Activatory | riboxswitch

Results: Activatory

Overall Aim

Can we assemble our switch as intended?

Hypothesis: PAGE can be used to demonstrate binding events because of the changing mobility of different structures. If we mix the strands that make up our switch we should see changes in mobility. Assembly may depend on the buffer conditions.

Results:

Gel 1a shows the positions of 5'FAM fluorophore strand, whereas Gel 1b shows positions of any dsDNA (labelled by SYBR Gold). The Switch and Fluorophore strands were present in the solution at equimolar concentrations. Annealed lanes refer to a mixture of strands that have been treated with thermal annealing to remove 'kinetic traps' that may occur during DNA hybridisation.

8% gel for 30mins

The Gels above are 10% PAGE gels, ran at 160V for 40 minutes.

Discussion:

Gels 1a and 1b enabled us to verify basic switch assembly and also find optimal salt concentrations for assembly. We can see that as the salt concentration increases (from left to right), the amount of unbound fluorophore strand (labelled), decreases and the fluorescence in the region of switch increases. This implies that more of the Fluorophore strand is hybridised to the switch and higher salt promotes a higher yield of hybridisation.

Because the higher salt concentrations (1350mM NaCl) are unlikely to occur in vivo, we have decided to continue with 1x buffer concentrations, which are more 'in vivo-like' and are optimal for fluorescence measurements. We then decided to proceed to verifying our switch assembly in a 96-well plate.

Can we detect fluorescence on a plate reader?

Hypothesis: We have chosen our buffer conditions and now we had to check if we can detect fluorescence and assembly by fluorometry on a 96-well plate. A 5' FAM Oligo on its own should have a higher Fluorescence Intensity (FI). Fluorescence intensity should decrease (Quenching) in the presence of Switch and Quencher strands due to co-localisation of the Fluorophore - Quencher pair upon successful assembly.

Results:

Our overall aim for this part of the project was to demonstrate a cooperative dose-response for the trigger sequences with our activatory switch design leading to opening of the switch and thereby allowing access to the Ribosome Binding Site.

We investigated this aim through a number steps:

Labelled Switch

Excess Fluorophore

H2O

1x Buffer

9x Buffer

Increasing Salt

9x Buffer

1x Buffer

H2O

Gel 1b (Stained)

Gel 1a (Unstained)

Introduction

Fluorescence

FRET

Dose-Response

Cooperativity

Can we detect change in structure by FRET?

Can we detect dose response changes in structure? What is our limit?

Top

Mechanism

What is the mechanism of this cooperativity?

Assembly

Hypothesis:

We’ve shown that Activarory switch assembly can be detected by measuring changes in Fluorescence Intensity (FI). In the following experiment, we proposed that this detection can be used to probe structural change. Upon addition of trigger strands in vast excess, we expected to see increase in FI as a result of Switch opening.

Results:

We added excess strands, about 50 times more concentrated than the switch.

Discussion:

There is a significant difference in Fluorescence Intensity between the proposed closed and open states of the switch. Therefore, we should be able to detect structural changes using our sensing system. In another words, our Fluorphore and Quencher oligos are suitable for detection and the FRET distance is sufficient. As we can see, this detection system also works for all types of our activatory switches.

Hypothesis:

Based on our Design, addition of increasing Trigger concentration should lead to switch opening upon successful assembly. The relative mobility of the switch should change as it transitions from open to closed state due to trigger binding.

Results:

Gel 2a shows the position of a switch labelled with 5'FAM Oligo. We can see that as the Trigger concentration increases, the mobility of the labelled Switch increases and the band appears to have travelled further down the gel, going from left to right.

Gel 2b shows all the present DNA strands. In lane 2, there are at least 2 distinct bands, which could correspond to labelled and unlabelled switch molecules. In lane 5, at equimolar trigger concentration, we can see appearance of a third band, that corresponds to an open labelled switch, in complex with Trigger oligos. In lane 6, the closed labelled switch band disappears and we can only see the labelled open switch band. This suggests that the switch has been completely opened at 2x Trigger concentration.

The above samples were run on 15% PAGE gels for 2.5 hourss and 2.5xBuffer.

Discussion:

Obtaining this gel was a little bit tricky and took a lot of trouble shooting. You may notice the odd percentage and run time in our gel conditions. Why is that?

First of all, PAGE allows the separation of macromolecules based on mobility, but there are multiple factors contributing to this. It created a problem for us, since our switch becomes 'more compact', but also heavier as it transitions from closed to open state upon trigger binding. Therefore, we had to find the right conditions that allowed us to distinguish the closed and open states. We used the highest suggested percentage (15%) for our gels. We imaged the gel every 30 minutes to look for the position of the labelled switch band and we stopped the run after 2.5 hours, as the band has reached 3/4 of the gel. As you may see, this allowed us to discriminate the two states with a very similar theoretical mobility.

From these gels, we can conclude that by adding increasing amounts of Trigger, we can see a clear transition of the switch from closed to open state and this transition appears to be very sharp, as we would expect it to be the case for a cooperative system.

Is the dose response cooperative? Can we distinguish it from a non-cooperative system?

Hypothesis:

FRET has proven to be a useful technique for structural change detection. Therefore, we asked if we can also use it to distinguish a system that is designed to have a cooperative dose response (i.e. nH between 1 and number of binding sites) from a system that is non-cooperative (i.e. nH = 1). We tested this by looking for correlation between change in Fluorescence Intensity due to Quenching and the trigger concentration. Fluorescence Intensity should increase as more trigger is added, since (in our design) quenching is reversed by opening the switch.)

Results:

Homo Sym:

Rsq = 0.9907

K = 81 +/- 9 nM

nH = 1.47 +/- 0.20

Single BS:

Rsq = 0.9975

K = 480 +/- 40 nM

nH = 0.99 +/- 0.07

Discussion:

The results show that our Homo Sym switch has a cooperative dose-response, as the nH is significantly greater than 1. Furthermore, our system can tell the difference between a cooperative and a non-cooperative system. The difference, however, becomes less significant at lower concentrations. These results are in line with our model, where adding two binding sites to a previously non-cooperative model with the right set up makes it more cooperative and creates a 'switch-like' behaviour.

Hypothesis:

After verifying that we can create a cooperative system, we then wanted to test the features of our mechanism and check that they truly contribute to cooperative behaviour. In general, if we perturb the system, such that the feature of cooperativity in question is minimized, we expect to see a decrease in the nH and/or expect not being able to distinguish between a cooperative and a non-cooperative system.

Results:

The graph shows three plots normalized based on the K1/2. The goodness of fit is not very good, based on the Rsq values. However, when only one trigger is added at increasing concentration, either 1* or 2*, the system appears to be non-cooperative (nH = 1). The two triggers also seem to have significantly different K1/2 which may suggest a different affinity of binding. When both triggers are added in equimolar amounts, the system appears to behave more cooperatively with nH around 2.

1*:

Rsq = 0.9884

K = 129 +/- 22 nM

n = 1.10 +/- 0.18

2*:

Rsq = 0.9605

K = 50 +/- 16 nM

n = 0.80 +/- 0.20

1* & 2*:

Rsq = 0.9696

K = 128 +/- 21 nM

n = 1.987 +/- 0.52

The graph above shows three experiments. 1* without 2* refers to FI change only related to increase in 1* concetration in the absence of 2* and 2* without 1* refers to the opposite case. 1*&2* refers to FI change when the triggers are added in equimolar amounts at increasing concentration.

The gel shows labelled switch molecules with different trigger conditions. All components were added at equimolar concentrations. We can distinguish three different states, based on different mobility. Lane 1 shows just the labelled switch. Lanes 2 and 3 show the labelled Switch with either 1* trigger or 2* trigger added to the solution. These lanes demonstrate a similar effect of increasing switch mobility, suggesting a switch opening event and increased mobility. Lane 4 contains the switch with both triggers present at equimolar concentration. We can see that the mobility is greater, compared to Lanes 2 and 3, suggesting a 'full' opening of the switch.

Discussion:

The temperature scan data shows that as we increase the temperature, the difference between our cooperative and non-cooperative system becomes much less significant. This is in line with our model, since by increasing the temperature, we destabilise the stem of the hairpin-loop and make the first binding event much easier, thus abolishing this feature of cooperativity. This has also been observed by Plaxco, et al.

The Figures showing Heterotropic Switch 'dissect' our mechanism into two separate binding events. These events can be visualised by PAGE and suggest a two step process of switch opening. The quality of the gel images is somewhat compromised due to running conditions needed to clearly separate the distinct switch states, which limited the imaging. Fluorescence data also suggests the occurrence of cooperative mechanism as we increase the number of binding steps. However, the data does not seem to fit the Hill function too well.

This data supports our model of creating a cooperative system, where the difficult first binding event makes the second binding event easier, which is consistent with the current models of cooperativity, including ours. For example, by increasing the temperature, we change the K1/K2 ratio and thus decrease cooperativity.

Is RBS accessibility affected by structural changes?

RBS accessibility

Gel 3: Unstained

Legend:

B = 1x Buffer

S = Homo Sym Switch

F = 5' FAM Oligo

Q = 3' BHG Oligo

1* = Trigger

NC = Negative control

Discussion:

Figure 1 shows that there a significant decrease in Fluorescence in the presence of all switch components, i.e. comparing columns 3 (F + B) and 8 (S+F+Q+B). Our control lanes suggest that there is an insignificant amount of 'non-specific' quenching that occurs in the solution due to hybridisation of labelled oligos and presence of DNA bases, such as Guanine, that could decrease the quantum yield of our fluorophore (Quenching). Furthermore, columns 9 and 10 show that binding of excess of Negative control strand and trigger strands increases fluorescence. However, note that the FI of Negative Control is much greater than excess of Trigger. Please, see our design page for possible explanations and suggestions of better Negative Controls.

Therefore, we can conclude that full switch assembly is possible and that we can detect change in FI as a result of assembly. We then decided to test our ability to use change in FI to probe structural change.

Can we improve upon our designs?

Hypothesis:

In the progress of carrying out our testing we should optimise our procedures and design so that we can improve on our previous work. Ideally we can investigate interesting behaviour in our switches without investigate lots more time and money by creative use of the strands we already have.

Results:

We designed our "Vrigg" strands, extended versions of the 1* trigger for our homo sym switch to investigate the stem-binding site overlap as a factor in cooperative opening. Our data ended up being a poor fit to a Hill curve however at high concentrations. perhaps indicating other structures such as inter-trigger dimers becoming dominant.

Evaluation

Conclusion

Gel 2a (Unstained)

1 2 3 4 5 6 7 8 9

Increasing 1*Trigger

Increasing 1*Trigger

Closed

Open

Gel 3a (Unstained)

Closed

Half-open

Gel 3b (Unstained)

Gel 3a (Unstained)

Hypothesis:

After proving that we can create cooperative systems, we asked if we can use this to use this for gene regulation by using a FAM- labelled Ribosome 16S rRNA 'mimic' (Design). We expect to see a decrease in Fluorescence as the concentration of Trigger increases and the RBS becomes more accessible. We expect the transition to have a cooperative behaviour.

Results:

Gel 2b (stained)

Labelled Mimic

Gel 4a (Unstained)

Half-open

Switch

Labelled Mimic

The above gels were 10% Acrylamide, run at room temperature for 40 minutes

Gel 4a (Unstained)

Gel 5a (Unstained)

Gel 5a (Stained)

Switch

Mimic

The above gels were 10% Acrylamide, run at 4ᴼC for 90 minutes.

Discussion:

From the gels we can see that the mimic fails to bind to a supposedly open switch at room temperature. This makes sense as the Mimic Oligo is only 8 nucleotides long (Tm ≈ 30ᴼC) and therefore does not hybridise very well. This, however, is not relevant in vivo as the Ribosome makes multiple extensive contacts with the target mRNA and scans for the RBS. In order to promote successful switch assembly, we decided to run this gel again at 4ᴼC. This resulted in increased hybridisation of the mimic to the Switch and successful assembly.

As well as temperature, salt concentration also affects DNA hybridisation. When the sample is loaded into a gel well, the salt concentration can change drastically, possibly leading to denaturation of our complex. This is not the case in 96-well plates used for Fluorescence measurements, where the salt concentration remains approximately constant. That is may be why we were able to see a successful assembly of Switch complexes at room temperature. From the graph, we can see a cooperative-like annealing of the Mimic-oligo to its target sequence. This sequence should be inaccessible in the closed state and become accessible in the open state upon trigger binding. Hence, we were able to indirectly detect the Switch structural change via our Mimic oligo, as the Switch and trigger sequences of the Mimic-Switch are identical to our homotropic symmetrical Switch designs.

Therefore, we can conclude that our design has the potential to cooperatively hide and expose the RBS and also potentially hinder Ribosome binding. However, this system is a very only an approximation to the in vivo Ribosome - mRNA interactions.

Fully-open

Mimic

We conclude the following our activatory in vitro testing:

We can assemble our desired switch strands
We can observe a change in fluorescence corresponding to a change in structure.
This structural change occurs in response to trigger concentration in a cooperative fashion.
This structural change is able to hide or expose an RBS sequence.
Altering the number of biding sites in a loop effects the dose-response relationship f switch opening.
We can demonstrate a heterotropic system where two different trigger sequences are more effective at opening the switch together than alone.

From these results we were confident that we have achieved all our core goals for this design. Future work that we would like to carry out would be to investigate other aspects of switch design include changing stem length and trigger length.