, we monitored the optical density (OD) during bacterial growth in the presence of xylose, as shown in the graph. The growth curves revealed no adverse effects on bacterial growth, ruling out the possibility of toxicity from the caffeine receptor.

Following the western blot results, we sent the PxylA-RBS-caffeine receptor construct for sequencing. Sequencing revealed a premature stop codon, or nonsense mutation, which explained the lack of protein detection. Unfortunately, due to time constraints, we were unable to reclone the construct and attempt a new PxylA-RBS-caffeine receptor build to achieve caffeine receptor production.

Characterization of the control receptor

We performed a western blot to assess the functionality of the PxylA-RBS-control receptor construct by inducing its expression using varying concentrations of xylose, similar to the experiment conducted with the caffeine receptor. The range of xylose concentrations used included 1%, 2%, 3%, 4%, 5%, and 10%, along with a non-induced control (0% xylose). The results showed the presence of the control receptor protein at the expected size of approximately 35.5 kDa. However, we observed no increase in protein production despite the increasing concentrations of xylose. This suggested that these xylose concentrations might be saturating the inducible promoter. Once this threshold is reached, additional xylose does not lead to further protein expression, indicating that the promoter has a limit beyond which it no longer induces increased protein production.To test this hypothesis, we conducted another experiment using a lower range of xylose concentrations: 0% xylose, 0.05% xylose, 0.2% xylose, 0.4% xylose, 0.6% xylose, and 0.8% xylose. The results of the second western blot confirmed an increase in control receptor protein production at these lower xylose concentrations. This validated our hypothesis that xylose concentrations above 1% indeed saturate the promoter, preventing further induction of protein expression.

As with the caffeine receptor, we conducted a toxicity test by measuring the optical density (OD) of each culture in the presence of xylose. Measurements were taken every hour over a 7-hour growth period. The results were consistent with those observed for the caffeine receptor, indicating that no toxic effects on cell growth were detected, even in the presence of xylose. These findings confirm that our constructs do not negatively affect the bacterial cells' growth under these experimental conditions.

Due to time constraints, we would have liked to test the induction of our promoters in the presence of xylose as well as another sugar, such as glucose, to observe its effects on the inducible promoter. This would have allowed us to evaluate the impact of catabolite control on protein production, providing insight into how different carbon sources affect promoter activity and gene expression.

   PHASE III : Characterization of the METALE Double-Level Expression System   

Characterization of the PcGMc construct


For this biological sample, six different fields of view (F.O.V) were captured to ensure a comprehensive representation of the entire sample, avoiding potential biases that could arise from examining only one region, which might be atypical or abnormal. By selecting two distinct fields of view (F.O.V 1 and F.O.V 2), we ensured that our observations were representative of the entire sample and not an isolated occurrence.

To present our results, we selected two fields of view from the six, demonstrating that our bacteria simultaneously produce the fluorescent proteins GFP and mCherry. The figure demonstrates the co-localization of GFP and mCherry fluorescent signals within the bacterial cells. This co-expression is visible in the "Merged" images (c and f), where the overlap between GFP's green fluorescence and mCherry's red fluorescent results in yellow regions, confirming simultaneous expression of both proteins in the cells. As a result, our construct successfully expressed the GFP and mCherry genes.

Additionally, to standardize our results and minimize noise contamination, we reduced the brightness and contrast of GFP fluorescence using the software ImageJ 1.54g Java 1.8.0_345 (64-bit) due to its overly strong intensity. Adjusting the brightness and contrast in the GFP channel prevented over-saturation, which could have obscured details or led to misinterpretation of the data. This highlights the importance of maintaining precise controls in fluorescence microscopy to ensure the accuracy and reliability of the results.

The consistent expression of both fluorescent markers across multiple fields of view further supports the conclusion that the genetic construct is stably expressed in the bacterial population, with minimal variation between cells.



The table data were generated after performing site-directed mutagenesis on the ribosome binding site (RBS) of the mCherry gene, with the intent to reduce its expression relative to GFP. The goal was to achieve differential expression between the two proteins, with mCherry being expressed at lower levels. However, the results show consistently low GFP levels and exceptionally low or negative RFP values across all E. coli strains, including the non-mutated wild type (WT). The near-zero RFP values, especially in mutants such as M2 and M5, suggest that the RBS modifications may have significantly disrupted the stability of mCherry mRNA, its translation efficiency, or overall gene expression.

In some mutants, GFP expression is still relatively close to the original construct (e.g., M3, M4, and M9), but others like M2 show very low GFP levels, indicating uneven effects of the mutations on GFP expression. The consistent lack of mCherry fluorescence suggests that the RBS mutations have had a more profound effect on its expression, possibly hindering translation initiation altogether.

Despite using a strong constitutive promoter, these data raise concerns about its effectiveness under these conditions. It seems that either the promoter is malfunctioning or the RBS mutations have had unintended consequences on mCherry production. Even the wild-type strain, which should show both GFP and mCherry expression, failed to produce detectable mCherry fluorescence, reinforcing the possibility of systemic issues with either the promoter or translation initiation.

Interestingly, prior fluorescence microscopy experiments showed clear co-expression of GFP and mCherry in these bacterial cells. However, the spectrophotometric data tell a different story, with no detectable mCherry across any strains. Both sets of data were collected from the same strains under nearly identical conditions, with only a short delay due to sequencing. This discrepancy between the microscopy results and the spectrophotometric readings suggests potential experimental inconsistencies or differences in sensitivity between the methods used.

To account for these unexpected results, several factors must be considered. Experimental variability, such as slight measurement inaccuracies or fluctuations in culture conditions, could have played a role. In complex systems involving multiple variables—like promoter strength, RBS efficiency, and protein folding—even minor differences can significantly impact gene expression.

Another possibility is that the RBS mutations affected mCherry mRNA stability. If these mutations caused structural changes in the mRNA, they could have made it more prone to degradation or less efficient in initiating translation, thereby reducing or eliminating mCherry protein production. Such effects could occur even if transcription proceeded normally, and this would explain why no mCherry fluorescence was observed.

Furthermore, interactions between the bacterial host and the synthetic construct might have contributed to the lack of mCherry expression. The host cells may have regulatory mechanisms that interfere with the expression of mCherry, particularly if the modified RBS reduced the efficiency of ribosome binding or mRNA processing.

Complementary investigation is necessary to clarify these findings. Testing alternative RBS designs or using different bacterial strains may provide insights into the lack of mCherry expression. Altering the promoter or modifying other regulatory elements could also help resolve the expression issues. As is often the case in synthetic biology, iterative design cycles, including testing and optimization, are critical for overcoming unexpected challenges and achieving the desired system functionality.