Liquid Chromatography Mass Spectrometry (LCMS)

Liquid chromatography mass spectroscopy (LCMS) is an analytical technique which combines the separation power of liquid chromatography, along with the detection capability of mass spectroscopy. In our project, we have utilised LCMS to determine the degradation of each of our standards. 4-Hydroxy-5-methyl-3-furanone was used to replicate the presence of AI-2 in the reaction. C4-HSL and N-(beta-Ketocaproyl)-L-homoserine lactone are examples of different AHLs, so we used these as a test for our AHL lactonase. First, we created a calibration curve for each of these standards.

4-hydroxy-5-methyl-3-furanone calibration curve

C4-HSL calibration curve

N-(β-Ketocaproyl)-L-homoserine calibration curve

These initial results confirmed that our standards could be detected on the LCMS and provided us with the necessary equations for calculating the concentration of each standard in the assay. In these equations, y is the area under the curve and x is the calculated concentration in micromolar.

4-hydroxy-5-methyl-3-furanone calibration:

$$ y = 68.727454*x - 70.794631 $$

C4-HSL calibration:

$$ y = 636.920486*x - 168.173302 $$

N-(β-Ketocaproyl)-L-homoserine calibration:

$$ y = 7852.257148*x - 6124.904398 $$

Following detection of LuxS and SP-AHL lactonase on the Western blot following the E. coli transformation we set up a 0-hour reaction, 6-hour reaction and a 24-hour reaction by adding the mix standard a then lysate of the enzyme. After the set duration time, we stopped the reaction by adding in a set volume of ice-cold methanol.

The LuxS was not expected to have any degrading action against any of our standards as it produces 4,5,-dihydroxy-2,3-pentanedione (DPD), a precursor of AI-2 ^[2]. As there are no other enzymes in the assay, DPD cannot be converted into AI-2 molecules. Therefore, we compared the activity of SP-AHL Lactonase to both the mCherry and the LuxS as controls.

We tested against a mixed sample, containing all three standards. As expected, the degradation of 4-hydroxy-5-methyl-3-furanone over time by AHL lactonase was negligible, shown in Graph A below. This was analysed by assessing the line for being non-zero and the p values, displayed on the Graph A, are greater than 0.05. The data presented shows the assay with AHL lactonase compared to both mCherry and LuxS. With mCherry and LuxS compared on Graph D.

The following graphs display the results obtained from measuring 4-hydroxy-5-methyl-3-furanone concentration:

A

B

C

(A) Scatter plot of 4-Hydroxy-5-methyl-3-furanone concenration over 24 hours for AHL lactonase and mCherry (B) Scatter plot of 4-Hydroxy-5-methyl-3-furanone concenration over 24 hours for LuxS and mCherry (C) Scatter plot of 4-Hydroxy-5-methyl-3-furanone concenration over 24 hours for AHL lactonase and LuxS normalised

The following graphs display the degradation of C4-HSL and N-(beta-Ketocaproyl)-L-homoserine lactone by AHL lactonase compared to mCherry and LuxS:

D

E

F

G

H

I

(D) Scatter plot of C4-HSL concentration over 24 hours for AHL lactonase and mCherry (E) Scatter plot of C4-HSL concentration over 24 hours for AHL lactonase and LuxS (F) Scatter plot of C4-HSL concentration over 24 hours for LuxS and mCherry (G) Scatter plot of N-(beta-Ketocaproyl)-L-homoserine concentration over 24 hours for AHL lactonase and mCherry (H) Scatter plot of N-(beta-Ketocaproyl)-L-homoserine concentration over 24 hours for AHL lactonase and LuxS (I) Scatter plot of N-(beta-Ketocaproyl)-L-homoserine concentration over 24 hours for LuxS and mCherry.

Graph D shows that the degradation was increased by approximately 2.5 times what was observed when cell lysate from the E. coli transformed with mCherry was used. Graph E shows that the degrading action of AHL lactonase compared to LuxS was over 4 times greater. A one-tailed Wilcoxon signed rank test was carried out to validate this and we obtained a p=0.027 for Graph D and p=0.002 for Graph E. As the p value is less than 0.05, we can reject the null hypothesis (that the samples follow the same distribution) and conclude that there is a significant difference in the distributions of the two samples.

From Graph G it can be concluded that the AHL lactonases were not degrading the N-(beta-Ketocaproyl)-L-homoserine lactone. This homoserine-lactone was isolated from V. fischeri ^[3]. Perhaps indicating that the enzymes we are producing might not be effective against certain bacterial strains.

Graph I shows that the mCherry had greater degradation than LuxS, confirming that the LuxS has no degrading action on N-(beta-Ketocaproyl)-L-homoserine lactone.

From Graph D and Graph E we have shown that there was a significantly increased rate of C4-HSL degradation by AHL lactonase, which is part of the quorum sensing (QS) system in P. aeruginosa ^[4] - a bacterium which is responsible for catheter-associated biofilm formation ^[5]. Meaning we have successfully shown that we can produce an enzyme which can target the QS system of a biofilm-forming species.

Western Blot

In the analysis of our DNA expression, we conducted several Western blots, yielding both negative and positive results. When testing whether the L. plantarum was expressing our DNA molecules, the blots failed to detect any expression. With the limited time scope, we took a step back, testing the transformed E. coli to verify the expression before we had transformed into L. plantarum.

The western Blot above is the for the lysates of L. plantarum after transformations. The blot revealed bands in all samples, but it did not show that the L. plantarum was expressing our desired proteins due to there being a band present in the wild type. This meant that the Western blot was showing a non-specific binding of a protein normally found within L. plantarum with the antibodies.

The above is the western blot for the supernatant of the signal peptide DNA samples to attempt to detect any expression. The membrane did not show any results with no bands on either the wild-type L. plantarum or the samples.

The next was a western blot for the L. plantarum whole cells to cover all avenues before preforming western blots on the E. coli to verify whether the transformation had been successful in E. coli before we had transformed into L. plantarum. This again did not yield any results.

Our next blots showed successful transformation for AHL lactonase and LuxS into E. coli.

A western blot using the plasmids containing the J23100 promoter, a constitutive promoter for E. coli DH5a was carried out. The positive control- His-mCherry - has a band in the correct place and the His-ladder is visible. There were bands in the genes: 5: His-LuxS and 8: Sp-His-AHL lactonase. 

Following this, we transformed our DNA into E. coli with T7 promoters and carried out western blots to verify the expression again. The image above shows that both J23100 (left) and the T7 (right) are expressing AHL lactonase and LuxS. There is a lack of expression within the wild type, validating the success of the assay.

Western blot showing protein expression of enzymes downstream of the strong constitutive promoter J23100. Lanes 1 and 2 are ladders (PageRuler Plus and BenchMark Histagged protein ladder), 3 showing a positive control (mCherry) whilst 4 is an E. coli ‘wild type’ of DH5a (negative control). Lanes ‘7-14’ are our enzymes (7 = his-QQ7, 8 = SP-his-QQ7, 9 = his-QQ5, 10 = his-LuxS, 11 = his-AHL lactonase, 12 = SP-his-AHL lactonase, 13 = SP-his-QQ5, 14 = his-CapA) Western blot showing protein expression of our enzymes downstream with the IPTG inducible ‘T7’ promoter. Lanes 1 and 2 are ladders (PageRuler Plus and BenchMark Histagged protein ladder), 3 showing a postive control (mCherry), whilst 4 is E. coli ‘wild type’ of BL21 (DE3) (negative control). Lanes ‘7-14’ are our enzymes (7 = his-QQ7, 8 = SP-his-QQ7, 9 = his-QQ5, 10 = his-LuxS, 11 = his-AHL lactonase, 12 = SP-his-AHL lactonase, 13 = SP-his-QQ5, 14 = his-CapA). It seems like there is a lot of non-specific binding or protein overloading which has led to smearing on this western blot, making it harder to read.

A band at 36KDa for mCherry, with no bands in our negative control validated our western blot.

The his-LuxS SP-his-AHL lactonase proteins expressed under control of the PJ21300 and T7 promoters were visualised at 18.8 KDa and 33.8 KDa respectively. A further band on lane 10 was seen at ~38 KDa suggesting dimerisation.

Reasons why the beginning western blots could have failed:

• His tag could be hidden in the tertiary structure, therefore cannot been detected by the blot.
• We initially used anti-his primary antibody raised in rabbit however, this was an old antibody and we saw much more success when we swapped to anti-his raised in mouse.
• Maybe L. plantarum was expressing the proteins but in very small concentrations, therefore the blot could not detect the proteins.

Biofilms

Biofilm formation is something that we have been working on in the lab from the very beginning. It started with trying to form biofilms with L. plantarum using both MRS and BHI media, to then forming biofilms with S. epidermidis and P. fluorescens using TSB media.

We used two different methods to verify our results. Firstly, we used the crystal violet (CV) assay to look at the OD₅₉₅, OD₅₂₅, and OD₄₇₅ (when necessary) of each of the wells in the 96-well plate. This allowed us to compare the OD (Optical Density) of our bacteria forming biofilms against a negative control (usually just pure media). As stronger biofilms would be thicker and contain more cells, more cells would be stained, therefore, the higher the optical density, the stronger the biofilm. Secondly, with the help of Christian Hacker, we were able to see our S. Epidermidis and P. Fluorescens biofilms by using the scanning electron microscope (SEM).

Here are the results:

L. plantarum

Bacterial adhesion plays a major role in biofilm formation as previously explained. See our project description. Thus, it was assumed that increased bacterial adhesion would lead to increased biofilm formation. Hence, to test whether overexpressing CapA increased adhesion, a L.plantarum biofilm formation assay (LINK to protocol on experiment page) was carried out.

Briefly, 1% and 5% of bacterial culture was innoculated onto 96 well-plates and left for 24h before a crystal violet stain was used to quantify biofilm presence. Optical density (OD) was measured at 595nm and to compare the data, the OD of the control wells were subtracted from all wells on the respective plate, giving numerical data for each well that could be normalised across different plates and repeats of the experiment.

All this data was then collated to produce this graph:

Two-sample t-tests were used to compare biofilms between WT and CapA bacteria. There was no significant difference in biofilm optical density between WT and CapA L.plantarum at 1% inoculation volume (p = 0.8909) nor 5% inoculation volume (p = 0.07628).

Therefore, our CapA L.plantarum did not have a significant effect on biofilm formation, and consequently adhesion. This is opposite to what has been suggested from literature (See our Description Page). This may be due to the fact that we used biofilm formation as a proxy to measure bacterial adhesion, assuming increased adherence directly correlated with increased biofilm formation, although this may not be the case if there is a trade-off between the two. The possibility that CapA does not affect adhesion at all should not be completely ruled out either, but further work is needed to warrant this. With more time, we would have wished to examine the catheter surface using SEM to quantify the number of bacteria that adhered, allowing us to more accurately infer whether overexpressing CapA increased adhesion.

Another idea we had while working on the adherence aspect was modifying the silicon surface of catheters as proposed by (1). This would involve modifying the silicon to contain mannose residues which L.plantarum can bind to with high affinity through their surface membrane mannose-specific adhesion (Msa) proteins. Although we did not have the time and resources to carry this out, this is something we hope will be tested in future work.

It was also noted from these experiments that the ratio of culture: media in each well also made no significant difference to the strength of the biofilms formed (p= 0.55198 and p= 0.26397 for WT and CapA L.plantarum respectively), which prompted the use of only a 0.01 ratio in future tests where biofilms needed to be formed.

S. epidermidis

After washing away the planktonic cells the optical density of the wells containing S. epidermidis was higher than the optical density of the wells containing just media.

Another set of plates containing the same set up underwent SEM imaging. These are just a few of the images obtained:

Control well:

Well containing S. epidermidis:

As seen in the images the control wells look a lot different to the wells containing S. epidermidis. This proved that the controls were uncontaminated and all the data on these plates was highly likely to be reliable.

To quote Christian Hacker - 'If that's not a biofilm I don't know what is', providing confidence in the method used to form them.

Additionally, the presence of the biofilm matrix seen between several of the cells displays the maturity of the biofilms. Here the extracellular matrix has facilitated in maintaining the adherence of the bacterial cells.

P. fluorescens

Although the P. fluorescens biofilms were slightly weaker and had more variance than those formed by S. epidermidis. The data collected was greater than the controls, validating biofilms had indeed formed, and that the method used worked.

Another set of plates containing the same set up underwent SEM imaging. These are just a few of the images obtained

Control wells:

Wells containing P. fluorescens

The SEM images confirmed successful biofilm formation in P. fluorescens validating the protocol used.

After carrying out these analysis', we had confidence in our biofilm formation protocols.

Inhibition of Biofilms

Biofilm assays allowed us to see if our enzymes were in fact inhibiting the biofilm formation. The same protocol was carried out for forming biofilms, however before the media and bacterial culture were added to the 96 well plate, a set volume (5µl or 10µl) of the enzyme lysate/supernatant was added onto the bottom of each well.

To quantify this, the crystal violet assay was used again to determine if the enzymes did inhibit the biofilm formation. We used pure media as a control and a wild type of S. epidermidis or P. fluorescens to compare against.

To quantify the effectiveness of the expressed molecules, the average of the data collected for the wild type of the P. fluorescens or S. epidermidis was subtracted from the data collected from the wells containing: media, bacterial culture and cell lysate/ supernatant. The remaining data sets were averaged and the standard deviations calculated allowing the graphs below to be plotted. Due to the nature of the data analysis, a relative optical density of below 0 represents a weaker biofilm was formed compared to the biofilm formed not in the presence of lysate/ supernatant. And a relative optical density above 0 represents a stronger biofilm was formed.

Original Promotors Lysate Assays

P32 His-QQ7, P32 His-LuxS, and P32 His-AHL lactonase all had an inhibitory effect of biofilm formation with a smaell variance However due to the very large variance when lysate from the wildtype was present we cannot demonstrate significant inhibition. However this does make them good candidates for further research.

This graph displays a large variance in the biofilms formed in the presence of lysate containing the Wildtype, PldhL1 His-AHL-lactonase, P32 SP-His-QQ5, P48 SP-His-QQ5 and PldhL1 His-CapA. Due to the overlapping of error bars between the wild type and subsequent data sets, we cannot draw substantial conclusions relating to biofilm formation.

As the error bars for both the 0.05 and 0.025 lysate ratios overlap with that of the wildtype, we cannot draw any conclusions as we cannot prove that the lysate assays result in a reduction of optical density, which is a measure of biofilm formation.

Supernatant Assays

The data we collected did not contribute any useful information to our project, so we decided to continue only using lysate assays.

PJ23100

As the error bars, with the exception of the 0.025 lysate ratio for His-QQ7 and SP-His-QQ7, for both the 0.05 and 0.025 lysate ratio overlap with that of the wildtype, we cannot draw any conclusions as we cannot prove that the lysate assays result in a reduction of optical density, which is a measure of biofilm formation.

Looking at 0.025 ratio:

Although the error bars overlap between the different lysates, the results for the 0.025 ratio do look promising as they show a mean decrease in relative optical density of biofilms formed compared to the wildtypes. For example, His-QQ7, SP-His-QQ7, His-LuxS and SP-His-AHL-Lactonase show a decrease in relative optical density and their error bars do not overlap with the wildtype's. These results are very promising.

Promising results

T7

As the error bars for both 0.05 and 0.025 lysate ratio overlap with that of the wildtype, we cannot draw any conclusions as we cannot prove that the lysate assays result in a reduction of optical density, which is a measure of biofilm formation.

As the error bars for both the 0.05 and 0.025 lysate ratios overlap with that of the wildtype, we cannot draw any conclusions as we cannot prove that the lysate assays result in a reduction of optical density, which is a measure of biofilm formation.

Summary

In some strains of bacteria, the presence of cell lysis/ cell lysate, can enhance the biofilm formation ^[6] . However this research seems contended by others where the addition of cell lysate did not affect biofilm formation ^[6]. There is a gap in literature about the effect cell lysate would have on S. epidermidis or P. fluorescens and thus it is difficult to explain the results we saw without further research into it. However it can be theorised that as the biofilm formation generally enhanced more when more lysate was added, there is a corelation between the amount of lysate and the effect it will have on the biofilm formation.

Further research would need to be done to find the optimum amount of lysate that would provide enough molecules to inhibit biofilm formation but not enhance biofilm growth too much.