Team:UPenn/Experiments

Light Theme

Proof of Concept

Fluorescence Calibration

In order to test the device’s ability to detect fluorescence, characterization experiments were conducted with the fluorescent dye, Lucifer Yellow, so that the concentration of dye and thus the expected relative fluorescence would be a known variable. Starting with the stock solution of the dye, serial dilutions of the dye by differing factors were created.


Calibration Curves

In order to create calibration curves for the fluorescence readout from the device, the stock solution was diluted with 3x dilutions until a dilution factor of 2916x was reached. Each dilution was plated in every well of the plate, and measurements were taken. The UV LED excited the fluorescent of the dye, and the photodiode measured the intensity of the fluorescent. The photodiode can take a single reading in about 10 ms. For this experiment, the photodiode took 100 readings per well at a time. These values were averaged to output a single, averaged reading per well for each dilution of the dye. After conducting the experiment thusly for all the dilutions, each well had a single photodiode reading value for each dilution. The different photodiode readings for a single well at different dilutions were then fit to a four parameter sigmoidal curve resulting in 96 different calibration curves.

To test the efficacy of the calibration curves, different dilutions of known concentration were created in random order and plated in all 96 wells. Photodiode readings were taken in a similar fashion where 100 readings per well were averaged to output a single reading per well. These readings were then compared to the calibration curve for the same well. Just by observing the placement of the readings, it seemed probable that the curve could accurately predict the concentration of the fluorescent dye given the photodiode reading.

For further confirmation, the photodiode readings were used to extrapolate the predicted concentration value. The spread of these predicted concentration values was compared to the spread of the initial photodiode readings to determine whether or not running the data through the calibration curve made the data more accurate and precise.


Optical Density Calibration

The device’s ability to read the optical density of a cell culture accurately and precisely was tested in a very similar way to how the fluorescence reading ability was tested. E. coli were first grown in LB broth and then concentrated to make a highly concentrated stock of bacteria. After consulting with collaborators, it was determined that they generally work with optical densities (OD) around 4-10. Thus, the bacteria was concentrated to an OD of about 10 in order to reach that high limit of detection. The actual ODs of the bacteria were measured with a Nanodrop machine and correlate to the concentration of the bacteria in a sample. The sample was then diluted to ODs ranging between 0 and 10. Each dilution was plated in every well of the plate and measurements were taken. Once again, 100 readings per well were taken and averaged so that the output from the experiment was one averaged reading per well for a given dilution.

Unlike the fluorescent reading experiments, the intensity of the red LED can be varied in OD measurements. The OD of each diluted cell culture was measured with the intensity of the red LED varying on a scale of 200-800 (the total range of the intensity of the LED runs from 0-4095 in arbitrary units). Based on the experiments, it was found that intensities below 200 did not accurately represent


Overnight in Bacteria

In order to test the robustness of the device to conduct experiments over long time scales, a series of experiments were conducted over a period of 18 hours. We collected both optical density and fluorescence measurements of arabinose-induced fluorescent bacteria. mAmetrine is a fluorescence protein that was cloned into E. Coli. Over 18 hours, we confirmed that bacteria could thrive in the device which was detected through optical density measurements that were corroborated by fluorescence measurements as well.


Stimulating an Opto-Tool

After validating that each part of the OptoReader worked independently (LEDs and photodiodes), we aimed to test our device in a biological system. Using the pDusk system, we were able to show light-dependent protein production in e. Coli.

pDusk is a system that activates transcription of a fluorescence protein (we cloned mAmetrine) in darkness and represses gene expression in light through a two-component system.

Over 18 hours, we tested two light conditions: darkness and 10% light dosage. Results showed a significant difference in mAmetrine production as indicated by fluorescence values normalized by optical density measurements. Thus, the OptoReader has proven to be able to monitor cell growth, read fluorescent output, and successfully activate/deactivate opto-tools via light-controlled gene expression taking high-throughput data at consistent reading intervals that could not have been achieved through manual reading methodologies that were previously available.


Feedback Control

Lorem ipsum