Overview

For our wet lab construct, our main objective is to achieve successful expression in Escherichia coli BL21 (E. coli BL21) of the Bacillus subtilis JA18 endo-β-1,4-glucanase gene (abbreviated as endo) fused to the secretion carrier protein YebF. IPTG-induced expression of the YebF–endo-β-1,4-glucanase fusion enables YebF-mediated secretion, which allows us to quantify the proteins from the periplasm and the culture supernatant. We then evaluate protein yield and catalytic performance using Bradford protein assays and DNS reducing-sugar assays. We will begin with the secretion of endo-β-1,4-glucanase, in future plans, generalize the YebF platform to different enzymes beneficial to treating food waste.

Throughout the experimental process, we follow the Design-Build-Learn-Test (DBLT) cycle. This process helps us rethink and troubleshoot issues we face, improving through every step in the cycle.

We successfully finished three cycles throughout the DBLT cycles, including the amplification of fragments, assembly and transformation (through Gibson assembly method), and assembly through overlap-extension PCR.

Cycle Overview

Plasmid Construct Design

Plasmid constructs were designed using SnapGene. The pET22b(+) vector was selected as the backbone, engineered to carry the endo-β-1,4-glucanase gene, both alone or fused with YebF. Both plasmid constructs insert sequences downstream to the native pelB sequence of the vector. The inclusion of YebF was intended to test whether extracellular secretion of endo-β-1,4- glucanase could be accomplished through the induction of YebF protein.

pET22b(+) - endo

pET22b(+) - YebF -endo

*Only the main features are included above for a clearer view; sizes do not correspond to actual size. For clear info, please check the SnapGene plasmid graphs below.

SnapGene view

Cycle 1

Design 1-1

In the first cycle, we aimed to successfully amplify the fragments that are needed throughout our experiment. The main objective of the whole experiment is to achieve the expression of endo-β-1,4-glucanase with the secretion-enhancing effect of adding YebF through Gibson assembly. To test out our claim, we designed three experimental groups, including:

1. Group X: pET22b(+)
2. Group Y: pET22b(+)+endo
3. Group Z: pET22b(+)+endo+YebF

In order to achieve the three experimental groups, the needed fragments to achieve this include the following:

1. pET22b(+)
2. endo-β-1,4-glucanase without an overlapping region (endo)
3. endo-β-1,4-glucanase with an overlapping region (endo-OR)
4. YebF

We plan to amplify all gene fragments needed through PCR using the HiFi Kapa enzyme kit. We amplified four targeted genes, gaining the linearized gene fragments for future assembly. We also designed eight sets of primers through SnapGene. The primers could be used both for PCR and Gibson assembly.

Build 1-1

The template DNA sequences were gained through two different methods, DNA extraction and DNA synthesis. The pET22b(+) backbone was obtained through plasmid extraction from E. coli provided by a certified Biosafety Level 1 (BSL-1) laboratory in National Taiwan University (NTU). The other DNAs and oligo sequences that were used are obtained through synthesis, as sourcing them from their original hosts would be hard to acquire and costly. We amplified all genes through PCR with an annealing temperature of 62°C and the PCR products were tested through agarose gel electrophoresis to verify fragment sizes. The desired bands were excised and purified for downstream cloning.

PCR time and temperature

Test 1-1

After running the Gel for PCR in Tm62, we realized there were many noise sidebands on the gel. We decided to extract the bands with approximately the right base pairs and do gel purification.

Test 2-0

*We skipped to the next cycle (assembly and transformation) to learn the protocols.

Using the products of gel purification, we then performed Gibson assembly for an incubation time of 1 hour, constructing the following experimental groups (shown in image below), further transforming the gene into E. coli DH5a. However, no visible colonies were shown in the experimental groups

Learn 2-0

Key takeaways:
Colonies that couldn’t grow out are caused by either of the two reasons:

• The fragments were not properly assembled during Gibson assembly. The product of Gibson assembly should be circular, keeping the Ampicillin resistance region, allowing the colonies to grow properly on plates.
• Transformation protocols weren't done well, maybe we didn’t heat shock successfully since it’s our first time doing the protocols.

Learn 1-1

Raising the temperature for PCR is crucial for having less noise bands. We decided to raise the temperatures for the next cycle for a better result. With one group being Tm 63 and another being 65 (only for groups pET22b(+) and endo due to their unclear bands). Also, to use a longer incubation time for Gibson assembly. It was also the first time we performed all protocols. Lots of trials and errors were made along the way and we spent more time revising the protocols afterwards.

Design 1-2

From our previous cycle, the bands are especially unclear in groups of pET22b(+) and endo. That is why we decided to redo PCR, raising the annealing temperature, separating it into 63°C and 65°C for groups pET22b(+) and endo. Proceeding with gel purification of PCR products.

Build 1-2

We redo PCR for groups pET22b(+) and endo, raising the annealing temperature to 63°C and 65°C. Afterwards, we performed gel electrophoresis to check the results for PCR and gel purification.

Test 1-2

After raising the annealing temperature, there were less noise bands this time, however, the bands were still not clear.

The results of PCR for group endo are almost not visible, but we decided to move forward to gel extraction with the remaining DNA from both temperatures.

*Gel purification check with pET22b(+) and endo. The other two groups(endo-OR, YebF) were remnants of the previous cycle.

Unfortunately, the bands faded more and we couldn’t move on to the Gibson assembly. The pET22b(+) also completely disappeared.

Learn 1-2

It is important to avoid proceeding to the next steps sometimes when the initial results are not satisfactory. In this case, the bands for both groups are clearly visible and gel purification would only decrease the amount of DNA, making the absence of visible bands a reasonable outcome.

Design 1-3

In the previous cycle, we decreased the amount of noiseband by raising the annealing temperature. However, the targeted band still isn’t clear enough and fades after gel purification. In order to address this, we decided to change the enzymes to Q5 DNA polymerase. Secondly, the issue seemed to also be associated with the choice of DNA polymerase. To address this, we decided to change to Q5 DNA polymerase from New England Biolabs. Lastly, we prepared the agarose gel with a lower concentration, reducing the concentration from 0.8% to 0.5% to address the lack of clarity of the DNA ladders. This adjustment was intended to improve the resolution of the ladder bands and enhance their visibility during electrophoresis.

Build 1-3

We performed PCR with new improvements and also tested our PCR and gel purification products through gel electrophoresis.

Test 1-3

Learn 1-3

After examining the PCR cycles closely, we realised we forgot to change the activation temperature of the PCR cycle. Q5 enzyme needs a higher activation temperature. Besides, we should separate our fragments into two PCR machines. Because with an overwhelming amount of annealing time, the shorter DNA fragments might start to replicate results we don’t want.

Design 1-4

Addressing the activation temperature of Q5 enzymes, we redesigned and tailored PCR methods regarding the previous challenges and designed many groups to try out the best combination to have purified DNA with a clear band.

We separated the groups into two types of enzymes: Q5 DNA polymerase and KAPA HiFi PCR Kit. Also, we separated the fragments into two machines. With the longer fragments (pET22b(+)) in one with a longer annealing time, and the shorter fragment (endo) in a PCR machine with a shorter annealing time. We only amplified group A [pET22b(+)] and group B (endo) for the KAPA HiFi PCR Kit, but we amplified all four fragments for Q5 DNA polymerase due to the lack of primers.

PCR table

Build 1-4

We performed PCR with the newly designed cycle, tested out the two different types of enzymes. Later, we performed gel extraction to purify the DNA fragments and reconfirmed their size through gel electrophoresis.

Test 1-4

*Extracted both endo upper and lower bands.
After gel extraction (Gel electrophoresis)

Learn 1-4

Through the first cycle, we learned how to properly perform protocols, including PCR, gel extraction, and gel electrophoresis. The importance of temperature control and enzymes is crucial.

Cycle 2

Design 2-1

The main objective of the second cycle is to successfully assemble the desired fragments and introduce the recombinant plasmid into E. coli DH5a. The Gibson assembly was chosen to ligate the fragments together.

We established the following three experimental groups:

One hour of incubation was done after applying the Gibson assembly enzyme kit. Further transformation was done by transmitting the recombinant plasmid into E. coli DH5a.

Build 2-1

We constructed the three groups using Gibson assembly with pET22b(+) as the backbone. Group X served as a vector-only control group, while Group Y carried the endoglucanase gene and Group Z carried both the YebF and endoglucanase gene. We expect to see a change in DNA size after miniprep, with Group Z being the largest, to Group X being the smallest, with no inserts of new fragments.

Each insert was amplified through PCR containing overlapping regions for further ligation. The assemblies were then carefully transformed into E. coli DH5α for cloning. Six colonies of each construct were then picked, carefully extracted DNA then tested out DNA size through gel electrophoresis. We ensured all junctions were scarless and ORFs were intact prior to BL21(DE3) expression.

Test 2-1

The transformed colonies were then plated, where we can see the plate with the most colonies being G$roup Y, followed by Group X, and lastly Group Z. We found that Group Y had large amounts of colonies due to unsuccessful transformation. The colonies are mostly background colonies.

We then further picked six colonies in both groups, extracted the DNA through miniprep, and ran gel electrophoresis to check the base pairs of the colonies.

After running the gel for both groups, we realized that there were some colonies on Group Y’s plate that were significantly larger than others, so we decided to pick some of the larger colonies and test out their DNA size again.

Learn 2-1

First, we identified an operational error in the miniprep step when we saw no visible bands on the gel and corrected the procedure in the next cycle. Even when transformation fails, there should still be a clear DNA band as the e.coli host contains plasmids. Among the bands that we could successfully visualize, there was no expected shift in the bands, indicating a failure of assembly in this cycle. In response, we decided to increase the incubation time of Gibson assembly, hoping to improve the success rate of assembly.

Design 2-2

In order to address the issue we faced at 2-1, we decided to prepare two different groups with different sets of incubation time. With both groups extending the incubation time to 90 min at 50℃, one of the groups will carry out one incubation overnight in the 37℃ incubator. This can potentially increase the success rate in the assembly of the fragments.

This indicates a total of six experimental groups to be tested in the following stages.

Build 2-2

We only successfully constructed four groups, leaving out the control groups due to the lack of enzymes (further explained in learn 2-2). The fragments were amplified throughout PCR, further ligated through Gibson assembly. The assembled fragments were then transformed into E. coli DH5a for further miniprep and gel electrophoresis to test out the ligated DNA.

Test 2-2

The assembled DNA fragments were then transformed into E. coli DH5α.

In the 1.5H incubation situation, Group Z shows an overabundance of colonies, likely due to background growth rather than successful transformation. Group Y(1.5H incubation) also shows numerous colonies despite unsuccessful plating. In contrast, both groups of the overnight incubation show a smaller number of colonies presented. We identify it as a possible chance of successful transformation. We picked out six colonies from both plates of the overnight incubation group and performed plasmid minipreps for downstream analysis.

We extracted the DNA of the four plates, each testing out 6 colonies.

However, there wasn’t a change in DNA size this cycle either.

Learn 2-2

This cycle is a total trial, error, and relearning experience for us, emphasizing the need for control groups. Under reagent constraints and deadline pressure, we omitted the vector-only control group and proceeded with the experiment with the other two groups. However, this action caused us to have an issue with interpreting the colonies. Without a control plate as a baseline, we couldn’t distinguish between transformants or background colonies. We mark this cycle as a learning round and redo it with controls and standardized plating in the next cycle.

Design 2-3

After rebuying our needed reagents, we carried on with the design again. This cycle, we keep the control group, crucial for the evaluation of the number of colonies.

Build 2-3

We performed Gibson assembly for experimental contrasts and transformed the assembled fragments into E.coli DH5α. After comparing the number of colonies with the control group.

We then selected 6 colonies for Groups Y and Z, carrying out plasmid minipreps to obtain DNA for downstream verification through gel electrophoresis.

Test 2-3

In both 1.5H and overnight incubation groups, the number of colonies successfully transformed was less than in the previous cycle, where it looked like background colonies.

We then selected 6 colonies per plate for Groups Y and Z under both incubation conditions (1.5H incubation and overnight incubation) Plasmids' size was then inspected through gel electrophoresis. In the 1.5H condition, Group Y showed a clear upward band shift relative to the vector control (circled in the picture below). We then submitted the corresponding DNA for DNA sequencing.

When we sequenced clones whose gels differed from the control, the results were unexpectedly poor. The T7 promoter–primed reads did not map across the designed junctions, and alignment suggested absent inserts together with compromised backbone regions (T7 terminator, f1 ori, AmpR promoter), pointing to assembly errors rather than true transformants.

Learn 2-3

The method we use to assemble our fragments is Gibson assembly, which means all the fragments are amplified through PCR. In order to decrease the risks of wrong sequences, it is best to decrease the amount of DNA amplified, since we can’t exactly check the products amplified without sequencing them.

We decided to switch to a different assembly method for the next cycle. Since the vector based on the sequencing results shows that the plasmid backbone is questionable, we decided to switch to restriction enzyme cloning methods for the next cycle.

Cycle 3

Design 3-1

After experiencing several failures through Gibson assembly, we decided to switch to classical restriction-ligation cloning with an overlap-extension PCR. We chose BamHI, HindIII enzyme sites to cut the pET22b(+) vector, making it linearized. We then generated a YebF–endo-OR fusion by overlap-extension PCR using Q5 and Taq polymerases (templates: YebF and endo-OR; primers: YebF-F and endo-R). The resulting fusion amplicon was then inserted into the BamHI/HindIII sites of pET22b(+) for downstream cloning.

Build 3-1

We first amplified endo-OR and YebF individually using Q5 high-fidelity polymerase. Next, we fused the two fragments by overlap-extension PCR (using the individual amplicons as templates) to generate the YebF–endo-OR insert. The PCR products were analyzed by agarose gel electrophoresis to verify the expected fragment/fusion size. The products were then purified for downstream BamHI/HindIII cloning into pET22b(+).

Test 3-1

Amplification of endo and YebF with Q5 polymerase was successful.

The assembly of the two fragments was not successful, resulting in a wrong basepair shown in gel electrophoresis.

Learn 3-1

Through this cycle and the previous cycles, we identified that the issue of the unsuccessful assemblies is probably due to the primers we designed. In this cycle, when we decide to assemble the short fragments together with PCR, the greatest determining factor of success is the design of the primer. Our primers have no issue in the amplification of fragments. However, when assembling the fragments together, we face unsuccessful assemblies in every attempt. In future cycles, we aim to redesign the primers for future assembly.

Conclusion

Due to financial and time constraints, we decided to stop our cycle until 3-1. Although we couldn't reach the final milestone, secreting the endo-β-1,4-glucanase protein to the outer membrane through YebF, each cycle allowed us to refine our designs, identify the limitations, and most importantly, to generate a new approach to reach our goal. Even without full success, our repeated DBTL cycles turned the experimental process into a strong foundation that teams could further build upon.