Engineering Success

The problem of synthetic dyes was identified with water quality being a predominant factor. A protein based dye that had an improved ability to cotton would provide an alternative to hazardous synthetic dyes.

There are a variety of naturally produced vibrant protein chromophores that are non-toxic to humans. Chromoproteins will be used as a chromophore due to their visibility under normal light and well established pathways in bacteria, specifically meffRed. A protein domain for cellulose binding has already been identified allowing for a potentially easy way to improve the cellulose binding of a protein based dye.

Our alternative dye would have to be produced in large quantities so a rapidly reproducing chassis is essential. The CIDAR MoClo Parts Kit has a wide array of established modular components designed to create tailor made expression vectors through Golden Gate cloning. Designed gene blocks need to be compatible with Golden Gate cloning in the MoClo scheme to fit the parts available in the CIDAR MoClo Parts Kit.

Our bacterial strain is easily manipulated for many iterative genetic experiments and is low maintenance. DH5-alpha E. coli will suffice for early experimentation.

Software for building gene blocks and testing their cloning digitally would be necessary. We will use Geneious software.

2. Build

The construction of our build will revolve around Golden Gate cloning in the MoClo scheme.

• Fast and efficient method of cloning that relies on type two restriction sites that are excised after ligation which prevents backwards reactions (Golden Gate cloning).
• Genetic segments for all parts can be obtained from online databases and sequences can be edited to contain Golden Gate cloning restriction and recognition sites.

The combination of meffRed and cellulose binding domain could cause misfolding during translation. Therefore multiple orientations were designed with two different types of cellulose binding domains to try and maintain the efficacy of the recombinant protein.

A histidine tag was included in the build to ensure easier separation of protein during purification. The position of the tag varied with the different orientations of each construct, on either the amino or carboxyl terminus.

3. Test

The construction of our build will revolve around Golden Gate cloning in the MoClo scheme.

The constitutive conjugants were observed for visible dye production. There was no visible color, even in the positive control with no cellulose binding domain. The inducible conjugants were grown on media with arabinose and also screened for visible dye production. There was no visible color.

Colonies were selected to be plasmid prepped, undergo restriction digest, and run on an agarose gel to verify correct construction.

4. Learn

After verification via plasmid digest we saw that none of the constructs had assembled correctly. To solve this issue we first decided to switch out some of our modular parts from the MoClo kit to see if that was the problem and had an easier time assembling.

The new construction switched RBS from BCD12_BC to BCD2_BC as seen in pGEC031 to pGEC035. This new RBS was higher strength, which would also increase expression of protein.

In order to see if our ordered coding sequence was the problem we used mScarlet-1 (c99) from the CIDAR MoClo Extension kit as a positive control to see if the cloning process wasn’t working or if our sequences were flawed. This is the pGEC037 construct. All parts in pGEC037 come from the established CIDAR MoClo kit.

Additionally we used the c99 mScarlet-1 control to test our arabinose inducible promoter sequence ((6)). This is the pGEC036 construct. All parts in pGEC036 come from the established CIDAR MoClo kit except the (6) arabinose inducible promoter sequence.

2. Build

The mScarlet-1 was already available and constructed to fit in our modular cloning system, so no new sequences had to be synthesized and ordered. Golden Gate cloning was used again to make the process faster.

3. Test

None of the colonies with meffRed expressed color. Both pGEC037 and pGEC036 (after adding arabinose) produced visibly pink cultures under normal light.

Colonies were again plasmid prepped, underwent restriction digest, and were ran on an agarose gel for assembly verification.

4. Learn

The agarose gel verification isn’t the most reliable method, so we also sent these constructs to Plasmidsaurus for next-generation sequencing. The results showed that the constructs pGEC031-pGEC035 did not assembled correctly, but the sequence of pGEC037 showed the expected components and layout. This means that we are correctly performing the Golden Gate reactions correctly, and that there is an issue with our designed coding sequences. The sequence of pGEC036 also proved to be a correct assembly. This means that the (6) arabinose sequence is correct. This leaves our CD sequences (1) through (5) as the reason for the unsuccessful cloning.

Due to the issues with the CD modules coding sequences (1) through (5), they were redesigned as (11) through (15). Golden Gate compatible Bpi1 sites were added on the outsides of (1) through (5) for the option to immortalize the parts as plasmids. Additionally, the extra nucleotides added to the ends of the DNA sequences should allow the restriction enzyme to work more effectively during cloning. The new designs were made with meffRed as well as mScarlet-1 and AmilCP as the chromophore. These designs include (16)-(21) in the constructs table. mScarlet was chosen because it was very vibrant when expressed as our positive control, and would be easier to use to test the binding domain. AmilCP was also included, which is a blue chromoprotein. This was just done to determine if meffRed specifically is a poor chromoprotein or if chromoproteins in general will be hard to express.

2. Build

The same building technique was implemented, using Golden Gate cloning for construct assembly. However, only a few constructs of each orientation were constructed to save time and streamline the testing process. Initially pGEC037, pGEC055, pGEC060, and pGEC62 were the only four assembled for the first round of testing. pGEC037 grew well which indicated that the cloning and transformation were successful. pGEC055 had small colonies that were very light pink, while pGEC060 also had small colonies but were vibrantly pink. pGEC062 had no color. This led us to move forward with mScarlet-1 as our protein based chromophore due to the success of pGEC060.

Next all of the inducible constructs using mScarlet-1 were cloned, which is pGEC046, pGEC047, pGEC048, pGEC049, and pGEC050. The inducible promoter was used over the constitutive because of the long growth period of the constitutive colonies due to constant expression of protein. The inducible colonies were originally grown without the inducer to encourage strong growth before protein expression. When grown with the inducer, constructs pGEC050 and pGEC046 were the most pink, pGEC049 was slightly pink, and pGEC047 and pGEC048 were not pink. This result led us to choose pGEC050, pGEC049, and pGEC046 for further testing.

All of the assembled inducible constructs were sent to Plasmidsaurus for next-generation sequencing. The constructs sent for sequencing all had correct assemblies except for pGEC048. pGEC048’s sequence has a meffRed sequence where a mScarlet sequence was expected. The lack of a vibrant pink color pGEC048 is then easily explained by the significantly worse meffRed sequence as seen in previous experiments. The remaining constructs also have large differences in vibrance of color. Since we know that they are assembled correctly from the Plasmidsaurus sequences, we attribute these changes in phenotype to the differences in orientation and cellulose binding domain.

3. Test

Since visibly red colonies were produced and the plasmid constructs were verified. The cellulose binding domain needed to be tested. Due to time constraints, we were unable to set up an assay to determine if the dissociation coefficient was the same as the 0.6 μM coefficient reported in the literature but we still wanted to find preliminary results to support or reject the idea that the binding domain was functional. Our initial method of testing involved creating handmade cotton filters to run cell lysate through and measure what was unable to bind to the cotton. Our crude handmade filters were not sealed properly, causing the solvent to leak everywhere but the collection tube. As a result, a new assay was created using t-shirt swatches to visually distinguish between mScarlet with and without a cellulose binding domain. While this experiment was only qualitative by nature, it provides preliminary information on whether or not the cellulose binding domain was functional until more thorough assays can be performed.

In order to have direct visual comparisons between the constructs with and without cellulose binding domains, serial dilutions of sample pGEC050 were performed and fluorescence was measured using a Cytation 5 plate reader. A 1:16 dilution of sample pGEC050 was found to be roughly equivalent to pGEC046. Sample pGEC046 and the 1:16 dilution of pGEC050 were compared by creating 3 replicates of cotton t-shirt pieces that were dyed with 1mL of each solution. They were incubated at room temperature overnight in falcon tubes. Samples from the previous night were imaged using a trans-illuminator. Then 40mL water was added to each sample and shaken at 250rpm for 30 minutes as a standardized wash procedure. Finally, the samples were re-imaged on the trans-illuminator and the images were compared.

4. Learn

This assay, while somewhat functional, could be vastly improved upon. First, since most of our protein was expressed in inclusion bodies, we had very low protein concentrations to test. This resulted in the need to adjust the contrast of photos after taking them to visually compare differences between the two test groups (adjusting the brightness of images was noted in the relevant results page). Re-solubilizing and then refolding inclusion bodies would result in higher concentrations of produced dye making evaluation easier. Furthermore, the experiment did not use purified protein, instead the cell lysate was used. The detergent in the B-PER lysing agent could interfere with the cellulose binding domain and as a result could skew results. While testing of protein binding under different conditions is necessary to create a fully characterized product, the test should be performed with purified protein first as a baseline. Future testing will take advantage of the 6x-His tag present on the constructs for IMAC and then desalting through dialysis. This will ensure that the charged residues responsible for specific binding to cellulose are not interfered with during testing. Finally, quantitative testing should be prioritized in the future over the qualitative method we used for these preliminary results.

We also learned that our constructs were able to express natural protein dyes with preliminary results indicating functionality and promise with further optimization.