Inspired by the success of overexpression for lsPETase with proper activity (Austin et al. 2018), the idea of creating our own construct with a different strategy was raised. The full sequence of lsPETase from the article was downloaded from NCBI (https://www.ncbi.nlm.nih.gov/protein/GAP38373).

In the article, the author used the pET21 as the backbone of the construct with T7 cassette and Lac operon. We instead chose BioBrick plasmid pET24a(+) which enables fast and efficient cloning (Du et al. 2012, Shepherd et al. 2017). Three biobrick restriction sites: EcoRI, ApoI, XbaI are available in the prefix part and two biobrick restriction sites: SpeI, PstI are available in the suffix part (sequence in Table 1). EcoRI and PstI are chosen to restrict both insert and backbone. T7 cassette with Lac operon is adopted. For easy protein purification, a pelB signal peptide was added at 5' (Shi et al. 2021) and a 6×His tag was added at 3' end (Hochuli et al. 1988).

Due to unexpected troublesome cloning issues with pET24(a)+, another biobrick plasmid pSb1c3 was also used for construct. Plasmid pSb1c3 contains chloramphenicol resistance gene and mRFP gene between biobrick restriction sites. mRFP as a chromoprotein can indicate if the constructs are false positives easily after transformation. However, although it is known that pSb1c3 lacks LacI compared to pET24(a)+, we still wanted to test if the construct can work properly.

Attempt 1

PETase was subcloned using pET24a plasmids and competent DH5α E. coli cells, prepared in the lab with CaCl₂ and heat-shock. DNA fragments carrying PETase were amplified through PCR (Figure 1). These PETase fragments and pET24a were digested with EcoRI and PstI (Figure 2), creating sticky ends that allowed for the ligation of the digestion products with T4 DNA ligase. Before ligation, the digestion went through gel purification to reduce the risk of religation. The ligation constructs, pET24a vector carrying a petase gene could then be used to transform DH5α cells. These would then be grown in a media containing ampicillin, as both the fragment containing PETase and the vector itself, pET24a, contained amp^r genes.

The DH5α allowed for the cloning of the PETase gene with high fidelity. And the resulting product would then be extracted through miniprep and analyzed through electrophoresis in an agarose gel.

However, this protocol ran into some issues. The product of the gel extractions after digestion was often very low in concentration, resulting in poor transformation efficiency.

The overall poor transformation efficiency could have been explained by different factors and, as such, transformation was attempted with the intact, non-digested, pET24a plasmid and compared to that of non-transformed DH5α cells and cells transformed with our construct, containing PETase. While cells containing the religated pET24a plasmid grew well (Figure 3), that did not happen with cells containing the ligation product with the PETase gene (Figure 4). As such, while transformation was possible, it did not seem to happen when a post-digestion gel extraction step was added.

Different strategies were attempted to correct the issue of low gel extraction product concentration:

1. Higher volumes of DNA were loaded onto the agarose gel for each gel extraction, but this did very little to improve concentration of DNA, as this also resulted in larger volumes to process during gel purification. This change to the protocol, alone, could not result in a higher transformation efficiency.

2. Gel extraction was dropped for a while, it seemed to result in too high of a loss of DNA after digestion and it seemed it would be preferrable to risk some colonies carrying religation product compared to not having any product at all. While colonies formed, none of them carried the PETase insert, all of the most likely just religation product (Figure 3). This strategy was immediately dropped afterwards.

Since changes to the gel extraction were not improving the overall efficiency of the experiment, new changes were proposed. Digestion time was changed from 1h, to 3h, to an overnight process. DNA amount used for transformation was doubled (check for amounts). A phosphatase was added to the digestion of pET24a, to reduce the likelihood of religation. These steps, however, did not improve transformation, as either no colonies were formed or the ones that were formed contained simply religated product.

This process was somewhat long due to the number of steps that required an overnight incubation, only made worse by the need to test any seemingly positive results post transformation, to see if they were just carrying religated pET24a, which was proven to be the case for all of the colonies tested. Often, to try to improve the chances of a good result, large amounts of cells were transformed, and even larger amounts of colonies were tested, resulting in an even more time and resource consuming process of elimination.

Attempt 2

The pSB1C3 plasmid contains an mRFP1 gene, which, when intact, gives its colonies a red color. This gene was, in the production of this plasmid, prior to our lab work, inserted into it through biobrick assembly. As such, the plasmid, when used for transformation of DH5α with PETase, using the protocol described above for pET24a, lost its mrfp1 gene and any colonies carrying the pSB1C3 plasmid with a PETase insert but no mrfp1 gene, therefore, lose their red color, taking on the usual “white” shade.

As such, the idea of using this gene to eliminate false positives more quickly, stemming from religation of the pET24a vector, was tested. Any red colonies after the transformation would most likely just be the result of religation while any white colonies, if present, should be indicative of positive results, DH5α cells carrying the PET24a gene.

Cells which carry additional genes on their plasmids, as a result of digestion and ligation, often grow more poorly, under the same, which means that even just a small population of cells carrying religated product, will normally be favored when in competition with the cells carrying the additional gene.

In the case of cells carrying the pSB1C3, as both the religated product, carrying mrfp1, and the intended construct, carrying PETase, carry an additional gene, when transformation is attempted, the small population of religated product should not have the same advantage as a cell carrying just pET24a had in comparison to the PET24a+PETase+ cells did.

Transformation of the same DH5α cells with pSB1C3+PETase constructs showed better results, as a small amount of colorless colonies formed. These were assessed through miniprep extraction of their plasmids, followed by gel analysis, as well as digestion and analysis for more exact results (Figure 5-6), as the mrfp1 was of similar size to the PETase fragment (picture of both gels). This showed that there was now, in pSB1C3, in mrfp1's place, a fragment of a size correspondent to that of the PETase fragment.

The gene was amplified through these DH5α cells and later the construct retransformed into a BL21 cell.

While pSB1C3 had shown positive results during the transformation of DH5α cells with pSB1C3+PETase, when it was retransformed into the BL21 cells, these did not overexpress PETase. This may have come as a result of pressure favoring cells where, for example, as was discussed, a frameshift mutation resulted in the silencing of the PETase gene, which would explain why these colonies were still colorless and carried a fragment of the same size as the gene for PETase but did not overexpress PETase when induced with IPTG.

The pSB1C3+PETase and pET24a+PETase constructs were very similar with a critical difference being pSB1C3’s lack of a lacI gene. While DH5α and BL21 cells have some level of LacI expression in other plasmids, it may not be able to fully stop PETase expression completely, and so it may have been that cells that expressed PETase properly were at a disadvantage by their constant waste of resources to express this enzyme.

Attempt 3

Due to the issues with pSB1C3, transformation with pET24a+PETase constructs was attempted again, this time with NEB5alpha cells, a commercially available strain of high efficiency competent cells. This time, even after gel extraction, a lot of colonies were formed after transformation.

Miniprep plasmid extraction was applied to some of these colonies and 2 colonies carrying PETase were identified (Figure 7). These colonies underwent colony PCR and sequencing. These were restreaked and their plasmid constructs used to transform BL21 cells.

To determine the optimal induction conditions of protein overexpression, two temperatures(37℃, room temperature) and two IPTG concentrations (0.1mM. 1mM) are tested for construct pSb1c3-PETase (Table 2). SDS-PAGE analysis of protein expression is shown in Figure 8. In lane 5, plasmid with mRFP was successfully overexpressed indicating induction was successful. However, possibly due to the lack of LacI, no intensive expression of PETase was observed (Figure 8). As a result of no uninduced sample as negative control, we can't really prove the weak band PETase or background. Considering how weak the band is, even though it is PETase, the amount of expressed PETase is negligible. So our plan was to take pET24a(+) back to the table again.

With the LacI gene in pET24a(+), it enables us to induce the protein expression at desired growing period with IPTG reducing the leaky expression. At the same time, T7 promoter initializes the transcription by T7 polymerase which is also regulated by LacI operon located in the chromosome of E. coli BL21. As a result of introducing LacI in plasmid, the growth and protein expression of transformants could be improved. The design of 6×His tag at C terminal with affinity selectively binds to immobilized metal ions which make purification with nickel column feasible.

Two colonies of BL21-pET24a-PETase transformant colonies were picked out and overnight cultured for inoculation on the second day. Both the overnight culture media and the overexpression media were LB liquid medium with ampicillin. IPTG was added to the culture when OD₆₀₀ reached 0.6 making IPTG final concentration at 1mM. To ensure that enough amount of protein for later purification, media volume was increased to 50ml for two induced BL21-pET-PETase cultures while the other four cultures were 10ml. Whole-cell protein and secreted protein in 80 uL culture media were collected and then analyzed with 12% SDS-PAGE (Figure 9). Induction conditions for all inductions were set with starting OD₆₀₀=0.6, IPTG 1mM, 37℃ overnight culture on a shaker.

It was at first concluded that protein was successfully expressed however the purification results afterward disagreed (Figure 10-11). Before purification, the cell pellet was collected by centrifuge, resuspended with sonication solution (buffer A), and then sonicated. Supernatant from sonicated samples was centrifuged and is then collected and filtered with 0.22μm filter before purification with a column. Purification was conducted with a nickel column and eluted with different concentrations of imidazole solution (see Figure 10). Surprisingly, no protein band is observed after purification. All experiments related to protein purification were conducted on ice to maintain its activity.

Later, in case of any mistake during purification, a repeat overexpression and purification of PETase with IF3 as positive control using the same conditions was carried out. Additionally, possible secretive expression because of pelB signal peptide was also analyzed with sample from supernatant of media. In Figure 12, both PETase and IF3 were successfully overexpressed and purified, however, only a neglectable amount of secretive expression was found. From SDS-PAGE, we can conclude that we managed to express the PETase through with other unexpected bands found after purification.

After proving that small-scale protein overexpression and purification are successful, an upscale experiment was conducted at Testa Center. A bioreactor with an automatic controlling system and 2.5L capacity was then used to upscale the bacteria culture. Anti-foam solution and ammonia were programmed to be added to prevent the over-formation of foam and maintain a stable pH level. O₂ and pH sensors were applied to detect dissolved O₂ level and pH while stir rate and temperature are also recorded. The induction condition was the same as in the small-scale experiment. The summary of growing status during the cultural process is shown in Figure 13.

Bacteria and media were separated by centrifuge at 4℃ and the bacteria pellet was resuspended with sonication buffer (buffer A). The bacteria pellet suspension was then crushed by Homogenizer and centrifuged again to collect the supernatant. To determine if any secreted protein in media, both cell protein and media were run through and purified by Chromatography Systems with a preset program. Flow through after sampling loading and fractions with UV absorption were collected. The purification result is shown in Figure 14 and Figure 15. The second peak (retention volume around 649.63 ml) in Figure 14 and the third peak (retention volume around 617.82ml) were believed to be PETase. Each elution volume is around 20ml.

The collected flow through and fractions were then analyzed with SDS-PAGE (Figure 16). The sample is overloaded thus it became impossible to identify each band separately, due to which, another SDS-PAGE analysis was carried out using diluted sample with uninduced sample and flow through as controls (Figure 17). Protein band of the correct size was observed and believed to be PETase. Compared to small scale protein expression, more unwanted protein and less target protein are found in SDS-PAGE analysis of purification in upscale experiments. One possible explanation of this phenomenon is that PETase is not completely soluble expressed in E. coli. When it is expressed on small scale, the protein production is not that fast so the majority of the protein was soluble, however, when reached to upscale, a bioreactor system can provide more optimal conditions for bacterial growth and accelerate the protein production process which may form more inclusion bodies that are not soluble after cell disruption and thus are not collected. Therefore, the protein yield in Testa Center is not as high as expected.

Activity assays were later conducted with the purified PETase.

Attempt 1 - Lysed cells

To inspect whether the produced PETase was active, we designed an activity assay based on found publications (Pirillo et al. 2021, Zhong-Johnson et al. 2021) and results from previous iGEM teams (TU Kaiserslautern 2019, TJUSLS China 2021). Since the PETase was not purified successfully in the beginning, only cell lysates were used in the first assay set up. Samples with cell lysates were tested for PETase activity in various conditions (Figure 18). Activity was measured at four different temperatures (room temperature, 30 °C, 37 °C and 42 °C), two types of treatment (with DMSO and with Triton X-100) and with two types of PET (powder and pellets). To measure potential activity of secreted PETase, unlysed cells in growth media were incubated at 37 °C and measured as the rest of the samples.

The measurement of activity was done spectrophotometrically at 260 nm. MHET emerging from the PETase reaction with PET absorbs the strongest between 240–244 nm, but DMSO could interfere in that range (Pirillo et al. 2021). Cell lysates however contained too much nucleic acids and other compounds absorbing at 260 nm, which made measurement of PETase activity impossible.

Attempt 2 - Purified protein

For the second attempt at activity assay, purified enzyme from scale up experiments at Testa center was used. An easier set up with less conditions was designed (Figure 19), where PETase activity was measured only at 37 °C and with triton treatment. Samples for measurement were taken and prepared in the same way as in the first attempt and then diluted 5 times with dH₂O before measurement. Spectrophotometric measurement was done at 242 nm since no DMSO was used this time and therefore could not interfere. Nevertheless the measured absorbances were always negative, no clear trend of absorbance change was observed (Figure 20) and therefore the activity of purified PETase could not be proven. The PETase samples were kept in the elution buffer from the purification process and the present imidazole and Tris could interfere, which would explain our inconclusive results.