Fig 1.
Schematic diagram and pictures of the components of device.
(A) General schematic diagram of the device and setup. (B) Photo of the setup using a small plastic food storage container. (C) Larger incubator setup using Styrofoam box that can be used with the same hardware. The latex balloon (orange) and a three-way connector covered with a flexible latex band (yellow) in the CO2 line acted as passive venting features. This ensures a small positive pressure in the thermos to supply the CO2, but without the risk of excessive pressure build up inside the thermos.
Table 1.
Bill of materials.
Fig 2.
Dry ice storage containers for providing CO2 to the incubator.
The thermos thus can store dry ice safely while providing CO2 to the incubator. The capacity to store more dry ice in a larger thermos allows uninterrupted use for over four days.
Fig 3.
Schematic showing how the key components are connected in this 3D architecture-based device.
Fig 4.
Screen display on Raspberry Pi.
The CO2 level went up to around 6% from a 5% setting, which may have occurred due to having slightly too much CO2 injected every time the valve opened. The sudden drop after 7.5 hr was from the lid being opened to remove the culture. The concentration of the CO2 was lowered without the cells and at the duration when the valve opened (achieving 5.2% at a 5% setting). We later optimized an algorithm (sampling time and valve opening duration) to better control the concentration. Consequently, we reduced the valve opening from 3 s to 1 s so that less CO2 entered the incubator.
Fig 5.
Web-based Graphana dashboard for monitoring CO2 levels.
(A) CO2 concentration is stable over 16 hr. Data collected by the device included current CO2 concentration, the CO2 concentration over time, the valve’s status (on/off), bed temperature, and incubator chamber temperature, and bed/chamber temperature over time. (B) The status of the solenoid valve when the valve is opened (we programmed the valve to open for a fixed duration). (C) Temperature of the heater bed (set to 38 °C), the air temperature of the incubator, and the probe measuring the air temperature (near 36 °C as measured). (D) The CO2 concentration (set point of 5%) as measured by the CO2 sensor. The difference between CO2 filtered and CO2 instant is that the filtered values are passed through a digital low-pass filter to reduce measurement noise. This element is a configurable option in the CO2 Meter firmware. We used the sensor program’s default setting of 32.
Table 2.
Primer and probe sequences.
Fig 6.
The culture conditions in the growth chamber reached equilibrium shortly after system start-up and stabilize within and between individual long-term experiments.
Changes in CO2 concentration within the growth chamber over three days are shown. A consistent 5% CO2 level was maintained by the dry ice in a thermos set up without any human intervention. The rise and drop were due to adding the dry ice into the thermos at the beginning and opening the container lid after 3 days.
Fig 7.
Picture of modified Thayer Martin agar plate incubated in three experimental conditions.
As shown in the figure, agar plates incubated in the regular non-CO2 incubator (top) did not have any bacterial colonies, whereas the commercial CO2 incubator (right) and the PrintrLab incubator (left), with CO2 supplied from dry ice stored in a thermos, both had bacterial colonies.
Fig 8.
Picture of modified Thayer Martin agar plate with colonies from serially diluted N. gonorrhoeae cells.
The serially diluted N. gonorrhoeae cells were incubated in the PrintrLab incubator for ~24 hr, with 5% CO2 supplied from dry ice stored in a thermos. The serial dilution was able to isolate and quantify the single colonies using the PrintrLab incubator.
Fig 9.
ATCC N. gonorrhoeae cells cultivated in 5% CO2 incubators proliferated well in 15 hr compared to the incubator providing atmospheric air without CO2 supplementation.
(A) Photos showing the phenol red-added media’s color change after overnight culture. The color change from pink to yellow in both the PrintrLab and commercial incubators suggests that bacterial growth in these two tubes were substantially higher than in the non-CO2-supplied incubator. Such observation can be confirmed by real-time PCR. Duplicate samples were taken from each tube for real-time PCR. (B) 1/10× diluted cells cultured in the incubators. (C) 1/100× diluted cells cultured in the incubators. As seen by the real-time PCR curves, cell growth led to lower Ct values. Cells placed in the nonregulated incubator showed little growth when compared to the sample at 0 hr incubation. Both the 1/10× and 1/100× dilution of cells showed the same trend.
Fig 10.
16s rRNA level in samples at two concentrations stored in a regular (non-CO2 supply), commercial, and PrintrLab incubator for 15 hr.
The sample at time zero was used as a reference. The culture conditions in the growth chamber reached equilibrium shortly upon system start-up and were stable within individual long-term experiments. The error bars of the plots showed the standard deviation based on samples taken from two duplicating tubes.
Fig 11.
Antibiotic susceptibility test performed using Etest strips.
Representative pictures of Etest for the determination of antibiotic susceptibility of ATCC N. gonorrhoeae cells (A) and UAB clinical isolate (B) with ciprofloxacin (CIP) and tetracycline (TET) Etest strips. The UAB clinical isolate was resistant to ciprofloxacin, confirming results from the previous characterization.