Continual improvement of CRISPR-induced multiplex mutagenesis in Arabidopsis.

CRISPR/Cas9 is currently the most powerful tool to generate mutations in plant genomes and more efficient tools are needed as the scale of experiments increases. In the model plant Arabidopsis, the choice of the promoter driving Cas9 expression is critical to generate germline mutations. Several optimal promoters have been reported. However, it is unclear which promoter is ideal as they have not been thoroughly tested side by side. Furthermore, most plant vectors still use one of the two Cas9 nuclear localization sequence (NLS) configurations initially reported. We genotyped more than 6000 Arabidopsis T2 plants to test seven promoters and six types of NLSs across 14 targets to systematically improve the generation of single and multiplex inheritable mutations. We found that the RPS5A promoter and bipartite NLS were individually the most efficient components. When combined, 99% of T2 plants contained at least one knockout (KO) mutation and 84% contained 4- to 7-plex KOs, the highest multiplexing KO rate in Arabidopsis to date. These optimizations will be useful to generate higher-order KOs in the germline of Arabidopsis and will likely be applicable to other CRISPR systems as well.

A robust high-throughput functional screening assay for plant pathogen effectors using the TMV-GFP vector.

Uncovering the function of phytopathogen effectors is crucial for understanding mechanisms of pathogen pathogenicity and for improving our ability to protect plants from diseases. An increasing number of effectors have been predicted in various plant pathogens. Functional characterization of these effectors has become a major focus in the study of plant-pathogen interactions. In this study, we designed a novel screening system that combines the TMV (tobacco mosaic virus)-GFP vector and Agrobacterium-mediated transient expression in the model plant Nicotiana benthamiana. This system enables the rapid identification of effectors that interfere with plant immunity. The biological function of these effectors can be easily evaluated by observing the GFP fluorescence signal using a UV lamp within just a few days. To evaluate the TMV-GFP system, we initially tested it with well-described virulence and avirulence type III effectors from the bacterial pathogen Ralstonia solanacearum. After proving the accuracy and efficiency of the TMV-GFP system, we successfully screened a novel virulence effector, RipS1, using this approach. Furthermore, using the TMV-GFP system, we reproduced consistent results with previously known cytoplasmic effectors from a diverse array of pathogens. Additionally, we demonstrated the effectiveness of the TMV-GFP system in identifying apoplastic effectors. The easy operation, time-saving nature, broad effectiveness, and low technical requirements of the TMV-GFP system make it a promising approach for high-throughput screening of effectors with immune interference activity from various pathogens.

A global LC-MS2 -based methodology to identify and quantify anionic phospholipids in plant samples.

Anionic phospholipids (PS, PA, PI, PIPs) are low-abundant phospholipids with impactful functions in cell signaling, membrane trafficking and cell differentiation processes. They can be quickly metabolized and can transiently accumulate at defined spots within the cell or an organ to respond to physiological or environmental stimuli. As even a small change in their composition profile will produce a significant effect on biological processes, it is crucial to develop a sensitive and optimized analytical method to accurately detect and quantify them. While thin-layer chromatography (TLC) separation coupled with gas chromatography (GC) detection methods already exist, they do not allow for precise, sensitive, and accurate quantification of all anionic phospholipid species. Here we developed a method based on high-performance liquid chromatography (HPLC) combined with two-dimensional mass spectrometry (MS2 ) by MRM mode to detect and quantify all molecular species and classes of anionic phospholipids in one shot. This method is based on a derivatization step by methylation that greatly enhances the ionization, the separation of each peak, the peak resolution as well as the limit of detection and quantification for each individual molecular species, and more particularly for PA and PS. Our method universally works in various plant samples. Remarkably, we identified that PS is enriched with very long chain fatty acids in the roots but not in aerial organs of Arabidopsis thaliana. Our work thus paves the way for new studies on how the composition of anionic lipids is finely tuned during plant development and environmental responses.

A rapid and highly efficient sorghum transformation strategy using GRF4-GIF1/ternary vector system.

Sorghum is an important crop for food, forage, wine and biofuel production. To enhance its transformation efficiency without negative developmental by-effects, we investigated the impact of GRF4-GIF1 chimaera and GRF5 on sorghum transformation. Both GRF4-GIF1 and GRF5 effectively improved the transformation efficiency of sorghum and accelerated the transformation process of sorghum to less than 2 months which was not observed when using BBM-WUS. As agrobacterium  effectors increase the ability of T-DNA transfer into plant cells, we checked whether ternary vector system can additively enhance sorghum transformation. The combination of GRF4-GIF1 with helper plasmid pVS1-VIR2 achieved the highest transformation efficiency, reaching 38.28%, which is 7.71-fold of the original method. Compared with BBM-WUS, overexpressing GRF4-GIF1 caused no noticeable growth defects in sorghum. We further developed a sorghum CRISPR/Cas9 gene-editing tool based on this GRF4-GIF1/ternary vector system, which achieved an average gene mutation efficiency of 41.36%, and null mutants were created in the T0 generation.

In planta imaging of pyridine nucleotides using second-generation fluorescent protein biosensors.

Redox changes of pyridine nucleotides in cellular compartments are highly dynamic and their equilibria are under the influence of various reducing and oxidizing reactions. To obtain spatiotemporal data on pyridine nucleotides in living plant cells, typical biochemical approaches require cell destruction. To date, genetically encoded fluorescent biosensors are considered to be the best option to bridge the existing technology gap, as they provide a fast, accurate, and real-time readout. However, the existing pyridine nucleotides genetically encoded fluorescent biosensors are either sensitive to pH change or slow in dissociation rate. Herein, we employed the biosensors which generate readouts that are pH stable for in planta measurement of NADH/NAD+ ratio and NADPH level. We generated transgenic Arabidopsis lines that express these biosensors in plastid stroma and cytosol of whole plants and pollen tubes under the control of CaMV 35S and LAT52 promoters, respectively. These transgenic biosensor lines allow us to monitor real-time dynamic changes in NADH/NAD+ ratio and NADPH level in the plastids and cytosol of various plant tissues, including pollen tubes, root hairs, and mesophyll cells, using a variety of fluorescent instruments. We anticipate that these valuable transgenic lines may allow improvements in plant redox biology studies.

Dot Scanner: open-source software for quantitative live-cell imaging in planta.

Confocal microscopy has greatly aided our understanding of the major cellular processes and trafficking pathways responsible for plant growth and development. However, a drawback of these studies is that they often rely on the manual analysis of a vast number of images, which is time-consuming, error-prone, and subject to bias. To overcome these limitations, we developed Dot Scanner, a Python program for analyzing the densities, lifetimes, and displacements of fluorescently tagged particles in an unbiased, automated, and efficient manner. Dot Scanner was validated by performing side-by-side analysis in Fiji-ImageJ of particles involved in cellulose biosynthesis. We found that the particle densities and lifetimes were comparable in both Dot Scanner and Fiji-ImageJ, verifying the accuracy of Dot Scanner. Dot Scanner largely outperforms Fiji-ImageJ, since it suffers far less selection bias when calculating particle lifetimes and is much more efficient at distinguishing between weak signals and background signal caused by bleaching. Not only does Dot Scanner obtain much more robust results, but it is a highly efficient program, since it automates much of the analyses, shortening workflow durations from weeks to minutes. This free and accessible program will be a highly advantageous tool for analyzing live-cell imaging in plants.

Ratiometric gibberellin biosensors for the analysis of signaling dynamics and metabolism in plant protoplasts.

Gibberellins (GAs) are major regulators of developmental and growth processes in plants. Using the degradation-based signaling mechanism of GAs, we have built transcriptional regulator (DELLA)-based, genetically encoded ratiometric biosensors as proxies for hormone quantification at high temporal resolution and sensitivity that allow dynamic, rapid and simple analysis in a plant cell system, i.e. Arabidopsis protoplasts. These ratiometric biosensors incorporate a DELLA protein as a degradation target fused to a firefly luciferase connected via a 2A peptide to a renilla luciferase as a co-expressed normalization element. We have implemented these biosensors for all five Arabidopsis DELLA proteins, GA-INSENSITIVE, GAI; REPRESSOR-of-ga1-3, RGA; RGA-like1, RGL1; RGL2 and RGL3, by applying a modular design. The sensors are highly sensitive (in the low pm range), specific and dynamic. As a proof of concept, we have tested the applicability in three domains: the study of substrate specificity and activity of putative GA-oxidases, the characterization of GA transporters, and the use as a discrimination platform coupled to a GA agonists' chemical screening. This work demonstrates the development of a genetically encoded quantitative biosensor complementary to existing tools that allow the visualization of GA in planta.

Degenerate oligonucleotide primer MIG-seq: an effective PCR-based method for high-throughput genotyping.

Next-generation sequencing (NGS) library construction often involves using restriction enzymes to decrease genome complexity, enabling versatile polymorphism detection in plants. However, plant leaves frequently contain impurities, such as polyphenols, necessitating DNA purification before enzymatic reactions. To overcome this problem, we developed a PCR-based method for expeditious NGS library preparation, offering flexibility in number of detected polymorphisms. By substituting a segment of the simple sequence repeat sequence in the MIG-seq primer set (MIG-seq being a PCR method enabling library construction with low-quality DNA) with degenerate oligonucleotides, we introduced variability in detectable polymorphisms across various crops. This innovation, named degenerate oligonucleotide primer MIG-seq (dpMIG-seq), enabled a streamlined protocol for constructing dpMIG-seq libraries from unpurified DNA, which was implemented stably in several crop species, including fruit trees. Furthermore, dpMIG-seq facilitated efficient lineage selection in wheat and enabled linkage map construction and quantitative trait loci analysis in tomato, rice, and soybean without necessitating DNA concentration adjustments. These findings underscore the potential of the dpMIG-seq protocol for advancing genetic analyses across diverse plant species.

Trait association and prediction through integrative k-mer analysis.

Genome-wide association study (GWAS) with single nucleotide polymorphisms (SNPs) has been widely used to explore genetic controls of phenotypic traits. Alternatively, GWAS can use counts of substrings of length k from longer sequencing reads, k-mers, as genotyping data. Using maize cob and kernel color traits, we demonstrated that k-mer GWAS can effectively identify associated k-mers. Co-expression analysis of kernel color k-mers and genes directly found k-mers from known causal genes. Analyzing complex traits of kernel oil and leaf angle resulted in k-mers from both known and candidate genes. A gene encoding a MADS transcription factor was functionally validated by showing that ectopic expression of the gene led to less upright leaves. Evolution analysis revealed most k-mers positively correlated with kernel oil were strongly selected against in maize populations, while most k-mers for upright leaf angle were positively selected. In addition, genomic prediction of kernel oil, leaf angle, and flowering time using k-mer data resulted in a similarly high prediction accuracy to the standard SNP-based method. Collectively, we showed k-mer GWAS is a powerful approach for identifying trait-associated genetic elements. Further, our results demonstrated the bridging role of k-mers for data integration and functional gene discovery.

Mapping of specialized metabolite terms onto a plant phylogeny using text mining and large language models.

Plants produce a staggering array of chemicals that are the basis for organismal function and important human nutrients and medicines. However, it is poorly defined how these compounds evolved and are distributed across the plant kingdom, hindering a systematic view and understanding of plant chemical diversity. Recent advances in plant genome/transcriptome sequencing have provided a well-defined molecular phylogeny of plants, on which the presence of diverse natural products can be mapped to systematically determine their phylogenetic distribution. Here, we built a proof-of-concept workflow where previously reported diverse tyrosine-derived plant natural products were mapped onto the plant tree of life. Plant chemical-species associations were mined from literature, filtered, evaluated through manual inspection of over 2500 scientific articles, and mapped onto the plant phylogeny. The resulting "phylochemical" map confirmed several highly lineage-specific compound class distributions, such as betalain pigments and Amaryllidaceae alkaloids. The map also highlighted several lineages enriched in dopamine-derived compounds, including the orders Caryophyllales, Liliales, and Fabales. Additionally, the application of large language models, using our manually curated data as a ground truth set, showed that post-mining processing can largely be automated with a low false-positive rate, critical for generating a reliable phylochemical map. Although a high false-negative rate remains a challenge, our study demonstrates that combining text mining with language model-based processing can generate broader phylochemical maps, which will serve as a valuable community resource to uncover key evolutionary events that underlie plant chemical diversity and enable system-level views of nature's millions of years of chemical experimentation.

Multi-class plant hormone HILIC-MS/MS analysis coupled with high-throughput phenotyping to investigate plant-environment interactions.

Plant hormones are chemical signals governing almost every aspect of a plant's life cycle and responses to environmental cues. They are enmeshed within complex signaling networks that can only be deciphered by using broad-scale analytical methods to capture information about several plant hormone classes simultaneously. Methods used for this purpose are all based on reversed-phase (RP) liquid chromatography and mass spectrometric detection. Hydrophilic interaction chromatography (HILIC) is an alternative chromatographic method that performs well in analyses of biological samples. We therefore developed and validated a HILIC method for broad-scale plant hormone analysis including a rapid sample preparation procedure; moreover, derivatization or fractionation is not required. The method enables plant hormone screening focused on polar and moderately polar analytes including cytokinins, auxins, jasmonates, abscisic acid and its metabolites, salicylates, indoleamines (melatonin), and 1-aminocyclopropane-1-carboxylic acid (ACC), for a total of 45 analytes. Importantly, the major pitfalls of ACC analysis have been addressed. Furthermore, HILIC provides orthogonal selectivity to conventional RP methods and displays greater sensitivity, resulting in lower limits of quantification. However, it is less robust, so procedures to increase its reproducibility were established. The method's potential is demonstrated in a case study by employing an approach combining hormonal analysis with phenomics to examine responses of three Arabidopsis ecotypes toward three abiotic stress treatments: salinity, low nutrient availability, and their combination. The case study showcases the value of the simultaneous determination of several plant hormone classes coupled with phenomics data when unraveling processes involving complex cross-talk under diverse plant-environment interactions.

A quantitative assay for the efficiency of RNA-guided genome editing in plants.

RNA-guided endonucleases originating from the bacterial CRISPR/Cas system are a versatile tool for targeted gene editing. To determine the functional relevance of a gene of interest, deletion of the entire open reading frame (ORF) by two independent double-strand breaks (DSBs) is particularly attractive. This strategy greatly benefits from high editing efficiency, which is strongly influenced by the Cas endonuclease version used. We developed two reporter switch-on assays, for quantitative comparison and optimization of Cas constructs. The assays are based on four components: (i) A reporter gene, the mRNA of which carries a hairpin (HP) loop targeted by (ii) the endoribonuclease Csy4. Cleavage of the mRNA at the HP loop by Csy4 abolishes the translation of the reporter. Csy4 was used as the target for full deletion. (iii) A Cas system targeting sites flanking the Csy4 ORF with a 20-bp spacer either side to preferentially detect full-deletion events. Loss of functional Csy4 would lead to reporter gene expression, allowing indirect quantification of Cas-mediated deletion events. (iv) A reference gene for normalization. We tested these assays on Nicotiana benthamiana leaves and Lotus japonicus calli induced on hypocotyl sections, using Firefly luciferase and mCitrine as reporter genes and Renilla luciferase and hygromycin phosphotransferase II as reference genes, respectively. We observed a >90% correlation between reporter expression and full Csy4 deletion events, demonstrating the validity of these assays. The principle of using the Csy4-HP module as Cas target should be applicable to other editing goals including single DSBs in all organisms.

Use of GRF-GIF chimeras and a ternary vector system to improve maize (Zea mays L.) transformation frequency.

Maize (Zea mays L.) is an important crop that has been widely studied for its agronomic and industrial applications and is one of the main classical model organisms for genetic research. Agrobacterium-mediated transformation of immature maize embryos is a commonly used method to introduce transgenes, but a low transformation frequency remains a bottleneck for many gene-editing applications. Previous approaches to enhance transformation included the improvement of tissue culture media and the use of morphogenic regulators such as BABY BOOM and WUSCHEL2. Here, we show that the frequency can be increased using a pVS1-VIR2 virulence helper plasmid to improve T-DNA delivery, and/or expressing a fusion protein between a GROWTH-REGULATING FACTOR (GRF) and GRF-INTERACTING FACTOR (GIF) protein to improve regeneration. Using hygromycin as a selection agent to avoid escapes, the transformation frequency in the maize inbred line B104 significantly improved from 2.3 to 8.1% when using the pVS1-VIR2 helper vector with no effect on event quality regarding T-DNA copy number. Combined with a novel fusion protein between ZmGRF1 and ZmGIF1, transformation frequencies further improved another 3.5- to 6.5-fold with no obvious impact on plant growth, while simultaneously allowing efficient CRISPR-/Cas9-mediated gene editing. Our results demonstrate how a GRF-GIF chimera in conjunction with a ternary vector system has the potential to further improve the efficiency of gene-editing applications and molecular biology studies in maize.

SPIRO - the automated Petri plate imaging platform designed by biologists, for biologists.

Phenotyping of model organisms grown on Petri plates is often carried out manually, despite the procedures being time-consuming and laborious. The main reason for this is the limited availability of automated phenotyping facilities, whereas constructing a custom automated solution can be a daunting task for biologists. Here, we describe SPIRO, the Smart Plate Imaging Robot, an automated platform that acquires time-lapse photographs of up to four vertically oriented Petri plates in a single experiment, corresponding to 192 seedlings for a typical root growth assay and up to 2500 seeds for a germination assay. SPIRO is catered specifically to biologists' needs, requiring no engineering or programming expertise for assembly and operation. Its small footprint is optimized for standard incubators, the inbuilt green LED enables imaging under dark conditions, and remote control provides access to the data without interfering with sample growth. SPIRO's excellent image quality is suitable for automated image processing, which we demonstrate on the example of seed germination and root growth assays. Furthermore, the robot can be easily customized for specific uses, as all information about SPIRO is released under open-source licenses. Importantly, uninterrupted imaging allows considerably more precise assessment of seed germination parameters and root growth rates compared with manual assays. Moreover, SPIRO enables previously technically challenging assays such as phenotyping in the dark. We illustrate the benefits of SPIRO in proof-of-concept experiments which yielded a novel insight on the interplay between autophagy, nitrogen sensing, and photoblastic response.

Mapping of cytosol-facing organelle outer membrane proximity proteome by proximity-dependent biotinylation in living Arabidopsis cells.

The cytosol-facing outer membrane (OM) of organelles communicates with other cellular compartments to exchange proteins, metabolites, and signaling molecules. Cellular surveillance systems also target OM-resident proteins to control organellar homeostasis and ensure cell survival under stress. However, the OM proximity proteomes have never been mapped in plant cells since using traditional approaches to discover OM proteins and identify their dynamically interacting partners remains challenging. In this study, we developed an OM proximity labeling (OMPL) system using biotin ligase-mediated proximity biotinylation to identify the proximity proteins of the OMs of mitochondria, chloroplasts, and peroxisomes in living Arabidopsis (Arabidopsis thaliana) cells. Using this approach, we mapped the OM proximity proteome of these three organelles under normal conditions and examined the effects of the ultraviolet-B (UV-B) or high light (HL) stress on the abundances of OM proximity proteins. We demonstrate the power of this system with the discovery of cytosolic factors and OM receptor candidates potentially involved in local protein translation and translocation. The candidate proteins that are involved in mitochondrion-peroxisome, mitochondrion-chloroplast, or peroxisome-chloroplast contacts, and in the organellar quality control system are also proposed based on OMPL analysis. OMPL-generated OM proximity proteomes are valuable sources of candidates for functional validation and suggest directions for further investigation of important questions in cell biology.

Characterization and development of a plastid genome base editor, ptpTALECD.

The modification of photosynthesis-related genes in plastid genomes may improve crop yields. Recently, we reported that a plastid-targeting base editor named ptpTALECD, in which a cytidine deaminase DddA functions as the catalytic domain, can homoplasmically substitute a targeted C to T in plastid genomes of Arabidopsis thaliana. However, some target Cs were not substituted. In addition, although ptpTALECD could substitute Cs on the 3' side of T and A, it was unclear whether it could also substitute Cs on the 3' side of G and C. In this study, we identified the preferential positions of the substituted Cs in ptpTALECD-targeting sequences in the Arabidopsis plastid genome. We also found that ptpTALECD could substitute Cs on the 3' side of all four bases in plastid genomes of Arabidopsis. More recently, a base editor containing an improved version of DddA (DddA11) was reported to substitute Cs more efficiently, and to substitute Cs on the 3' side of more varieties of bases in human mitochondrial genomes than a base editor containing DddA. Here, we also show that ptpTALECD_v2, in which a modified version of DddA11 functions as the catalytic domain, more frequently substituted Cs than ptpTALECD in the Arabidopsis plastid genome. We also found that ptpTALECD_v2 tended to substitute Cs at more positions than ptpTALECD. Our results reveal that ptpTALECD can cause a greater variety of codon changes and amino acid substitutions than previously thought, and that ptpTALECD and ptpTALECD_v2 are useful tools for the targeted base editing of plastid genomes.

Investigating the dynamics of protein-protein interactions in plants.

Both stable and transient protein interactions play an important role in the complex assemblies required for the proper functioning of living cells. Several methods have been developed to monitor protein-protein interactions in plants. However, the detection of dynamic protein complexes is very challenging, with few technologies available for this purpose. Here, we developed a new platform using the plant UBIQUITIN promoter to drive transgene expression and thereby to detect protein interactions in planta. Typically, to decide which side of the protein to link the tags, the subcellular localization of the protein fused either N-terminal or C-terminal mCitrine was firstly confirmed by using eight different specific mCherry markers. Following stable or transient protein expression in plants, the protein interaction network was detected by affinity purification mass spectrometry. These interactions were subsequently confirmed by bimolecular fluorescence complementation (BiFC), bioluminescence resonance energy transfer and co-immunoprecipitation assays. The dynamics of these interactions were monitored by Förster resonance energy transfer (FRET) and split-nano luciferase, whilst the ternary protein complex association was monitored by BiFC-FRET. Using the canonical glycolytic metabolon as an example, the interaction between these enzymes was characterized under conditions that mimic physiologically relevant energy statuses.

A protocol for rapid generation cycling (speed breeding) of hemp (Cannabis sativa) for research and agriculture.

Hemp (Cannabis sativa) is a highly versatile crop with a multitude of applications, from textiles, biofuel and building material to high-value food products for consumer markets. Furthermore, non-hallucinogenic cannabinoids like cannabidiol (CBD), which can be extracted from female hemp flowers, are potentially valuable pharmacological compounds. In addition, hemp has high carbon sequestration potential associated with its rapid growth rate. Therefore, the hemp industry is gaining more traction and breeding hemp cultivars adapted to local climate conditions or bred for specific applications is becoming increasingly important. Here, we present a method for the rapid generation cycling (speed breeding) of hemp. The speed breeding protocol makes use of the photoperiod sensitivity of Cannabis. It encompasses vegetative growth of the plants for 2 weeks under continuous light, followed by 4 weeks under short-day conditions, during which flower induction, pollination and seed development proceed, and finally a seed ripening phase under continuous light and water stress. With the protocol described here, a generation time of under 9 weeks (61 days) from seed to seed can be achieved. Furthermore, our method synchronises the flowering time of different hemp cultivars, thus facilitating crosses between cultivars. The extremely short generation time will enable hemp researchers and breeders to perform crosses in a time-efficient way and generate new hemp cultivars with defined genetic characteristics over a short period of time.

M2 plants derived from different tillers of a chemically mutagenized rice M1 plant carry independent sets of mutations.

Generation of mutant populations with high genetic diversity is key for mutant screening and crop breeding. For this purpose, the single-seed descent method, in which one mutant line is established from a single mutagenized seed, is commonly used. This method ensures the independence of the mutant lines, but the size of the mutant population is limited because it is no greater than the number of fertile M1 plants. The rice mutant population size can be increased if a single mutagenized plant produces genetically independent siblings. Here, we used whole-genome resequencing to examine the inheritance of mutations from a single ethyl methanesulfonate (EMS)-mutagenized seed (M1 ) of Oryza sativa in its progeny (M2 ). We selected five tillers from each of three M1 plants. A single M2 seed was selected from each tiller, and the distributions of mutations induced by EMS were compared. Surprisingly, in most pairwise combinations of M2 siblings from the same parent, ≥85.2-97.9% of all mutations detected were not shared between the siblings. This high percentage suggests that the M2 siblings were derived from different cells of the M1 embryo and indicates that several genetically independent lines can be obtained from a single M1 plant. This approach should allow a large reduction in the number of M0 seeds needed to obtain a mutant population of a certain size in rice. Our study also suggests that multiple tillers of a rice plant originate from different cells of the embryo.

Early detection of late blight in potato by whole-plant redox imaging.

Late blight caused by the oomycete Phytophthora infestans is a most devastating disease of potatoes (Solanum tuberosum). Its early detection is crucial for suppressing disease spread. Necrotic lesions are normally seen in leaves at 4 days post-inoculation (dpi) when colonized cells are dead, but early detection of the initial biotrophic growth stage, when the pathogen feeds on living cells, is challenging. Here, the biotrophic growth phase of P. infestans was detected by whole-plant redox imaging of potato plants expressing chloroplast-targeted reduction-oxidation sensitive green fluorescent protein (chl-roGFP2). Clear spots on potato leaves with a lower chl-roGFP2 oxidation state were detected as early as 2 dpi, before any visual symptoms were recorded. These spots were particularly evident during light-to-dark transitions, and reflected the mislocalization of chl-roGFP2 outside the chloroplasts. Image analysis based on machine learning enabled systematic identification and quantification of spots, and unbiased classification of infected and uninfected leaves in inoculated plants. Comparing redox with chlorophyll fluorescence imaging showed that infected leaf areas that exhibit mislocalized chl-roGFP2 also showed reduced non-photochemical quenching and enhanced quantum PSII yield (ΦPSII) compared with the surrounding leaf areas. The data suggest that mislocalization of chloroplast-targeted proteins is an efficient marker of late blight infection, and demonstrate how it can be utilized for non-destructive monitoring of the disease biotrophic stage using whole-plant redox imaging.