Results

In the same operon as ismA, the enzyme that catalyzes the first reaction of the cholesterol to coprostanol pathway, we discovered an MFS transporter that we believed could be involved in cholesterol transport. We decided to use integrative vectors and promoters from LMU-Munich 2012's Bacillus subtilis BioBrick Box to clone the MFS transporter into B. subtilis under the control of constitutive promoters of varying strength. We assembled the full MFS transporter-promoter-backbone combinations in B. subtilis first as it is much easier to work with. Promoters were combined with backbones first in six different combinations. We used EcoRI and PstI to combine pBS1C (BBa_K823023) and pBS4S (BBa_K823022), amyE and thrC integration vectors, respectively, with Pveg (BBa_K823003), PlepA (BBa_K823000), and PliaG (BBa_K823002), very strong, strong, and moderate constitutive promoters, respectively. Expected ligation products were pBS1C + Pveg (BBa_K4348106), pBS1C + PliaG (BBa_K4348107), pBS1C + PlepA (BBa_K4348108), pBS4S + Pveg (BBa_K4348109), pBS4S + PliaG (BBa_K4348110), and pBS4S + PlepA (BBa_K4348111). Transformed colonies were screened using restriction digestion with NcoI and XhoI and comparing band patterns with simulated ones on agarose gel electrophoresis.

We first ligated his-tagged ismA, AKR1D1, and AKR1C4 (catalyze sequential steps of the cholesterol to coprostanol pathway) into the pPROEX backbone, a protein expression vector with an ampicillin resistance selection marker. The MCS of pPROEX is downstream of the lac promoter, making our proteins IPTG inducible. gBlocks^TM were obtained for ismA, AKR1D1, and AKR1C4 and digested with NcoI and NotI alongside pPROEX. These were combined in a ligation reaction and transformed into BL21 E. coli. For AKR1D1 and AKR1C4, colonies were screened by miniprepping and performing a double digest with EcoRI and NcoI. For ismA, colony PCR was used with these primers:

ismA_N-HisRemoval_F 5'-TCGACGTGTTGGCTTCAAGG-3'
pPROEX-gBlock_seq_R 5'-TTTCACTTCTGAGTTCGGCATGG-3'

Agarose gel electrophoresis was used to separate resulting DNA fragments by length, and samples whose band pattern matched the virtual digest run on Benchling were deemed positive. These cells will be later IPTG-induced and used for protein expression.

To test the efficacy of our system in our chosen chassis, we wanted to integrate our enzymatic pathway into Bacillus subtilis. The transformed B. subtilis chassis should theoretically express all the proteins necessary to process cholesterol into coprostanol. This moves us one step closer to the ideal final product, a B. subtilis probiotic bacterium that is able to be used for lower cholesterol.

To do this, we designed an integrative vector with all of our genes encoding for ismA, AKR1D1, and AKR1C4 under the control of Pveg (BBa_K823003), a very strong constitutive promoter, and subsequently transforming it into mannitol-induced supercompetent B. subtilis REG19. The vector was created using Gibson assembly, with ribosome binding sites and a terminator being introduced by overhangs in the primers used for the assembly. We first transformed into OneShot TOP10 B. subtilis cells as B. subtilis transformation is much less efficient. Once a colony containing the correct plasmid was identified through colony PCR, its miniprepped DNA was used for B. subtilis transformation.

B. subtilis transformants were screened using colony PCR for presence of the correct insert. The primers used were the following:
Pveg_seq_F 5'-gtgtgcagccgctgaagaata-3'
pBS1C_seq_R 5'-caattttcgtttgttgaactaatgggtgc-3'
The expected amplification length was 3644 bp.

The colony containing the correct insert was cultured in 30 mL of LB broth overnight then lysed to extract the proteins. We tested for the presence of our proteins using anti-his tag western blotting, as all three of our proteins have his tags. AKR1D1 and AKR1C4 have a size of around 37 kDa, while ismA is slightly smaller at 30 kDa. As such, we expect to see two bands at these two sizes if all three proteins are expressed.