Team:UNESP Brazil/Hardware



     The designed hardware aims to fluoresce RNAs in real time within the cell, without the need for constant sample manipulation.

     Our biological circuit intends to create a programming language for cells, an autonomous and temporal regulation system of gene expression. Cells that will have the ability to process a message and transmit it in binary code. Binary Code is a coding system where all values ​​are represented by 0 and 1. The fluorescent RNA 'Broccoli' will represent the 0s and the fluorescent RNA Corn will represent the 1s.

     The designed hardware aims to fluorescently quantify these RNAs in real time within the cell. A plate reader is the device commonly used to detect fluorescence, but to perform this function a sample processing time is required that can compromise the coding of the transmitted message, as bacterial RNAs are synthesized and matured quickly, and have, in their most have a half-life of 3 to 8 min.

Back to top


     The device built will allow continuous measurement of the sample using a detection system of the luminous intensity of the sample under excitation. The system consists of the following modules:

- Excitation Module

- Capture Module

- Circulation Module

Back to top


Image 1. Representative schematic of the continuous fluorescence detection device.


     As shown in Figure Z, the excitation module is formed by light-emitting LEDs in the 460-465 nm range and a transparent cuvette; the capture module consists of a 532nm optical bandpass filter and the BH1750-FVI GY-30 luminous intensity sensor; the circulation of the medium is made using the circulation module consisting of the reaction vessel and a 12V peristaltic pump controlled by Arduino.

     The 12V peristaltic pump is used due to its precision and simple flow control (up to 100 mL/min). Silicone tubes enable the transfer of fluid from the reaction medium to the emission module, where it is connected to a glass cuvette for excitation. Blue LEDs are used for excitation of the sample at wavelengths from 460-465 nm. More than one LED can be added to the device to enable detection by the light sensor if the sample contains low concentrations or has low emission.

     The BH1750-FVI GY-30 light sensor built into the Arduino is used to detect light intensity. Enabling detection in a wide range of wavelengths with high resolution (1 – 65535 lx), the light sensor used can be found in many devices, such as smartphones, LCD TVs, car navigation and digital cameras.

     As shown in Figure X, the sensor is ideal for detecting light intensity at visible range wavelengths.


Image 2. Spectral response of the BH1750-FVI GY-30 light sensor.


     Observing the response curve of the sensor, about 40% of the light intensity coming from the LED - 465 nm will be identified by the sensor. However, due to its high intensity compared to the fluorescence response of the sample, it is necessary to block the entire wavelength coming from the LED. To ensure light blocking, the optical filter Bandpass FL05532-10 has a CWL (Center Wavelength) of 532 ± 2 nm, with a FWHM (Full Width Half Max) of ± 10 nm allowing only the passage of wavelengths coming from the fluorescence of the sample , as shown in Figure Y.


Image 3. Transmission band of optical filter FL05532-10.


Back to top


Back to top







Back to top


     The presence of interfering fluorescence in a given sample is considered a big problem in fluorometric methods due to the possible spectral overlap between it and the fluorescent emission of the medium. Therefore, it's important to know if there's fluorescence interfering in a given sample and substances of interest under the intended working conditions.

     The tests were performed with green fluorescent protein (GFP), as it accumulates stably in the cell while bacterial RNAs are synthesized and matured quickly. They mostly have a half-life of 3 to 8 min, which can make device analysis difficult.

     GFP has wavelength very similar emission and absorption properties to Broccoli fluorescent RNA. So we chose it as a proof of concept.

Image 4. Broccoli emission and absorption.
Image 5. GFP emission and absorption.

     We chose to use the power medium as it has a low fluorescence interference rate.

     The fermentation media was adjusted to pH 7.0

     Composition of power medium:

     - 70 g/L sucrose

     - 1 g/L yeast extract

     - 25 g/L NaNO3

     - 0.333 g/L KH2PO4

     - 1 g/L Na2HPO4·12H2O

     - 0.15 g/L MgSO4·7H2O

     - 7.5 mg/L CaCl2

     - 6 mg/L MnSO4·H2O

     - and 6 mg/L FeSO4· 7H2O

     Six tests were performed in order to analyze the degree to which the culture medium can interfere with the reading of the samples.

     - Power medium

     - Power medium + cells

     - Power medium + cells + GFP

     - Concentrated GFP

     - 1% saline solution

     - GFP diluted in 1% saline solution

     The Arduino IDE (Integrated Development Environment) is a cross-platform application written in C and C++. It is used to write and upload programs on Arduino compatible boards. We used this program to control and adjust all the functions of our hardware.

     The code is set to take measurements every 0.5s, that is, there is a 0.5 step formation in the graph, but the device can be configured for 25/3 measurements per second. We decided to keep this setting due to the sensor sensitivity.

     The graph above shows the analysis of the cell-free culture medium. The capture signal remains at 0 for almost 120 readings in a row, indicating that the medium does not emit a fluorescence signal that the sensor can detect. It's a good result for our purposes, as the medium does not interfere with data capture.

     When there are cells in the culture medium, the device detects fluorescence signals.

     The code is set to take measurements every 0.5s, resulting in a step around 32s. If we set it to take 25/3 readings per second, we would see curves and not straight lines.

     The chart above is very similar to the previous one. The difference consists in changing the signal from 0.5s to 1.5s, indicating that the basal fluorescence of the GFP-producing cell may interfere in the data capture.

     When we analyze the concentrated GFP, we notice that the signal increases reasonably from 8*10-1 xl to 41*10-1 xl. With concentrated GFP, the excitation module was more effective in amplifying the received signal.

     The capture signal remains at 0 for almost 400 readings in a row, indicating that the saline solution does not emit a fluorescence signal that the sensor can detect.

     The graph shows GFP diluted continuously in 1% saline. It starts at 50x and reaches 1000x until the signal is lost.




Back to top


     The tests showed that the culture medium (power) does not emit a signal that the device can detect, but when the sample of interest is inside the cell in low concentration, there's a possible spectral overlap between GFP and the basal fluorescence emission of the cell.

     The dilution tests showed that the sensor has a high capacity for detecting diluted substances in a medium that doesn't interfere with the signal. It's possible to make adjustments to eliminate basal cell fluorescence overlap by decreasing the pump speed and flow or by adding a powerful laser and mirrors to the excitation module.

Back to top