Overview

Tomato is one of the most important crops worldwide that can be cultivated in a broad range of climates. In 2023, tomato continued to confirm its powerful agricultural and economic dominance with a production of about 192 million tons grown on 5.41 million ha, ranking in second place among the most cultivated vegetables, right after potato. Considering the total global tomato production from 2000 to 2023, global tomato production was about 3.75 billion tons, while China alone contributed about 30.2％ of the global production, providing the world's top output. (Food and Agriculture Organization of the United Nations [FAO], 2025)

However, tomatoes together with many other economically important vegetable crops, are highly vulnerable to bacterial infections, and it is estimated that yield losses caused by bacterial infections range from 5％ to 10% (Holtappels, Fortuna, Lavigne, & Wagemans, 2021). In a poll published on Molecular Plant Pathology, Gram-negative bacteria of the Pseudomonas were considered the most destructive among many pathogenic bacteria (Mansfield et al., 2012). Among these, Pseudomonas syringae pv. tomato (Pst) is pathogenic to both tomatoes and cruciferous plants. Infection in the early growing season leads to bacterial speck disease, which severely impairs the photosynthesis of leaves, thereby greatly reducing the yield and rendering them unsuitable for the fresh market (Louws et al., 2001; Wilson, Campbell, Ji, Jones, & Cuppels, 2002). Depending on weather conditions (temperature and relative humidity), the actual yield losses vary from 75% of plants infected in the early growth stage to 5% of those infected in the later stage of the season (Yunis, Bashan, Okon, & Henis, 1980), which cannot be ignored in the context of large-scale tomato cultivation worldwide.

Table 1. Top 10 bacterial pathogens voted for by plant bacteriologists associated with the journal Molecular Plant Pathology

Once a tomato plant is infected with Pst, it is difficult to be cured. Common treatment methods include deep burial and burning of infected plants, which could result in massive economic losses if the disease has spread on a large scale. Another treatment method is applying copper-based fungicides to control the progression of the disease; however, for maximum effectiveness, this chemical spraying must be applied before symptoms develop (University of Wisconsin-Madison, 2017). Thus, due to the uncontrollability of disease transmission after infection and the importance of treatment in early stage, it is precisely that a set of methods for early detection of bacterial speck disease or Pst are in urgent need. Such methods would enable farmers to take those preventive measures in advance and avoid significant economic losses.

To address this need, our team has developed Proteolysis-based Test for Pseudomonas syringae pv. Tomato Detection (Protato). This system detects fluorescence or glucose signal produced through the binding of split protein fragments, which is jointly activated by AvrRpt2 and COR—effectors secreted into tomato plants by Pst. By packing all the protein fragments needed into our designed hardware, we successfully built a cell-free detection biosensor that could be triggered by an AND-Gate signal pathway, which theoretically allowed for early detection of infection approximately 10 days before the appearance of visible leaf speck symptoms. (Shakeel Muzaffar, 2024)

Why cell-free protein biosensor

Table 2. Current Limitations of Pst Detection Methods

As shown in table 2, current detection methods are either too slow, too costly, or dependent on laboratory infrastructure and trained personnel, and thus none are suitable for rapid, on-site, early detection of Pst. This highlights the urgent need for a tool that is fast, safe, inexpensive, stable, and farmer-accessible — namely, our all-protein cell-free biosensor.

Cell-free protein biosensors are composed entirely of purified or recombinant proteins, without the need for living cells. In our field-deployable tomato detection system, this design minimizes biosafety concerns during on-site use — preventing the release of genetically modified materials and ensuring safe operation even in open farm environments.

Because all sensing and reporting functions are carried out by proteins in vitro, cell-free protein biosensors exhibit rapid response kinetics, as they are not constrained by cell growth or metabolism. They can also be precisely tuned and standardized for consistent performance.

Furthermore, protein-based biosensors are highly stable and can be lyophilized or stored at ambient temperatures, making them ideal for field deployment and long-term storage. Their modularity allows easy customization to detect a wide range of targets while maintaining a safe, compact, and user-friendly diagnostic format.

AND-Gated Biosensor for Highly Specific Pathogen Detection

This detection system is based on a dual-input activated logic-gating mechanism, enabling highly specific recognition of target pathogens. Two key input signals are defined:

• a pathogen-specific protease: AvrRpt2
• a characteristic toxin molecule secreted by the pathogen: phytotoxin coronatine (COR)

Only when both signals are present can they jointly activate the downstream AND logic gate to trigger the output signal.

The system incorporates a positive-feedback design to enhance signal intensity and detection sensitivity. The final output is fluorescent signal or glucose concentration, facilitating qualitative readouts. Entirely implemented in a cell-free setup, the system offers rapid response, ease of operation, and avoidance of biosafety risks, making it well suited for on-site rapid detection scenarios.

Drylab

• Our model series, POLARTO (Practical Optimization with Linked Affinity, Reaction and Transmission Oversight), is designed to rigorously validate the practical feasibility of our experimental pathway and hardware design, and to provide actionable strategies for its optimization.

  To achieve this, we developed an integrated computational suite that bridges scales: we used GROMACS for molecular dynamics to verify the thermodynamic and kinetic viability of each reaction step; built Seq2Affinity, a prediction and design model, to predict the binding strength of peptides especially for coiled-coil(CC) domains and attempt to optimize the binding affinity of CC domains in our system; employed COMSOL Multiphysics to simulate and validate the signal detection capability of our hardware. Furthermore, the Cellular automaton & SEIR model (CS model) combined with Q-Learning was developed to optimize economic utility by finding the best detection strategy in the field. We also developed a system of ordinary differential equations (ODEs) to simulate the dynamic changes of molecular species throughout the reaction pathway. The ODEs were used to estimate the time required for the system to reach equilibrium, providing references for both wetlab and hardware.

  Collectively, POLARTO integrates multi-scale computational approaches—from molecular interactions to system kinetics and field-level detection strategies, forming a closed-loop framework that not only validates the core functionality of our biosensing system but also drives continuous refinement across molecular, hardware, and strategic dimensions. See model for more details.



Fig.3 Our Model Series POLARTO: Integrating Models from Wetlab and Hardware to Agricultural Applications and the iGEM Community

• For project construction, we developed the Protato Kit biological detection kit (See Hardware for more details), aimed at enabling Protato to more easily transition from the biology lab to broader usage environments, providing low-cost reliable experimental testing environments for users, experimenters and developers.

  

• To complement our Protato hardware system, we developed a comprehensive mobile application designed to serve as an all-in-one monitoring and analysis tool. The application connects seamlessly to the hardware via a local Wi-Fi network, allowing it to fetch and visualize real-time plant health data and display a live video feed. Its core functionality is a powerful, on-device artificial intelligence model that automatically analyzes images of tomato leaves to diagnose nine different types of diseases, providing users with immediate and actionable insights into their crop's condition(See Software for more details).

  

References

1. Food and Agriculture Organization of the United Nations (FAO). (n.d.). FAOSTAT: Production quantities (QCL). https://www.fao.org/faostat/en/#data/QCL (Accessed [October 1, 2025])
2. Holtappels, D., Fortuna, K., Lavigne, R., & Wagemans, J. (2021). The future of phage biocontrol in integrated plant protection for sustainable crop production. Curr Opin Biotechnol, 68, 60-71. (Accessed [October 1, 2025])
3. Louws, F. J., Wilson, M., Campbell, H. L., Cuppels, D. A., Jones, J. B., Shoemaker, P. B., . . . Miller, S. A. (2001). Field Control of Bacterial Spot and Bacterial Speck of Tomato Using a Plant Activator. Plant Dis, 85(5), 481-488. (Accessed [October 1, 2025])
4. Mansfield, J., Genin, S., Magori, S., Citovsky, V., Sriariyanum, M., Ronald, P., . . . Foster, G. D. (2012). Top 10 plant pathogenic bacteria in molecular plant pathology. Mol Plant Pathol, 13(6), 614-629. (Accessed [October 1, 2025])
5. Wilson, M., Campbell, H. L., Ji, P., Jones, J. B., & Cuppels, D. A. (2002). Biological control of bacterial speck of tomato under field conditions at several locations in north america. Phytopathology, 92(12), 1284-1292. (Accessed [October 1, 2025])
6. Yunis, H., Bashan, Y., Okon, Y., & Henis, Y. (1980). Weather dependence, yield losses, and control of bacterial speck of tomato caused by Pseudomonas tomato. Plant disease, 64(10), 937-939. (Accessed [October 1, 2025])
7. University of Wisconsin-Madison Extension. (2017, February 17). Bacterial speck of tomato [UW Plant Disease Facts]. https://pddc.wisc.edu/2017/02/17/bacterial-speck-tomato/ (Accessed [October 1, 2025])
8. Tomato Answers. (2024, July 11). Bacterial speck: Prevention and treatment tips. https://tomatoanswers.com/bacterial-speck-on-tomatoes/ (Accessed [October 1, 2025])
9. Peňázová, E., Dvořák, M., Ragasová, L., Kiss, T., Pečenka, J., Čechová, J., & Eichmeier, A. (2020). Multiplex real-time PCR for the detection of Clavibacter michiganensis subsp. michiganensis, Pseudomonas syringae pv. Tomato and pathogenic Xanthomonas species on tomato plants. PloS one, 15(1), e0227559. (Accessed [October 1, 2025])
10. Furci, L., Pascual-Pardo, D., & Ton, J. (2021). A rapid and non-destructive method for spatial-temporal quantification of colonization by Pseudomonas syringae pv. Tomato DC3000 in Arabidopsis and tomato. Plant methods, 17(1), 126. (Accessed [October 1, 2025])
11. Reis Pereira, M., Dos Santos, F. N., Tavares, F., & Cunha, M. (2023). Enhancing host-pathogen phenotyping dynamics: early detection of tomato bacterial diseases using hyperspectral point measurement and predictive modeling. Frontiers in plant science, 14, 1242201. (Accessed [October 1, 2025])