Team:Lund/Results
Introduction
To prevent the formation of the bacterial amyloid curli in the gut, we aimed to express various inhibitors. As the lab work of the project progressed and results of the experiments were obtained, we continuously adjusted the course of our project. The DNA and cloning work was divided into three cycles; in the first two cycles we worked with the plasmid pTRKH3-ermGFP and in the third cycle with pET-11a. To read more about the purpose of each cycle, see Engineering. For protocols, see Experiments.
Curli Detection Work
Thioflavin T Fluorescence Assay
The assay was performed to detect curli in different strains, but also to confirm the purification of curli after sonication from the different strains. E. coli MC4100, E. coli BL21(DE3) and E. coli TG1 were tested. Further, the temperature 26°C and 37°C as well as different incubation times were tested. Figure 1-4 highlights the results of the fluorometry measurements of E. coli MC4100 which was incubated at 26°C for 138 h. The samples were cellbroth mixed with Thioflavin T (ThT) (Figure 1), cells concentrated and resuspended in PBS buffer mixed with ThT (Figure 2), sonicated cells in PBS buffer with ThT (Figure 3) and the supernatant without cells after sonication (Figure 4).
Figure 1 showed no significant peak. A small peak could be detected between 500 and 550 nm. The specific ThT peak at approximately 482 nm was not present. This could be explained by the fact that the cells are very diluted and therefore the peak is not visible. The reason for the small peak is most likely due to ThT interacting with components in the cellbroth.
Figure 1. Fluorometry measurement of 1 ml of cellbroth of E. coli MC4100 (26°C, 138 h) with 40 µl of ThT. Wavelengths (nm) plotted against intensity (a.u.).
In Figure 2 a peak could be detected at 485 mn at an intensity of 130 a.u. This Indicates that ThT has been bound to amyloids. However the intensity is rather low which could also indicate that the fluorescence is created by cells interacting with ThT and not necessarily by amyloids.
Figure 2. Fluorometry measurement of 1 ml of harvested E. coli MC4100 (26°C, 138 h) resuspended in PBS buffer with 40 µl of ThT. Wavelengths (nm) plotted against intensity (a.u.).
In Figure 3, a peak at 486 nm could be detected. The peak presented an intensity of 526 a.u., making it higher than the peak for the sample where the cells were only resuspended in PBS (Figure 2). While the total amount of amyloids was expected to be the same, a possible explanation for the increase in intensity could be that when the cells are sonicated the amyloids are more easily accessible.
Figure 3. Fluorometry measurement of 1 ml sonicated E. coli MC4100 (26°C, 138 h) resuspended in PBS buffer with 40 µl of ThT. Wavelengths (nm) plotted against intensity (a.u.).
A peak at 487 nm with an intensity of 408 a.u. could also be observed in Figure 4.That indicates that after sonication amyloids were dissolved in the supernatant. A survival test of the soniciated cells confirmed that the cells survived.
Figure 4. Fluorometry measurement of 1 ml supernatant of centrifuged sonicated E. coli MC4100 (26°C, 138 h) resuspended in PBS buffer with 40 µl of ThT. Wavelengths (nm) plotted against intensity (a.u.).
Apart from 26°C and 138 h of incubation time with the strain E. coli MC4100, we also attempted to detect curli of different E. coli strains with different incubation times at 37°C. These results were however inconclusive. Variations of the results when using the same conditions were very high. The same goes for the results with the conditions shown above (Figure 1-4). Therefore, a correlation between temperature, incubation time and different strain could not be observed. Despite the fact that such a correlation could not be drawn, an amyloid specific peak could still be detected at around 482 nm which strongly suggests the presence of curli. In order to interpret a conclusive correlation, the experiments should be repeated with optimized parameters, and more time would be necessary. Also, it would be of great advantage if a negative control would be available in order to subtract background fluorescence. This negative control should be a CsgA- strain, so the cell does not produce any curli fibers. In that way it would be possible to exclude background fluorescence which might be created by the cells or components in the buffer giving fluorescence intensity at 482 nm. This is further discussed under “Future experiments”. Despite that, it would be helpful to have pure curli as a positive control. Additionally, a standard curve could be created which would allow to determine the produced curli concentration of the different E. coli.
Bromophenol Blue Assay
The experiment performed with bromophenol blue (BPB) did not give any conclusive results. So it can be said that by using BPB, no curli could be detected under conditions tested in this experiment and no positive results were obtained. The reason for this might be that bromophenol is not very sensitive, meaning that a very high concentration of amyloids is required to detect it. ThT is more sensitive and can therefore detect lower concentrations of curli.
Fluorescence Imaging
The stained plates with E. coli MC4100 showed red fluorescence when placed under a green light wearing filter glasses while the plate with E. coli BL21 (DE3) presented no red fluorescence. The results were conclusive since amyloids fluoresce red under green light when stained with BPB. Literature confirms that E. coli MC4100 forms amyloid curli and is a well known strain to study the production of curli [4]. It needs to bear in mind that this fluorescence imaging application is qualitative and not quantitative. It only indicates that E. coli MC4100 forms more amyloids than E. coli BL21(DE3).
DNA and Cloning Work
Cycle 1
Plasmid Prep
Initial experiments testing what erythromycin concentration was optimal for our pTRKH3-ermGFP vector indicated that cells should be inoculated overnight in LB medium with 10 µg/ml erythromycin. Further experiments testing the pTRKH3-ermGFP plasmid extraction showed that streaking out the bacteria on LB agar plates containing 50 µg/ml erythromycin provided the optimal plasmid yield, as demonstrated in Figure 5 (FEM50, lane 4). Extracted plasmid was then digested with BamHI for added confirmation for extraction success, which is also demonstrated in Figure 5 (lanes 3 and 5).
Figure 5. 0.6% agarose gel electrophoresis of extracted plasmid and BamHI digested plasmid (linearized). The following amount of samples were added into the wells (from left to right): 2.5 µl 1kbp gene ruler; 10 µl FEM10; 10 µl BamHI digested FEM10; 5 µl FEM50; 5 µl BamHI digested FEM50; For FEM10, 3x10 ml E. coli pTRKH3-ermGFP were streaked out on 10µg/ml erythromycin plate from original vial on 20210704; For FEM50, 3x10 ml E. coli pTRKH3-ermGFP were streaked out on 50 µg/ml erythromycin plate from original vial on 20210704. Bands circled with blue frames were our desired plasmid. Visualization on geldoc showed that plasmid prep and restriction enzyme digest was successful for all samples.
PCR Amplification of Inserts
Vector inserts containing the curli inhibitors were amplified via PCR reaction with temperature gradient from 50ºC to 70ºC to determine optimal conditions of the primers. Figure 6 shows the gel electrophoresis of the PCR products for the amplification of three of the inserts (each coding for the inhibitor tANK6, but with different signal peptides L. lactis Ups45, SLPmod and B. licheniforms amyL).
Figure 6. 1.0% agarose gel electrophoresis of PCR product for the amplification of 3 insert (tANK6 with L. lactis Ups45, SLPmod and B. licheniformis amyL) with temperature gradient from 50ºC to 70ºC with Phusion High-Fidelity DNA-Polymerase. The well order (from left to right): 1-5: 5x tANK6_L. lactis Ups45, PCR reaction temperature gradient from 50ºC to 70ºC (50, 55, 60, 65, 70), 6-10: tANK6_SLPmod, PCR reaction temperature gradient from 50ºC to 70ºC, 11-15: tANK6_B. licheniformis amyL, PCR reaction temperature gradient from 50ºC to 70ºC.
Table 1 shows the expected length of the target PCR products from Figure 6. PCR product gel electrophoresis showed clear bands at ~600 bp, indicating that the PCR reaction was successful. PCR reaction temperatures of 60 and 65 ℃ worked for all inserts.
Table 1. Expected PCR product length for Figure 6.
| Sequence | length |
| tANK6_L. lactis Ups45 | 594 bp |
| tANK6_B. licheniformis amyl | 606bp |
| tANK6_SLPmod | 585 bp |
Table 2. Numbering and labeling for inserts during Cycle 1
| Construct | Insert |
| A10 | DB3DB3_SLPmod |
| A20 | DB3DB3_L. lactis Mub2 |
| A30 | DB3DB3_B. licheniformis amyL |
| A40 | DB3DB3_Lb. Reuteri Mub2 |
| A50 | tANK6_SLPmod |
| A60 | tANK6_L. Lactis Ups45 |
| A70 | tANK6_B. licheniformis Mub2 |
| A80 | tANK6_Lb. reuteri Mub2 |
| A90 | CsgF_SLPmod |
| A100 | CsgF_L. lactis Ups45 |
| A110 | CsgF_B. licheniformis amyL |
| A120 | CsgF_Lb. reuteri Mub2 |
| A130 | DegP_SLPmod |
| A140 | DegP_L. lactis Ups45 |
| A150 | DegP_B.licheniformis amyL |
| A160 | CsgC_SLPmod |
| A170 | CsgC_L. lactis Ups45 |
| A180 | CsgC_B. licheniformis amyL |
Optimization of erythromycin concentrations
The erythromycin concentrations in both the liquid medium and on the solid plates were attempted to be optimized as mentioned above. The erythromycin concentration in the liquid medium was tested by testing different concentrations and measuring the plasmid concentration. A plasmid preparation was made and the extracted plasmids were quantified on a spectrophotometer using a microlitre cuvette. The double stranded DNA concentration was 32.3 ng/µl for 10 µg/ml erythromycin which we deemed sufficient and which prompted the use of 10 µg/ml in the liquid medium throughout the project. The concentration in the liquid medium was enough to have the plasmid retained and this was shown throughout the many times plasmid preparation was performed.
As described before, extracted plasmid from original viall of E. coli pTRKH3-ermGFP streaked out on 10µg/ml erythromycin plate was called FEM10, and extracted plasmid from original viall of E. coli containing pTRKH3-ermGFP streaked out on 50 µg/ml erythromycin plate was called FEM50. The measured concentration of FEM10 is 72.1ng/µl while that of FEM50 is 144ng/µl. As a result, the e.coli containing pTRKH3-ermGFP is better grow on 50 µg/ml erythromycin plate before innoculation into 10 µg/ml erythromycin medium.
Different erythromycin concentrations on the solid plates were tested multiple times. There were no conclusive results since there was at some points growth of negative controls (bacteria without the resistance plasmid) on plates that contained 100 µg/ml erythromycin and at other points no growth of the negative control on 50 µg/ml plates. Efforts were made to use as low a concentration of erythromycin as possible since any successful transformation was desired to be viable on the plate but at the same time the concentration should not be so low as to give rise to many false positives. Most of the time transformations were performed on 100 µg/ml erythromycin plates but it is unclear if this was a good concentration.
First Gibson Assembly and Transformation Using Ligation Mixture
A Gibson assembly was performed using extracted pTRKH3-ermGFP and PCR amplified inhibitor constructs which had been digested with MscI (Table 2). In Table 3, the observations of a transformation using these ligation mixtures is described. What is described as possible contamination in the controls was determined to be air bubbles in the following weeks as these appeared in later experiments when the plates were exposed to excessive heat. Transformants with plasmids containing genes for curli inhibitors showed some colonies. These were determined to be false positives by the colony PCR described in the next section.
Table 3. Plate Observations from Transformation using first Gibson assembly ligation mixture.
| Sample | Plate observation |
| Transformants A10-A60, A80-A110 | Enough colonies |
| Transformants A120-A130 | Enough colonies |
| Transformants A140-A180 | Suspected to be contaminated with TG1 cells due to incorrect addition of medium from previously used tube |
| Transformant A70 | Had very few colonies |
| Transformants Negative Ligation mix | Might be contaminated, later determined to be air bubbles |
| Transformants F10 plasmid (positive control) | Might be contaminated, later determined to be air bubbles |
Colony PCR After Transformation Using First Gibson Assembly Ligation Mix
In total, 72 samples were analyzed with colony PCR, including 9 positive controls (plasmid, 55-57, 61-66) and 9 negative ligation mixtures (58-60, 67-72). The numbers of constructs corresponding to different inserts in Gibson assembly are presented in Table 2, and the numbers on the section on master plates corresponding to the colony PCR number are presented in Table 4.
Table 4. Samples in colony PCR after transformation using first Gibson assembly ligation mix.
Figure 7a. 0.6% agarose gel electrophoresis of colony PCR (gel 1)
Figure 7b. 0.6% agarose gel electrophoresis of colony PCR (gel 2) 1 % agarose gel.
Figure 7c. 0.6% agarose gel electrophoresis of colony PCR (gel 3)
Unfortunately there was no expected band in gel electrophoresis, neither samples nor positive control, which is shown in Figure 7. We suspected at the time that the lack of bands to be due to errors in the Gibson assembly, as it is relatively error prone. As such, this Gibson assembly was repeated, followed by a transformation and the results of which are described below.
Second Gibson Assembly and Transformation Using Ligation Mixture
During the second attempt of Gibson assembly in cycle 1, only a limited number of inserts were used in order to reduce the workload. Given that, we were also able to investigate the effect of different molar ratios between insert and plasmid in the ligation mixture. The inserts used were DB3DB3_SLPmod, CsgF_SLPmod and CsgC_SLPmod. For each insert, three molar ratios between insert and plasmid were tested: 2:1, 3:1 and 5:1. Transformation was performed using the ligation mixtures and we observed growth on all transformant plates, including positive control. However, the control had only very small colonies and as such, we determined the difference in size between positive and negative control to indicate some selection pressure (Table 5).
Table 5. Plate Observations from Transformation using second Gibson assembly ligation mixture.
| Sample | Plate observation |
| DB3DB3 2:1 | Mostly empty. One small colony. (Undiluted) |
| DB3DB3 3:1 | Looks OK. No overgrowth. Could potentially pick several colonies. (Diluted x10) |
| DB3DB3 5:1 | More growth than DB3DB3 3:1. Bigger colonies. (Diluted x10) |
| CsgC 2:1 | A lot of growth, hard to pick a single colony. Moderate size of colonies. (Concentrated) |
| CsgC 3:1 | Few colonies ≈20, varying in size. (Diluted x10) |
| CsgC 5:1 | Carpet of bacteria. Hard to pick a single colony. (concentrated) |
| CsgF 2:1 | Looks OK. Some colonies are bigger and easier to pick. Smaller carpet areas. (Concentrated (less volume)) |
| CsgF 3:1 | Same as CsgF 2:1. (Undiluted) |
| CsgF 5:1 | Carpet of bacteria. Hard to pick a single colony. (Concentrated) |
| Negative control (negative ligation mix) | Colonies varying in size (Undiluted) |
| Positive control (0720 plasmid) | Big colonies. Carpet of smaller colonies in the middle. (Concentrated) |
Colony PCR After Transformation Using Second Gibson Assembly Ligation Mix
Colony PCR was performed to confirm whether observed colonies were false positives or not. The primers were expected to anneal on the homology sequences used in Gibson assembly. The results of the colony PCR, seen in Figure 8, once again showed no bands of expected size for plasmids containing our inserts, meaning the cells had not been successfully transformed.
However, there was an expected band in the positive control, which means that our colony PCR protocol probably worked. We doubted the lack of correct transformants was caused by the position of inserts in plasmid or the Gibson assembly. Additionally, we expected better expression results in the next cycle and the methods were the same, so we decided to proceed to cycle 2.
Table 6. Numbering of samples in colony PCR, see Figure 8.
Figure 8a. 1.0% agarose gel electrophoresis of colony PCR (gel 1), see table 6 for explanation of labelling.
Figure 8b. 1.0% agarose gel electrophoresis of colony PCR (gel 2), see table 6 for explanation of labelling.
Cycle 2
PCR Introduction of Homology Sequences to Inserts
In order to proceed with the new Gibson assembly in cycle 2, PCR was performed to introduce new homology sequences to the inserts. Figure 8 shows the introduction of the homology sequences to three inhibitors: DB3DB3, CsgF and CsgC, all with SLPmod as a signal peptide. Fewer inserts were used in this cycle to allow for variation of experimental parameters. In this step, a temperature gradient was used in the PCR machine to test five different annealing temperatures for each insert in order to find the most suitable one.
Figure 9. 1.0% agarose gel electrophoresis of PCR product for the introduction of new homology sequences of 3 inserts (CsgC, CsgF and DB3DB3; signal peptide SLPmod) with temperature gradient from 50ºC to 70ºC with Phusion High-Fidelity DNA-Polymerase. The well order (from left to right): 1-5: 5x CsgC_SLPmod, PCR reaction temperature gradient from 50ºC to 70ºC, 6-10: CsgF_SLPmod, PCR reaction temperature gradient from 50ºC to 70ºC, 11-15: DB3DB3_SLPmod, PCR reaction temperature gradient from 50ºC to 70ºC.
Table 7 shows the expected length of the target PCR products from Figure 9. PCR product gel electrophoresis showed clear bands at ~600-700 bp, indicating that the PCR reaction was successful. Considering the intensity of the bands, an annealing temperature of 56.6°C seemed to be suitable for all three inserts since it produced strong and clear bands. In addition, all samples presented a fainter band at the bottom of each line (smaller than 100bp). This is probably primer-dimers. The negative controls did not produce any bands except those from the suspected primer dimers.
Table 7. Expected PCR product length for Figure 9.
| Sequence | length |
| CsgC_SLPmod | 705 bp |
| CsgF_SLPmod | 666 bp |
| DB3DB3_SLPmod | 621 bp |
Taken together, these results suggest that PCR was successful and that subsequent PCR analyses with these inserts were performed with the annealing temperature of 56.6°C.
Colony PCR After Transformation Using Gibson Assembly Ligation Mixture
Transformation was performed using the ligation mixture from a Gibson assembly of the constructs using a newly introduced homology sequence. To evaluate the success, colony PCR was performed again and the visualization of the gel electrophoresis results can be seen in Figure 10. Since no bands indicating plasmids containing our inserts could be seen in the gel electrophoresis, the observed colonies were determined to be false positives.
Table 8. Numbering of samples in colony PCR, see Figure 10. 6:1, 4:1 and 10:1 indicate what plasmid to insert molar ratio was used in Gibson assembly.
Figure 10. 1.0% gel electrophoresis of colony PCR after transformation from cycle 2 Gibson assembly, see Table 8 for explanation of labelling.
Restriction Cloning: PCR Introduction of New Restriction Enzyme Site to Inserts
The negative results from the second Gibson assembly attempt lead us to an attempt for genetic transformation using restriction enzyme cloning. PCR was performed on the original inserts for DB3DB3_Usp45 and CsgC_Usp45 to introduce a new restriction site (EagI). BamHI restriction site was already present on them as the contingency plan for cycle 1. A smaller collection of samples were selected, again to allow for flexibility to change experimental parameters. A temperature gradient was used in the PCR machine to test 5 different annealing temperatures for each insert. Clear bands corresponding to their theoretical values (~700-800 bp) at all temperatures can be seen in Figure 11, meaning that all the annealing temperatures work.
Figure 11. 1.0% agarose gel electrophoresis of PCR product for the introduction of EagI restriction enzyme site to inserts (DB3DB3, and CsgC; signal peptide Usp45) with temperature gradient from 50ºC to 70ºC with Phusion High-Fidelity DNA-Polymerase. The well order (from left to right): 1-6: 5x DB3DB3_Ups45, PCR reaction temperature gradient from 50ºC to 70ºC; 7-8, 10-12: CsgC_Ups45, PCR reaction temperature gradient from 50ºC to 70ºC, 13-14 negative control. Sample 9 was lost during the sample preparation process.
Digestion and Ligation
PCR amplification was followed by double digestion of inserts using EagI and BamHI, thus removing the remaining old homology from cycle 1. As the restriction enzyme sites used for restriction cloning were identical to those used for Gibson assembly, the same linearized vector was used. Digestion was followed by ligation, in which two different molar ratios of insert to plasmid were attempted, 3:1 and 5:1. These ligation mixtures were then used to perform transformation on competent E. coli XL1-Blue cells.
Transformation and Colony PCR
All of the transformation plates had growth, including the negative control control. However, those transformed with ligation mixture had larger colonies, so we still determined that the erythromycin might have an inhibitory effect on the untransformed cells. As such a colony PCR was performed, the results of which are pictured in figure 12 below.
As can be seen in figure 12, the bands did not appear at the expected weight of 801 bp for correct insertion of csgC_Usp45 and 717 bp for a correct insertion of DB3DB3_Usp45, but at 1 891 bp which was expected for undigested plasmid.
Figure 12. 1.0% agarose gel electrophoresis of colony PCR of transformants made using restriction enzyme cloning. Wells contain, from left to right: 1: 100 bp ladder, 2-4: DB3DB3 3:1, 5-7: DB3DB3 5:1, 8-10: csgC 3:1, 11-13: csgC 5:1, 14-16: Negative ligation control, 17-19: Positive transformation controls , 20-22: Negative transformation controls from a previous transformation, 23-25: positive transformation controls from a previous transformation, 26-28: Negative transformation control, 29: Negative PCR control, 30: 100bp ladder.
Termination of Cycle 2
By the end of cycle 2, we had performed multiple transformations with false positives, shown by occasional growth on negative controls, and negative colony PCR results, presented above. As such we deemed it appropriate to switch the cloning system to pET-11a, as it has the robust selection marker ampicillin and was readily available in the lab.
Cycle 3
After switching to use the pET-11a plasmid as a vector, transformation of competent E. coli TG1 cells using a plasmid with CsgC inserted was eventually successful. Similarly to cycle 2, the initial stages of cycle 3 included PCR amplification of inserts to introduce new restriction sites, double digestion of inserts, plasmid prep and plasmid double digestion. These are followed by ligation and transformation. However, this took several attempts. An overview of the attempts is listed below, followed by an extensive description of results from the successful transformation.
Failed Transformations
In cycle 3 we did a series of experiments, changing different parameters at different steps of the process. These experiments are compiled in Table 9, where each ligation was used for one transformation, except the first ligation. Changes involved switching to a different E. coli strain for transformations, increasing the molar ratio of insert to plasmid during ligation. 30:1 was used in the last two ligations since our DNA concentration measurements were in the lower range of the spectrophotometer and might have been faulty. In the final three transformations a positive control for digested plasmid and digested insert was obtained from Lovisa Wallman, PhD-student at Pure and Applied Biochemistry, LTH. As transformants with control plasmid grew, whereas those with the control insert did not, we suspected the fault to be due to the plasmid gel clean up or plasmid digestion. Therefore, the next adjustments were the performance of elution using water in the gel clean up, as well as the addition of a heat inactivation step for the dephosphorylation of the plasmid during digestion (see figure 15 for visualisation of gel electrophoresis result). The transformation using the ligation mixture containing this plasmid was the first to be deemed successful. The remainder of the results will focus on the successful results.
Table 9. Compilation of experiments, the parameters used and observed results. Molar ratios refers to the insert to plasmid molar ratio.
| Plasmid digest | Insert digest | Ligation | Transformation | Results |
| 1st plasmid | 1st PCR and digest of inserts | 3:1 molar ratio, 2h incubation | E. coli XL1-blue | Carpet of cells, suspected to be due to expired ampicillin |
| Same ligation as row above | Same ligation as row above | Same ligation as row above | E. coli XL1-blue | Carpet of cells on plates with expired ampicillin, |
| 1st plasmid | 1st PCR and digest of inserts | 3:1 molar ratio, 2h incubation | E. coli XL1-blue | Only growth on positive control, but only around 20 colonies which was deemed insufficient |
| 1st plasmid | 1st PCR and digest of inserts | 3:1 molar ratio, 2h incubation | E. coli XL1-Blue and E. coli TG1 controls | Only growth on the positive controls: 30 colonies on XL1-Blue and carpet of cells on TG1 |
| 2nd plasmid | 2nd PCR and digest of inserts | 3:1 and 5:1 ratio 2h incubation | E. coli TG1 | Only growth on the positive control |
| 2nd plasmid and control plasmid | 2nd PCR and digest of inserts, and control insert | 30:1 ratio 2h incubation, | E. coli TG1 | Only growth on positive control and ligation with control plasmid. No growth on insert control. |
| 3rd plasmid and control plasmid | 2nd PCR and digest of inserts | 30:1 ratio, overnight incubation at +4 °C | E. coli TG1 | Only growth on positive control and ligation with control plasmid. |
| 4th plasmid with heat inactivation of phosphatase and control plasmid | 2nd PCR and digest of inserts | 30:1 ratio, 2h incubation | E. coli TG1 | See Table 11 |
Introduction of Restriction Enzyme Sites to Inserts
For cycle 3 restriction sites for both BamHI and NdeI were introduced to the beginning and end of the coding inserts, excluding the signal peptide. Since pET-11a contains all other necessary parts for expression, these were the only parts still required. All inhibitors were amplified from templates with SLPmod as the signal peptide. The restriction sites were successfully introduced using PCR amplification, suggested by the appearance of bands at expected weights (Table 10) in gel electrophoresis performed on PCR products.This can be seen in figures 13 and 14 below, which show the results of the 1st PCR and digest of the inserts respectively.
Table 10. Expected PCR product length of inserts in Figures 13 and 14.
| Sequence | length |
| DB3DB3_SLPmod | 240 bp |
| tANK6_SLPmod | 206 bp |
| CsgC_SLPmod | 326 bp |
| CsgF_SLPmod | 283 bp |
| degP_SLPmod | 1442 bp |
Figure 13. 1.0% agarose gel electrophoresis of PCR amplified inserts, to introduce the new restriction sites. Results are from the first attempt. Wells contain, from left to right: 1: 100bp ladder, 2-3 DB3DB3 SLPmod, 4-5: tANK6 SLPmod, 6-7: csgC SLPmod, 8: empty, 9: csgF peptide SLPmod, 10-11: degP SLPmod.
Figure 14: 1.0% agarose gel electrophoresis of double digested PCR products from fig 13. Wells contain, from left to right: 1: 100 bp ladder, 2: DB3DB3_SLPmod, 3: CsgF_SLPmod, 4: degP_SLPmod, 5: tANK6_SLPmod, 6: CsgC_SLPmod, 7: NC (degP_SLPmod) -NdeI, 8: 100 bp ladder.
Due to the similarity in size to undigested inserts, it cannot be determined from this control whether the digestion was successful. However, it can be seen in figure 14 that the inserts still appear to be of expected size, suggesting that NdeI nor BamHI do not appear to be cutting unspecifically.
Digestion of Plasmid
Digestion was done in four attempts, as can be seen in the overview table 9 above. The results of the final plasmid digestion with the added heat inactivation step can be seen in figure 15 below. The expected size of the linearized plasmid is 5637 bp. This band appears in wells with double and and single digested plasmid, but not in the well with undigested plasmid. Similar results were obtained for the first plasmid digests which are not shown here, however some previous attempts showed some impurities.
Figure 15. 0.6% agarose gel electrophoresis of double digested dephosphorylated pET-11a plasmid. Wells contain, numbered from left to right: 1 and 15: 1 kbp ladder, 2-3, 5-6, 11: Double digested pET-11a, 12: Negative control, single digested plasmid using only BamHI 13: Negative control, single digested plasmid using only NdeI 14: Negative control, undigested plasmid 15: 1 kbp ladder
After gel electrophoresis, the bands were cut and the DNA purified. In this purification, some samples were eluted with NE-buffer as done previously, some were eluted with Milli-Q water. This was done in case something in the NE-buffer was compromising the DNA.
Transformation, Colony PCR and Sequencing
Digestion of inserts and plasmid was followed by a ligation made in the molar ratio of insert to plasmid of 30:1 and with a two hour incubation at room temperature. Positive controls for digested plasmid and digested inserts made by a non-team member were used in some of the controls. The resulting ligation mixtures were used in transformation of E. coli TG1 and the contents of ligation mixtures are described in the table 11 below, beside the observation from the plates from transformation.
Table 11: Final transformation of E. coli TG1. PC stands for positive control, NC for negative control. MQ and NE indicates if Milli-Q water or NE-elution buffer was used in the gel clean-up of the plasmid. “NC ligation MQ” refers to a negative control in the ligation where only digested pET-11a was added.
| Ligation mixture | Observation | Ligation Mixture | Observation | Ligation mixture | Observation |
| PC digested insert, PC digested pET-11a | Over 10 colonies grown very close | DB3DB3 NE | No growth | DB3DB3 MQ | No growth |
| PC digested insert MQ | One colony on edge | tANK6 NE | No growth | tANK6 MQ | No growth |
| CsgC, PC digested pET-11a | One colony on edge | CsgC NE | Two colonies grown together,Satellite colonies,One colony on edge of plate | CsgC MQ | No growth |
| NC ligation, MQ | No growth | CsgF NE | No growth | CsgF MQ | No growth |
| PC Undigested pET-11a | Many colonies | ||||
| NC no added DNA | No growth |
The observed growth described in table 11 indicated that some of the inserts had successfully ligated to the plasmid and transformation had been successful. This was confirmed using colony PCR, shown in Figure 16, with labelling explained in Table 12. For samples with positive controls for digested insert a range of expected PCR product length is given, as we were not given the relevant information.
Table 12. Numbering of samples in colony PCR, see figure 16. Three colonies were taken from each plate containing transformants. CsgC is one of our chosen inhibitors, NE/MQ signifies whether NE-buffer or milli-Q water was used for the gel clean up of the used plasmid. PC undig signifies the positive control for the transformation made with undigested plasmid.
| Number on section | Sample | Expected PCR product length | Number on section | Sample | Expected PCR product length |
| 1 | CsgC NE | 465 bp | 9 | PC digested insert MQ | 400-600 |
| 2 | CsgC NE | 465 bp | 10 | PC digested insert, PC digested pET-11a | 400-600 |
| 3 | CsgC NE | 465 bp | 11 | PC digested insert, PC digested pET-11a | 400-600 |
| 4 | CsgC LW | 465 bp | 12 | PC digested insert, PC digested pET-11a | 400-600 |
| 5 | CsgC LW | 465 bp | 13 | PC Undigested pET-11a | |
| 6 | CsgC LW | 465 bp | 14 | PC Undigested pET-11a | |
| 7 | PC digested insert MQ | 400-600 | 15 | PC Undigested pET-11a | |
| 8 | PC digested insert MQ | 400-600 |
Figure 16. 0.6% agarose gel electrophoresis of colony PCR products from transformed TG1. Labeling is according to table 11 (20210916).
Correct insertion of our inhibitor code was also confirmed for clones 1, 2, 3, and 6 from table 12 using the Eurofins Genomics custom DNA sequencing service Mix2Seq [1]. Alignment with csgC inserted in pET-11a was done in benchling confirming the correct insertion of the correct sequence. An example chromatogram for the clone 3 is shown in figure 17.
Figure 17. Chromatogram of sequencing of csgC inserted in pET-11a. The plasmid used in the sequencing was purified from a transformed E. coli TG1 colony E.Z.N.A. plasmid prep protocol.
The plasmid was then purified from an ON culture of transformed TG1 clones and subsequently used to transform competent E. coli BL21(DE3) cells. All plates from transformation of E. coli BL21(DE3) had a lot of growth (>30 colonies) except the negative control.
Expression Study: Cultivation and SDS-PAGE
ON cultures of two BL21(DE3) clones, labelled 7 and 8, were made. Each clone was later used to inoculate three shake flasks. These cultures were later induced with IPTG to initiate the expression of the CsgC inhibitor. This process was attempted twice, as the first set of cultures were grown without the addition of ampicillin as selection marker. The growth of the cells were tracked with OD measurements, shown in figure 18. Three different times of induction were tested for each clone as the optimal induction time was not known. It can be noted that those induced at the same time (7A and 8A; 7B and 8B; and 7C and 8C) appear to have grown in a similar manner after induction. Before induction, OD measurements were only performed for one of each clone (7C and 8C).
Figure 18. Growth of BL21(DE3) clones 7 and 8. 7A and 8A were induced at the same time, 7B and 8B 30 minutes later and 7C, and 8C another 30 minutes later. The control pET-11a culture was induced at the same time as 7B and 8B.
SDS-PAGE was used on samples from different times after induction to determine if the CsgC inhibitor was successfully expressed. If expression had been successful, a band corresponding to the size of CsgC (11.7 kDA) would have been expected to grow more intense in the later samples as the protein would accumulate after induction. However, as seen in the visualisations of the gels in figure 19, no such band can be seen suggesting no expression. However all of the bands in all of the lanes which means that unfortunately the results from the SDS-PAGE were inconclusive. Due to time constraints, we were not able to conduct any additional repeats of the experiments.
Figure 19a. SDS-PAGE gel following expression experiment. Wells contain, from left to right: 1:7A 0h after induction, 2: 7A 1h after induction, 3: 7A 2h after induction, 4: 7A 4h after induction, 5: 7A overnight cultivation 6: undigested pET-11a overnight cultivation, 7: marker, 8: 7B 1h after induction, 9: 8B 2h after induction, 10: 7B 4h after induction, 11: 7B ON cultivation, 12: 8A 1h after induction, 13: 8C 1h after induction, 14: 8A overnight cultivation, 8C: 4h after induction.
Figure 19b. SDS-PAGE gel following expression experiment. Wells contain, from left to right: 1: Empty 2: 7C 0h after induction, 3: 7C 1h after induction, 4: 7C 2h after induction, 5: 7C 4h after induction, 6: 7C overnight cultivation 7: undigested pET-11a overnight cultivation, 8: marker, 9: 7B 1h after induction, 10: 8B 2h after induction, 11: 7B 4h after induction, 11: 7B ON cultivation, 12: 8A 1h after induction, 13: 8C 1h after induction, 14: 8A overnight cultivation, 8C: 4h after induction
SDS-PAGE was performed using samples from different time intervals after induction in order to determine if the CsgC inhibitor was successfully expressed, the results of which are shown in figure 19. Successful expression of CsgC should result in a band corresponding 11.7 kDA from single CsgC and or 23.4 kDA as a result of dimerized CsgC. The gels have been visualized but no bands formed at the expected size range which suggests no expression. Furthermore, according to the figure, all the bands formed after SDS-PAGE were all the same throughout different time intervals when the sample was taken. As we expected the bands to grow more intense in the later samples as the protein would accumulate after induction, this also suggests a lack of expression. However the bands are smeared in the expected size range and all non-sonicated samples in both samples were extremely faint, making it difficult to distinguish one band from the other. It might be possible that expression occurred but is not discernable in these results, as such we interpreted these results as inconclusive. Due to time constraints, we were not able to conduct any additional repeats of the experiments.
Future Experiments
Expression of Inhibitors
If our team had been able to continue the lab work, our first priority would have been to repeat the attempts of expressing our CsgC inhibitor in E. coli BL21(DE3). Suggestions of improvements to that process could for instance include optimizing the induction time. It would also have been interesting to continue doing transformations of E. coli TG1 to try to obtain cells containing plasmids with the inserts for the remaining inhibitors. This could also require repeating earlier steps, such as ligation or even as far back as digestion of the inserts.
Another future task related to the expression of the inhibitors would be to improve the purification of the inhibitors. This could involve optimizing the lysing of the cells by adjusting the amplitude and duration of the sonication. Additionally, the lysate could be filtered to concentrate the sample. Improving the purification would allow for better analysis of both the expression and the effectiveness of the inhibitors.
Curli Assay
Further developing the detection method for the curli protein would allow for testing of the efficacy of the inhibitors, upon successful expression of them. This could be accomplished by continuing to fine-tune the purification process or by exploring other opportunities such as different dyes or forms of microscopy. This new method should provide us with a way to quantify curli reduction over time and as a function of inhibitor concentration.
First of all, while E. coli produces curli in static conditions, it is not triggered at a specific time point. As a result, it’s not discernable for our experiments if curli has grown for the same amount of time in different samples even if inoculation occurs simultaneously. In order to successfully trace curli formation, an inducible expression system is desired. A possible such system would be to introduce the CsgA gene in a knockout strain via an inducible plasmid. Additionally, a knockout strain lacking CsgA could be used as a negative control. After corresponding with Dr. Anne S. Meyer, who was able to offer the LRS10 strain (MC4100 csgA-) at the end of the summer. However, due to the time constraints, these experiments are not included and it is instead something that could be incorporated in future curli studies. This would also provide greater assurance that detected peaks may specifically be due to curli and not ThT binding to other structures in the samples.
Furthermore, curli is not formed immediately in an amyloid state, which complicates planning of the detection, and at what state curli becomes amyloid is still unknown. Again, an inducible system is desired, but a system specifically for amyloid formation is not not yet developed. Instead, experimental design needs to be adjusted accordingly.
An alternative method to observe the presence of curli would be to use a specific antibody for curli subunits CsgA or CsgB after a high performance purification. Once the presence is confirmed, fluorometry using ThT or Bromophenol blue could more readily be used to quantify the amyloid. Alternatively, Congo Red dye could be used to confirm the presence of amyloid as its red staining of amyloids is easy to visually detect [3]. However, Congo Red is known to be mutagenic and should only be handled with the required protection and care.
In addition, we only used one curli strain producer. Needless to say that, more strains would have been better for the flow of the experiments. However, after months of searching for suitable strains for our project, E. coli MC4100 was the only strain producing curli at 37°C that we could get our hands on given our biosafety level.
Expression in Different Chassis
The shuttle vector pTRKH3-ermGFP is appropriate for work in L. reuteri due to it carrying the resistance gene for the selection marker erythromycin. However, from our results from working with pTRKH3-ermGFP, we learned that using erythromycin resistance as a selection marker for E. coli is quite difficult. If given the opportunity, we would have liked to explore other shuttle vectors for L. reuteri or perhaps other chassis all together. A more well-studied bacteria used for secretion of recombinant proteins Bacillus subtilis could have made for a good alternative, as could the strain E. coli Nissle since it is also known to have probiotic properties, which would be in line with the team’s original vision of using a probiotic bacteria to inhibit curli formation.
Testing the System
If a method to detect and measure the amount of curli present in our sample had been successfully developed, that could have been used to measure the effect of our inhibitors once they had been successfully expressed. This would have allowed us to characterize and compare the curli inhibiting abilities of the different inhibitors.
One simple assay for studying inhibitor effectiveness would be to add the purified inhibitor to a culture of bacteria producing curli. By adding different concentrations of inhibitor, the effect of dose dependent inhibition could be studied and compared.
If a secretion system, such as the one our team designed for L. reuteri, was successfully constructed, that system could be tested as well once a curli detection method had been established. This could be accomplished through a co-culture of L. reuteri and E. coli. It must then be kept in mind that the optimal growth conditions of the two bacteria are different. With a co-culture you could study the generic effect of the probiotic, as not only the inhibitor will act but also the reuterin, a protein secreted by L. reuteri [2]. This would serve as a general test of the curli combatting abilities of the engineered probiotic.
Gut-on-a-chip
An idea we had in an early stage of the project was to use the technique gut-on-a-chip to visually study how the curli inhibition would function in a gut environment. Gut-on-a-chip is essentially a microfluidic chip that serves as a model for the whole intestine which recreates multiple dynamics, physical and functional features of the human intestine [4]. It also provides a maintained environment for intestinal absorption and toxicity studies.
Due to time constraints, we were unable to perform any experiments using gut-on-a-chip. In our project, we planned on designing a gut-on-a-chip containing two plates made up of polycarbonate to create two different microchannels aligned on the lower and upper section chambers separated by a tissue patch. The two chambers were to be perfused with culture media required for functioning of the chip. The tissue patch contains a porous membrane with extracellular matrix coated alongside a human epithelial intestine cell and this has similar physiological and dynamical characteristics of a living intestine [5]. A C2BBe1 clone of Caco-2 from ATCC was planned to be used as the human cell line for this experiment. These cells form an apical brush border monolayer which is comparable to that of a human colon morphologically [6].
In conclusion, gut-on-a-chip is an effective model for studying the effect of our intended product in the intestine, and forms a dependable part of our future experiments.
Summary
In summary, there are several future experiments that we would like to conduct to improve and develop our project. They include improvements of both curli and cloning work, as well as further testing of other functional approaches.
References
[1] 1. Mix2Seq - Smart • Easy • Quick [Internet]. Eurofinsgenomics.eu. 2021 [accessed 11 October 2021]. Available from: https://eurofinsgenomics.eu/en/custom-dna-sequencing/eurofins-services/mix2seq/
[2] Cleusix V, Lacroix C, Vollenweider S, Duboux M, Le Blay G. Inhibitory activity spectrum of reuterin produced by Lactobacillus reuteri against intestinal bacteria. BMC Microbiology. 2007;7(1).
[3].Reichhardt C, Jacobson AN, Maher MC, et al. Congo Red Interactions with Curli-Producing E. coli and Native Curli Amyloid Fibers. PLoS One. 2015;10(10):e0140388. Published 2015 Oct 20. doi:10.1371/journal.pone.0140388
[4]. Grandfils E, Casquillas GV. Gut-on-chip: keeping up with the technology. Elveflow. 2020 Dec 2.
[5].Poceviciute R, Ismagilov RF. Human-gut-microbiome on a chip. Nature biomedical engineering. 2019 Jul;3(7):500-1.
[6].Kim HJ, Huh D, Hamilton G, Ingber DE. Human gut-on-a-chip inhabited by microbial flora that experiences intestinal peristalsis-like motions and flow. Lab on a Chip. 2012;12(12):2165-74.