In this article I will summarise the tests that the previous teams did with the different prototypes and the one I did last week. The first set of tests was done by the 2013 Biodesign team. They used their prototype that had a 475 nm LED to excite the GFP and a digital camera. The aim of these tests was to know if it could detect different levels of fluorescence. It was found that the prototype was sensible to light that was too concentrated. They found that for small concentrations of dextran ( 0-0.02 [g/L] ) the prototype was able to precisely distinguish the different concentrations, but with higher concentrations the image was saturated and the different samples could not be differenced. Afterwards the bacteria were tested in a laboratory fluorometer, and as expected we found an absorption peak at 488 nm and an emission peak at 508 nm. The second prototype was also tested with a dextran solution, whose intensity is 4 times greater than eGFP, and a linear curve proportional to the dextran concentration was found. There was no problem regarding the saturation of fluorescent light.
They then followed the protocol to detect the Arsenic by using the biosensor, as expected the results gave a linear fluorescence intensity curve proportional to the Arsenic concentration. The two prototypes were also tested with the Arsenic and the bioreporter. The camera prototype could not distinguish at all the different samples, and the second prototype did not even produce results of the eGFP signal because it could not be distinguished from the LED light despite the filter. See here for the complete report of the lab session.
The 2014 Biodesign team also went into the lab to test their prototype, which we will call v1.0. The prototype v1.0 was tested and the results were compared to the results obtained by using a laboratory range fluorometer. One notable result is the fact that the prototype detects much more fluorescence in the most concentrated bioreporter solution than the lab equipment, which could be due do the light scattering. The prototype and the fluorometer did not detect any signal after a certain level of dilution was reached, therefore it was concluded that the tests have to be began with enough bacteria. The tests were not done with a known concentration of bacteria, this has to be done in the future to be able to proceed to a precise calibration of the prototype. See here for the complete report of the lab session.
Last week, I went into the lab myself to test my upgraded prototype v1.0. As the last session of test, I worked with three different solutions. One of LB medium, one of GFP bacteria and the bioreporter (which would work as the blank and to establish the extent of the light scattering due to the bacteria).
The first task of the day was for me to get familiar with laboratory methods as I had not set foot in a lab for more than one year. That being done we grew 2 different strains of bacteria. Our bioreporter and one strain that always expresses GFP.
The bacteria being grown we than tried to adapt the Arsenic assay protocole from the laboratory of environmental and evolutionary microbiology from the University of Lausanne. Our aim was to adapt the protocol in a DIY way. We ware able to test our prototype with three different samples, one filled with Luria-Bertani medium, one with the bioreporter and a last one with GFP producing bacteria. We also ran the test a second time with a 510 nm with a 50nm bandwidth filter.
The immediate conclusion from these experiments is that we can clearly see that the red light from the turbidity measurement is removed by the filter, which what was expected from the theory. We can also conclude that we have a lot of light scattering due to te blue LED, the measurements increase by 300% from the LB sample to the Ars sample. We have to find the right concentration of bacteria to have enough signal from the GFP, but the least light scattering possible (we need a better SNR). We have to investigate alternating the LED and the measuring device, and observe the results. My aim would be to investigate precisely the effect of the filter in every possible situation and then implement that in the Arduino code, in a way to implement a digital filter that does not distort the measurements.
From these measurements we can also see that the concentration of the bacteria was roughly the same in the Ars sample and the eGFP sample. One thing that could have disrupted the measurements is that I forgot the cardboard cache that I had built, maybe the mesures taken with the cash would have less rogue rays that disrupt the measurements.
For the next laboratory sessions special attention has to be given to a certain number of points.
– A number of measurements with different concentrations have to be done with this prototype, as was done during the other lab sessions. Each measurements has to be done 3 to 5 times to have a precise result and eliminate possible one-time errors.