All reagents and chemicals used were of analytical grade. Nuclease-free water from Integrated DNA Technologies, Australia, was employed for preparing all aqueous solutions. The PureLink^™ Microbiome DNA purification kit was purchased from Thermo Fisher Scientific, Australia. Additionally, all designed primers and the synthetic target (4.2 × 10⁷ copies/μL) were purchased from Integrated DNA Technologies, United States. The WarmStart^® Colorimetric LAMP 2X Master Mix and WarmStart^® Multi-Purpose RT-LAMP 2X Master Mix, as well as the 2X SensiFAST SYBER No-ROX Master Mix, were obtained from New England Biolabs, Ipswich, MA, United States, and Meridian Bioscience, Cincinnati, Ohio, United States, respectively. Furthermore, all other reagents were purchased from Sigma-Aldrich, United States.

The Xalb strain 3/14/9 bacterium was obtained from the Sugar Research Australia (SRA), Indooroopilly Research Station, Indooroopilly, Queensland, Australia (S 27.50°, E 152.98°). It was then cultured in the modified liquid broth Wilbrink’s media at a temperature of 28°C for 5 days, following the protocol of Davis et al. (1994). Leifsonia xyli subsp. xyli (Lxx), the bacterium responsible for sugarcane ratoon stunting disease (RSD), was used as a negative control. Lxx was isolated from an infected sugarcane plant at the SRA Woodford Pathology Research Station (S 26.93°, E 152.78°) in Queensland, Australia. It was cultured in a modified liquid S8 medium and incubated at 28°C for 4 weeks with gentle shaking at 200 rpm in the dark, following the protocols of Davis et al. (1980) and Chakraborty et al. (2024).

Three separate field trials with the same cultivars were conducted at the SRA Pathology Research Station in Woodford (S 26.93°, E 152.78°) for this study from 2022–2023. For this, 10 sugarcane cultivars were obtained from a disease-free propagation block at the same research facility, which was established using sugarcane setts treated with long hot water (50°C for 2 h) with a sterile cane knife to eliminate systemic Xalb infections (Table 1). Two-month-old plants were inoculated with Xalb 2–3 months before sampling using the protocol of Koike (1965). On a cloudy day, the cane tops of six plants per cultivar were decapitated using sharp short, handled cane knives above the growing point between the 3rd and 4th dewlap. Subsequently, the Xalb bacteria were “painted” onto the cut tissue with a brush dipped in the inoculum (~10⁸ cells/mL). Each replicated plot comprises 3 m (length) × 1.5 m (width) with a 1 m gap between two plots. The soil preparation for the trial involved maintaining good soil structure (red and grey soils) and ensuring moisture conservation through minimum tillage practices. The region receives an annual rainfall of 1,200–1,400 mm, supporting optimal soil conditions for the experiment. The trials were established using a randomized complete block (RCB) design, with three replications, each consisting of six plants.

Sugarcane leaf tissues, meristematic tissues, and xylem sap were sampled from three stalks (one stalk/replication of each cultivar) at different time points of post-inoculation from the SRA Woodford LS field trial in July 2022, October 2022, and January 2023 for the first trial. For xylem sap, vascular extracts (2 mL/cultivar) were obtained using positive air pressure extraction, as outlined by Croft et al. (1994). The extracts were kept on ice during transport to the laboratory and subsequently stored at −20°C until further processing. To validate the assay’s performance using field-derived samples, sugarcane samples were collected from eight cultivars (Q68, Q208, Q124, Q133, Q96, Q63, Q87, and Q44) with various disease ratings (Table 1) initially included in the first trial. In subsequent trials, two additional highly susceptible cultivars (Q189 and Q199) were included to obtain more consistent results (Table 1).

To determine the best leaf position for detecting Xalb presence, leaf tissues were collected from three locations on each of the three plants across all 10 sugarcane cultivars (Table 1). These positions were chosen based on their relevance to Xalb detection, ensuring consistent and reliable sampling. Leaf tissue samples were collected monthly from the second LS trial between May 2023 and December 2023 to observe consistency at different time points. Samples were taken from the top visible dewlap (TVD) leaf, the topmost fully expanded leaf with a visible dewlap, as well as from TVD +1 and TVD −1, the leaves immediately above and below the TVD leaf, respectively. These three positions served as standard reference points for consistent sampling. To track Xalb population development in the leaf tissues, xylem sap, and meristematic tissues of the highly susceptible (Q44) and resistant (Q208) cultivars over time, samples were collected monthly from October 2022 to October 2023 from the third LS field trial at SRA Woodford following the same protocols described above. LS disease resistance was assessed over multiple years based on visible symptoms for most of the cultivars (Rott et al., 1997), as detailed in Table 1.

Genomic DNA from cultured Xalb cells and LS-infected sugarcane samples was extracted for qPCR analysis using the protocol provided in The PureLink^™ Microbiome DNA purification kit manual (Thermo Fisher Scientific, Australia). Each experiment was performed in triplicate with three replications to ensure consistency.

For the detection of the Xalb bacteria via LAMP assay, a rapid, reagent-free method for DNA isolation was utilized, as detailed by Chakraborty et al. (2024). The concentration of bacteria was estimated subsequently via comparison against a standard curve of known Xalb amounts and consequential LAMP detection limitation was determined. To establish the standard curve, a series of titrated cultured Xalb cells (ranging from 10⁷ to 10⁰ cells/μL) was introduced into disease-free sugarcane xylem sap samples (100 μL). These sap samples were collected from SRA’s disease-free propagation block, confirmed as Xalb-free by qPCR, and then subjected to boiling at 95°C for 2 min in a heat block. Subsequently, 2 μL of the boiled suspension was pipetted directly into LAMP or qPCR reactions. Each concentration was replicated three times.

Following the establishment of standard detection curves, field collected leaf tissue and meristematic tissue samples were cut into small pieces and incubated in 100 μL of distilled water for 10 min. The sample solution was then heat lysed by boiling at 95°C for 2 min in a heat block, and 2 μL of the suspension was used for further analysis. Similarly, for the xylem sap samples, 100 μL of sap was directly boiled at 95°C for 2 min without any incubation, and 2 μL of the suspension was used for analysis. Each experiment was conducted three times to ensure accuracy and reliability.

Primers were designed to target a specific 196 bp region within the XALB1 albicidin pathotoxin biosynthesis gene cluster between positions 43,128 and 43,323 of Xanthomonas albilineans (GenBank accession no. AJ586576.1) (National Center for Biotechnology Information, 1988). The selection was based on the crucial role that albicidin plays in the pathogenesis of Xalb, as previously highlighted (Royer et al., 2004; Pieretti et al., 2009). Additionally, this region was not homologous to any other species sequenced to date in the GenBank database.

For LAMP analysis, six primers were designed to recognize a total of eight distinct regions within the target sequence (Table 2 and Figure 1). These included two loop primers (XalbLF and XalbLP), two outer primers (XalbF3 and XalbB3), and two inner primers (XalbFIP and XalbBIP). The design process was carried out using the NEB LAMP primer designing tool available at https://lamp.neb.com. For qPCR analysis, forward and reverse primers (XalbFP and XalbRP) (Table 2) were designed using the NCBI primer blast web tool (Ye et al., 2012). To ensure the specificity of each primer, the corresponding sequences were screened against the NCBI nucleotide and genome databases using the BLASTn tool (Chen et al., 2015). All primer sequences showed 100% homology to the corresponding Xalb sequences. Furthermore, the possibility of hairpin and dimer formation was assessed using the OligoAnalyzer^™ Tool provided by Integrated DNA Technologies Inc., United States to ensure the primer design robustness and avoid any potential issues with primer self-interactions.

The standard LAMP mixture conditions were followed (Yang et al., 2022), with slight adjustments following the manufacturer’s instructions. For WarmStart fluorescent LAMP/RT-LAMP reactions, a 25 μL mixture was prepared, consisting of 2.5 μL of a 10X LAMP primer concentration (2 μM each of XalbF3 and XalbB3, 16 μM each of XalbFIP and XalbBIP, and 4 μM each of XalbLF and XalbLB). To this, 12.5 μL of 2X WarmStart Multi-purpose master mix, 0.5 μL of SYBR Green fluorescent dye, 2 μL of DNA template, and 7.5 μL of nuclease-free water were added. For the fluorescent LAMP assay, real-time amplification was performed at 65°C for 40 min using a CFX96 Touch Real-Time PCR Detection System (Bio-Rad Laboratories Pty Ltd., Australia). In this LAMP assay, fluorescence threshold time (Tt) is used to quantify amplification, analogous to Cq (quantification cycle) in qPCR. Unlike qPCR, LAMP amplification is isothermal and proceeds continuously, yet Tt represents a measurable and reproducible point where fluorescence exceeds a defined threshold (Schmidt et al., 2021). For WarmStart colorimetric LAMP reactions, a 25-μL mixture was prepared, comprising 2.5 μL of a 10X LAMP primer concentration (2 μM each of XalbF3 and XalbB3, 16 μM each of XalbFIP and XalbBIP, and 4 μM each of XalbLF and XalbLB). To this, 12.5 μL of 2X WarmStart colorimetric LAMP master mix, 2 μL of DNA template, and 8 μL of nuclease-free water were added. The mixture was incubated at 65°C for 40 min using the CFX96 Touch Real-Time PCR Detection System (Bio-Rad Laboratories Pty Ltd., Australia). Subsequently, the amplified products were stored at 4°C prior to further analysis. The colorimetric LAMP products were visually inspected, with samples turning yellow classified as Xalb-positive, and those remaining pink considered Xalb-negative, following the manufacturer’s instructions (New England Biolabs, United States). Sterile distilled water (ddH₂O) or fresh sap were used as no-target controls (NTCs). Meanwhile, Leifsonia xyli subsp. xyli (Lxx) cells or purified DNA from Lxx cells were utilized as negative controls (NG). Each sample, NTC or NG was analyzed in triplicate to evaluate intra-assay variability which was determined by calculating the SD (standard deviation) between the Tt (threshold time) values for each replicate.

The qPCR experiments were conducted following the guidelines provided by the manufacturer (New England Biolabs, United States). Each qPCR mixture, with a volume of 20 μL, consisted of 10 μL of 2X SensiFAST SYBER No-ROX Mix, 0.8 μL each of XalbFP (10 μM) and XalbRP (10 μM), 2 μL of DNA template, and 6.4 μL of nuclease-free water. To perform the qPCR, a CFX96 Touch Real-Time PCR Detection System (Bio-Rad Laboratories Pty Ltd., Australia) was utilized, applying the following reaction conditions: initial denaturation at 100°C for 1 min, 40 cycles at 98°C for 15 s, 52°C for 30 s, and 72°C for 30 s. The reaction was then terminated with a heating step at 72°C for 2 min, followed by a hold at 4°C for 5 min. At the completion of the reaction, the Cq value for each dilution was analyzed. A positive result for Xalb was recognized if it was observed within less than 40 cycles. Each assay was performed in triplicate for each repetition to ensure accuracy and reliability. For gel documentation, 5 μL of each LAMP or PCR product was loaded onto a 1% agarose gel and electrophoresed in 1X Tris-acetate-EDTA (TAE) buffer at 90 V for 40 min. The gel was then stained with SYBR Safe and visualized using UV light on a gel documentation system. The sizes of the amplified products were estimated using a 100 bp GeneRuler (Thermo Fisher Scientific, Australia).

Data analyses were conducted using OriginPro 2022 v.9.9.0.225 (OriginLab, Northampton, Massachusetts, United States), the R programming language (version 4.2.1), and Microsoft Excel 365 (United States). To analyze the sensitivity, serial dilutions of known Xalb target sequence, Xalb cells, and purified genomic DNA concentrations were utilized to create a standard curve for the absolute quantification of Xalb. This standard curve was represented as a semi-log regression line plot, with Tt and Cq values plotted against the (−log) of the input DNA template amount. The efficiency (E) of the fluorescence LAMP and qPCR assays was calculated using the formula E = (10^−1/slope) − 1. Both the fluorescence LAMP and qPCR results obtained with the target-specific primers were validated within the range of 0.9 < E < 1.1, with an E value closer to 1.0 indicating higher amplification efficiency. To estimate the development of bacterial population over time in leaf tissue, meristematic tissue, and xylem sap of two sugarcane cultivars (Q44 and Q208) using LAMP, a logistic model was used as:

where, P is estimated development of Xalb population, d is estimated maximum population, b is estimated rate of development, c is inflection point or estimated median Xalb population development time, and t is Xalb population development time (month).

To convert the LAMP data (Tt values) to Xalb population or Cells the following equation was utilized:

where, x is the logarithm of the initial cell number a is the y-intercept, b is slope of the standard curve obtained from the Xalb cell number, and y is the Tt value.

To predict the cultivar disease ratings for Xalb, Tt and Cq values from the analysis of sugarcane samples were considered, whereby a correlation between Cq values and pathogen titer, was previously shown to correspond to disease ratings of cultivars (Davis et al., 1988; Rott et al., 1997). To analyze the Tt and Cq datasets, a linear mixed model was fitted using proc mixed in SAS version 9.4 (SAS Institute, Cary, NC). This model enabled the evaluation and interpretation of Tt and Cq data in relation to the disease ratings of sugarcane cultivars. Cultivars were considered as fixed effects, while block (replication) and the error term (residual) were regarded as random effects in the statistical analysis. To determine appropriate significance factors, protected-mean comparisons of all possible pairwise differences in Tt and Cq values were analyzed at alpha = 0.05 using Fisher’s protected LSD test. The PDMIX800 SAS Macro was employed to convert the mean separation outputs into letter groupings (Saxton, 1998). Spearman rank correlation was conducted to assess the relationship and accuracy of the newly developed method by comparing the outputs of the LAMP and qPCR techniques. Data visualizations were created using BioRender (BioRender, 2022), SnapGene software,¹ and Microsoft PowerPoint 365 (United States). The color change observed in the colorimetric LAMP reaction was captured using a mobile camera (Samsung Galaxy A52s).

The newly designed LAMP primer sets successfully amplified the targeted XALB1 albicidin pathotoxin biosynthesis gene cluster region (196 bp) in both colorimetric and fluorescent LAMP assays using 10 pM (equivalent to 10⁷ copies/μL) of synthetic target sequences. This confirmed the efficacy of the primers in detecting Xalb (Supplementary Figure S1A). In the colorimetric LAMP assay, a distinct color change from pink to yellow indicated a positive response when Xalb targets were present, while the no-template control (NTC) and negative control (NC) remained pink (Supplementary Figure S1B), highlighting the specificity of the primers and the suitability of the assay for naked-eye detection of Xalb. In the gel electrophoresis, the LAMP amplification products exhibited a typical ladder-like pattern, with the primary band corresponding to the expected 196 bp amplicon, along with other bands resulting from the amplification properties of LAMP (Supplementary Figure S1C).

Furthermore, the positive color change to yellow was consistently observed across all eight serially diluted concentrations of synthetic target sequences, from 10⁷ to 10⁰ copies/μL, with the lowest detection limit of this assay of 1 copy/μL (1 ag/μL) (Figure 2A). Similarly, in the fluorescent LAMP assay, detection of the synthetic targets was achieved solely in the infected sap samples, with no signal observed in the NTC (Supplementary Figure S2). The fluorescent LAMP signal exhibited a direct and reproducible correlation with the target DNA concentration, as indicated by the Tt values obtained at various time points throughout the reaction (Figure 2B), yielding a high correlation coefficient (r) of 0.99. The lowest detection limit for the LAMP assay remained at 1 copy/μL (1 ag/μL), and gel electrophoresis confirmed the amplification of the expected 196 bp product, along with additional bands, exclusively in positive reactions (Figures 2A; Supplementary Figure S2).

To further validate the results of the LAMP assay, a standard curve for qPCR was established using titrated synthetic targets ranging from 10⁷ to 10⁰ copies/μL of Xalb (Figure 2C), with a lowest concentration detected was 10 copies/μL (100 ag/μL) or 1 nM at an annealing temperature of 52°C (Supplementary Figure S5A), which was 10 times less sensitive than the LAMP assay (Figure 2B). Moreover, a single amplicon of the correct size (196 bp) was visualized exclusively from the positive samples (Supplementary Figures S5B, S6B), further supporting the robustness and reliability of the LAMP assay for detecting Xalb.

Using the standard curve generated from predetermined titrated concentrations of Xalb cells (10⁷–10⁰ cells/μL), the lowest detection limit of the colorimetric LAMP assay was approximately 1 cell/μL (Figure 3A), consistent with the fluorescence LAMP assay results (SD ≤5% across three replicates, r = 0.99; Figure 3B). Additionally, the expected LAMP amplicon of 196 bp, along with a ladder-like pattern, was visible on agarose gel for positive samples only (Supplementary Figure S3).

To further validate the LAMP assay with Xalb cell samples, a qPCR standard curve was created by plotting Cq values against the log-transformed quantities of purified DNA extracted from titrated Xalb cell concentrations. The qPCR assay reliably amplified the target region from as few as 100 Xalb cells (Figures 3C; Supplementary Figure S6B), with the expected 196 bp amplicon observed on agarose gel for positive samples (Supplementary Figure S6D). A strong correlation (r = 0.99) was observed between Tt and Cq values from LAMP and qPCR assays. Notably, the LAMP assay showed a sensitivity 100 times greater than qPCR for in-situ detection and quantification of Xalb bacteria (Figure 3B vs. Figure 3C).

Target DNA was amplified from all the field samples collected from the first LS field trial, indicating varying levels of pathogen load (Figures 4, 5). In colorimetric LAMP, a yellow color development was observed in all field samples, providing a rapid and consistent qualitative indication of bacteria in the sample (Figure 4).

The fluorescent LAMP assay effectively amplified Xalb DNA from all collected samples across different Tt values, reflecting varying levels of pathogen load (Figures 5A–C). Most samples exhibited Xalb DNA concentrations within the assay’s lowest detection limit of 1 aM, highlighting its suitability for practical field applications. The detected Xalb DNA level in the different sugarcane samples correlated with bacterial load. For example, the Q87, Q63, Q68, and Q208 cultivars showed relatively higher Tt values (11.36–22.62) for all the analyzed sample types collected at different time points, suggesting lower Xalb presence (approximately 10⁴–10⁷ cells/μL) in xylem sap, leaf tissue, and meristematic tissue (Figures 3B, 5A–C). Conversely, cultivars Q44, Q96, Q124, and Q133 displayed progressively lower Tt values (4.69–19.33), indicating higher bacterial load (around 10⁷–10⁹ cells/μL) in these samples (Figures 3B, 5A–C). This data demonstrates that the LAMP method is effective for detecting and quantifying Xalb titres in infected sugarcane samples across various cultivars, irrespective of their resistance status. Notably, significantly different amounts of the amplified product of expected size (196 bp) for Xalb DNA were detected among them, with no amplification observed in control reactions (Supplementary Figure S4).

The target qPCR region was successfully amplified from all infected field samples (Figure 6), revealing a significant correlation (r = 0.9, p < 0.001) between the qPCR values and the results of the LAMP assay, as outlined in Table 3. Despite the differences in amplification properties between LAMP and qPCR, strong correlations (r > 0.9) were observed between LAMP and qPCR results (Table 3), regardless of sample collection date or site. Additionally, a single amplicon was observed on the gel from positive samples only (Supplementary Figure S7), confirming the specificity of the assay.

When analyzing the Tt values from the LAMP experiment across different sample types using the SAS Proc Mixed Model, significant differences (p < 0.0001) were observed in Tt values (indicating Xalb load) among clones, sample origins (leaf tissue, meristematic tissue, or xylem sap), sampling times, or their interactions (Supplementary Table S1). Strong correlations were identified between leaf tissue samples and disease ratings when collected at 5 months (r = 0.75) and 8 months (r = 0.8) post-inoculation. However, no differences in Tt values were found between cultivars when samples were taken 11 months after inoculation (Figure 7A). A moderate correlation (r = 0.53) was observed between varietal resistance and Tt values at around 8 months post-inoculation, while low or no correlation was found at 5- or 11 months post-inoculation (Figure 7B). Additionally, there was no meaningful relationship between varietal resistance and Tt values in xylem sap (Figure 7C). These results suggest that leaves may be the suitable plant part for effective sampling for leaf scald diagnosis and varietal disease rating.

Significant differences were observed in Tt and Cq values among cultivars collected from the second LS trial and between the methods used (LAMP and qPCR). Specifically, in the interaction between cultivar and method (p < 0.001) and between the two methods themselves (p < 0.001) (Table 4). However, no significant differences were found between leaf locations within a variety, indicating a similar Xalb load among the leaf positions (TVD, TVD +1 and TVD −1) on a sugarcane stalk (Table 4).

Following inoculation, the development of the Xalb population over time in leaf tissue, xylem, and meristematic tissues were analyzed for two sugarcane cultivars collected from the third LS trial, described by the regression equation in Table 5 and illustrated in Figure 8. The initial Xalb population, as estimated from the regression equations, was approximately 2.4 × 10³ Cells/μL for the susceptible cultivar Q44 and less than 60 Cells/μL for the resistant cultivar Q208. The maximum Xalb levels in the leaf tissue of Q208 and Q44 were estimated to be approximately 1.1 × 10⁸ Cells/μL and 1.8 × 10⁹ Cells/μL, respectively, indicating that the levels in Q44 were more than 17 times higher than those in Q208. Indeed, the rate of Xalb reproduction was slower in Q208 compared to Q44; taking about 146 days for Q208 to reach 50% of its maximum Cells, while Q44 reached it in approximately 131 days. Similar trends were observed among the meristematic tissues and xylem saps between the two cultivars.