Team:SDU-Denmark/Demonstrate

Demonstration & Results

Overview

Bacteriocin purification (Top)

Figure 1 shows results of a cPCR originating from respective colonies from transformation of K2018011/pTXB1/E.coli Top10 (ThuricinS), K2018012/pTXB1/E.coli Top10 (LacticinQ) and K2018014/pTXB1/E.coli Top10 (Laterusporulin-ThuricinS). From right; Well 1-3 + 5-7 + 13-14= K2018011, Well 4+8 = K2018014, Well 9-12 = K2018012. The expected length of ThuricinS is 304 bp. GeneRulerTM 100 bp DNA ladder were used ThermoFisher scientific. .

Successful cloning of the bacteriocins into the IMPACT vector pTXB1 were a key step in purifying the bacteriocins. Cloning into the IMPACT vector pTXB1 were shown by gel electrophoresis (Figure 1). A successful cloning was verified from the results of a cPCR performed with specially designed primers, according to the theoretical expected length (304 bp) of ThuricinS with IMPACT overhangs. The results show bands of similar lengths, thus showing a successful ligation of ThuricinS into pTXB1.

Determination of bacteriocin concentration (Top)

Table 1 shows the respective concentrations of the purified bacteriocins. The concentrations are calculated according to the equation y = 0.0011x + 0.0202 derived from linear regression of the Bradford Standard Protein Assay. The protein concentration (x) is calculated as x = (y - 0.0202)/0.0011.

BACTERIOCINS	OD (595 nm)	µg/mL
IMPACT ELUATE (Laterosporulin-ThuricinS)	0.078	52.55
IMPACT ELUATE (LacticinQ)	0.069	44.36
IMPACT ELUATE (Laterosporulin)	0.094	67.09
IMPACT ELUATE (ThuricinS)	0.057	33.45
IMPACT ELUATE (LacticinQ-LacticinZ)	0.054	30.36

We purified the bacteriocins using the IMPACT Method. To assay the concentration of the purified bacteriocins, we used a Bradford standard protein assay with known concentrations of BSA (Bovine serum albumine) to compare measured absorbance of the bacteriocin to the relative absorbance of known BSA concentrations. The following results were obtained due to the measured OD values.

Bacteriocin effect towards S. aureus MRSA and P. aeruginosa (Top)

Figure 2: The graphs shows the results of MIC tests performed on cell lysates containing respective bacteriocins. The graphs are plotted as OD/hours where SN = Supernatant. The color code indicates the concentrations used [µg/mL]. SDU11 = ThuricinS K2018011 tested on E. Cloacae, SDU14 = Laterusporulin-ThuricinS K2018014 tested on P. aeruginosa.

From the collaboration with the iGEM Stockholm team 2016 we got an indication of the effect of our bacteriocins. The team performed MIC tests using cell lysates that contained our bacteriocins, towards the strains S. aureus, P. aeruginosa and E. cloacae. Part of the MIC results from iGEM Stockholm team (Figure 2).

The red/brown line (Figure 2) indicates gentamycin as a positive control which is shown to inhibit the growth of Enterobacter cloacae. E. cloacae is considered to be an important cause of nosocomial infections Keller, R., Pedroso, M. Z., Ritchmann, R., & Silva, R. M. (1998). Occurrence of Virulence-Associated Properties in Enterobacter cloacae. Infection and Immunity, 66(2), 645–649. . The cell lysate containing ThuricinS (SDU11-SN) with a concentration of 30 µg/mL elicit a similar and better effect towards E. cloacae compared to the traditional antibiotic, gentamycin. The lysate containing ThuricinS is still effective after 40 hours of incubation. SDU14-SN represents a MIC test performed towards P. aeruginosa with cell lysates containing our hybrid bacteriocin Laterusporulin-ThuricinS. The cell lysate shows to have similar effect towards P. aeruginosa when compared to gentamycin. The results from iGEM Stockholm team 2016 thus suggest ThuricinS and the hybrid Laterusporulin-ThuricinS to be potent inhibitors of the growth of E. cloacae and P. aeruginosa, respectively.

We purified our bacteriocins in order to specify the effect of our bacteriocins as single acting proteins. To assay the effect of our purified bacteriocins we performed a MIC test following SOP0027_v01. The bacteriocins were dissolved by a two-fold dilution in water from wells A-G having the largest concentration possible at column A. The maximal concentration used is the half of the calculated concentration of the bacteriocin. We used different MRSA strains to test the effect of the bacteriocins in order to state the possibility for the bacteriocins as substitute for traditional antibiotic of which the bacteria have evolved resistance towards. Following strains where tested:

MRSA:CC398 – a methicillin and tetracyclin resistant S. aureus isolated from a patient, which have gained it from a pig. The MRSA:CC398 is called pig-MRSA, which is found in many pigs in the Scandinavian and is a huge contributor to the resistance development.
Hetero - VISA – a S. aureus strain. The strain is capable of growing at vancomycin concentration of >4 mg/L and have thereby gained relative vancomycin resistance. This is due to thickened walls of the strain, which is suggested to catch the vancomycin before entry Conly, J. M., & Johnston, B. L. (2002). VISA, hetero-VISA and VRSA: The end of the vancomycin era? The Canadian Journal of Infectious Diseases, 13(5), 282–284. .
MRSA:USA300 – the S. aureus strain is a community associated MRSA and is a common cause of skin infections Alam, M. T., et al. (2015). "Transmission and microevolution of USA300 MRSA in U.S. households: evidence from whole-genome sequencing." MBio 6(2): e00054..
P. aeruginosa:PAO1 – the strain is commonly spread at hospitals thus given rise to re-hospitalization of people with open wound infection, which we would like our bacteriocins to face.

MIC test (Top)

Figure 3 shows visual MIC test results of K2018010/Laterusporulin. Well A(1-10) contain 33.5 µg/mL as final maximal concentration. A two-fold dilution is performed in well B-G(1-10). Well H(1-10) does not contain bacteriocin, but contain bacterial culture. Well A-H(11-12) is BLANK.

The MIC plates were performed with a two-fold dilution of the respective concentrations shown in Table 1. The concentrations in table 1 indicate final concentrations i.e. the concentration of the bacteriocin in the well after adding bacterial culture. An example of the bacteriocin K2018010/Laterosporulin plated in the MIC wells is shown in Figure 3. To validate growth or inhibition we used background absorbance from dH₂O+Mueller Hinton (MH) Media plus 10% deviation as a reference value for estimation of growth. In Figure 3 a visual inhibition of growth of the listed strains by Laterosporulin is shown.

Table 2 and 3 shows the respective MIC values according to the tested strains and the bacteriocins/traditional antibiotic they were exposed to. N.I = No inhibition. Thus no inhibition was observed at the measured maximum concentrations (Table 1).

Bacteriocin/MIC [µg/mL]	hVISA	USA300	CC398	PAO1
ThuricinS	16.7	16.7	16.7	16.7
LacticinQ	22.2	22.2	22.2	N.I < 22.2
Laterosporulin	33.6	16.8	16.8	33.6
LacticinQ-LacticinZ	7.6	7.6	7.6	15.2
Laterosporulin-ThuricinS	13.1	13.1	6.6	N.I <13.1

Traditional Antibiotic/MIC [µg/mL]	hVISA	USA300	CC398	PAO1
Ampicilin	<16.0	<16.0	1000.0	N.I<1000.0
Chloramphenicol	78.0	N.I<1000	N.I<1000	39.0

Effect of bacteriocins (Top)

Figure 4 visualizes MIC values for each bacteriocin towards each tested strain. The respective strains are listed in the x-axis and the MIC values are plotted logarithmic(2) on the y-axis. The lowest bar indicates the bacteriocin eliciting the strongest effect.

MIC values are visualized in the bar-chard in Figure 4. The bacteriocin eliciting the strongest effect is indicated by the lowest bar – thus the lowest MIC value. Compared to MIC values of traditional antibiotics (Table 3) i.e. ampicillin and chloramphenicol, the bacteriocins show similar effect. HOWEVER, the bacteriocins also show better effect due to inhibition of growth of strains, which the traditional antibiotics does not inhibit.

In Figure 4 it is seen that ThuricinS affect all of the strains at a concentration of 16.6 µg/mL. LacticinQ shows to affect only the S. aureus strains at a MIC value of 22.2 µg/mL. Laterosporulin shows inhibition of growth towards all of the strains with the lowest MIC being towards the MRSA strains, USA300 and CC398, with a MIC at 16.8 µg/mL. Even though the laterosporulin as a single bacteriocin affect P. aeruginosa, the Laterosporulin-ThuricinS hybrid does not elicit effect towards P. aeruginosa. However, it is seen that the MIC values towards hVISA, USA300 and CC398 (13.1 µg/mL, 13.1 µg/mL, 6.6 µg/mL respectively) are lower than the MIC value found in their single bacteriocin form (ThuricinS MIC 16.7 µg/mL and Laterosporulin MIC 16.8 µg/mL). This indicates a synergistically effect of hybrid bacteriocins compared to their effect as single bacteriocins.

LacticinQ shows no inhibition towards P. aeruginosa. However, the absence of inhibition could also be due to a higher MIC value i.e. need of a higher concentration to inhibit growth than used in the MIC test. The hybrid LacticinQ-LacticinZ is seen to inhibit the growth of all the tested strains with a MIC equal to 7.6 µg/mL against the S. aureus strains and a MIC corresponding to 15.2 µg/mL towards P. aeruginosa (Figure 4). In addition, the MIC is decreased compared to the MIC estimated for LacticinQ alone (22.2 µg/mL). This could be due to the presence of LacticinZ. However, LacticinZ is not tested as a single bacteriocin and a direct comparison cannot state a decrease in MIC for LacticinZ but only a decrease in MIC for LacticinQ. However two bacteriocins working in concert (a hybrid) seem to contribute to a decrease in MIC and thus a stronger effect, indicating synergy. It should thus be noted that the hybrid Laterosporulin-ThuricinS “looses” its effect towards P. aeruginosa, compared to their effect as single proteins.

The MIC plates were incubated for 28 hours with equal OD values for each bacterial strain i.e. equal amount of bacteria were added. For replicating the MIC test, the respective generation times of the bacterial strains tested should be considered. The generation times can diverge between the bacterial strains thus representing diverse conditions, which is reflected in the OD measured. Therefore, the OD value between the strains cannot be compared, but the MIC values can. Furthermore instead of using duplicates, multiple replications could have been made for specification of the OD value. Thereby we could determine the OD value due to calculating a median, instead of calculating a mean value and thereby be able compare more OD values. In addition by use of duplicates, pipette errors cannot be neglected. A new MIC test could therefore be performed with multiplicates, thus eliminating some of the stated source of error.

Silk assembly

Insertion of MaSp gene fra into pSB1C3 (Top)

Figure 5 shows the cPCR products of MaSp gene fragments with overhangs. GeneRulerTM 100 bp DNA ladder were used.

In order to achieve a renewable source of the genes a key step was to ligate each gene fragment into their separate pSB1C3 vectors. Recall that the MaSp genes contain 3 gene fragments, which together make up a monomer and thus a functional MaSp protein. The ligated gene fragment/pSB1C3 was transformed into E. coli. Successful ligations were verified by colony PCR (Figure 5). The results from cPCR appeared to be 400 bp, which corresponds to one MaSp gene fragment with cPCR overhangs.

The ICA technique allows us to ligate different pieces of DNA together stepwise due to matching overhangs. Eco31I recognizes four nucleotides, but cuts four randomly base pairs at the location + 1bp upstream from the recognition site. By making specific primers that add restriction sites, we were able to create compatible overhangs between silk fragments. This approach was inspired from the iGEM UCLA team 2015.

Ligation test (Top)

Figure 6 shows a ligation test. Bands at 106 bp are the unligated silk gene. Bands at 212 bp represent the ligated product, and thereby verifies a successful ligation test.

We tested the ability of the genes to ligate by performing a ligation test. The digested gene is 106 bp. A bond at 212 bp verified that the gene fragments can be ligated together (Figure 6).

Preparing silk by the ICA method (Top)

Figure 7 shows a monomer of MaSp1 and MaSp2 made by ICA.

It was possible to proof the concept of preparing silk from the ICA method. One monomer of MaSp1 and MaSp2 was made by the ICA method (Figure 7). The monomer is expected to have a length of 406 bp as one monomer of a MaSp gene consists of the gene fragments AB, BC and CA with the respective lengths of 106 bp. Additional base pairs originates from primer overhangs.

We tried to make constructs containing 4 MaSp gene monomers. However, we struggled succeeding in making a longer construct. The closest result we achieved was to make a gene containing 700 bp. In the end we found that our magnetic beads had little to no streptavidin bound, alas it was not possible to use the ICA method for consistent results. There is however a great potential in the method, but one should test each individual component before attempting the ICA.

The dream of creating a silk-bacteriocin hybrid (Top)

Figure 8 shows the result of a fusion PCR. The result indicates that we succeeded getting silk-overhangs on ThuricinS.

We wanted to create the hybrid silk of a bacteriocin fused with one or multiple monomers of MaSp1, MaSp2 with different proportions of the genes on DNA level. This was tested by using recombinant DNA technology. Two methods were tested: ICA, and restriction cloning.

We managed to prepare a bacteriocin with specific overhangs, matching the overhangs present on the MaSp genes, in order to make them compatible. Lane no. 5 (Figure 8) shows the bacteriocin ThuricinS with DA-overhangs. The band is expected to be 217 bp. ThermoFisher scientific GeneRuler 50 kb ladder was used.

Despite various attempts we never succeeded in cutting and ligating ThuricinS together with a silk monomer. It thereby all fit in theory, but when it came to reality, it was difficult to perform in practice. The bacteriocin with specific silk overhangs are therefore left to further research.

ICA troubleshooting

Due to the multiple ligation steps in ICA, individual ligation test were made, in order to establish the viability of the constructs. Individual ligation tests were time consuming, but a necessity since a singular failed ligation can create negative results. As the ICA method is heavily reliant on everything to work as planned, alternative methods were tried. In addition, it must be noted that the small overhangs are seldom fit to ligate at room temperature. Likewise, redesigning the gene towards a larger overhang, might create easier ligations. Furthermore during ligation, constant turbation of the fluids must be applied, as the beads subside in the bottom of the tube without creating unfavorable conditions for ligations.

Alternative method

We tried ligating constructs, which did not have overlapping overhangs eg. Initiator-AB-BC, CA-AB and BC-CA-Terminator. This were obtained by ligating in series at 16 degrees celsius overnight. The lower temperature is suggested to increase the possibility of ligation. However the lower temperature also decrease the activity of T7 ligase. Thus we added a small amount of 40 mM ATP to the sample at each ligation step hoping to create a silk construct. Unfortunately we ran out of time due to expired streptavidin beads. The experiment are thus left for further trial.

Missing amplification of ThuricinS with silk overhangs

We noticed that using the standard primers from iGEM, VF2 and VR, our pPCR on plasmids containing the bacteriocin/silk-overhangs, the following digestion leaded to two bands that only differed by 1 bp from each other. We therefore decided to amplify our bacteriocin with silk-overhangs. However, after changing primers we experienced that the pPCR gave unexplainable results which included a lot of smear. Various pPCR programs were tested but we never ended up with a product that could be purified and used in the ICA method. We also tried to digest the pSB1C3 plasmid and the ThuricinS/silk-overhang pPCR product to perform a ligation into the pSB1C3 directly. This result was verified by gel electrophoresis. However, it was not possible to digest the plasmid containing the ThuricinS with silk overhangs and a purification could not be performed. Ligation between the ThuricinS and a silk monomer were therefore not achieved. For future testing it would be interesting to digest with restrictions enzymes for longer periods, or design new primers for the pPCR, necessary for ICA method.

Production of Polyhydroxybutyrate (PHB)

The following section contains our results for optimizing and characterizing the production of PHB in E. coli.

PHB producing cells (Top)

Table 4 shows the content of the biobrick constructs tested.

In order to determine the effectiveness of our library of PHB producing cells, we used flow cytometry to get a picture of how effective the different strains were at producing plastic. By staining the cell cultures with nile red, we created a relative quantitative measure for how much plastic that were present inside each cell. The following results were obtained with 4 biologic replicates of each E.coli strain (Table 4). The E.colistrains were incubated for 28 hours before the measurements were made.

Seperation of strains due to PHB production (Top)

Figure 9 display the number of events on the y-axis and the corresponding intensities on the x-axis. Cells of the different strains are displayed in different colors. Each event represents a cell and the intensity represents the amount of PHB in the respective cell.

Figure 9 shows a clear separation of the different strains. This we would expect for strains with different promoter and ribosomal binding sites ahead of the phaCAB genes. We included the BioBrick created by Tokyo Tech in 2009 (K934001) and one by Imperial College (K1149051) in 2012 as references for our new constructs. The strain containing the reference constructs are displayed in green in figure 9. However, as seen on the figure 9, the events with the largest intensities, originates from the E.colistrain containing our construct (K2018036). This indicates that our construct generates more plastic, than any other hybrid promoter phaCAB tested in the figure.

Unlike our initial expectations, the results of our flow cytometry analysis strongly suggest that the strongest additional promoter does not produce the largest amount of plastic. However, without further qualitative data, we can only speculate as to why, this is the case.

Additional RBS - largest increase in PHB production (Top)

Figure 10 display the different E.colistrains on the x-axis. The promoter and RBS are marked with corresponding affinity: “s” = strong, “W” = weak and “MS”=medium. The y-axis displays the average intensity on the flow cytometer detected by red fluorescence. The intensity of the red fluorescence is calculated due to a mean value of the intensities detected in figure 9.

Equally surprising is the fact that the additional RBS is responsible for the largest increase in PHB. The additional RBS for the PhaCAB gene is not followed by a start-codon and as a result, ribosomal attachment to the specific RBS will not be able to initiate transcription. We suggest that a strong RBS will bind the ribosome with high affinity and thus it will not detach readily. For a medium RBS, the affinity will be high enough for ribosomes to be attached to the RNA frequently, but due to the lower affinity, they will also detach more frequently. The event results in an increase in the local concentration of ribosomes. For the weak RBS the affinity will be too low for the local concentration of ribosomes to be effected.

The strong and weak RBS thereby generates less plastic, relative to the medium strong RBS. We propose that this tendency is caused by changes to the local concentration of ribosomes, due to the presence of an additional RBS, upstream from the transcriptional RBS.

Examining methods of PHB extraction

The method of PHB extraction is an important part of the production that should be taken into account, as it has a large impact on the purity and yield of PHB. We have chosen four methods of extraction; Hypochlorite digestion, Hypochlorite digestion with Triton X-100 pre-treatment, Chloroform solvent extraction and Ethyl acetate solvent extraction.

HNMR as qualitative measure of purity for extration methods (Top)

In order to determine whether the different methods of PHB extraction we have used were effective, we analyzed the purified material with HNMR.
We determined the peaks caused by PHB by analyzing a sample of pure PHB as a reference. The sample of pure PHB can be held against the spectra from extracted PHB.

Figur indsættes - T--SDU-Denmark--PHB_ren_HNMR.png

Figure: Each peak is caused by hydrogen atoms. The area under each curve is proportional to the number of chemical equivalent hydrogen atoms causing the peak.

Since this spectrum is based on pure PHB, all the peaks present are not considered contaminations in the other spectra observed in Figure X. For a more detailed analysis of the spectrum click here.

Figur indsættes - T--SDU-Denmark--PHB_extractions_HNMR.png

Figure X shows the illustration of the spectra for the methods of PHB extraction of which we have tested. Top left is for Chloroform extraction, top right is for Ethyl acetate extraction, bottom left is for Hypochlorite extraction and bottom right is for Hypochlorite extraction with Triton X-100 pre-treatment.

Extraction with Chloroform

The spectrum that shows chloroform extraction has very clearly defined PHB tops and no indications of impurities. The spectrum indicates that the specific method of extraction, provides PHB with a high level of purity.

Extraction with Ethyl acetate

The spectrum that shows ethyl acetate extraction has well defined PHB peaks, but an unknown compound in the sample cause a peak at approximately 2.17 ppm. This indicates that the PHB purity is lowered with this extraction method, when compared to chloroform extraction.

Extraction with Hypochlorite

In the spectrum showing hypochlorite extraction the PHB peaks are clear, but once an impurity is registered in the spectrum, there is a broad peak at 2.3 ppm. The area under the two peaks near the impurity are almost doubled compared to the peak at 5.2 ppm. This inequality, however, is not significant as it is the impurity in the sample, that contribute to the additional area under the peaks.

Extraction with hypochlorite and Triton X-100

The peaks representing PHB peaks are present but the peak for water is absent. The event is not surprising as the PHB sample was dried out. (Der skal billeder ind af prøverne et sted, evt. Når musen holdes over titlen) There is no indication of impurities, although these could have evaporated along with the water. This specific method is similar to the Hypochlorite extraction method. The Triton X-100 pre-treatment is performed to increase the digestion of cellular content, which could explain the absence of impurity. For the hypochlorite PHB extraction with Triton X-100 and the ethyl acetate purification, the entire content of the purification, was not soluble in chloroform. This indicates that large impurities present in the sample were not visible in the HNMR spectra.

Determination of intracellular PHB concentration(Top)

Figure x+1 shows the mirrored spectrum from a gas chromatography analysis of 20 mg cells containing (K2018036) (blue) and a 2 mg pure PHB standard (red). PHB causes a top at 1.150 min. and a top at 2.135 min. The top at 2.135 min is an internal methyl benzoate standard.

With gas chomatography the amount of PHB inside the cells can be determined. In order to do this, PHB standards containing 2 mg of pure PHB were created to compare and determine the amount of PHB in the tested cell samples. Cell samples of 20 mg were analysed and quadruplets of both standards and samples were made.

The area under the curve illustrates the amount of the substance analyzed at the respective time. The amount of PHB in the figure is significantly lower in the tested cell sample than observed for the standard sample containing 2 mg pure PHB. The observation indicates that there is less than 2 mg PHB in the tested cell sample. When the data was normalized according to the methyl benzoate standard, it showed that there were 0.7 mg PHB in 20 mg cell sample. This suggest that 3.5% of the cell mass is PHB. Compared to the amount of PHB we purifirf from each gram of biomass (to be continued)

iTRAQ analysis

The iTRAQ labelling was only partial, resulting in compromised data. However, the up- and down regulation of several proteins, was identified in the cells where pantothenate kinase II was expressed. In the MS/MS analysis a total of 2402 proteins was identified in our bacteria. Out of these we determined 21 of these were upregulated and 11 were downregulated.

String analysis

We performed a string analysis on the proteins that had been upregulated in order to determine the effect of pantothenate kinase II on a protein level. In figure x the proteins involved in metabolism are highlighted. It is not surprising that a large amount of the proteins are involved in metabolism, as pantothenate kinase itself is a metabolic protein. The three central proteins in the figure; tpiA (triose phosphate isomerase), PGK (phosphoglycerate kinase) and pykF (puryvate kinase) are all involved in glycolysis. NuoG (NADH dehydrogenase subunit G) is involved in the electron transport chain, another mechanism by which the cell generates energy. The rest of the upregulated proteins did not have as clear a correlation.

The increased amount of glycolytic enzymes indicates an increased energy consumption in the cells, due to the presence of pantothenate kinase II. The fact that coenzyme A is utilized in the oxidation of pyruvate, fits with the increase in glycolytic enzymes.

Identification of phosphorylations

Due to the titanium dioxide purification the phosphorylated proteins could also be analyzed in the sample. It is however important to note that this purification is meant for analysis of eukaryotic cells and therefore the amount of identified proteins will be low. There were no identified phosphorylations that were more or less abundant in cells that expressed pantothenate kinase II. Since none of the upregulated proteins where protein kinases this result is not surprising. Although the phophoprotein analysis did not provide any data relevant for our project, a phosphorylation on serine 5 of 2,3-bisphosphoglycerate-independent phosphoglycerate mutase that is not registered on uniprot, was identified in each of our samples.