1.1 Salkowski assay

The Salkowski assay is a colourimetric assay that detects IAA with high specificity among other indoles. There are many different types of Salkowski reagents which work at different concentration ranges of IAA and with varying specificity. They all vary slightly in composition and measurement method. We used the most specific reagent according to a paper which works at a concentration range of 0 to 260µM^[5]. Modelling of the Auxin Xpress construct informed us that IAA production would be within this range. This standard assay is the simplest way to determine whether there is IAA present in solution. First we created a standard curve with increasing IAA concentration in LB broth using synthetic IAA (Figure 1&2). This was used to determine IAA concentration from OD measurements of IAA-producing E. coli.

The aim of this experiment was to determine whether our construct is working and roughly how much IAA we are producing when compared to control cells without the construct.

Figure 1: Standard curve of Salkowski assay made with synthetic IAA in LB. (Data by Imperial College London iGEM team 2011).

Figure 2: Cuvettes used to measure OD for the standard curve. As IAA concentration increases, the solution progresses towards red. (Picture by Imperial College London iGEM team 2011).

We did preliminary tests with our IAA producing cultures with a spectrophotometer. Once IAA presence was confirmed by colour change, we set up a more thorough assay with a BMG Omega plate reader. The first assay we did was with E. coli DH5α (Figure 3&4), the results of which were positive for our IAA construct. The IAA producing E. coli were producing approximately 55 µM of IAA. From modelling, we have determined that our construct would be able to produce 72.25 µM IAA, which shows that we were in the correct order of magnitude. E. coli are known to naturally produce IAA, which is why all of our controls show moderate levels of IAA production^[1]. However, cells containing the Auxin Xpress construct have repeatedly shown to produce almost twice as much IAA as control cells without.

Figure 3: Results from trial 1 of Salkowski assay with cell filtrate of IAA producing E. coli DH5α. Filtered through a 0.2 µm pore filter. (Data by Imperial College London iGEM team 2011).

Figure 4: Visual results correlating with OD measurements. The eppendorf on the right contains IAA producing E. coli DH5α and the eppendorf on the left contains control E. coli DH5α cells. (Data by Imperial College London iGEM team 2011).

Due to inconsistent results with colour change on various occasions, we redid the Salkowski assay with repeats where we varied certain conditions. We tested different growth media, tryptophan concentration and light exposure to optimize IAA production.

Interestingly, from testing the Salkowski assay on synthetic IAA in LB left overnight in dark versus light suggests that light exposure does lead to IAA degradation (Figure 5).

Figure 5: Testing the effect of light exposure on synthetic IAA stability. The cuvette on the left shows the colour change at point zero. The three middle cuvettes were left in the dark overnight and the three on the right were left exposed to light, after which Salkowski reagent was added to all samples to observe colour change. (Picture by Imperial College London iGEM team 2011).

Figure 6: Salkowski assay performed on IAA producing E. coli and control E. coli incubated for 20 hours in different media and at two different temperatures. All samples were incubated in the dark. (Data by Imperial College London iGEM team 2011).

Due to the results of the light exposure test, all future cultures were incubated in the dark. In the assay comparing growth media (LB and tryptone broth) the OD of each sample, after 20 hours of incubation, was measured at 600 nm and appropriate dilutions of cell filtrate were made to normalise against the different cell densities of samples. The OD of cell filtrates after the addition of Salkowski reagent was measured at 530 nm and blanked with the appropriate growth medium for each sample. The OD measurements were then converted to IAA concentration with the standard Salkowski curve. Surprisingly, the results (Fig. 6) suggest that IAA production was optimal in LB, although the OD at 600 nm of cultures grown in tryptone broth (very nutrient rich) was much higher. We cannot draw a conclusion from this data. However, it seems that the IAA producing pathway endogenous to E. coli is much more complicated than anticipated. We postulate that IAA is not produced when growth conditions are very favourable and cell density is high.

The final Salkowski assay we tried was with E. coli BL21 transformed with the Auxin Xpress construct. This strain has two proteases knocked out and is favourable for protein over-expression. We left the liquid cultures to incubate in the dark for 20 hours and then measured OD as previously described. From this assay we got extremely high yields of IAA (shown by the bright red colour after addition of Salkowski reagent to cell filtrate, Figure 7) in the Auxin Xpress and control cells. We think that the assay was saturated and therefore we will need to test the supernatant with HPLC.

Figure 7: Salkowski assay of filtered supernatant from IAA expressing E. coli BL21 and control competent E. coli BL21 that have not been transformed. The extremely high levels of IAA produced seem to have saturated the reagent and therefore a conclusion cannot be drawn.

1.2 HPLC

Since the BL21 strain of E. coli seemed to be saturating the Salkowski reagent we looked for alternative quantification methods to measure the amount of IAA being produced by BL21 cells containing the Auxin Xpress construct. The extraction method resulted in a large loss in IAA from the cell filtrate and no peaks were detected by HPLC, therefore we ran filtered cell supernatant directly through the column and the peak characteristic to IAA at 227 nm was visible.

From the Salkowski results, we assumed that cells containing the Auxin Xpress construct were producing about 50 uM IAA and therefore the positive control used for HPLC was at the same concentration. From the positive control, only one peak at 227 nm showed up at about 15 mins so we knew that this corresponded to IAA and looked for the same peak in our samples between 15 and 16 minutes.

Considerably smaller IAA peaks were present in the sample of interest and negative control and a difference in magnitude between the two could not be determined (fig 8). Therefore we needed an even more sensitive approach to accurately determine how much IAA our engineered bacteria were producing.

Figure 8: HPLC peak corresponding to IAA. Positive control is 50 uM IAA in acetonitrile. Auxin Xpress is filtered supernatant of E. coli BL21 transformed with the Auxin Xpress construct. Negative control is filtered supernatant of E. coli BL21 transformed with a vector containing only a promoter and antibiotic resistance.

.

1.3 LCMS

The third test we did to determine the concentration of IAA secreted by our engineered bacteria was liquid chromatography mass spectrometry. This highly quantitative method combines HPLC and mass spectrometry making nano-molar concentrations detectable. Luckily Dr Colin Turnbull and his team were running IAA samples on LCMS and generously ran our samples for us. E. coli strain BL21 were used to maximise IAA output.

The LCMS peak confirms that IAA is present in our sample (fig. 9)and the data shows that E. coli transformed with our Auxin Xpress construct produce more IAA than cells without the construct (Fig. 10). Unfortunately the extraction method resulted in a high loss of IAA in the sample(post extraction concentrations were nearly 12 fold lower than pre-extraction). Although the relative levels of IAA are still informative, we hope to optimise the extraction in the future to improve the yield of IAA.

From this data we know that E. coli engineered with the Auxin Xpress construct, produce 37 percent more IAA than negative control E. coli.

Figure 9: Peaks produced from LCMS, the peak area of the MRM transition 176-130 was used to quantify IAA. The bottom window shows the peak for authentic IAA run in the same batch compared to the test sample extracted from our engineered bacteria. The peak around four and a half minutes corresponds to another unkown metabolite. The peak at about six and a half minutes corresponds to IAA.

Figure 10: LCMS results show that cells containing the Auxin Xpress construct produce substantially more IAA than control E. coli. Positive control 1 was LB spiked with 50 µM IAA, positive control 2 was the negative control spiked with 50 µM IAA, and the negative control was E. coli BL21 transformed with a vector only containing a promoter and antibiotic resistance.

Collapse

2.1 Confocal imaging with fluorescent reporter cell lines

We are not only interested in constructing the IAA-producing pathway in our bacteria but we also want to investigate what effect the IAA has on plants to verify our assumptions about the effects of indole 3-acetic acid (IAA). This will help us with the human practices aspect of our project, as it is part of product safety testing, and it will also provide a good assay for the functionality of IAA-secreting bacteria.

To observe how IAA influences plants, we will be working with the plant model organism Arabidopsis thaliana. Arabidopsis is well-established for research into plant biology and researchers have established lines that respond to IAA exposure by expressing reporter genes, which are particularly useful for our project.

We used DR5:3XVENUS plants that respond to IAA by expression of YFP to look at the plant response to synthetic IAA and bacteria-secreted IAA. This will allow us to monitor how much IAA is taken up and which cells respond to it. We will be using confocal microscopy to image and evaluate the relative strength of fluorescence expressed by the plant. This type of imaging is especially useful as it makes three-dimensional imaging of samples possible thanks to precise imaging of individual points. The fluorescence acted as an indirect reporter on the IAA concentration supplied as it relies on the plant expressing fluorescence in response to stimulation by the hormone.

2.2 Effect of IAA produced by our bacteria

Once we had engineered our bacteria to express IAA, we an experiment to investigate the IAA's effect on the roots. This was done by supplying the plants with bacteria that produce IAA. Plants that were not exposed to bacteria showed strong autofluorescence in response to the IAA naturally produced in the roots. However, the roots supplied with bacteria unable to express IAA only exhibited very weak fluorescence. Plants supplied with IAA-expressing bacteria were brighter than either of the two controls. Imaging was carried out at precisely the same settings to ensure that the data could be compared and is representative of actual fluorescence.

In order to quantify the expression of fluorescence in the plants exposed to different treatments we processed the images in ImageJ and analysed fluorescence intensity in individual particles. The relative fluorescence intensity of roots exposed to our auxin-secreting bacteria is 1.93x that of roots exposed to no bacteria and 2.9x that of roots exposed to non-auxin-expressing bacteria. It therefore appears that bacteria normally suppress production of IAA in root tips. However, the bacteria we have engineered to express IAA produce enough of the compound to not only overcome this limitation but to also increase expression of the reporter gene.

2.3 Effect of synthetic IAA

We exposed Arabidopsis seedlings to different concentrations of synthetic IAA.

Plants incubated with 0.1mM of IAA showed strongly enhanced lateral root growth but also stunted growth. This does not come as a surprise as concentrations this high are usually considered detrimental for the plant.

At lower concentrations, stunted growth was not observed. However, root tip cells were very brightly fluorescent and increased root branching could be observed.

At even lower concentrations, fluorescence was almost impossible to detect and much less branching of the root can be observed.

2.4 Effect of IAA concentrations on Arabidopsis root growth

To assess the effect of IAA on plant roots, we applied different concentrations of synthetic IAA to Arabidopsis seedlings in liquid culture. The results of this experiment have given us an idea of the ideal concentration of IAA that our engineered bacteria should be producing to optimise root growth.

Figure 11: Arabidopsis grown in liquid media (A) 3 flasks of Arabidopsis in liquid media containing 3 different IAA concentrations : control (left), 0.1 pM (middle) and, 0.1 nM (right) (B) Set up of Arabidopsis grown in liquid media. (Photos by Imperial College London iGEM team 2011).

The literature states that Arabidopsis is sensitive to IAA concentrations of 0.1 nM - 10 mM with increased branching seen between 0.5 µM - 20 µM^[2]. In our experiments, we initially tested the effect of IAA concentrations of 0.1 nM - 10 mM. Unfortunately, our negative control (seedling with no additional auxin) got infected by a fungus preventing us from determining whether IAA could enhance root growth. Analysis of the uninfected seedlings revealed that increasing IAA resulted in decreased root growth. In response to this finding, we repeated the experiment with an extended IAA range of 0.01 pm - 10 mM. Within this extended range, we were able to find the optimum concentration of IAA for enhanced root growth. Seedlings were grown for three days and imaged using a Fujifilm LAS3000 Imager. Data analysis was conducted using ImageJ (Figure 12).

Figure 12: The effect of IAA on root length. (A1) shows root growth of plants supplied with 10¹¹ fM IAA and (A2) with 10⁹ fM IAA (B1) shows the root growth with 10⁷ fM IAA and (B2) with 10⁵ fM IAA. (C1) shows root growth of plants supplied with 10 ⁴ IAA and (C2) with 1,000 fM IAA (D1) shows root growth of plants supplied with 100 fM IAA and (D2)with 10 fM IAA (E) Comparing the effect of root growth with 10¹¹ fM (0.1 mM), 10⁹ fM (1 µM), 10⁷ fM (10 nM), 10⁵ fM (0.1 nM), 10 ⁴ fM, 1,000 fM, 100 fM, 10 fM and control (no IAA) from left to right. (Data by Imperial College London iGEM team 2011).

Plotting average root lengths (Figure 13) shows that IAA concentrations between 10 fM to 0.1 mM promoted root growth, whereas concentrations greater than 10 nM stunted root growth. These results provide us with an approximate threshold IAA concentration of 10 nM, past which Arabidopsis' roots growth will be negatively affected. We can use this value to optimise IAA production by our Auxin Xpress construct by switching the promoter and RBS without surpassing the threshold value. The root branching could not be determined by the Imager since the plants were to small to observe. However, the confocal microscopy shows the root tip morphology at higher IAA concentrations (from 10⁹ fM onwards) to be much more fibrous than the control.

Figure 13: Plot of root length data collected demonstrating root length of Arabidopsis as a bell-shape curve with increasing IAA concentrations in liquid medium. (Data by Imperial College London iGEM team 2011).

The biomass of the plants were also determined using the dry weight (figure 14). The plot of the biomass of 20 Arabidopsis seedlings against the different IAA concentrations in liquid media they were grown in show the similar pattern with the root length. This confirms that IAA could increase the plants' biomass and the increasing degree corelates well with the increase in root length.

Figure 14: Plot of biomass of 20 Arabidopsis plants. Results show biomass of Arabidopsis as an approximate bell shape curve with increasing IAA concentration in liquid media. (Data by Imperial College London iGEM team 2011).

2.5 The effect of IAA distance from the root tip on root growth.

The effect of IAA on root growth was also tested in phytogel by dotting it at varied distances from the root to simulate growth. On these gels, individual seedlings grow horizontally into a gel containing plant nutrients. These gels enable us to supply the plant with IAA at set distances from the seedling itself. From these experiments we hoped to determine whether our IAA producing bacteria will have an effect on plant root growth from around the root in the soil without being taken up by the root. We used DR:3VENUS seeds. These germinate into plants whose roots respond to IAA by expressing YFP.

From modelling we estimated the level of IAA secretion from our bacteria to be 10¹¹ fM. Accordingly, we added three different IAA concentrations ranging from 10⁸ fM to 10¹⁰ fM IAA at five distances ranging from 2 to 10 cm from the seed.(IAA was applied on the left side) Five replicates were set up for each concentration. The roots grew perpendicular to the line on which IAA and Arabidopsis seeds were applied. The phytogel plates were kept in the dark and at 4°C for three days in order to prevent the degradation of IAA. They were then transferred into light for another six days to observe root growth. After nine days total, the phytogel plates were imaged and root lengths measured.

Figure 15: Phytogel plates applied with IAA concentrations of (a) 10¹⁰ fM (10 µM), (b) 10⁹ fM (1 µM), (c) 10⁸ fM (0.1 µM) and (d) control (0M) from left to right. (Data by Imperial College London iGEM team 2011).

Figure 16: Plot of root length data showing the effect of IAA concentration on root growth when applied at different distances from the seed. (Data by Imperial College London iGEM team 2011).

These results (Figure 16) suggest that IAA can have a promoting effect on root growth in close proximity to the seed, as well as at distances further away. Accordingly, whether our system promotes or inhibits root growth is influenced both by the position of the bacteria relative to the root and the amount of IAA produced by the bacterial population. Future optimisation of IAA production levels therefore depends on the affectivity of the Phyto-Route module and how close to the root our bacteria would swim in the soil and whether uptake into the roots would take place in nature. In future iterations of the design cycle, this data could influence both Phyto-Route and Auxin Xpress.

2.6 The split root experiment

To visually compare the difference of the root growth with and without IAA supplementation, we set up a split root experiment as recommended to us by Dr Alexandru Milcu. This experiment extends beyond the scope of normal controls because the same plant is subjected to two different treatments in phytogel. We grew one half of the seedlings in phytogel containing three different concentrations of IAA and the other half in phytogel without IAA. Three replicates of seven day old Arabidopsis DR:3VENUS were set up for each concentration and grown for two weeks before measuring root length (Figure 18)

Figure 17: The split plates applied with IAA concentration of (a) 10¹⁰ fM (10 µM), (b) 10⁹ fM (1 µM), (c) 10⁸ fM (0.1 µM) and (d) and 10⁵ 0.1 nM. (Data by Imperial College London iGEM team 2011).

Figure 18: Curves showing the decrease of root length ratios by increasing IAA concentrations in split root experiment. (Data by Imperial College London iGEM team 2011).

In figure 18, the ratio is shown to fit logarithmically with increasing IAA concentrations. However, the inhibiting effect found in phytogel is higher than the liquid medium. We then carried out the experiment with lower IAA concentrations to observe the positive effect of IAA on root length. (Data by Imperial College London iGEM team 2011).

Figure 19: The split plates applied with IAA concentration of (a) 10⁴ fM, (b) 1,000 fM, (c) 100 fM pM, (d) and 10 fM The right half of the plates shows the control. (Data by Imperial College London iGEM team 2011).

The ratios between the IAA samples and the controls at each concentration were calculated and plotted into a graph as the same as mentioned above

Figure 20: Curves showing the decrease of root length ratios by increasing IAA concentrations in split root experiment. (Data by Imperial College London iGEM team 2011).

Even though in phytogel, the root length decreasing trend is shown quite clearly and reach the maximum at 10 fM, however, all of the concentrations lower than 10⁵ fM obviously shows higher root length than the control.

2.7 Data fitting to support the IAA uptake and root growth modelling

A root growth rate of 0.96 cm/day with the external IAA concentration 5x10^-5 mol/L is selected from the literature in order to simulate and visualise the root system (Click here to see the details of the root growth simulation). To get an accurate growth rate which is particularly fitting to our project, we decided to do data fitting analysis to the Arabidopsis we plant. The root length and number of branches of the Arabidopsis were recorded every two days from day 0 to day 9. Root length, daily root growth rate and number of branches are plotted against time and IAA concentration. Then the plots are analysed to give an approximation of the relationship between IAA concentration and root growth. The following graph gives an example of root length against time^[6].

Figure 21: Root growth speed decays over time. This graph gives a prediction of the root growth speed (cm/day) for 20 days. The exponential decay constant is 0.048. (Data by Imperial College London iGEM team 2011).

Figure 22: Primary root length (mm) vs. time (day) and external IAA concentration (mol/L). The data fitting result gives this plot of the Arabidopsis primary root length. The length increases as the growth time increases, and reaches a maximal depth when the IAA concentration is approximately 1 pM. (Data by Imperial College London iGEM team 2011).

Figure 23: Primary root growth rate (mm/day) vs. root growth time (day) We used the data fitting toolbox of Matlab to obtain this figure. The relationship between the growth rate and the IAA concentration can be approximated by a Gaussian equation. The abnormality of the 0.1 nM curve is due to two contaminated samples which stopped growing at 7 mm after day 5. Figure 9 is consistent with the prediction of the decay of the root growth speed given by Figure 18. (Data by Imperial College London iGEM team 2011).

Figure 24: Number of lateral branches vs. external IAA (log) concentration. The optimal concentration for lateral root branching = 1 uM-10 nM, at this concentration, the Arabidopsis root gained the most lateral branches. (Data by Imperial College London iGEM team 2011).

The result of the data fitting does not coincide with our latest wet lab experimental results, since the optimal external concentration for Arabidopsis is found to be 0.1 nM. The failure of the previous experiment is analysed and gives the guidance for the new experiments in the following ways:

1. Instead of measuring the primary root length and number of lateral branches of Arabidopsis, the growth rate should be tested by measuring the dry mass of the roots.

2. At the early stage of planting, the number of lateral branches is very small, therefore massive percentage errors and standard derivation are produced.

3. Huge error is also introduced by the dilution process of the IAA solution to reach a low concentration level of 10^-13 mol/L.

4. The wet lab result shows that the IAA concentration in phytogel media is negligible after 9 days, the degradation of IAA is an important factor when the root growth is modelled.

Collapse

3.1 Soil erosion experiment

Promoting root growth to anchor top soil is the main objective of our project. We set up a soil erosion experiment to evaluate the effect of IAA on Arabidopsis thaliana's ability to hold down soil. In particular, we wanted to know which concentration of IAA causes A. thaliana to retain the most soil. We exposed plants to differe IAA concentrations ranging from 0.1 mM (10¹¹ fM) to 10 fM. Fifteen plants were seeded per pot and two replicates were set up for each concentration. We chose to plant our seeds in M2 compost which consists of organic compost and sand. This compost was chosen to minimise errors resulting from different soil mixtures.

Figure 25: Experimental set up (A) 3 pots of Arabidopsis watered with 3 different IAA concentrations : 0.1 pM (left), 10 nM (middle), and 1 mM (right) (B) Set up of Arabidopsis growing in preparation for the erosion experiment.(Data by Imperial College London iGEM team 2011).

The effect of IAA can be seen at a very early stage. We observed the germination of the seeds that had been exposed to different IAA concentrations (Fig. 26 and 27).

Figure 26: Germination speed and effectivity is affected by the by external auxin concentrations. (Data by Imperial College London iGEM team 2011).

Figure 27: Graph showing the correlation of the plants' ability to grow more than two leaves related to the external IAA concentration. (Data by Imperial College London iGEM team 2011).

Plants watered with IAA concentrations ranging from 10 pM to 10 nM germinated first and produced a higher number of leaves than plants exposed to concentrations higher than 10 nM and lower than 10 pM. The growth trend caused by differing IAA concentrations in soil is quite similar to that observed in liquid media. However the ideal IAA concentration for plant growth is higher in soil than in liquid media (1 nM rather than 0.1 nM). This might result from higher dispersal and degradation of IAA in soil.

We allowed the plants to grow for four weeks and then looked at the effect of IAA on Arabidopsis. More specifically, we wanted to look at whether there would be a difference in water retention within the soil. The water mass was obtained by measuring the difference between the ratio of the wet soil mass and the dry soil mass right after the Arabidopsis was watered and 3 days after being watered. The results are shown below .

Figure 28: Water retainment by the soil with plants watered in different IAA concentration, control is soil without any Arabidopsis. (Data by Imperial College London iGEM team 2011).

The results correlate with the results obtained from the experiments on Arabidopsis in liquid culture. The optimal IAA concentration which gives the best water retetntion is between 0.1 nM - 1 nM.

While Arabidopsis is considered a fast growing plant, it takes several weeks to grow large enough to hold down soil effectively. This effect is exacerbated by the fact that we are growing several plants in one pot, thereby increasing the competition for nutrients. The plants took a total of five weeks to reach a size suitable for conducting the final erosion experiment. We then set the experimental setup to imitate water erosion as shown in the picture below.

Figure 29: Experimental set up of the erosion experiment. (Picture by Imperial College London iGEM team 2011).

The pots were cut open at one end to allow water with eroded soil to flow into the collecting basin. The slope was constructed by placing the pots on top of a 20 cm x 10 cm slanted platform to create a 30 degree slope. We applied water with a hose to imitate heavy rain fall. The pressure of water was kept constant by an adaptable shower head which imitated the average speed of rainfall (9 m/s) as suggested by Dr. Martin Selby. The shower head was placed about 90 degree to the ground and approximately 30 cm above the plant pot and water was applied for 10 seconds.

The eroded soil was collected from the basin and heated to evaporate the water so that we could measure the dry mass. The amount of dry mass collected gives us an indication of the amount of soil that would be swept away by water. Our results therefore show that soil erosion can be prevented through the use of vegetation that has been supplemented with IAA.

Figure 30: Mass of Soil eroded in different IAA concentration, control is soil supplemented with IAA but without Arabidopsis. (Data by Imperial College London iGEM team 2011).

The results show a similar pattern with the water retainment results. In bothe of these results, 0.1 nM appears to be the magic number. At this concentration, the mass of soil eroded by water was the least and the water retention was the highest. Compared to the control with Arabidopsis and no IAA supplementation, the optimal IAA could improve soil retention by a factor of 100. Moreover, the optimal IAA concentration could decrease the mass of the soil eroded by a factor of 200 when compared to a control where soil in the absence of Arabidopsis was irrigated with IAA supplemented water.

After the experiment, the plants were collected from each IAA concentration and incubated to obtain the dry mass. The results are shown below.

Figure 31: Biomass of plants grown in different IAA concentrations (Data by Imperial College London iGEM team 2011).

The results of the biomass show a similar trend with the water loss, mass of the soil eroded and the biomass in liquid media. This allows us to support the viability of our project. This proof that IAA does increase plant biomass at optimal concentrations has huge implications in the use of our system in other applications such as improving crop yields.

Collapse

2.1 Behavioral analysis

During the planning stage of our project we discussed several ways that the IAA could be delivered to the plant. In the end, we decided to either have the bacteria chemotax to the root and secrete the IAA around it or to have them be uptaken by the root through a natural process that has been recently documented in 2010^[7]. In order to determine whether the latter option is viable we have to look at the cell viability once it has been taken up into the root. In order to study that, we decided to use a photoconvertible protein called Dendra2.

Dendra2 is a photoconvertible protein that normally exhibits green fluorescence but can be irreversibly switched following irradiation at 405 nm or 488 nm to red fluorescence. This conversion cannot be reversed^[4]. Dendra2 is therefore very useful for observing the metabolism of cells: if all protein in a cell is converted to red fluorescence and it subsequently expresses green fluorescence, the cell has to be metabolically active.

4.1 Photoconversion

First we had to determine whether the excitation wavelength of the green fluorescense could photoconvert Dendra2. Since the excitation of the absorbance prior to photoconversion is 486nm and the photoconversion wavelength according to the literature is 488nm we had to make sure that a wavelength of 486nm would not cause photoconversion. In order to test for this, we excited cells with 486 nm and 558 nm wavelengths and measured the fluorescence at the emission wavelengths 505 nm and 575 nm.

Figure 32: Graph showing the photostability of Dendra2. The excitation wavelength of 486 nm does not cause photoconversion even though the wavelength of 488 nm does (click on graph to enlarge). (Data by Imperial College London iGEM 2011).

The results show no detectable photoconversion meaning that the excitation wavelength of 488nm can be used. Once this was verified we attempted to look at the photoconvertability of Dendra2. We used a wavelength of 405nm and 505nm and then measured the emission at 575nm. Since the FluoroMax-3 machine we intended to use could not emit both of these wavelengths at the same time we had to seek the help of Tim Wilson who pried open the machine and modified it with his own 405 nm laser pen and a filter from our gel imager. These modifications allowed us to obtain the following results.

Figure 33: Graph to show the increase in fluorescence at 575 nm after excitation with 405 nm and 505 nm. (Data by Imperial College London iGEM 2011).

We found that Dendra2 will fluoresce red upon conversion by light at a wavelength of 405 nm. Dendra2 reached its maximum conversion after only 200 seconds making it suitable for in vivo studies. We think that the degradation after the maximum conversion is due to photobleaching. We then decided to move on to something more interesting: testing whether Dendra2 will convert within a bacteria that has been taken up by a root cell!

4.2 In vivo photoconversion

Following the successful uptake of our Dendra2 expressing bacteria we conducted an in vivo photconversion study on the confocal microscope. Conversion with the 405 nm laser was completed after about 15 rounds of irradiation at 50% laser intensity with the pinhole set to 3 airy units.

Figure 34: Dendra2 photoconversion in bacteria taken up inside plant roots. The graph on the top displays averaged fluorescence over the entire photoconverted area and the amount of brightfield light (the background light used to see the outline of the roots) recorded is therefore very high. The graph on the bottom displays emission at green and red fluorescence over the same time span.

Video 1. This video shows the photoconversion of Dendra2 within E. coli cells that have been taken up into the plant roots as a time-lapse of pictures taken after each round of irradiation at 405nm. The targeted area of cells being photoconverted corresponds to the top graph in Figure 30. There is a single bacterium visible on the right that was not targeted for photoconversion and serves as a control. (Data and imaging by Imperial College London iGEM team 2011).

As is visible from Video 1, protein in bacteria that are inside the area targeted by the laser switches from green to red fluorescence after about fifteen rounds of irradiation. One can also see a single bacteria that was left outside the targeted area that was not converted red. The switch from green to red fluorescence can be visualised much more acutely when observing the change in fluorescence in a single bacterium.

4.3 Tracking metabolic activity

We did a follow up experiment to trace the metabolic activity of the cells inside the plant roots by photoconverting Dendra2 in E. coli taken up by the roots and periodically image the same cells to assess whether they remain metabolically active once they are inside the plant. This is especially important as we would need to know whether the cells would be able to actively produce IAA inside the root cells.

Before carrying out this experiment in Arabidopsis we photoconverted cell culture producing Dendra2 and measured fluorescence over a period of incubation to mimic how the cells will behave inside the plant. As expected, greed fluorescence decreased after photoconversion and red fluorescence increased. As the sample was left to incubate at 30˚C, green fluorescence re-emerged as the bacteria are producing more Dendra2. This proves that the cells remain metabolically active and that metabolic activity can be tracked using Dendra2. Surprisingly red fluorescence also increased over time without additional photoconversion. We suspect that this may be a delayed response from the initial photoconversion.

Figure 35: Experimental set up for photoconverting cell culture of E. coli transformed with the Dendra2 construct. A laser was used to excite the protein with blue light to photoconvert Dendra2 from green to red.

Figure 36: Green and red fluorescence emitted by E. coli expressing Dendra2, before and after photoconversion with a blue laser.

Figure 37: Re-emergence of green fluorescence over time as cells produce new Dendra2 protein that has not been photoconverted.

4.4 Tracking metabolic activity in vivo

To assess whether the bacteria were metabolically active inside Arabidopsis roots, we photoconverted dendra2 inside Arabidopsis roots as before, but this time left the roots in cygel over night so that they could be re-imaged after 24 hours. The bacteria were shown to express green Dendra2 protein, proving that they are metabolically active inside the roots (Video 2).

ImageJ was used to analyse the relative fluorescence intensities in the different samples. This analysis showed that the amount of red fluorescence increased after photoconversion and decreased after 24 hours' incubation. There was a high background fluorescence in the red spectrum as roots are highly autofluorescent in the same spectrum. The amount of green fluorescence had decreased after photoconversion but not completely disappeared, as not all protein was photoconverted by the laser. However, green fluorescence increased after over night incubation, showing that new protein was produced in the cells (Figure 36).

Figure 38: Relative fluorescence of Dendra2-expressing bacteria inside Arabidopsis roots.

Collapse

Since we have been working with the new photoconvertible fluorescent protein Dendra2 in order to do the in vivo imaging studies, we thought that it might be a good idea to characterize and compare other characteristics of this protein to the other fluorescent proteins in the registry. Since we were already working with sfGFP for our soil survivability experiments and root uptake experiments as well as mRFP1 for the Gene Guard we decided to use these as well as the standard GFPmut3b to compare their thermostability values. This study will give a good indication of which proteins will work best in thermophiles which will be useful for any future project involving thermophiles.

To characterise these parts, we determined the temperature at which these proteins denature and cease to fluoresce. The data was collected in two overlapping sets, ranging from 35°C to 66°C, and from 57°C to 96°C. This was because it was only possible to fit a maximum of eight samples into the thermocycler at once.

Once the data was collected by taking fluorescence readings from a 96-well plate, it was normalised to a sample of untreated cell lysate containing the fluorescent protein. This gave a value for Relative Fluorescence, which was plotted against temperature to create a scatter plot. A line of best fit was applied, and the mid point of the sigmoid region of the line was taken as the denaturation point of the protein.

Figure 39: Results of the heat denaturation experiment. The temperature at which half of the protein is denatured measured by looking at its fluorescence (PTm50) mRFP1: 82.2°C; GFPmut3b: 61.6°C; Dendra2: 89.1°C; sfGFP: 75.0°C.

The sigmoidal curves that were calculated gave us the following function in order to find K which we also call PTm50 (temperature at which half of the protein is denatured measured by looking at its fluorescence):

Collapse

[1] Ball, E (1938) Heteroauxin and the growth of Escherichia coli. Journal of Bacteriology 36(5): 559-565.

[2] Spaepen S, Vanderleyden Jos and Remans R (2007) Indole-3-acetic acid in microbial and microorganism-plant signaling. FEMS Microbiology Reviews 31(4): 425-448.

[3] King JJ, Stimart DP, Fisher RH and Bleecker AB, (1995) A mutation altering auxin homeostasis and plant morphology in Arabidopsis. The plant cell 7: 2023-2037.

[4] Gurskaya N et al. (2006) Engineering of a monomeric green-to-red photoactivatable fluorescent protein induced by blue light. Nature Biotechnology 24: 461-465.

[5] Gordon, S. & Weber, R. (1951) Colorimetric estimation of indoleacetic acid. Plant Physiology 26: 192-195.

[6] Leitner D, Klepsch S, Bodner G, Schnepf A (2009) A dynamic root system growth model based on L-Systems: tropisms and coupling to nutrient uptake from soil. Plant and Soil 332(1-2): 177-192.

[7] Paungfoo-Lonhienne C et al. (2010) Turning the table: plants consume microbes as a source of nutrients. PLoS One 5(7): e11915.

Collapse