Expression of subtilisin E in E. coli

Our aim here was to express the natural subtilisin E in E. coli and to determine the optimal conditions.

Building the expression system

Day 1

Our aim was to create a functional expression system for subtilisin E in Escherichia coli. At first, we had to optimize the sequence of the subtilisin E gene from Bacillus subtilis for Escherichia coli codon usage. For this purpose, our DNA sequence was improved with the DNA and protein sequence analysis tool ‘Geneious’ and the ‘Codon Optimization Tool’ from IDT. In the following, we proceeded to design the expression system digitally.

The expression system consists of commonly used BioBricks:

This construct was ordered at IDT as a gblock and named ‘EI0002’.

We decided to use the pSB1C3-backbone from iGEM. To amplify the backbone, we transformed BBa_J04450 (RFP expression system in a pSB1C3 backbone) in Escherichia coli BL21.

Day 2

The ordered DNA sequence EI0002 was resuspended to a concentration of 20 ng/µL and amplified via PCR with the primers shown below.

Multiple identical samples were prepared.

Afterwards, the size of the PCR products was examined via agarose-gelelectrophoresis.

As all samples showed the correct size, the DNA concentration was increased by purification of multiple samples through one column during PCR clean-up.
In addition to this, multiple overnight cultures of 5 mL were prepared with the plates of the transformation from BBa_J04450 in E. coli BL21 from the day before. As controls, overnight cultures with E. coli BL21 wildtype were prepared, with and also without addition of antibiotic.

Day 3

We prepared cryo cultures of the overnight cultures of BBa_J04450 in E. coli BL21 and isolated the plasmids from the cells for further experiments.

Day 4

In the following, we cloned the PCR-amplified, linear expression system EI0002 from day 1 into the pSB1C3 backbone of BBa_J04450. Both samples were digested with the enzymes EcoRI and PstI. After that, the samples were ligated for one hour at room temperature and then transformed into E. coli BL21. All controls were prepared according to the protocol.

Day 5

To find a successfully ligated plasmid, we screened 24 colonies via Colony PCR with primers that bind outside of the insert to pSB1C3 and in parallel prepared overnight cultures of the same colonies. The samples were tested via agarose-gelelectrophoresis after the PCR.

A few colonies showed the correct size and were therefore used for further experiments. The overnight cultures of all negative colonies were discarded.

Day 6

We prepared cryo cultures of the overnight cultures of the ligation product in E. coli BL21 and isolated the plasmids from the cells. The purified plasmids were sent in for sequencing.

Day 7

All sequenced samples show fatal mutations. Either frameshifts in the coding sequence or multiple deletions in the promoter occurred.
Despite repeating the experiment above multiple times, we were not able to obtain an expression system for subtilisin E without fatal mutations. Combined with the experience of other iGEM teams that the promoter BBa_R0011 shows some degree of expression before the addition of the inducer, we assumed that subtilisin E prohibits cell growth of E. coli if expressed in its active form.
As a result, we progressed by exchanging the promoter.

Day 8

To exchange to promoter of our expression system, we started with extracting parts of it via PCR. Hence, we used the primers shown below to obtain a sequence with the secretion tag, the subtilisin E gene and a terminator that is flanked by the BioBrick prefix and suffix.

Afterwards, we cloned this sequence in the BioBrick BBa_J04500 with the restriction sites EcoRI and PstI. BBa_J04500 is a protein expression backbone which consists of the lacI promoter BBa_R0010 and the ribosome-binding site BBa_B0034 integrated in the BioBrick standard vector pSB1C3. Then, we transformed the obtained construct in E. coli BL21 together with all needed controls.

Day 9

We prepared 5ml overnight cultures of the colonies and the usual controls.

Day 10

At first, we prepared cryo cultures of the overnight cultures and then isolated the plasmids from the cells. In the following, we linearized the purified plasmids with EcoRI and tested their size via agarose-gelelectrophoresis.

The samples with the correct size were sent in for sequencing.

Testing the Expression

Day 1

For the analysis of the expression level we needed an empty backbone control. Hence, we digested BBa_J04450 with XbaI and SpeI and ligated the sample without addition of a further insert. The ligation product was then transformed in E. coli BL21.

Day 2

Some colonies of the ligation product appeared to be pink. This was only possible due to a re-ligation of the RFP-expression cassette into pSB1C3. Therefore, these colonies weren’t used for further experiments. Only white colonies were used for the preparation of 5ml overnight cultures.

We prepared cryo cultures of the overnights of the empty backbone and purified the plasmids from the cells. Afterwards, we checked the size of the plasmids via agarose-gelelectrophoresis.

One of the samples that showed the correct size was sent in for sequencing.

Day 4

One of the samples of the PCR extracted expression system in BBa_J04500 that was sequenced beforehand and also the sample with the empty backbone showed the correct sequence and could therefore be used for further experiments.
To analyze the expression level of subtilisin E we prepared 20ml pre-cultures of the positive sample. In addition, we prepared pre-cultures of the E. coli BL21 wildtype and the pSB1C3 empty backbone in E. coli BL21 as controls.

Day 5

The pre-cultures from the day before were used to inoculate two 50ml main-cultures each to an OD₆₀₀ of 0.1. This main-cultures were then incubated at 37°C, 250 rpm until OD 0.3 was reached. Then, the main-cultures were induced by adding 0.5mM IPTG. Afterwards, one of the main-cultures was incubated at 30°C and the other at 25°C. We took samples of 500µL hourly and continued the incubation overnight.

Day 6

We took one sample of each main-culture from the day before and then lysed all samples in a sonication bath for about 30 minutes. The supernatant and the pellet of these samples were then tested separately via SDS-gelelectrophoresis.
Unfortunately, we were not able to obtain any clear results from the SDS gel and thus could not verify the expression of subtilisin E.

Testing the folding and activity

Day 1

We continued our experiments by performing a skim milk assay on agar plates. This assay utilizes the clearance of skim milk due to proteolysis to detect proteolytic activity. Therefore, we poured LB skim milk agar plates containing IPTG and the needed antibiotic and streaked the E. coli BL21 cells containing the plasmid with the expression system.

Comparing the clearance of the skim milk plates, a proteolytic activity could be observed for the cells containing the expression system for native subtilisin E. As a result, we concluded that within three days these cells are able to produce the native protease which will then digest the skim milk in the agar plates, resulting in a clearance.

Conclusion

We were able to express subtilisin E in E. coli in amount high enough to be detected via skim milk assay.

Mutating the expression system

This phase of the project had two aims:
At first, we wanted to show that both the active site and the pro-peptide cleavage site of subtilisin E are essential for the activity of the enzyme. Thus, we planned to simulate the incorporation of ONBY and DMNBS by exchanging both amino acids at the target sites against bigger amino acids and then to analyze the activity of the mutated enzyme as a proof of principle.
Our second goal was to mutate our current expression system in order to exchange serine in the active site and tyrosine in the pro-peptide cleavage site against the UAG amber stop codon in preparation for the expression of a photocaged enzyme.

Site-directed mutagenesis

Day 1
To prepare the expression of a photocaged protease and to prove our principle, we performed a site-directed mutagenesis. Four different strategies were pursued:

As a proof of principle we exchanged serine in the active site against tyrosine and tyrosine in the pro-peptide cleavage site against tryptophan. In preparation for the expression of a photocaged enzyme we exchanged the codons for the targeted amino acids against the UAG amber stop codon.
To achieve this, we simultaneously performed the site-directed mutagenesis three times (SDM 1, 3 and 4) with fitting primers and our expression system after the exchange of promoters as a template. SDM 2 had to be performed separately and with SDM 1 as a template at a later time.
For SDM 1, 3 and 4 we executed the following PCR:

Multiple samples were prepared for each SDM. After completion of both PCR programs the size of all samples was verified via agarose-gelelectrophoresis

As all samples showed the correct size, identical samples were purified through one column during PCR clean-up to increase the DNA concentration.
After that, the DpnI digest was set up and incubated at 37°C overnight.

Day 2

In the beginning we performed a heatkill with the DpnI-digested samples of SDM 1, 3 and 4. Then, all samples were transformed in E. coli BL21 together with all needed controls.
Day 3

Multiple 5ml overnight cultures were prepared for each SDM approach together with all needed controls.

Day 4

We prepared cryo cultures of the overnight cultures of SDM 1, 3 and 4 in E. coli BL21 and isolated the plasmids from the cells to use them for further experiments. A few samples were sent in for sequencing.

Day 5

The sequencing results showed that we were able to obtain samples with the desired, modified sequence. Hence, we used one sample with a confirmed sequence of SDM 1 as a template for SDM 2. In the following, we executed the following PCR:

Multiple identical samples were prepared. After completion of both PCR programs the size of all samples was verified via agarose-gelelectrophoresis.

As all samples showed the correct size, identical samples were purified through one column during PCR clean-up to increase the DNA concentration.
After that, the DpnI digest was set up and incubated at 37°C overnight.

Day 6

We performed a heatkill with the DpnI-digested samples of SDM 2. Then, the sample was transformed in E. coli BL21 together with all needed controls.

Day 7

	Growth?
SDM 2	Yes
Control 2: RFP	Yes
Control 3: BL21 WT without insert	No

Multiple 5ml overnight cultures were prepared together with all needed controls.

Day 8

	Growth?
SDM 2	Yes
Control 1: BL21 WT with antibiotic	No
Control 2: BL21 WT without antibiotic	Yes

We prepared cryo cultures of the overnight cultures of SDM 2 in E. coli BL21 and isolated the plasmids from the cells. Some samples were sent in for sequencing.

Day 9

The sequencing results showed that in some samples we successfully modified our expression system with the desired mutations.

Testing the expresion

In the following we wanted to analyze the expression level of the different, mutated versions of subtilisin E.
To express a photocaged enzyme not only a stop codon at the target site is needed but also an orthogonal tRNA/synthetase pair that incorporates the desired non-canonical amino acid in response to the amber stop codon. While targeting the pro-peptide cleavage site, we wanted to use the ONBY-tRNA/synthetase pair for E. coli which derives from the tyrosyl-tRNA/synthetase pair from Methanococcus janaschii. This year, the DNA sequence for the ONBY-tRNA/synthetase-pair was available on the iGEM kit plate as BioBrick BBa_K1416000. However, this part was only available in pSB1C3. This means, that the plasmids for both our expression system and the tRNA/synthetase-pair have the same origin of replication and therefore are competing against each other in terms of which of them will be replicated. Therefore, we weren’t able to use the BioBrick BBa_K1416000 for our experiments.
Fortunately, we were able to contact members of the former iGEM team Texas, Austin from 2014 who originally created the part BBa_K1416000. They sent us the ONBY-tRNA/synthetase pair in a backbone that carries a gentamicin resistance and is replicated with an ori that is not part of the same incompatibility group as the ori of pSB1C3.

This allowed us to use the ONBY-tRNA/synthetase pair that is described as BBa_K1416000 without the need of cloning it into a different backbone.

Moreover, the experiments to test the incorporation efficiency of the ONBY-tRNA/synthetase pair showed that this pair is also able to incorporate DMNB-serine to a certain degree. This is why we decided to try to incorporate DMNBS in our SDM 4-modified enzyme with this tRNA/synthetase pair.

Day 1

We received the ONBY-tRNA/synthetase pair from team Texas, Austin 2014 as dried DNA on filter paper. At first, we cut the marked region from the filter paper, put it in a 1.5ml tube and immersed it in 100µl Tris-HCl pH 8.0. After incubation for 10 minutes, we used 2µl for a transformation in E. coli BL21 and in parallel transformed the usual controls.

Day 2

	Growth?
ONBY-tRNA/synthetase pair	Yes
Control 2: RFP	Yes
Control 3: BL21 WT without insert	No

Multiple 5ml overnight cultures were prepared for each SDM approach together with all needed controls.

Day 3

	Growth?
ONBY-tRNA/synthetase pair	Yes
Control 1: BL21 WT with antibiotic	No
Control 2: BL21 WT without antibiotic	Yes

Cryo cultures of the ONBY-tRNA/synthetase pair in E. coli BL21 were prepared and we isolated the plasmids from the cells. One sample was sent in for sequencing.

Day 4

The sequencing results showed that the sample had the correct sequence and could therefore be used in further experiments.
In the following we executed a co-transformation in E. coli BL21:
SDM 2-modified expression system and ONBY-tRNA/synthetase pair
SDM 4-modified expression system and ONBY-tRNA/synthetase pair
During the transformation 2µl of each plasmid sample were used and the cells were plated on LB agar plates with two antibiotics.

Day 5

	Growth?
Co-transformation	Yes
Control 2: RFP	Yes
Control 3: BL21 WT without insert	No

We prepared multiple 5ml overnight cultures of each co-transformation together with the needed controls.
As the transfer from the solid complex medium to liquid minimal medium causes a high stress-level for the E. coli cells, we intended to change at first from solid complex medium to solid minimal medium and then to liquid minimal medium. Hence, we streaked the same colonies that we used for the overnight cultures on M9 agar plates.

Day 6

	Growth?
Co-transformation	Yes
Control 1: BL21 WT with antibiotic 1	No
Control 2: BL21 WT with antibiotic 2	No
Control 3: BL21 WT with antibiotic 1 + 2	No
Control 4: BL21 WT without antibiotics	Yes

To back up the co-transformed cells in M9 medium we prepared cryo cultures of the overnight cultures. As the cells were able to grow in medium with two antibiotics, both plasmids must have been successfully transformed into them. As a consequence, these cells can be used for further experiments. Continuing the transfer to liquid minimal media, we prepared 5ml overnight cultures with the positive clones from the M9 agar plates in liquid M9 medium.

Day 7

In the following we wanted to start with the analysis of the expression level. For that reason, we prepared different pre-cultures:
SDM 1-modified expression system in complex media
SDM 2-modified expression system + ONBY-tRNA/synthetase pair in minimal media
SDM 3-modified expression system in complex media
SDM 4-modified expression system + ONBY-tRNA/synthetase pair in minimal media
Empty backbone pSB1C3 in complex media
Empty backbone pSB1C3 in minimal media
E. coli BL21 wildtype in complex media
E. coli BL21 wildtype in minimal media
All pre-cultures were incubated at 37°C, 250 rpm overnight.

Day 8

We then proceeded to prepare two 50ml main-cultures for each pre-culture from the day before. All main-cultures were prepared with the same medium as the pre-culture. To the cultures with minimal medium we added the non-canonical amino acid solved in 50% DMSO. Also, these cultures had to be handled in the dark after addition of the non-canonical amino acids due to their light-sensitivity.

Main-culture	Non-canonical amino acid
SDM 2 + ONBY-tRNA/synthetase pair	ONB-tyrosine
SDM 4 + ONBY-tRNA/synthetase pair	DMNB-serine
Empty backbone pSB1C3	ONB-tyrosine
E. coli BL21 wildtype	ONB-tyrosine

All main-cultures were inoculated to an OD₆₀₀ of 0.1 and incubated at 37°C, 250 rpm to OD₆₀₀ 0.3. Then, the main-cultures were induced by adding 0.5mM IPTG. Afterwards, one of each type of main-culture was incubated at 30°C and the other at 25°C. We took five identical samples of 500µL of each main-culture hourly and continued the incubation overnight.

Day 9

We took five identical samples of each main-culture from the day before and then lysed all samples with different cell lysis methods:
Lysozyme solution (1h, 37°C, 250 rpm)
Lysozyme solution (1h, 37°C, 250 rpm) + sonication bath (30 minutes)
Sonicator (3 minutes)
Glass beads mill (10 minutes)
The supernatant of these samples was then tested via SDS-gelelectrophoresis. Unfortunately, we were not able to obtain any clear results from the SDS gel and thus could not verify the expression of subtilisin E.

Testing the folding and activity

Day 1

As we were not able to detect the expression of the modified proteases via SDS gel, we proceeded by executing a skim milk assay on agar plates containing IPTG and the needed antibiotics. Therefore, we streaked the cells containing the modified expression systems on these plates and incubated at 30°C for three days.

Neither the empty backbone nor the SDM 1-modified expression system did seem to cause a proteolytic activity. A clearance and therefore a proteolytic activity could only be observed for the native protease (as demonstrated before) and the SDM 3-modified expression system.
Via SDM 1 we exchanged the targeted serine in the active site against a bigger amino acid, namely tyrosine. By demonstrating that this modification doesn’t result in a clearance of the skim milk plates, we were now able to prove that serine is essential for the proteolytic activity of the protease and that exchanging it would inactivate the enzyme.
SDM 3 has been executed to exchange tyrosine in the pro-peptide cleavage site against tryptophan, a bigger amino acid. As seen on the pictures above, a clearance has occurred for those modified cells and a proteolytic activity could be assumed. Contrary to our former beliefs, it could now be deduced that exchanging tyrosine doesn’t result in a change of activity. Consequently, tyrosine in the pro-peptide cleavage site is not essential for the activity of subtilisin E.

Conclusion

We were able to proof that serine in the active site would result in a loss of activity. Hence, we demonstrated that exchanging serine against a photo-labile, non-canonical amino acid will inactivate subtilisin E and therefore proved the principle of our project.
Unfortunately, we were not able to prove the principle of our project regarding tyrosine in the pro-peptide cleavage site, as it seems to be not essential for the proteolytic activity. Thus, exchanging tyrosine against ONB-tyrosine will not influence the activity of the enzyme.

Photoclevage

This part of the project was dedicated to the actual photocaging of the protease. Unfortunately, we were not able to execute all of the experiments that we planned due to a lack of time. But as we were able to prove the concept of our project, we will not only show our progress so far but also our plans for the experiments that we weren't able to conduct.

Irradiation

Day 1

To characterize O-(2-nitrobenzyl)-L-tyrosine (ONBY), we decided to test the photocleavage with the light box provided by the iGEM team Düsseldorf 2016 and to analyze it, using gas chromatography-mass spectrometry (GC-MS). To achieve this, we dissolved 5mg of ONBY powder in 100µl pyridine and incubated at 37°C for 10 minutes. Then, the analysis was executed via GC-MS with the following conditions:

1. Injection Temperature: 80°C; Holding Time: 2 minutes
2. Final Temperature: 250°C; Temperature Ramp: 8°C

The sample was then exposed to UV light of intensity 14.8 mW/cm² with different exposure time of 1, 3, 5 and 10 minutes using the light box and then analyzed using the same GC-MS method described above. The exposure resulted in photocleavage, and produced 2-nitroso benzaldehyde which has a molecular weight of 132.122 g/mol. Compounds with similar molecular weight have an elution time of 5.3 minutes.

The chromatogram of the analyzed samples shows the appearance of a peak at 5.22 minutes close to the expected elution time of 5.3 minutes. We observed that, as the exposure time was incremented from 1 minutes to 10 minutes (figure 1, graph C-F), the intensity of the peak also amplified indicating an increase in the amount of cleavage product. Also the absence of this peak in pyridine (blank) and the sample without exposure (figure 1, graph A and B), supports the absence of cleavage product. However, due to the unavailability of standard 2-nitroso benzaldehyde, we could not quantify the amount of cleaved product produced on UV exposure. The spectrum of the cleaved product show peaks at 51, 52, 73 and 79 m/z (figure 2) which is almost similar to the data available. [1]

A rough estimate of the change in concentration was made by measuring the area under the curve for the peak at 5.22 minutes (figure 3).

Also based on similar measurements, we observed a decrease in area for the peak at 19.02 minutes, indicating the presence and decrease in the concentration of ONBY (figure 4).

As a conclusion, we deduced that the amount of cleavage product increases as the exposure time is extended.

Day 2

We intended to use the results of the ONBY photocleavage characterization to irradiate our samples S221X- and Y77X-mutated proteases which were obtained through cultivation of the SDM 2 and SDM 4-modified expression system in E. coli BL21. Furthermore, we planned to perform a skim milk assay with these samples before and after irradiation and then the degradation efficiencies of the proteases were compared by the extent of clearance. With these results we wanted to prove that due to the incorporation of DMNB-serine in the catalytic triade of the protease would lose its proteolytic activity and would only re-gain it after irradiation which would result into photocleavage.