Overview

The results of or project are divided into four phases. Firstly, comprehensive literature research in which we looked at scientific papers and consulted experts with varying expertise in order to refine our project idea and find promising enzymes. Secondly, subcloning of enzyme candidates as an intermediate step to generate template plasmids which can be used in following project phases. Thirdly, results of our reference approach, induced expression, are explained in which enzyme candidates were cloned into inducible expression vectors and activity assays were carried out to identify enzymes functioning well in the host Bacillus subtilis. Ultimately, promising enzymes were chosen for our final spore surface display strategy and analyzed by activity assays as well.

---

Phase 1: Literature Research

Decision of Enzyme Candidates

As soon as our team decided on the project topic “ReFiBa – Recycling of Textile Fibers using Bacillus subtilis”, we reviewed the literature, looked at the work of previous iGEM teams and consulted local experts, all of which shaped our project (see Design and Human Practices page).

Our project is divided into two main strategies:
1) Induced expression as a reference approach, in which we aimed to find active cellulases and PETases for working with B. subtilis as expression host.
2) Spore surface display as final strategy, in which chosen enzymes are immobilized on the surface of B. subtilis spores.

First, we needed to decide on enzyme candidates which are capable of degrading cellulose and PET. All details and information, which influenced the selection of enzyme candidates, can be reviewed on the Design page. Finally, cellulase candidates out of the phylum Firmicutes (Bacillota) were chosen as they are most promising for functioning well in B. subtilis. The PETase candidate was already codon optimized for Bacillus (Xi et al. 2021). The following table (Tab. 1) displays all ten enzyme candidates, the species in which they were found and our abbreviation which will be used from now on.

Thereby, we achieved our first milestone:

---

Phase 2: Subcloning

Subcloning of Enzyme Candidates into psB1C3

After choosing ten enzyme candidates, we designed biological parts functioning as translational units containing the ribosome binding site for B. subtilis as well as the gene of interest. The parts were flanked by the BioBrick prefix and suffix sequences allowing BioBrick standard assembly and restriction-ligation-cloning. All information on construct design can be found on the Design page.

In order to create a template from which these parts could be amplified, we aimed to subclone the parts into a small vector pSB1C3. For that purpose, we isolated the plasmid from E. coli DH10β provided by the laboratory collection of Prof. Thorsten Mascher, yielding a DNA concentration of 431.0 ng/µl. Afterwards, we performed a Backbone PCR of pSB1C3 (Fig. 1) and subsequently digested the amplified vector backbone and the ordered parts with EcoRI and PstI, which were purified via the HiYield® PCR Clean-up/Gel Extraction Kit (see Experiments page for detailed protocols).

After ligation, the plasmids were transformed into chemically competent E. coli DH10β cells. Transformants were selected by chloramphenicol resistance (35 μg/ml chloramphenicol) encoded on the pSB1C3 backbone (Fig. 2). For the negative control, no DNA was added during the transformation procedure leading to no cell growth on selection plates. For the positive control, cells were transformed with the vector pSB1C3 resulting in a pink bacterial lawn due to the RFP insert. On the selection plates of the target constructs, white colonies were tested by Colony PCR and agarose gel electrophoresis whether they contain the correct insert (Fig. 3).

If they showed a band with the correct size of the insert, these colonies were chosen for plasmid isolation according to the HiYield® Plasmid Mini DNA Kit (Tab. 2). Finally, ten plasmids containing one of our ten enzyme candidates each were verified via sequencing by Microsynth Seqlab GmbH. These plasmids served as templates for PCR amplification of parts required for the next phases of our project.

---

Phase 3: Induced Expression

Cloning Enzyme Candidates into Inducible Expression Vectors

After subcloning our biological parts into a small vector pSB1C3, we started project phase 3 focusing on testing functionality and activity of chosen enzyme candidates. For that purpose, the parts were cloned into xylose inducible expression vectors in order to overexpress our target genes. We chose both replicative and integrative vectors, pBS0E-xylR-PxylA and pBS2E-xylR-PxylA respectively (Popp et al. 2017). Details and vector maps are documented on the Design page.

For that purpose, the vectors were isolated from E. coli DH10β provided by the laboratory collection of Prof. Thorsten Mascher, resulting in DNA concentrations of 67.5 ng/µl pBS0E-xylR-PxylA and 165.9 ng/µl pBS2E-xylR-PxylA. These vectors were digested with EcoRI and PstI (Fig. 4) and purified via gel extraction using the HiYield® PCR Clean-up/Gel Extraction Kit.

Biological parts, already digested with EcoRI and PstI in the subcloning phase, were ligated into the digested expression vectors. In case there was not enough restriction product left, parts were amplified via PCR (Fig. 5) with the plasmids from project phase 1 as template, e.g. pSB1C3-BhBglA, and subsequently digested and purified via PCR clean up (see Experiments page for detailed protocols).

After ligation, the plasmids were transformed into chemically competent E. coli DH10β and transformants were selected by ampicillin resistance (100 μg/ml ampicillin) encoded on the vector backbone. However, no colonies grew on the plates (LB + ampicillin) for some constructs (Fig. 6 A), probably due to very low transformation efficiency. For those constructs, either the entire vector or the amplified vector backbone was digested, purified by PCR clean up instead of gel extraction and finally led to successful transformation (Fig. 6 B). Using this approach, we did not get rid of the original vector insert containing RFP. However, colonies with religated plasmids appear pink on the plate due to the RFP insert and can therefore be distinguished from transformants with the target insert which appear white.

White colonies transformed with the expression plasmids were analyzed by Colony PCR and agarose gel electrophoresis (Fig. 7). If they displayed a band with the correct size of the insert, these colonies were chosen for plasmid isolation. Finally, ten replicative and ten integrative expression plasmids containing one of our ten enzyme candidates each were verified by sequencing.

Ultimately, these expression plasmids had to be transformed into our target host B. subtilis. Since our reference strategy focused on the secretory expression of target enzymes (except for β-glucosidases), we chose WB800N as expression strain, a genetically engineered variant of B. subtilis 168 in which all extracellular proteases are disrupted. This eight-extracellular-protease-deficient mutant is widely used in industry as it increases the stability of secreted proteins (Jeong et al. 2018).

Since we discovered background activity in initial quality activity assays (see Activity assays of secreted and intracellular enzyme candidates), we aimed to generate B. subtilis WB800N deletion strains (ΔeglS, ΔbglH) in order to remove the background signal caused by the native endoglucanase EglS and β-glucosidase BglH. For that purpose, we wanted to replace the gene with a resistance cassette, as a simple way of knockout. First, we amplified the spectinomycin resistance cassette and an up and downstream sequence of the gene to be knocked out by PCR. These fragments were then joined together in an Overlap PCR (see Experiments page for detailed protocols). Here, the first difficulties arose, as the ratio and concentration of the different fragments to each other are crucial for efficient amplification of the target product. Several ratios and concentrations were tested (see Experiments), with 10 ng each of up and down fragment and 30 ng of spectinomycin cassette being the best tested option (Fig. 8).

However, transformation of the constructs into B. subtilis WB800N failed several times with no growth or a few false colonies. Even an attempt to use a different resistance cassette (tetracycline) resulted in problems with the efficiency of construct amplification and no transformation success. After multiple failed attempts to generate those deletion strains, we decided to discard this idea and to transform the expression plasmids directly into WB800N. Being aware that there is some background activity, this strain will be included as control in all activity assays. The transformants were selected by MLS resistance (1 µg/ml erythromycin and 25 µg/ml lincomycin) encoded on the vector backbones ( ).

Since the Bacillus transformation did not work out for all plasmids, the transformation was repeated with earlier addition of DNA to growing WB800N cells (at OD600 ~ 0.7 instead of 1.1) to not miss the timepoint of competence. Afterwards, cells were grown until OD600 ~ 1.1 and the same procedure was followed as in the initial protocol (see Experiments page).

Finally, all remaining plasmids could be transformed into WB800N and colonies were verified by Colony PCR, whereas both upstream and downstream integration into the lacA locus was checked (Fig. 10). Two colonies each with correct insert size were chosen for cryo conservation serving as biological duplicates.

In the end, we successfully generated all 20 expression plasmids, both integrative and replicative, as well as their corresponding E. coli and B. subtilis strains (Tab. 3, Tab. 4), whereby the latter were used for subsequent activity tests.

Additionally, we noticed that our replicative vector pBS0E-xylR-PxylA does not contain a terminator after the insertion site of our constructs. Therefore, we cloned all replicative plasmids with an additional terminator as a backup option in case there were problems with the expression.
For this purpose, the terminator B0014 was amplified from a plasmid template, kindly provided by the laboratory collection of Prof. Thorsten Mascher, and purified by gel extraction. All generated replicative plasmids (Tab. 3) were digested with SpeI and PstI. The terminator was digested with XbaI and PstI. Both the plasmids and the terminator were purified via PCR clean up. The terminator was then ligated into each replicative plasmid due to the complementary sticky ends of XbaI/SpeI and PstI/PstI (see Experiments page for detailed protocols).
The replicative plasmids containing the terminator B0014 were transformed into chemically competent E. coli DH10β and transformants were selected by ampicillin resistance (100 μg/ml ampicillin) encoded on the vector backbone. White colonies were tested for the presence of the terminator by colony PCR and agarose gel electrophoresis (Fig. 11). If they showed a band of the correct size, these colonies were selected for plasmid isolation using the HiYield® Plasmid Mini DNA Kit (Tab. 5). Finally, the integration of the terminator into all replicative plasmids was verified via sequencing by Microsynth Seqlab GmbH.

Activity Assays of Secreted and Intracellular Enzyme Candidates

---

Phase 4: Spore Surface Display

Cloning Enzyme Candidates fused to Anchor Protein into pBS1C

After choosing enzymes for spore surface display, we finally entered project phase 4 in which we aimed to immobilize enzymes on Bacillus spores. In order to express target genes only under sporulation, a sporulation-dependent promoter PcotYZ was chosen. In order to anchor the target enzymes on the spore surface, they were fused to the N-terminus of the anchor protein CotY. We first tested N-terminal fusions, but due to limited time capacities we did not manage to test C-terminal fusions as well. Moreover, different linkers were analyzed: L1 – a short flexible linker, L2 – a long flexible linker and L3 representing a rigid linker. All details about construct design are documented on the Design page.

First, all biological parts including enzyme candidates with varying linkers (Fig. 12) as well as the promoter PcotYZ, the terminator B0014 and the anchor gene cotY were amplified by PCR (Fig. 13). Plasmids generated in phase 2 (Subcloning, Tab. 2) served as templates for PCR of enzyme candidates with addition of linkers by oligonucleotides. Promoter and anchor genes were amplified from genomic DNA of B. subtilis W168 and the terminator from a plasmid provided by the laboratory collection of Prof. Thorsten Mascher.

Subsequently, parts were assembled via Overlap PCR by complementary overhangs which were designed and added by oligonucleotides (Fig. 14). All Overlap PCR products of BsEglS, BhBglA, PpBglB and BhrPET could be constructed. However, difficulties occurred during cloning of larger enzyme candidates AtCelO and AtCelS, whose parts could not be assembled in time, probably due to wrong assembly caused by similarity of overhangs with some sequences in the parts. Out of the exoglucanases only one Overlap PCR product was successfully constructed: PcotYZ-AtCelO-L2-cotY-B0014.

Afterwards, Overlap PCR products were cloned into the vector backbone pBS1C enabling genome integration into the amyE locus in B. subtilis (Radeck et al. 2013). For that purpose, inserts as well as pBS1C were digested with EcoRI and PstI, whereas inserts were purified by PCR clean up and the vector by gel extraction (Fig. 15) using the HiYield® PCR Clean-up/Gel Extraction Kit (see Experiments page for detailed protocols).

After ligation, the plasmids were transformed into chemically competent E. coli DH10β cells and transformants were selected by ampicillin resistance (100 μg/ml ampicillin) encoded on the vector backbone (Fig. 16). However, no colonies grew on the plates (LB + ampicillin), probably due to very low transformation efficiency. As we already encountered these problems during phase 3 of our project, we repeated the restriction of pBS1C with purification by PCR clean up instead of gel extraction (DNA concentration: 25.0 ng/µl, 29.2 ng/µl). Using this approach, we did not get rid of the original insert containing RFP, but colonies with religated plasmids appear pink on the plate due to the RFP insert and can therefore be distinguished from correct transformants. Transformation plates of spore display plasmids based on vector purification by PCR clean up are shown below (Fig. 16).

White colonies were analyzed by Colony PCR and agarose gel electrophoresis (Fig. 17). If they displayed a band with the correct size of the insert, these colonies were chosen for plasmid isolation and verification by sequencing. All plasmids of BsEglS, BhBglA and BhrPET with three varying linkers were successfully generated. For PpBglB, only the constructs with linkers L1 and L3 could be built. The PpBglB-L2 assembly product resulted in white transformants with correct insert size which were subsequently sequenced, but none of the inserts had the correct sequence. After repetition of ligation, E. coli transformation and Colony PCR, no correct colonies could be found anyway. Unfortunately, we could not manage to generate this construct in time. Upon exoglucanases, only the AtCelO-L2 plasmid could be constructed, due to cloning difficulties already explained above.

Ultimately, these plasmids had to be transformed into our target host B. subtilis using the normal wildtype W168 this time. The protease deficient strain WB800N was not required for the spore display approach since proteins are not secreted in this case. Transformants were selected by chloramphenicol resistance (5 μg/ml) encoded on the pBS1C vector backbone (Fig. 18).

The successful integration into the genome was verified via starch assay (Fig. 19). Four colonies per construct were transferred onto replica and starch plates. In case of successful integration into the amyE locus, the native amylase of Bacillus is not produced correctly resulting in the inability to degrade starch. Correct transformants displayed no halo around the cells, in contrast to the wildtype W168 as control whose amylase is functional leading to a clear halo. Two biological duplicates were chosen for cryo conservation.

In the end, we successfully generated 12 of 18 planned spore display plasmids (six enzyme candidates with three linkers each) as well as their corresponding E. coli and B. subtilis strains (Tab. 6), whereby the latter were used for subsequent activity tests.