Synthetic biology is a combination of biology and engineering that attempts to gain new understanding and utilization of the natural world by applying engineering principles to design genetic elements for cellular control. Common examples of synthetic biology include memory switch [1] and oscillator [2] circuits that enable cells to switch between distinct states based on chemical or other input. This ability to sense an input from the environment enables cells to serve as a biosensor. A synthetically engineered biosensor [3–6] takes advantage of the versatility of living cells to detect and process inputs into analyzable outputs. Living cells, especially bacteria, can survive under many conditions [7,8] and can be engineered to detect and diagnose infectious biomarkers in clinical settings.

Genetically modified organisms (GMOs) are regulated by many laws and governmental bodies [9–13], making them difficult to deploy outside of laboratory settings. To allow engineered biosensors to be deployed for disease diagnosis in the field, these regulations must be satisfied. Additive manufacturing, in the form of three-dimensional (3D) printing, can serve as a solution to isolate GMOs from the environment when used in the field.

Critical additive manufacturing approaches were first commercialized three decades ago [14]. Since then, additive manufacturing technology has evolved rapidly, making 3D printers widely available. The 3D printers utilized for this work function by successively depositing layers of extruded filaments to build a 3D structure [15]. These types of 3D printers have already been proven to successfully design, print, and test medically relevant devices [16,17].

In this work, we chose to employ a robust approach for detecting the quorum signal of Pseudomonas aeruginosa (P. aeruginosa) that is frequently used in synthetic biology (Fig. 1(a)). The living biosensor consists of engineered Escherichia coli (E. coli) that are designed to detect the presence of this quorum signal, N-3-oxododecanoyl homoserine lactone (3O-C12). The living biosensor was constructed by introducing components of P. aeruginosa’s native 3O-C12 signaling pathway to E. coli. Here, a transcription factor, LasR, acts as a signal transducer by binding 3O-C12 and dimerizing. This complex then binds to a transcription activation domain found within its cognate promoter, PLas [6,18,19], which is also transplanted into E. coli. In our design, once bound, LasR promotes the expression of the gene for mCherry, a red fluorescent protein. We specifically chose 3O-C12 as the chemical to be detected due to its relevance in medical settings [20–25].

As an example of the clinical relevance of this approach, in rural and less-developed areas, lung infections from the organism P. aeruginosa can be acquired in local clinics. Upon lung infection by P. aeruginosa (Fig. 1(b)), the organism may be found in lung sputum samples. This sputum can be collected and placed within commercially available sample tubes. As a result, an approach enabling these samples to be immediately analyzed in the field, using the tools of synthetic biology, would be ideal for early diagnosis.

In the solution presented here, the sample tube is mated with a unique cap containing the synthetic biological sensor (Fig. 1(b)). This unique 3D-printed device completely encases the living biosensor, allowing its transport from the laboratory, while also allowing the sensor to be introduced directly to a sample. When the sputum sample is ready to be analyzed, force is applied to the device by the user’s finger, causing the bacterial biosensor to break out of the housing and fall into the sample. Detection of 3O-C12, which will be produced by P. aeruginosa in the sample or is already present in the sample, causes the living biosensor to produce mCherry (Fig. 1(c)) and turns the sample red; this color can be detected through fluorescence analysis or visually with the human eye. Thus, a colorimetric change in the sample can indicate the presence of P. aeruginosa in the lungs.

P. aeruginosa functions as an opportunistic pathogen in patients with cystic fibrosis, causing recurring infections of the lungs and respiratory tract. Current diagnostic technology typically takes the form of enzyme-linked immunosorbent assays (ELISAs) [26] and real-time polymerase chain reaction (PCR) assays [27], which are either expensive or require extensive training to administer. The sensing approach described here could serve as a simple and inexpensive complement to current diagnosis techniques for detecting P. aeruginosa. Since the cellular biosensor can be grown within a secure, 3D-printed device for containment, this device can be used as a platform for the deployment of synthetically engineered biosensors in the field.

The 3D-printed device was designed to attach to a common laboratory sample tube form factor (Fig. 3(a)) to ensure ease of use. The device consists of two components (Fig. 3(b)), referred to herein as the “housing” and “cover.” The cover serves to isolate the contents enclosed in the housing interior from the surrounding environment. In addition, the top of the cover is designed to be ergonomic for pressing with a thumb or forefinger. Once assembled, the device can be enclosed within a rectangular prism with dimensions of approximately 0.026 m × 0.026 m × 0.023 m. A cross-sectional schematic of the assembled device (Fig. 3(c)) shows the housing and cover held together by a press-fit. The housing holds a bacterial well, which contains rich agar with a small lawn of bacteria; this serves to sustain and contain the bacterial biosensor within the device. Next, we added small mechanical features to the device, the design of which was based on modeling with a finite element method software suite [33], COMSOL Multiphysics®. It is important to note that our models were based on construction with ABS, a thermoplastic. Thermoplastic polymers become pliable after heating, harden upon cooling, and are frequently used in 3D printing. ABS is often used because its glass transition temperature of 105 °C makes it well-suited for 3D printing, as the extruder temperature in many 3D printers is set at 230 °C [34]. This high printing temperature also enhances the sterility of the printed device.

In addition to the cover, housing, and bacterial well features discussed above, we designed a small mechanical feature that enables an axial force (i.e., a force applied directly downward on the cover, along the major axis of the sample tube) to cause the bottom of the housing to quickly separate from the housing, thus releasing the biosensor into the sample during a (designed) catastrophic failure. In order to cause this fast yielding, we created two depressed annuli features in the housing bottom (on both the interior and exterior, shown in Figs. 4 and 5 with additional detail). Herein, we refer to these annuli as “stress-focusing cutouts.” These cutout features serve to concentrate the stresses that develop from applied axial force on the cover, thus allowing simple activation of the device by means of a user’s thumb using a pinch grip.

Finite element analysis was used to numerically evaluate the stresses that developed within the device when force was applied during its use. The stress profile through the stress-focusing cutout was modeled using COMSOL for different designs in order to arrive at the final version, which is presented in this paper [33]. An evenly distributed pressure load was applied to the surface of the housing by the cover to generate stress in the cutout. Three different designs of the stress-focusing cutout were modeled in order to identify the stress magnitude and concentration in the cutout for a given applied force. Relative to other preliminary designs, the final design presented here (Fig. 3(d)) has both the highest maximum stress developed within the cutout for a given applied force and the best spatial distribution of the high stress concentration through the cutout. This design of the cutout allows our device to be constructed from a relatively high-strength material, while still requiring a relatively small force to activate the sensor by breaking the cutout.

The device was also designed to be simple to assemble and use, as shown in Fig. 6. A 40 µL agar plug was added to the bacterial well and the 3O-C12 living biosensor was plated on its surface. The device cover was then fit onto the housing and attached to a culture tube, which isolates the biological sample from the environment. The user’s thumb is then used to activate the device. This applied force on the cover develops stress in the stress-focusing cutout within the housing. The bacterial well then breaks away from the housing and falls into the sample. After analysis, the cover can be pushed into the housing to create the second stage of the press-fit seal, thus preventing the sample within the test tube from escaping during a leak, in the event that the device is inverted or agitated. This seal also allows the user to transfer the biohazardous waste back to a biosafety lab for proper disposal.

Upon immersion in the sample, the biosensor was grown for 8 h. The 3O-C12-activated biosensor then produced mCherry protein. As shown in the flow cytometry analysis (Fig. 7(a)), the induced biosensor showed a shift in fluorescence emitted at 590 nm, which was distinctly separate from the un-induced biosensor control fluorescent peak. In our dose–response experiment (Fig. 7(b)), 0.1 µmol·L⁻¹ of 3O-C12 was sufficient to induce the living biosensor to produce visible fluorescence in the tube after 8 h growth. When quantified with a flow cytometer, the dynamic range of the sensor was observed to be in the range of 0.01–1 µmol·L⁻¹ 3O-C12.

Although synthetic biology-based biosensors can be versatile in the detection of chemical inputs, many are currently confined to a laboratory setting due to regulations that restrict the use and transport of GMOs. It is worth noting that a paper-based synthetic biosensor was recently developed by Pardee et al. [35]. This paper-based biosensor was developed by freeze-drying cell-free synthetic gene networks on paper to create materials that had the fundamental transcription and translation properties of a cell yet remained abiotic. The paper-based approach offers faster detection than cell-based living biosensors and eliminates the use of living bacterial cells for detection, which makes it suitable for the safe deployment of engineered gene circuits beyond the lab. However, the paper-based biosensor requires different sample processing and a different set of specialized materials. Thus, the system we describe here allows the complementary use of living biosensors, which can potentially be integrated in parallel with cell-free detection methods in the future. We envision our system being especially beneficial in areas that lack significant medical infrastructure or access to healthcare.

The 3O-C12 biosensor is able to detect 3O-C12 concentrations ranging from 0.1 to 10 µmol·L⁻¹. It is well documented [36–39] that a P. aeruginosa infection in the respiratory tracts can cause tissue inflammation. Although the specific method of action of chemical 3O-C12 remains unknown, it is believed to occur through interactions with toll-like receptors on the cell’s surface [40], which produce an inflammatory response. This inflammation is triggered by a 3O-C12 concentration in the range from 10 to 100 µmol·L⁻¹. The range of concentrations found in infected lungs lies outside of the dynamic response range of our 3O-C12 biosensor. However, a straightforward series of dilutions (e.g., 1/2, 1/4, …) with media—as is done for other quantification assays, such as the Bradford Assay [41]—can be performed before analysis. Diluting the sputum collected in media will also allow the cells to divide faster with a more suitable environment for growth.

The bioassay device can be fabricated anywhere a 3D printer is available. The 3D printer that we used for this research is sold for less than 2000 USD, and many new 3D printers can be purchased for prices ranging from 200 to 500 USD. Our bioassay device can also be fabricated quickly, as needed, with devices ready after less than 2 h (Fig. 8(a)). The material cost to manufacture one of the 3D-printed devices described here is approximately 0.50 USD, using 8 g of ABS. In comparison with the biosensor housing and cover, the biosensor itself contributes only a small portion of the overall cost of the device. The printer can be used at any location, making the deployment of this device flexible. Biosensors can also be readily grown, freeze-dried, and shipped worldwide. With a small quantity of freeze-dried biosensor loaded into the devices (Fig. 8(b)), they can be sealed and be ready for use. The devices can then be directly transported to the point-of-care for use (Fig. 8(c)). In the case of an accident, in which a device comes into contact with the environment, it is important to quantify how well the device isolates the biosensor from the environment. An experiment using nutrient-rich media (Fig. 8(d)) was used to simulate a worst-case scenario. When shaken at 140 r·min⁻¹ in 100 mL of media (Fig. 8(e)), the bacterial well that had been introduced into the media was able to propagate cells immediately. When the entire intact device was submerged in media, the sealed device was able to isolate the biosensor from the environment for a period of 14 h before the cells escaped and showed visible turbidity in 100 mL of media.

In the field, a sputum sample (Fig. 9(a)) can be collected from the patient and diluted into a sample tube with cell media. To ensure that no bacteria outcompete the kanamycin-tolerant living biosensor in the sample, a kanamycin antibiotic could be added, making the bacterial samples from the patient’s sputum unable to survive. The biosensor cover would then be pressed firmly and actuated, forcing the freeze-dried bacterial pellet to enter the sputum sample. Next, the tube could be incubated and analyzed in the field (Fig. 9(b)). The biosensor can simultaneously quantify specific levels of 3O-C12 exposure when maintained in exponential state or can switch into an on/off behavior by simply adding the biosensor to the sample (Fig. 9(c)). A 9 h incubation allows the biosensor to turn the liquid sample visibly red, thus serving as a two-state system that is useful to visually determine in the field whether a patient has a P. aeruginosa infection.

The device was designed to be fabricated and used by personnel or volunteers with minimal training. After attachment to a culture tube, the device and culture tube can be gripped by the hand, and the thumb can be used in a lateral pinch grip to activate the sensor. The average lateral pinch strength of an adult using the dominant hand ranges from approximately 5.0–7.5 kg of force [42], which is sufficient for the use of this device. Due to the design of the stress-focusing cutout presented here, the lateral pinch force generated by an adult will allow the device to reliably function in the field.

The experimental results presented here indicate that the 3O-C12 biosensor is able to sense inducer concentrations when P. aeruginosa biofilms are present, as in the case of recurring infections. The living biosensor, while useful for the purpose of detection, has the potential to face fewer regulatory hurdles due to its enclosure within the bioassay device, which permits its contained deployment outside the lab. The 3D-printed device turns the GMO into a component of the overall device and ensures that the biosensor is isolated from the environment. Thus, 3D printing this device enables our living biosensor, with its versatility to process environmental input, to be field-deployed as a novel, cost-effective diagnostic technology.