Results

Gibson Plasmid A

Figure 1. Gibson Plasmid A containing parts 1A (pancreatic lipase gene), 2A (pancreatic colipase gene), 3A (T7 RNA polymerase gene), and 4A (cre binding sites + lacI gene).

On July 22nd, we attempted Gibson Assembly with our “A” parts, specifically 1A (lipase), 2A (colipase), 3A (T7 RNA Polymerase), and 4A (cre-binding site component) (Figure 1). After combining the inserts with the vector backbone (the pBS1C plasmid, digested from EcoRI to PstI) and the Gibson HiFi Master Mix (New England Biolabs, E2611S), we transformed NEB® 5-alpha Competent E. coli (High Efficiency) (New England Biolabs, C2987H) NEB Competent E. coli cells with our Gibson Assembly plasmid. The pBS1C backbone permits red-white screening to ensure that the transformation was successful. The digest from EcoRI to PstI should remove the RFP gene from the backbone, and thus any recircularized plasmids should appear white. Any red/pink plasmids are undigested plasmid and thus do not contain the desired inserts.

On July 23rd, we observed growth of one white colony on the LB-ampicillin plate (Figure 1C). This indicates that the Gibson Assembly may have been successful, as the cells need a recircularized vector containing the ampicillin resistance gene to successfully grow on the selective media. We established an overnight culture with this colony, and then mini-prepped the DNA (Zymo Research, ZymoPURE Plasmid Miniprep Kit, D4211). The DNA was used for analytical digests in the laboratory, but was also sent to GenScipt for sequencing.

Figure 2. Gibson Assembly constructs assembled into NEB super competent DH5A. All colonies grown on LB-ampicillin medium. A. Transformed with pUC19 control vector. Several colonies present. B. Transformation with no insert. No colonies present. C. Transformation with Gibson Assembly A constructs. 1 white colony present (red box). D. Transformation with Gibson Assembly B constructs, 3 white colonies present (red box).

Figure 3. A. 1% agarose gel with ten lanes of linearized DNA. Lanes (from left to right) loaded with: Quick-Load® 1 kb Plus DNA Ladder (New England Biolabs, N0469S), blank, pBS1C, blank, A1: Gibson Assembly A colony 1, blank, B1: Gibson Assembly B colony 1, B2: Gibson Assembly B colony 2, B3: Gibson Assembly B colony 3, blank. All DNA samples were digested with SmaI. B. Expected results for gel shown in A when DNA is fully digested with SmaI.

Figure 4. Gibson Assembly (from July 22nd, 2024) data and hypothesis. A. pBS1C with part “1A” corresponding to human pancreatic lipase (PNLIP). B. Diagram hypothesizing how only part 1A was inserted during Gibson Assembly. The end Gibson sequence of 1A was able to bind to an overhang created by a PstI cut site.

We attempted multiple additional Gibson Assembly methods to successfully assemble all four of our desired parts in one construct. These included various methods, such as inverse PCR of the pBS1C-1A plasmid, reattempting Gibson Assembly with sequential addition of inserts after incubation periods, and adding the vector last to permit attachment of inserts first, but none of these were successful.

Gibson Plasmid B

Throughout the month of July, we performed many attempts at Gibson Assembly with our “B” parts, but these proved to be unsuccessful. On July 29th, we reattempted Gibson Assembly with the “B” parts, specifically 1B (lipase), 2B (colipase), 3B (terminator/anti-terminator sequence), and 4B (cre-binding site component). After combining the inserts with the vector backbone (the pBS1C plasmid, digested from EcoRI to PstI) and the Gibson HiFi Master Mix ((New England Biolabs, E2611S), we transformed NEB® 5-alpha Competent E. coli (High Efficiency) (New England Biolabs, C2987H) with our Gibson Assembly plasmid. As this is the same vector used in the Gibson Plasmid A assembly, we were able to use red-white screening to determine if transformation and assembly were successful.

On July 30th, we observed growth of over 25 colonies on the LB-ampicillin plates. This transformation efficiency and growth was much higher than results seen in previous Gibson Assembly attempts, and thus had high potential for success. We established an overnight culture with this colony, and then miniprepped the DNA (Zymo Research, ZymoPURE Plasmid Miniprep Kit, D4211). The DNA was used for analytical digests in the laboratory, but was also sent to PlasmidSaurus for sequencing.

Figure 5. Colonies transformed with Gibson B construct (from July 30th, 2024) on an LB-ampicillin plate.

On August 10th, we received our sequencing results from PlasmidSaurus. We were able to successfully clone all four “B” parts into the pBS1C backbone, producing a pBS1C-1B-2B-3B-4B construct.

Figure 6. Gibson B construct with all four parts: 1B (lipase), 2B (colipase), 3B (RAT-T), and 4b (cre-binding site).

Inverse PCR

After successfully assembling Gibson plasmid B, we still lacked a plasmid containing all our “A” parts. This plasmid, named “Gibson plasmid A,” needed to include parts 1A, 2A, 3A, and 4A. Instead of starting Gibson Assembly from scratch, we attempted to use an existing plasmid containing part 1A to assemble all A parts into one plasmid. We designed primers (International DNA Technologies) to linearize the plasmid containing 1A in a way that exposed the 3’ Gibson Assembly sequence of part 1A so it could be used in a secondary Gibson Assembly reaction. After PCR amplification with the new primers, we attempted Gibson Assembly (using protocol recommended by New England Biolabs) of Gibson plasmid A on August 12th. Several colonies of transformed E. coli JM109 grew, and 18 colonies were selected for screening. The 18 colonies were grown overnight, miniprepped (Zymo Research, ZymoPURE Plasmid Miniprep Kit, D4211), then digested with KpnI. The digested constructs were run on an agarose gel, which was deemed inconclusive due to control bands not aligning with expected band sizes.

Ligation

With Gibson plasmid B, the team first attempted to make a usable construct by inserting part 3A with traditional ligation. Using restriction enzyme sites designed into the Gibson Assembly sequences, we attempted to create constructs: 1B + 3A, 2B + 3A, and 1B + 2B + 3A. After digesting Gibson plasmid B with a combination of RsrII, PstI, and BspDI, and part 3A with the same enzymes, parts were attempted to be ligated together following New England Biolab’s protocol for T4 DNA Ligase. However, after transforming the ligated constructs, we observed no growth of cells on the plates. Thus, it is likely that the ligation was unsuccessful. We repeated the procedure using different ligases (Hi-T4) and different incubation periods/temperatures (overnight, at room temperature, etc.). We also attempted to perform ligation in-agarose. All attempts at ligation were unsuccessful, and we are still troubleshooting why this may have occurred. It is possible that our 3 kb insert (part 3A) was too large to be inserted into the vector without overburdening the cell. However, we have not yet arrived at any conclusions and need to conduct further testing to do so.

Gibson 4A and 4B

To test our glucose switch, we needed to isolate parts 4A and 4B, which correspond to two parts with different cre sites under the control of a constitutive promoter. To simplify assays and testing, we decided each part needed a visible reporter to qualitatively confirm function. Green fluorescent protein (GFP) was chosen as a reporter due to its good characterization in fluorescent assays. To insert GFP into parts 4A and 4B, custom PCR primers for parts 4A, 4B, and GFP—each including complementary Gibson overhangs—were ordered and synthesized by Integrated DNA Technologies (IDT). After PCR amplification with the new primers, the parts and GFP (now with Gibson overhangs) were run through a 1% agarose gel to isolate the desired DNA. The bands were then gel extracted (Zymo Research, Zymoclean™ Gel DNA Recovery Kit, D4007). Parts 4A + GFP and 4B + GFP were then assembled with Gibson Assembly, following New England Biolabs’ recommended protocol. After transformation with competent E. coli JM109 cells, several colonies were selected for screening. An analytical digest using restriction enzyme, BamHI-HF (New England Biolabs, R3136S), was performed on several colonies. The resulting gel aligned with expected results, suggesting Gibson Assembly was successful (Figure 7). Two colonies were sent to Plasmidsaurus for sequencing.

Figure 7. Constructs (4A + GFP and 4B + GFP) from Gibson Assembly attempt were digested with restriction enzyme BamHI.

Sequencing revealed that the selected colonies contained the correctly assembled construct (Figure 8). The construct was miniprepped from overnight cultures of the selected colonies (Zymo Research, ZymoPURE Plasmid Miniprep Kit, D4211), then immediately transformed into B. subtilis 168. If the parts are working correctly, transformed B. subtilis should express green fluorescent protein constitutively, and in the presence of glucose, fluorescence should diminish or disappear. We have yet to observe fluorescence in transformed B. subtilis grown in glucose-free medium, but testing remains ongoing.

Figure 8. Sequencing results from Plasmidsaurus. A. Plasmid with part 4A and GFP. B. Plasmid with part 4B and GFP.

Biocontainment Switch

Our biocontainment switch contains three unique subparts: a xylose induced antitoxin, a temperature sensitive component, and a xylose repressor. The xylose induced antitoxin corresponds to a coding sequence for toxin MaxE under the control of a xylose-sensitive promoter, the temperature sensitive component corresponds to a coding sequence for antitoxin MazF controlled by a temperature-sensitive RBS, and xylose repressor corresponds to the coding sequence for a xylose operon. We initially designed our parts so they could be assembled through traditional ligation; however, all attempts at ligation have failed. When troubleshooting ligation, we recognized a shortcoming in our design: we neglected to include a six base pair overhang after the last restriction enzyme cut site in each synthesized part. We hypothesize this oversight is the reason our ligations have been unsuccessful, as the lack of extra base pairs prevents the restriction endonuclease from binding and cutting appropriately.

To continue assembling our biocontainment switch, we shifted our method of construction to Gibson Assembly. To do this, we ordered custom PCR primers for each part from International DNA Technologies (IDT). The new primers contain overhangs corresponding to Gibson Assembly sequences. After PCR amplification, the parts will be compatible with Gibson Assembly, and we will be able to construct the kill switch. As of October 2nd, Gibson Assembly has not been completed, but preparative steps are being taken. The team has successfully amplified all parts with the new primers, and we will begin assembly soon.