Experiments

Designing Bliss

BLISS’s entire genetic circuit consists of over 7000 base pairs, meaning it could not be synthesized in entirety due to size and monetary constraints. As a workaround, we broke our circuit into four unique parts, each under 3000 base pairs. Variations of the four parts were also created and denoted as “A” or “B.” Table 1 (below) describes what each number-letter combination represents.

Table 1. Number-letter names for each part and their corresponding significance. More information about each part can be found under the Engineering page of our wiki.

Modifications—such as rearrangements of genes, signal peptides, and purification tags in parts 1A, 2A, 1B, and 2B—were made to improve chances of the circuit functioning as intended. Parts 4A and 4B serve identical functions—to bind carbon catabolite complex, HPrSerP-CcpA, and halt expression of lacI—however, literature and database coding sequences conflicted with each other. Both variations were synthesized, and the functionality of each will be tested in the future with a fluorescence assay.

All parts were ordered as gene fragments from International DNA Technologies (IDT). Each part was synthesized with 50 base pair Gibson Assembly overhangs meticulously designed to include internal restriction enzyme cut sites. This design allows all parts to be assembled in one step with Gibson Assembly into two “storage plasmids,” designated “storage plasmid A” and “storage plasmid B,” where storage plasmid A contains all four “A parts” and storage plasmid B contains all four “B parts” (Figure 1). From these storage plasmids, individual parts can be isolated with the internal restriction enzyme cut sites. This modular design allows us to create entirely new and customized constructs suitable for the various assays needed to test functionality of individual parts.

Figure 1. Fully assembled Gibson A and Gibson B storage plasmids. A. Gibson A storage plasmid with parts: 1A, 2A, 3A, and 4A. B. Gibson B storage plasmid with parts: 1B, 2B, 3B, and 4B.

Gibson Assembly

Gibson Assembly was chosen to construct our storage plasmids because it allowed us to construct each plasmid in one reaction, saving time and resources. Plasmid pBS1C was chosen as the backbone for our storage plasmids because it contains genes encoding for both ampicillin and chloramphenicol resistance. pBS1C was also selected because it can be integrated into the chromosomal DNA of B. subtilis, preventing the plasmid from being lost or excised. Another unique feature of pBS1C is that it contains a coding sequence for red fluorescent protein (RFP). When RFP is cut out from the backbone and DNA parts are successfully cloned in, transformed E. coli colonies will appear white. Any red colonies present after transformation are immediate feedback cloning wasn’t successful.

During our first attempt at Gibson Assembly, we sought to create both A and B storage plasmids. After digesting pBS1C with EcoRI and PstI to remove the RFP, we combined the digested pBS1C with all A parts and all B parts in two separate reactions. Gibson Assembly products were transformed into NEB high competency E. coli DH5α cells, then plated on selective LB-ampicillin medium. Several resulting colonies on each plate were grown in overnight cultures, miniprepped, then analytically digested and screened on an agarose gel. The gel results, viewable on the Results page, weren’t exactly what we expected, however, one A and one B colony were selected for sequencing. Sequencing performed by GeneWiz (Azenta Life Science) revealed parts 1A and 1B had been successfully cloned into pBS1C. The remaining parts (i.e. 2A, 2B, 3A, 3B, 4A, and 4B) were not assembled because the 3’ sequences of 1A and 1B had a six base pair region of complementarity with the cut site created by PstI digesting pBS1C. This allowed parts 1A and 1B to ligate to the backbone instead of the 5’ Gibson sequences of the consecutive 2A and 2B parts. A more detailed explanation of this outcome can be found within the Results page of our wiki.

While cloning parts 1A and 1B into pBS1C was a success, they are unusable without part 3A: T7 RNA Polymerase, which serves to activate the promoters regulating the lipase and colipase genes. In following weeks, we attempted several ligations to create constructs containing 1A, 2A, and 3A. A more in-depth explanation of the ligation process is found under the Ligation section of this page. Alongside ligation, we also reattempted to create storage plasmids A and B using Gibson Assembly. Eventually, we were able to fully assemble storage plasmid B containing parts 1B, 2B, 3B, and 4B. To date, we have not been able to assemble storage plasmid A.

One Gibson Assembly attempt, we utilized an inverse PCR technique on our plasmid containing part 1A. We designed PCR primers to amplify and linearize the entire plasmid in a way that exposed the 3’ Gibson Assembly sequence of part 1A so that we could add in parts 2A, 3A, and 4A. The resulting constructs were digested and run on an agarose gel. Several bands were observed; however, the quantity and size of the bands did not align with anticipated results. For this reason, this attempt was ultimately considered unsuccessful. We have not reattempted Gibson Assembly for storage plasmid A due to the redesign of parts 1 and 2.

Several times, we attempted to clone part 3A into our pBS1C-1A and pBS1C-2A constructs with traditional ligation methods. With time running out before the Jamboree, we decided to redesign parts 1B and 2B so that they are constitutively active, rather than dependent on part 3A to activate. These new parts have been named parts 1C and 2C. Each contains the same coding sequences for lipase and procolipase, however, the promoter controlled by T7 RNA Polymerase was swapped with a constitutively active one. The new parts have yet to arrive, and the team anticipates attempting Gibson Assembly with them soon.

Ligation

BLISS’s modular design allows individual parts to be removed from storage plasmids with restriction enzymes. Enzyme cut sites were strategically placed so that we could cut out and ligate parts together, creating new constructs with specified functions. We prioritized creating constructs 1A + 2A + 3A and 1B + 2B + 3A, as these combinations of parts would allow us to test for enzyme secretion in transformed B. subtilis.

Each ligation began with digesting the storage plasmid containing the parts we needed and digesting our synthesized stock solution of part 3A with the same restriction enzymes. After digestion, the DNA was run through an agarose gel, and DNA bands of interest were gel extracted. Once isolated, the DNA was ligated on several occasions with different ligases including T4 and T7 DNA ligases, as well as the NEB Quick Ligation Kit. To increase chances of success, the ligation reaction mixture was incubated at varying times and temperatures. After the ligation reactions were transformed into competent E. coli, analytical digests of transformed colonies were performed. No positive results were obtained.

In-agarose ligation was also attempted. Different than traditional ligation, in-agarose ligation involves ligating digested DNA directly from an agarose gel—no extraction necessary. After transformation of the ligation product, no colonies (with the exception of a positive control) were yielded.

After several attempts and many rounds of troubleshooting, the team has yet to successfully create a new construct through ligation. Efforts remain ongoing to create a construct capable of secreting enzymes lipase and procolipase.

Biocontainment Switch Mechanism

We planned to assemble the three parts – xylose induced antitoxin, temperature sensitive component, and xylose repressor – with traditional assembly. We digested the pHY300PLK backbone, the xylose induced antitoxin part, and the xylose repressor part and ran through a gel. We then gel extracted and attempted ligation, transformed into E. coli, and plated on 1%, 2%, and 4% xylose-amp plates. We attempted this twice with both times failing.

Thus, we re-evaluated our parts and concluded that a design error may be causing failed ligation. Therefore, we designed PCR primers with flanking Gibson overhangs with the goal of adding these sequences to the part and then reattempting assembly via the Gibson assembly method. We have attempted PCR with the new primers which we verified on a gel, and it appears to be successful. Additionally, we digested the pHY300PLK backbone in preparation for Gibson assembly which we then ran through a gel and extracted. We plan to attempt Gibson assembly in the following week and if successful, we will transform into B. subtilis and perform viability assays.