When plants light up: fungal bioluminescence pathway–based sentinel systems for autonomous virus diagnostics

Abstract

Sensor-based systems for detecting plant pathogens may provide field-ready alternatives to conventional, destructive, laboratory-based methods. This highlight examines a recent study that engineered a sensor system for detecting potyviruses via a bioluminescent color shift in Nicotiana benthamiana. This system incorporates a viral protease–responsive fusion between a fluorescent protein and the luciferase Luz, which does not require an exogenous substrate and served as the energy donor for bioluminescence resonance energy transfer (BRET). The authors combined this sensor with components of the fungal bioluminescence pathway. Together, these components produced a yellow signal in healthy plants; upon infection and cleavage of the fusion protein, this signal shifted to green, a change that is detectable with standard consumer cameras. Moreover, when sentinel N. benthamiana plants were intercropped with tomato plants and inoculated with tobacco etch virus (TEV), this system detected infection days before symptoms became visible in the tomato plants and distinguished TEV from other tested viruses. This work introduces a field-deployable diagnostic platform for viral diseases and expands the biosensing repertoire of the fungal bioluminescence pathway in precision agriculture.

Diseases caused by plant pathogens result in major yield losses worldwide every year and early detection of these infections remains a formidable challenge in plant health management. Traditional detection methods such as polymerase chain reaction (PCR) and enzyme-linked immunosorbent assays (ELISA) are accurate but require sample collection, laboratory infrastructure, and trained personnel, making them unsuitable for real-time, in-field monitoring [1]. Passive spectral techniques such as infrared thermography offer non-invasive alternatives, but rely on indirect physiological signatures that often appear only after symptom development [2]. These techniques, along with simple visual inspection, also do not diagnose infection during the presymptomatic incubation period of many diseases, when plants are infectious but show no visible symptoms, allowing pathogens to spread undetected.

Plant synthetic biology offers a fundamentally different route: embedding detection and reporting functions directly into the plant genome, transforming the crop itself into a phytosensor. For example, some approaches employ pathogen-inducible native promoters to drive the expression of reporter genes (e.g., the red fluorescent protein gene ppoRFP from the coral species Porites porites, β-glucuronidase, or the betalain biosynthesis pathway gene RUBY) [[3], [4], [5]]. Activation of the reporter gene upon infection produces a visible output. However, many native defense-related promoters have weak activity and are insufficient to drive a robust, remotely detectable signal and require amplification circuits [6]. Moreover, to date, most phytosensors use fluorescence as their output. Detecting a fluorescent signal requires excitation of the fluorescent protein with light, which introduces background autofluorescence, can lead to photobleaching, and complicates long-term, continuous monitoring under field conditions. Bioluminescence, light produced by an enzyme-catalyzed reaction without external illumination, offers an alternative. However, conventional luciferase systems (e.g., from firefly [Photinus pyralis] or the soft coral Renilla reniformis) require continuous external supply of their substrates, which are costly, impractical at scale, and often phytotoxic.

In this context, the fungal bioluminescence pathway (FBP) has emerged as a transformative enabling technology. The FBP is a self-sustaining metabolic cycle that converts the endogenous plant phenylpropanoid intermediate caffeic acid into green light without any added substrate [7,8]. In a recent article published in Nature Communications, Calvache and colleagues reported an elegant strategy for creating bioluminescent sentinel plants that enable specific, autonomous, and low-cost diagnosis of potyvirus infection [9]. The authors engineered N. benthamiana plants with a bioluminescence-based switch that shifts in color from yellow to green upon infection with potyviruses, using a protease-responsive bioluminescence resonance energy transfer (BRET) module as part of the FBP. This work represents a significant step toward deployable, reagent-free phytosensors for precision agriculture.

The team first established a versatile approach to visualize viral spread using autoluminescence. They generated transgenic N. benthamiana lines constitutively expressing three of the four FBP components (hispidin-3-hydroxylase [H3H], luciferase [Luz], and caffeylpyruvate hydrolase [CPH]), leaving out hispidin synthase (HispS), which encodes the rate-limiting enzyme of this pathway. Instead of HispS, the authors delivered compact, plant-derived substitutes (hydrangea macrophylla synthase [HmS], polyketide synthase 2 [PKS2], anther-specific chalcone synthase-like [ASCL]) via recombinant viral vectors, each reconstituting the complete bioluminescence cycle specifically in infected tissues of N. benthamiana plants expressing H3H, Luz, and CPH. This strategy allowed the authors to spatially and temporally track viral proliferation and systemic movement in whole plants in real time, using consumer-grade cameras. This first system also revealed an uncoupling between local infection and systemic spread of different viruses, laying a solid methodological foundation for subsequent sensor development. Beyond serving as a proof-of-concept for FBP-based tracing, this part of the study offered a flexible toolkit to investigate virus movement at the whole-plant scale without the need for exogenous substrates.

The authors then engineered the FBP luciferase Luz into a protease-responsive component of a BRET pair by placing a flexible linker between Luz and the red fluorescent protein (DsRed). BRET occurs by proximity-dependent energy transfer between a bioluminescent donor (Luz) and a compatible fluorescent acceptor (DsRed). Luz oxidizes its substrate via an enzymatic reaction, emitting light. A portion of the energy released by the reaction is non-radiatively transferred to the acceptor, which then emits light at a longer wavelength. In the intact fusion, energy transfer from Luz to DsRed produces a yellow glow due to the combined green luminescence from Luz and red fluorescence from DsRed; this is the “Yellow-ON” state of the sensor. When the linker contains a cleavage site for the potyviral protease Nla, viral infection triggers site-specific proteolysis, releasing DsRed from Luz, disrupting BRET, and shifting the emission to green and the “Green-ON” state of the sensor (Fig. 1A). Extensive experiments were performed to optimize this genetic circuit, including testing DsRed and another acceptor with compatible excitation in the blue range of the spectrum, testing N-terminal and C-terminal fusions, and screening multiple linker sequences. The optimized Luz-DsRed fusion protein was further modified to include a cleavage site for a viral protease. The resulting sensors responded specifically to their cognate protease: a tobacco etch virus (TEV)-responsive switch was not activated by the related tobacco vein mottling virus (TVMV) protease, and vice versa. This ‘yellow-ON/green-ON’ logic mimics an electronic status indicator, wherein a steady yellow luminescence confirms that the detection system is intact, but a shift to green luminescence signals that the plant is infected. Such dual-output design will likely substantially decrease false negatives, a critical requirement for field-deployable sensors.

Fig. 1.

Schematic overview of FBP-derived BRET genetic circuit for autonomous potyvirus surveillance in sentinel plants. A Core molecular BRET switching mechanism. Top panel: Intact Luz-Linker-CS^TEV-DsRed fusion protein mediates BRET and weak yellow luminescence in healthy cells; Bottom panel: TEV Nla protease cleavage disassociates the donor–acceptor pair, eliminates BRET, and generates strong green Luz-only bioluminescence. B Schematic representation of the genetic modules used to generate the sentinel plants. Two independent modules were integrated into N. benthamiana via Agrobacterium-mediated co-transformation. Module I is the eFBPΔLuz module that constitutively expresses H3H, HispS, CPH, and NpgA, providing the enzymatic activities required for caffeic acid recycling and substrate biosynthesis. Module II represents the BRET switch module in which Luz is fused to DsRed through a protease-sensitive linker containing the TEV Nla protease cleavage site. All expression cassettes are driven by the CaMV 35S constitutive promoters. C Dark-field camera imaging of plant-level luminescence phenotypes. Virus-free sentinel tobacco displays faint yellow BRET-derived glow, whereas TEV-infected plants show vivid green bioluminescence after protease-mediated BRET disruption. D Prospective field application of the sentinel plants for precision agriculture. Sparse sentinel tobacco rows are intercropped within mainstream tobacco fields. Distinct yellow/green luminescence signals from healthy/infected sentinel plants are remotely surveyed by imaging drones to achieve autonomous pre-symptomatic viral disease surveillance across large farmland areas. The image was generated using the GEMINI 2.5 PRO MODEL (https://gemini.google.com/). The figure was created in BioRender.

Finally, the authors generated stable transgenic N. benthamiana lines co-expressing an FBP system lacking only Luz (eFBPΔLuz) and the BRET switch driven by the constitutive cauliflower mosaic virus (CaMV) 35S promoter (Fig. 1B). After selecting lines with bright, homogeneous yellow background luminescence, they challenged plants with TEV, TVMV, potato virus X (PVX), or tobacco mosaic virus (TMV). Only TEV triggered a reproducible green shift (Fig. 1C), detectable by both a charge-coupled device camera and a low-cost digital single-lens reflex camera equipped with a purple filter. Confocal microscopy and immunoblotting confirmed that TEV infection released DsRed from the endomembrane compartments into the nucleocytoplasm, whereas the fusion remained intact in plants infected with other viruses. In an intercropping experiment with tomato (Solanum lycopersicum) plants, TEV-inoculated sentinel plants developed green luminescence days before visible symptoms appeared on the TEV-inoculated tomato plants. This result is particularly encouraging for agricultural applications, where early warning is key to containment.

Despite this remarkable progress, there is room for improvement of this bioluminescent sentinel system for large-scale agricultural applications. First, signal robustness in complex field environments needs to be enhanced. Although the current system performs well in greenhouses and low-light rural areas, strong ambient light and light pollution in farmlands near cities might interfere with the detection of luminescence signals. Recent advances in FBP engineering provide a clear roadmap for improvement. Metabolic engineering strategies, such as boosting precursor supply and redirecting metabolic flux away from competing pathways, have substantially increased plant luminescence derived from FBP [10,11]. Directed evolution of Luz and H3H has yielded brighter and more thermostable variants [12]. Combining these optimizations could boost the luminescence emitted by sentinel plants to levels suitable for outdoor use. Second, diversification of the sensing host is required. The current chassis is N. benthamiana, a laboratory-friendly host with broad susceptibility. Transferring the sensor circuit into major food and cash crops such as potato (Solanum tuberosum), tomato, and tobacco (Nicotiana tabacum) would allow the crops themselves to serve as sentinels, eliminating the need for intercropping. Third, strategies for detecting multiple pathogens with one or more sensors should be explored. The protease-responsive design described in the highlighted study is inherently limited to pathogens that encode sequence-specific proteases, primarily potyviruses and certain bacterial effectors. For other pathogens, alternative sensing mechanisms may be needed. In addition, long-term field evaluation of transgenic sentinel plants should be carried out to investigate the risks of transgene flow and the effect on plant growth and development, to promote its safe application.

The FBP has rapidly moved from a subject for basic research into a versatile biotechnological tool. For example, Chiu and colleagues engineered tobacco plants in which Luz expression is driven by the promoter from Tomato phosphate starvation–induced gene 1 (TPSI1), creating ‘BioLumiPi’ plants that autonomously report phosphorus deficiency [13]. Balakireva and colleagues generated autoluminescent Arabidopsis thaliana and N. benthamiana lines using promoters responsive to the phytohormones jasmonic acid or salicylic acid, enabling non-invasive imaging of defense hormone activity over weeks [14]. These examples highlight a modular design principle: the luminescence emitted by the FBP can be coupled to virtually any inducible promoter or posttranslational switch, turning plants into living sensors for a wide range of environmental and biological stimuli.

Looking further ahead, artificial intelligence–driven protein design promises to overcome current brightness and spectral limitations. Deep learning has already produced de novo luciferases (e.g., LuxSit-i) with high thermal stability and substrate specificity [15]. Applying such approaches to FBP-related enzymes could yield brighter, more stable variants, improving tissue penetration and compatibility with multiplex detection. In addition, multicolor FBP variants were recently reported, achieved by inserting fluorescent proteins into Luz loops guided by AlphaFold predictions [16]. These variants open the door to encoding recognition of multiple pathogens in different colors within the same plant. Combined with intelligent agricultural equipment such as drones and automated monitoring robots (Fig. 1D), luminescent sentinel plants can form an early warning network for diseases in the field, enabling automatic sampling, real-time alerts, and data-driven disease prevention and control.

In conclusion, this study by Calvache and colleagues [9] delivered a compelling proof-of-concept for autonomous, protease-activated bioluminescent sentinel plants that report viral infections through a visible color change. Their work bridges fundamental plant virology, synthetic biology, and practical diagnostics, illustrating how FBP-based circuits can be engineered to process molecular information and produce a clear, remotely detectable visual signal. With further technical optimization and commercial development, FBP-based bioluminescent sensing systems could become an important part of modern precision agriculture, helping transform agriculture from reactive disease management to proactive early detection, and supporting sustainable global crop production.

CRediT authorship contribution statement

Peng Zheng: Writing – original draft, Formal analysis, Data curation, Conceptualization. Yan Liu: Writing – original draft. Enhui Shen: Writing – original draft. Hao Du: Writing – review & editing, Formal analysis, Data curation. Shengchun Xu: Writing – review & editing, Conceptualization.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Data availability

No data was used for the research described in the article.