Team:Cambridge/Ribosense/Results

From 2012.igem.org

Parts for a reliable and field ready biosensing platform

Implementation of biosensors in real world situations has been made difficult by the unpredictable and non-quantified outputs of existing solutions, as well as a lack of appropriate storage, distribution and utilization systems. This leaves a large gap between a simple, functional sensing mechanism and a fully realised product that can be used in the field. We aim to bridge this gap at all points by developing a standardised ratiometric luciferase output in a Bacillus chassis. This output can be linked up with prototyped instrumentation and software for obtaining reliable quantified results. Additionally, we have reduced the specialized requirements for the storage and distribution of our bacteria by using Bacillus' sporulation system. To improve the performance of our biosensing platform we have genetically modified Bacillus’ germination speed. Lastly, we demonstrated the robustness of our system by testing it with a new fluoride riboswitch, providing the opportunity to tackle real life problems.

One minute tour! :)

>> Return to page
>> Return to page


Contents

Judging Form

  • Please help the judges by filling out this form. Tell them what medal you think you deserve and why. Tell them which special prizes you should win. Help them find your best parts. Show them how you thought about the safety of your project. Helping the judges will help you too.

  • Team: Cambridge
  • Region: Europe
  • iGEM Year:2012
  • Track:Foundational Advance
  • Project Name:Parts for a reliable and field ready biosensing platform
  • Project Abstract: Implementation of biosensors in real world situations has been made difficult by the unpredictable and non-quantified outputs of existing solutions, as well as a lack of appropriate storage, distribution and utilization systems. This leaves a large gap between a simple, functional sensing mechanism and a fully realised product that can be used in the field.

    We aim to bridge this gap at all points by developing a standardised ratiometric luciferase output in a Bacillus chassis. This output can be linked up with prototyped instrumentation and software for obtaining reliable quantified results. Additionally, we have reduced the specialized requirements for the storage and distribution of our bacteria by using Bacillus' sporulation system. To improve the performance of our biosensing platform we have genetically modified Bacillus’ germination speed. Lastly, we demonstrated the robustness of our system by testing it with a new fluoride riboswitch, providing the opportunity to tackle real life problems.

Back to wiki


iGEM Medals for non-software teams

  • We believe our team deserves the following medal:
    • Bronze
    • Silver
    • √Gold

Because we met the following criteria (check all that apply and provide details where needed)

Requirements for a Bronze Medal

  • √Register the team, have a great summer, and plan to have fun at the Regional Jamboree.
  • √Successfully complete and submit this iGEM 2012 Judging form.
  • √Create and share a Description of the team's project using the iGEM wiki and the team's parts using the Registry of Standard Biological Parts.
  • √Plan to present a Poster and Talk at the iGEM Jamboree.
  • √Enter information detailing at least one new standard BioBrick Part or Device in the Registry of Standard Biological Parts. Including:
    • √Primary nucleaic acid sequence
    • √Description of function
    • √Authorship
    • Safety notes, if relevant.
    • √Acknowedgment of sources and references
  • √Submit DNA for at least one new BioBrick Part or Device to the Registry.

Additional Requirements for a Silver Medal

  • √Demonstrate that at least one new BioBrick Part or Device of your own design and construction works as expected; characterize the operation of your new part/device.
  • √Enter this information and other documentation on the part's 'Main Page' section of the Registry
    Part Number(s): BBa_K911004

Additional Requirements for a Gold Medal: (one OR more)

Back to wiki

iGEM Prizes

All teams are eligible for special prizes at the Jamborees. more... To help the judges, please indicate if you feel you should be evaluated for any of the following special prizes:

  • √Best Human Practice Advance
  • √Best Experimental Measurement
  • Best Model

Please explain briefly why you should receive any of these special prizes:

Best Human Practice Advance:

We feel that we deserve this prize for three reasons:

  1. We explored the impacts, *both positive and negative*, of synthetic biology as a solution to real world problems, through interviewing professionals working in a relevant field, namely the impact of arsenic water contamination in Bangladesh.
  2. We recognized existing problems with the way the current direction of synthetic. On going through the registry we found that most of the characterization data for biosensing parts is often neither comparable nor replicable. We have worked to solve this issue, for example with our ratiometric dual channel output.
  3. *Our project doesn’t stop here*, in Chanel number 6 (Team:Cambridge/HumanPractices/FutureDirections) we considered the future implications and technological applications of our project, as well as the means by which it could be improved by subsequent users. We feel that the end to an iGEM project should not be the conclusion of an idea, but the start of it.

Best BioBrick Measurement Approach:

It is absolutely vital that a quantitative, numerical, robust, and flexible measurement approach exists to relay information to a user that is an accurate representation of the input processed by a biological device. Working from these principles, the following was done:

  1. We designed and built Biologger, a *cheap, arduino-based, fully functional automatic rotary device* that has an incorporated ratiolumnometer
  2. Our project is entirely open-sourced and open-platform. We have published source code for the two applications which serve to operate the device, one for PCs and the other for Android devices, as well as the open source circuit design that provides this ratiometric reading. Furthermore, the Android app is able to receive its data wirelessly, which we feel is a great advance in BioBrick measurement.
  3. Our dual-channel luciferase reporter was successfully tested with a dilution series of E.coli transformed with the Lux Operon (under pBAD) biobrick (Part BBa_K325909) of the Cambridge iGEM 2010 team. It can detect, with good accuracy, both different light intensities, as well as the percentages of blue or orange frequencies in a sample.
  4. Our device was successfully tested using artificial light to detect different frequencies (colours) as well.

Having done all the above, we believe that this fully open-sourced instrumentation kit (mechanical) chassis, electronics, software code), estimated at *$35.00* (or $85.00 if a Bluetooth modem is required), is a complete BioBrick measurement solution for any and all BioBricks with a light output.

Back to wiki

Team_Parts

To help the judges evaluate your parts, please identify 3 of your parts that you feel are best documented and are of the highest quality.

  • Best new BioBrick part (natural)
    BBa_K911003
    Best new BioBrick part (engineered)
    BBa_K911004
  • Best improved part(s): None

List any other parts you would like the judges to examine:BBa_K911001, BBa_K911009, BBa_K911008

Please explain briefly why the judges should examine these other parts:

  • Magnesium Sensitive Riboswitch BBa_K911001
    As a riboswitch sensing construct, this part is an entirely new type of biosensor (along with the fluoride construct) that could potentially change the way we think about designing input genetic circuits. Unlike the fluoride riboswitch, it is a derepression system and therefore serves to demonstrate the principle that riboswitches can be used regardless of whether they turn on or off their reporter.
  • Fluorescent ratiometric construct for standardizing promoter output BBa_K911009
    Fluorescence is a major cornerstone for biosensors in the registry, however, most parts do not involve the use of a ratiometric output, which has been shown in the literature to provide much more reliable and meaningful data. This part not only furthers the development of ratiometric measurements in molecular biology but due to the choice of promoters and terminators it can be used to characterize the difference in activity between E. coli and B. Subtilis
  • Fast Germination (B. subtilis) BBa_K911008
    This part is entirely novel for the registry and fully utilizes the recombination machinery inherent in the Bacillus chassis. Have spores that can germinate at a faster rate is certainly a worthy achievement and could help with experiments with B. Subtilis that any future iGEM teams may wish to perform.

Back to wiki

iGEM Safety

For iGEM 2012 teams are asked to detail how they approached any issues of biological safety associated with their projects.

The iGEM judges expect that you have answered the four safety questions Safety page on your iGEM 2012 wiki.

Please provide the link to that page: Page name: Team:Cambridge/Safety

Attribution and Contributions

For iGEM 2012 the description of each project must clearly attribute work done by the team and distinguish it from work done by others, including the host labs, advisors, and instructors.

Please provide the link to that page, or comments in the box below: Page name: Team:Cambridge/Attributions

Comments

If there is any other information about your project you would like to highlight for the judges, please provide a link to your wiki page here: Team:Cambridge/Overview/DesignProcess

Ribosense Results

Fluoride Riboswitch

Having obtained our DNA, the first test was a rather crude β-Galactosidase assay using cells cultured from the plates Yale sent us, the results of which are shown below, the protocol for which is found here.


A β-galactosidase assay of the fluoride riboswitch, the numbers represent fluoride concentration in mM

These results were very positive as it hinted at a good correlation between fluoride concentration and reporter output.


The next step was a more quantitave assay using a 96 well plate reader. This was in line with the assays done in the original paper by Breaker et al. In the paper they carried out a Miller assay, and whilst we have included the full and correct protocol on our protocols page, including the appropriate calculations, we lacked the filters and so were only able to record the A450nm. We feel that this was close enough in wavelength to be a suitable proxy for the A420nm reading required to assay the o-nitrophenol produced. We ran an initial assay and indicative results are below. All the A450 bar charts are based on two repeats, error bars are not included as they would not be statistically significant.

The timecourse of the first run of the β-galactosidase assay using crcB knockout B. subtilis transformed with the fluoride riboswitch
A450nm readings after 420 minutes with crcB knockout B. subtilis transformed with the fluoride riboswitch
A450nm readings after 420 minutes with B. subtilis strain 168 transformed with the fluoride riboswitch
A450nm readings after 420 minutes with E. coli carrying a plasmid containing the fluoride riboswitch

This data seems to show a trend, that at low fluoride concentrations there is a positive relationship between concentration and A450, the A450 then plateaus, and then at high fluoride concentration the A450 drops again, this gives the overall data a bell shape. A possible explanation for this is that there are two conflicting effects, an increase in fluoride leads to an increase in β-galactosidase expression which, in turn, leads to an increased A450. However, the increased fluoride concentrations begin to become toxic and so limit the cell’s ability to produce β-galactosidase, leading to a reduction in A450. It would also seem that the linear region of the initial positive relationship seems to lie between 0μM and 100μM. We therefore repeated the assay, with a longer time course, and with greater resolution in this range, in order to see if the original trend could be reproduced, and also to better quantify this initial correlation. Included below are the results for the repeated assay.

The timecourse of the second run of the β-galactosidase assay using crcB knockout B. subtilis transformed with the fluoride riboswitch
A450nm readings after 790 minutes with crcB knockout B. subtilis transformed with the fluoride riboswitch
A450nm readings after 790 minutes with B. subtilis strain 168 transformed with the fluoride riboswitch
A450nm readings after 790 minutes with E. coli carrying a plasmid containing the fluoride riboswitch

For clarity, and to act as an internal control, the bar chart results are expressed as a percentage change from the same culture, after the same length of time, but with no fluoride added. This assumes that the cultures behaved the same across all fluoride concentrations over the assay, i.e. cell density and productivity were unaffected by the fluoride. This sort of assumption would no longer be needed if a ratiometric system were used, part of our motivation for developing ratiometrica. This data seems to confirm the initial findings, as well as giving greater resolution in the 0μM to 100μM range. The results for higher fluoride concentrations do not seem to quite fit the trend, but as this is consistent across all the cell types it is possible that this was due to errors in making up the solutions, rather than actually being representative of the riboswitch. This data is very encouraging as the WHO safe limit for fluoride is 1.5mg/L (according to Farooqi et al) which works out at 79μM. This means our riboswitch is at its most accurate around the safe level for humans. For more information on the implications of this, have a look at our human practices page.

It should be noted that we expected a greater difference in sensitivity between the knockout Bacillus strain provided by Yale, the 168 strain we were using, and the E. coli we used. In the original paper there was a significant, nearly 100-fold, difference in sensitivity between the knock and wild type bacillus strains, but our data did not reproduce this. However, we feel it is actually very positive as it means our part is consistent across multiple chasses, and that it has the desired sensitivity to fluoride, without the need for a special mutant to be ordered. This fits with our goal for our kit to be a standard which is widely compatible.

The next step was to place this construct in a ratiometric fluorescent construct, before eventually testing our luciferase construct with it. But, due to delays in developing the ratiometric construct, we only had one attempt at inserting and testing it and it proved unsuccessful. We shall carry on testing it and, whilst we cannot put our results on here due to the wiki freeze, we hope to present them to you at the jamborees. However, we would like to include here what we plan to do, in order that future teams might benefit. We plan to insert the fluoride riboswitch into the ratiometrica construct and check that they work together as expected, i.e. get consistent readings for various concentrations, in spite of varying cell densities.

Magnesium Riboswitch

For the magnesium riboswitch, we successfully made the following construct:

The construct made in the pJS130 vector for the detection of changes in Mg2+ ion concentrations

The sequence of this construct was verified via Sanger sequencing.

To characterize this construct, we used a 96-well plate reader to assay the effects of the different concentrations of Mg2+ and IPTG on the levels of GFP. We expected the presence of either to allow the expression of sfGFP, however because transcriptional attenuation by the riboswitch occurs before expression of the repressor protein, it was expected that Mg2+ would somehow demonstrate a dominant effect.

Normalized final fluorescence readings from our construct. Readings were adjusted against final OD620. The construct did not work as expected.

As can be seen, our initial assay did not give the expected results. At differing IPTG concentrations, the response to magnesium seems to have inverted.

Typical growth curves of our construct. Magnesium concentrations varied from 1 μM (most red) to 5mM (most green).

We repeated this experiment, using a different range of magnesium concentrations (1μM - 5mM) and doing duplicates in alternate rows. The cells grew well, although they did not appear to do so in an exponential fashion. It is believed that this may be a feature of the minimal medium in which we were performing the assay, as failed tests using a rich defined medium (which, unfortunately, autofluoresced at GFP wavelengths due to the aromatic amino acids it contained) showed normal exponential growth with the same construct and at identical magnesium concentrations.

Having collected OD620 data as well as fluorescence data, we hoped to generate graphs containing fluorescence readings normalized to cell density. We used the following formula to normalize our data:

NormalisationFormula.png

Initial analysis produced fairly convincing data, as shown in the graphs below (top images). Certainly, the part appeared to produce a convincing response to magnesium between 1 μM and ~10μM, within the sensitivity range of the original paper. A more modest increase can be seen past this point. More careful analysis of the raw data indicated that this apparent trend may have been an artifact of the normalization formula, as no particularly convincing trend can be seen in the raw final fluorescence data (bottom images).

Surface plots of our plate reader assay data. Top images have been normalized to cell population, while bottom images have not. Images on the right leave out the outlier at 1μM.
Line graphs of our plate reader assay data. Top images have been normalized to cell population, while bottom images have not. Images on the right leave out the outlier at 1μM.

However, visual inspection under the fluorescence microscope demonstrated that the low magnesium cultures had very different fluorescent properties, both in quantity and quality of the light produced (higher magnesium cultures had a more yellow hue). We took this as a sign that our plate reader may not be producing reliable data, perhaps not surprising given that we were having to use non-optimal emission and excitation filters. Additionally, the final fluorescence was lower than the initial fluorescence despite the fact that visual inspection demonstrated that it increased considerably from start to finish. This just goes to show the dangers of normalizing data without looking at it first! It also shows the importance of an internal ratiometric control channel which is measured in the same way - this should eliminate the normalization artifacts that we have found from using OD620 as our normalization channel. If the plate reader was unable to measure one fluorescence channel, it is unlikely that it would be able to measure the second, eliminating the artifacts.

We then attempted to transform the construct into Bacillus subtilis, with a view to repeat the same assay in this new chassis. Because the part originally comes from bacillus, we hoped that this might give more useful data. Unfortunately, we did not have time to characterise the part in this chassis, as our transformation attempts failed. This may be an excellent starting point for any future teams seeking to explore this part.