2014 Final Report

From BioDesign for the Real-World
Jump to: navigation, search

2014 Team at EPFL

Spring 2014

Falco Enzler is a third year student in Life Sciences at the EPFL.
He's been looking forward in joining the project as it is a multi-cultured and multi-domained experience.
This will thus allow him to finally put at use knowledge accumulated during his studies. On the side, Falco enjoys playing music in a rock band!

Bastien Orset studies life sciences at the EPFL.It's his third year of bachelor.
He is very interested in biology, engineering and molecular processes.
For him, this project seems to be a great mix of the learned knowledge and can help him in future orientation. He likes the humanitarian side of this project and thus finds it even more motivating to take part in it.


Welcome to our final report page. This report will show you how the Biodesign project made progress this year. This year our team was composed of two students.

Our work was to continue what the first team began last year. Thus, we worked on their second prototype of Arsenic detection with GMO bioreporters.

An important aim of this semester was to validate our prototype of arsenic detection by the authorities.

We made some improvements to the prototype too:

  • A screen interface for an easier reading of the results in the field.
  • An extra LED in order to measure the turbidity of our sample.

We will come back on these modifications later.

Also, this wiki might make people aware that the use of GMO in the field isn’t as controversial as one could think. This wiki will show you why.

Here, you can find the plan of our report. Please click on the titles to read the complete paragraph. We wish you a pleasant reading.

The prototype

Principle of our prototype

Our prototype is based on a fluorescent method where we use GMO bacteria expressing GFP (Green Fluorescent Protein) in presence of Arsenic in a water sample. This fluorescence is detected optically and can be quantified in order to measure the concentration of Arsenic.

A vial containing the water sample we want to test is positioned on a socket through which a blue LED passes. GFP absorbs this blue light (λ=475 nm) and emits green light (λ=504 nm) which is detected by a photosensor on which the light is concentrated with the help of two lenses that avoid loss of intensity and that allow more precise results.

Moreover, a red LED was added to measure the turbidity. Turbidity is the cloudiness or haziness of a fluid caused by individual particles. In our case, if there are a lot of GFP, one GFP can absorb light emitted previously by another GFP and then emit light itself and thus creating a chain reaction falsifying the results. The measurement of turbidity will allow us to normalize our results.

With these measures, one can determine the concentration of arsenic in the sample and so know if the water is drinkable or not.


Arduino Uno R3: Microprocessor needed to control the prototype.
Where to get it: e.g. arduino at boxtech, software
27.80 CHF
LCD Arduino Shield: Screen to display the results.
GitHub with schematics, PCB, how to use it and how to build it.
20.00 CHF
XPE Series Cree Blue LED: Emits a blue light at 475nm. Used to quantify the fluorescence of the sample via GFP excitation.
12.95 CHF
3mm Red High Flux LED: Emits a red light at 630nm. Used to quantify the turbidity of the sample.
0.25 USD
Light To Frequency ConverterTSL235R: Photosensor that measures the intensity of light that goes through the sample.
2.95 USD
2x P-channel Mosfet Transistor: Pull-up, PMOS transistor
at Akizuki Denshi
~2.00 CHF
2x Plastic Biconvex Lenses: To concentrate the light coming from the sample onto the photo detector for a higher precision of counting.
at Knight Optical
We’d like to thank Valentin Simeonov at ENAC for providing these for free.
~7.00 CHF
Resistors, wires, solder, button etc…
  • 1x 100 ohm resistor
  • 1x 47 ohm resistor
  • 2x 10 ohm resistors
~5.00 CHF
Total price: ~85.00 CHF

Let’s note that this price can still be further reduced by, for example, choosing a cheaper Arduino microcontroller.

Build it step by step

Our prototype is contained in a wine box. The idea is to have a prototype easily reproducible so you can use any closable box.

To build it, follow the following instructions.

Step 1: The optics

To make this part, you need a wood board where you have to foresee:

  • A socket of 15 mm in diameter where you can put the future sample that we want to test.
  • A square piece of wood to stick the photosensor on.
  • Two trenches for two biconvex lenses that focus the light on a target point.

The distance between the two lenses should be optimal for the focusing of light.
By experimenting, we concluded that this distance was 12,2 cm.

2014 prototype img1.jpg

Step 2: Add the photosensor and the blue LED

For the photosensor, you have to stick it on the square block of wood after connecting its anode and cathode to the wires which will be connected themselves to the circuit.

2014 prototype img2.jpg

For the blue LED, inside the socket, you have to make another hole that allows the light of the LED to go through. Then after connecting the anode and cathode as before, you stick the LED in order to have the light passing through the hole.

2014 prototype img3.jpg

2014 prototype img4.jpg

Step 3: Add the red LED at the end of the box

Connect one wire for the cathode and one for the anode and stick the LED with tape. Be careful not to hide the LED with the tape. One way to do this is to cut a square inside the tape and put the LED in this square.

2014 prototype img5.jpg

Be sure that the red LED is correctly aligned with the sample and the lenses of the board inside the box so they're at the same height. Place the board as close as possible to the red LED.

2014 prototype img6.jpg

Step 4: Make the circuit

The following sketch explains the wiring required. Note that the LED screen is not seen on it and that it’s an interface between the arduino and the circuit. However, when you incorporate the LCD screen on the Arduino board, you will have to solder the cables for the different pins onto the LCD board and then reconnect it to the circuit in the same place. This modification can be seen on the following pictures.

2014 prototype img7.jpg

2014 prototype img8.jpg

2014 prototype img9.jpg


To use the prototype, you’ll need to use this Arduino code so you can start measuring! Comments are added to the code so you can understand it and modify it if you need to. The program is uploaded onto the Arduino microcontroller (via a USB cable) so it can function autonomously without the computer. However, you still need to keep the computer connected to power the microcontroller; a future prototype design will be able to run on a battery.

In short, the program setups the screen and LEDs and then awaits the push of the button before starting the measurement processes. On the push of the button, the blue light will turn on (for 30 seconds) in order to measure the fluorescence of the sample. Once the measurement is done, the blue light is switched off, the result is printed on the screen and the system then pauses for 3 seconds before turning the red light on (for 300 milliseconds) to measure the turbidity. Once the turbidity measurement is done, the red light switches off and the result is printed on the screen. The system then awaits the push of the button before starting the aforementioned process again. Let’s note that the time lengths can be changed in the code if necessary. Here’s a diagram illustrating the code’s loop function that controls the above process:

2014 prototype code diagram.jpg

Global view

Once all these steps are done, assemble everything together and you'll have (hopfully) functioning prototype! To use it, upload the code, place the sample in the socket, close the box and press the button to start measuring.

This is what you should get:



Here you can see a little video giving a brief explanation about the prototype.

One word about GMO

Our prototype uses GMO bacteria for arsenic detection. The detection of Arsenic by fluorescence was inspired of the defense mechanisms of bacteria against Arsenic. Because of recent research, we know that bacteria like E.Coli have two mechanisms of defense:

  • An arsenic pump which expulses the arsenic from the cell.
  • An arsenate reductase which reduces arsenate in arsenite.

ArsR is a repressor of the arsenic defense genes coding for arsenite pumps and arsenate reductase. It acts as a detector of Arsenic. Indeed when arsenic is present, ArsR loses its affinity to the promoter thus leading to the transcription of the genes. So the idea was to express GFP instead of these defense mechanisms.

This diagram illustrates this:

Arsenic reporter how it works.png

To make these GMO, we used a specific laboratory bacteria strain called K-12 E. coli DH5α, which isn’t pathogenic, has a very low survival rate in the environment and that cannot do conjugation (i.e. transfer of genetic material between bacterial cells). The plasmid has a resistance to Kanamycin for easy selection during growth of the bacteria.

But what happens if the GMO are released in the field?

Well, not much actually. If they are still alive, they won’t be able to transfer their plasmid by itself to other bacteria as they cannot do conjugation. Thus no genetic material would be transferred. If the cells die and lyse, we could imagine that the plasmid DNA could be liberated and theoretically be taken up by naturally transformable strains. However, this usually only happens in case of selective pressure; which isn’t the case in the environment. Also, GFP is nontoxic and isn’t dangerous. Concerning Kanamycin resistance, lots of other antibacterial drugs exist to treat bacteria if needed. This resistance could be removed; however we’d lose a lot of efficiency in our measurements.

Due to the GMO character of our bacteria , the question of their confinement was very important and this is why we use a special vial to avoid contamination.

Handling and confinement

In order to deal with this issue of confinement, we use a vial having a silicon septum, thus allowing us to inject a water sample inside the vial containing the bioreporter while keeping the system confined. Indeed, the hole resorbs after the needle is retracted, thus leaving no room for contact between the environment and the GMO; this diagram illustrates this:

2014 GMO img1.jpg

Here’s how this vial actually looks like:


This vial was chosen because of its confinement properties and because it has already been used in another similar project called ARSOlux which was approved by the German authorities. We’ll come back on how we used this as an important argument for the validation of our prototype by the Swiss authorities.

Extra care will be taken to avoid contact between the tip of the syringe and the bioreporter to avoid contamination of the syringe. After use, for safety reasons, the syringe is put in a syringe container like this one:


After the measurement of arsenic concentration, the cells in the vial are, at first, partially neutralized with alcohol before being brought back to the lab for destruction by autoclaving or inactivation by a hypochlorite solution.

Also, all manipulations with the vials containing the GMO are done over a retention tank to avoid any contamination of the environment if the vial would fall and break.

With all the aforementioned precautions taken, we can say that the use of our system is not considered to present a potential hazard to humans, animals and the environment.

Here’s a list of the elements you’ll need:

PK100 screw top clear vial 4ml Sigma Aldrich/Fluka: Vials containing the bio reporter in which we will transfer the water sample into.

Price for 100 vials:

26.88 CHF
Assembled screw cap with hole and PTFE/silicone septum: Caps which allows the injection of the water sample into the vial while keeping the content of the vial confined.

Price for 100 caps:

92.40 CHF
Safe Syringe: So we can safely inject the water sample into the vial.

Price: Varies depending on the model chosen (for 100 syringes):

30.00 USD
- 80.00 USD
Syringe Container; E-safe 0.5 l Universal A: To dispose of used needles.

You can find several online. We’d like to thank Dr. Sabrina Leuenberger for supplying one for free.

~15.00 CHF
Retention Tank: Over which we do the manipulations. It’s just a plastic box, look around you. ~0.00 CHF
Total Price for 100 samples: ~180.00 CHF

Let’s note that this amount can be reduced by choosing cheaper elements. There might also be a way to reuse some of these elements if they are properly cleaned from all GMO after usage.

Lab session and results

The next logical step was to test the prototype in the lab. The goal of the experiment was to measure with our prototype the fluorescence of GFP and the turbidity of some samples on 4 decades.

To do so, we prepared:

  • A solution of bacteria expressing GFP
  • A solution of bacteria expressing GFP in presence of arsenic. So this is the bioreporter we’d use in the field. However, no arsenic was added as we just wanted to measure the background fluorescence generated by bacteria.
  • A solution of LB medium

We measured, with lab equipment, their GFP fluorescence by exciting them at a wavelength of 488nm and by measuring the emission at a wavelength of 525nm. We noted that there was more fluorescence (30561) in the first sample than in the second one (1281) as the first is full of GFP. However, the second sample still has more fluorescence than the background LB medium measure (716), thus indicating that bacteria do indeed slightly fluoresce naturally.

We also measured the absorbance of each sample in order to see the amount of bacteria in each sample. Here are the results:

Wavelength: GFP Bioreporter LB
470 0.4023 0.4589 0.0637
480 0.3951 0.4445 0.0579
490 0.3846 0.4342 0.0535
500 0.3792 0.4273 0.0524
510 0.3688 0.419 0.0508
520 0.356 0.409 0.0484
530 0.3484 0.3984 0.046
540 0.3428 0.394 0.046
550 0.3377 0.3862 0.0453
560 0.3346 0.3812 0.0442
570 0.3268 0.3719 0.0418
580 0.3219 0.3666 0.0417
590 0.3206 0.3622 0.0422
600 0.3139 0.3564 0.0413
610 0.3089 0.3518 0.0413
620 0.3057 0.347 0.0419
630 0.3013 0.3409 0.0392
640 0.2963 0.3361 0.0394
650 0.2932 0.332 0.039
660 0.2879 0.3268 0.0385

2014 labo1.jpg

We noticed that the bioreporter sample had more bacteria. So we had to calculate how much of each sample had to be taken in order to have solutions of each sample with the same amount of bacteria. So we took the above results at 630nm (emission wavelength of the red light we use in our prototype) and subtracted the background absorbance (LB) in order to obtain 0.3013 – 0.0392 = 0.2638 OD for GFP and 0.3409 – 0.0392 = 0.3051 OD for the bioreporter. The bioreporter sample has thus 0.3051/0.2638 = 1.15 more bacteria than the GFP sample.

Having 4ml vials, we prepared a first solution for GFP containing 2ml of the bacteria and 2ml of LB medium. To have the same concentration of bioreporter bacteria, we prepared a solution for the bioreporter containing 2*1.15 = 2.3ml of the bioreporter bacteria and 1.7ml of LB medium.

Having our two main solutions with the same amount of bacteria, we did a four decade dilution of each. To do so we followed these steps:

  • Take 0.4ml of your GFP solution and dilute it in 3.6ml of LB. This solution is thus a 10x dilution.
  • Vortex this new solution.
  • Take 0.4ml of this solution and dilute it in 3.6ml of LB. This solution is thus a 100x dilution.
  • Continue this way until you have a total of 5 solutions:
  • The main solution i.e. the most concentrated
  • 10x dilution solution
  • 100x dilution solution
  • 1000x dilution solution
  • 10000x dilution solution
  • Repeat all the above steps for the bioreporter solution. You should thus have a total of 10 solutions at the end.

Below, you can see the results that we have obtained for the absorbance and the fluorescence for our prototype compared to the true values measured by the lab equipment. (1 indicates the most concentrated solution, 0.1 the 10x dilution, 0.01 the 100x dilution etc…).

For the absorbance, as our prototype measures the turbidity, we have to inverse the obtained values by the prototype in order to correlate them with the others. Indeed the absorbance is a measure that indicates the amount of bacteria in the sample. In our prototype, we detect the light of the red LED that goes through the sample, so the more bacteria there are (so the higher the absorbance); the less light will be detected. At the end, we can observe that the two curves have the same shape and there is only one point which doesn’t correlate with the lab results. Let’s note that there’s no unit of absorbance for the prototype yet which means that we can’t compare the results amplitude wise.

For the fluorescence, let’s first note that our prototype seems to have background fluorescence at roughly 3000. Indeed, for very small dilutions, this background is always present. Where does this come from? We think that the light from the blue LED might manage to reach the detector without passing through the sample. So even though our sample would contain just plain water, light would still reach the detector. To fix this we should add some sort of cardboard which would only let light emanating from the sample reach the detector. One other notable result is that our prototype detects much more fluorescence in the most concentrated bioreporter solution than the lab equipment. Which isn’t normal, indeed this solution contains very little GFP (at least compared to the most concentrated solution of GFP). So where is this false result coming from? It might be because of the fact that the solution contains a relatively high amount of bacteria; these particles might reflect the light coming from the LED onto the detector. We’re not too sure about this and thus further investigations in future tests should be done.

In general, we can conclude that we’ve probably diluted our samples too much. Indeed, when the dilutions are over 100x, both the prototype and the lab tests fail to detect a difference! What does this mean? Well, that we should’ve prepared initial solutions with more bacteria before doing the dilutions. Because of this mistake, our results are partially inconclusive.

Next time, one should test the prototype with higher amounts of bacteria in order to get some concise results. If the results are then not precise, one could consider getting a more precise photo sensor.

2014 labo2.jpg
2014 labo3.jpg
2014 labo4.jpg
2014 labo5.jpg

Validation by the authorities, a step by step to go in the field

As we want to use it in the field, our prototype should be approved by the authorities as we’re using GMOs. This is a delicate social issue and it’s important to consider the media pressure that it could generate. Having an official agreement from the authorities is a very strong argument against any trouble that may come our way.

The following step by step is one way to do that:

Step 1: Get in touch with a biosafety expert

We first met Dr. Sabrina Leuenberger who explained us the legal aspect of our project. She gave us the important law texts we had to consider. For our kind of prototype, we had to refer to the Ordinance on Handling Organisms in Contained Systems» 814.912

Also, she explained us how one can interpret the law and helped us to prepare for all eventualities.

Step 2: Be informed

We had to read some law texts such as the Ordinance on Handling Organisms in Contained Systems. It is important that you come across as someone that knows what you’re talking about. Background reading on all subjects of your project (bioreporter functioning etc…) is an important part of the preparation leading to the meeting with the authorities.

Step 3: Contact the authorities

We sent an email to Basil Gerber from the Federal Office for the Environment (FOEN); he is a deputy head of section and is responsible for the implementation of the Biodiversity Convention in the area of genetic resources. Thus, we organized a meeting with him in Bern to talk about our project. This first contact was required because of the novelty of our project and the problem concerning the use of GMO.

Step 4: Make a powerpoint presentation

During three weeks, we prepared our presentation. This presentation contained some explanation about how our prototype works and targeted the problem of confinement and GMO. We also added some slides about the legal aspects of our prototype. An important fact, is that we compared our prototype to ARSOlux. ARSOlux has a kit called AQUA-CHECK3 that detects the concentration of Arsenic in a water sample by bioluminescence which also needs GMO bioreporters. So the idea was to show them that our prototype was not so different from this already approved kit. Indeed, as we said, we use the same vial. Also, we use a similar strain of bacteria that also can’t do conjugation. With this comparison, one of our strong arguments to show that our prototype is safe, was to give them the ARSOlux risk assessment done by the ZKBS in Germany.

At the end of this presentation, we managed to get a verbal agreement by Basil Gerber. He gave us his point of view about our project and talked about the possible issues concerning our prototype. The main issue was to prevent that the user hurts himself with the syringe and advised us to use of a safe syringe.

Here's the powerpoint presentation we did in Bern with all the main keypoints you should think of pdf

Step 5: Send a notification to the authorities to have a written agreement

In order to have a written agreement, we sent them several files: We filled in the official notification form, added a document where we explained our project and added the powerpoint presentation that we showed to Basil Gerber.

Future Directions and Conclusion


What does the future hold for our prototype?

We envision several improvements that would make it an overall better system:

  • Code a formula that converts the results received by the photo sensor into an actual real arsenic concentration and turbulence indicator.
  • Have a built-in system :
    • Closed box with only one opening to insert the vial. The top lid of the box could have screws so you could remove it if the system inside needs maintenance.
    • Electronic system (i.e. Arduino and electric wiring) incorporated inside the box in order to keep it isolated from the environmental disturbances like water, sun etc…
    • Screen with buttons on the side of the box to start the measurements and read the results.
    • Battery pack to power the prototype so we don’t need a computer in the field.
  • Design a smartphone application that could communicate with the prototype and receive its results. We could then map where the water measurements were done using GPS and save the results to that position.
  • Adapt it to the detection of other harmful elements like Mercury etc…


In conclusion, our prototype is still in its early stages and needs aforementioned improvements to be more efficient and field testable.

Finally, we’d like to thank Sachiko Hirosue and Robin Scheibler for taking us on board on the Biodesign team. We had a fun hands-on, challenging and interesting experience building this prototype while improving our overall skills as future engineers! We hope that our work had an important impact on the advancement of the prototype and we look forward to see how it will evolve with the next team.

We’d also like to thank Prof. Jan Van Der Meer, Sabrina Leuenberger, Siham Beggah, Basil Gerber, Urs Gaudenz and Yashas Shetty for their precious input and help.