A novel Leifsonia xyli subsp. xyli quantitative LAMP-based diagnostic correlated with sugarcane ratoon stunting disease rating

Source and production of inoculum

Leifsonia xyli subsp. xyli (Lxx) was isolated from naturally infected sugarcane plants grown at the Sugar Research Australia (SRA), Woodford Pathology Research Station, Woodford, Queensland, Australia (S 26.93°, E 152.78°). For this, stalks were cut at the base using sterilised secateurs and scrubbed to eliminate any extraneous material. Afterward, a small air compressor (Boss Air Suspension, Queensland, Australia) with an attached soft rubber cup was used to provide positive pressure to the stalk from the basal nodal and internodal regions to push the sap out of the vascular bundles. The sap was collected in 15 mL centrifuge tubes and immediately kept on ice during transport to Griffith University, Nathan campus. The extracted sap was then filtered using a 0.2 μm syringe fitted filter and aliquoted into 1.5 mL microcentrifuge tubes. For Lxx culturing, 2 mL of the sap was also added to 100 mL of modified liquid broth S8 medium and incubated at 28°C for 4 weeks with gentle shaking at 200 rpm in the dark according to Sugar Research Australia’s (SRA) industry protocol (Davis et al. 1980; Brumbley et al. 2002).

Sugarcane leaf scald (LS) disease-causing bacteria, Xanthomonas albilineans (Xalb), were used as a negative control. Xalb strain 3/14/9 was obtained from SRA, Indooroopilly Research Station, Queensland, cultured in modified Wilbrink’s broth media and incubated at 28°C for 5 days according to the industry protocol (Dawson 1957). The culture flasks were incubated in temperature-controlled shakers with gentle shaking at 250 rpm until the OD600 reached approximately 0.5, where Lxx and Xalb cell densities were calculated to be 4 × 10⁸ cfu/mL and 16 × 10⁸ cfu/mL, respectively (Monteiro-Vitorello et al. 2004; Mira et al. 2022). Each measurement was conducted in triplicate for three repeated experiments.

Establishment of field trial, inoculation, and planting

An RSD field trial was established at SRA’s Pathology Research Station, Woodford, Queensland, Australia in September 2020. Ten sugarcane genotypes with known disease ratings for RSD (Ngo et al. 2023) were sourced from a disease-free propagation block in Kallangur (S 27.235°, E 153.01°), Queensland, approximately 50 km south of SRA’s Pathology Research Station. Stalks were cut into one-budded setts using an electric saw and washed in tap water to remove dirt and debris (Table 1). Sugarcane setts were then treated in 52°C hot water for 30 min and dried before inoculation with Lxx. Inoculation of the one-budded setts was done by soaking them in a suspension of 16 × 10⁸ cfu/mL of lab-grown Lxx cells measured at OD600 following the method described by Ngo et al. (2023). The inoculated setts were then kept in trays under LED lights covered with moistened vermiculite at 30°C with 60–80% humidity for 2 weeks in a germination chamber. Subsequently, they were planted in a field at the SRA Pathology Research Station, Woodford in a randomised complete block design, with three replications consisting of six plants each.

Sugarcane field samples

Xylem sap samples were collected from three stalks (i.e. one stalk/replication) harvested at 53 weeks after inoculation, when the stalks produced at least two to three visible internodes above-ground. Xylem sap extracts (2 mL/variety) were collected as previously described (Croft et al. 1994) and stored at −20°C until processing. Sap samples were collected from the following genotypes: CP72-2086, Ho06-537, Q232, Q253, SRA20, Q208, WSRA24, Q242, SRA26, and SRA22, all with previously designated and industry accepted RSD disease ratings based on ELISA and/or qPCR studies (Croft et al. 1994; Croft 2002; Croft and Johnson 2013; Ngo et al. 2023). The disease ratings employed in this study were on a 0 to 9 scale and adapted from Ngo et al. (2023; Table 1).

Reagents and materials

All reagents and chemicals used in this study were of analytical grade. Nuclease-free water (Integrated DNA Technologies, Australia) was used to prepare all aqueous solutions. PureLink™ Microbiome DNA purification kits were purchased from Thermo Fisher Scientific (Australia). Designed primers and synthetic targets (4.2 × 10⁷ copies/μL) were purchased from Integrated DNA Technologies (USA). WarmStart® Colorimetric LAMP 2X Master Mix and WarmStart® Multi-Purpose RT-LAMP 2X Master Mix, and 2X SensiFAST SYBER No-ROX Master Mix were purchased from New England Biolabs (Ipswich, MA, USA), and Meridian Bioscience (Cincinnati, Ohio, USA), respectively. All other reagents were purchased from Sigma-Aldrich (USA).

Reagent-free DNA isolation method

To develop a standard detection curve using a heat-induced reagent-free rapid DNA isolation method, a range of known concentrations of cultured Lxx cells (10⁷−10⁰ cells/μL) were inoculated into 1 mL of freshly extracted and ‘clean’ sugarcane xylem sap. Subsequently, the mixtures were boiled at 95°C for 2 min in a digital heat block (Morganville Scientific, USA). Then, 2 μL aliquots of the cooled supernatant were used as templates for LAMP or qPCR assays. Once the detection thresholds were established, 100 μL volumes of field-collected and naturally infected sap samples were subjected to the same method and 2 μL of the resulting supernatant was again used for LAMP or qPCR. Each experiment was performed in triplicate and repeated three times to ensure accuracy and reliability.

Commercial kit for DNA extraction

Genomic DNA of Lxx was extracted from the known numbers of Lxx cultured cells (10⁷−10⁰ cells/μL) and RSD-infected field samples for qPCR analysis according to the described protocol of a commercial kit (The PureLink™ Microbiome DNA purification kit manual, Thermo Fisher Scientific, Australia). Each experiment was conducted in triplicate and repeated three times.

Target selection and primer design

Primers were designed targeting a 210 bp conserved section of the Lxx intergenic spacer (IGS) region between the 16S and 23S rRNA genes (GenBank accession no. AE016822.1) corresponding to positions 1,35,389–1,35,598 in L. xyli subsp. xyli strain CTCB07 complete genome (National Center for Biotechnology Information (NCBI) 1988). This locus was chosen because of its single copy number status (Supplementary Fig. S1) (Monteiro-Vitorello et al. 2004, 2013) and had no significant similarity to any other sequences available in GenBank (searched on 26/04/2022). For qPCR analysis, forward and reverse primers (LxxFP and LxxRP respectively, Table 2) were designed using the NCBI primer blast web tool with standard parameters (Ye et al. 2012). For LAMP analysis, two loop primers (LxxLF and LxxLP), two outer primers (LxxF3 and LxxB3), and two inner primers (LxxFIP and LxxBIP) were designed using the NEB LAMP primer designing tool, again with standard parameters (https://lamp.neb.com) (Table 2) (Fig. 1). The specificity of each primer was confirmed by screening each sequence using the BLASTn (Chen et al. 2015) tool against the NCBI nucleotide and genome databases. All the sequences were found to be 100% homologous to the corresponding Lxx sequences (Chen et al. 2015) and no close matches (E < 10) was found. The OligoAnalyzer™ Tool (Integrated DNA Technologies Inc., USA) was used to assess the possibility of the formation of hairpins and dimers.

Fig. 1.

Schematic illustration of the primer design for the proposed LAMP assay representing the position of the eight primers spanning the target gene and the nucleotide sequences of the intergenic spacer (IGS) region between 16S and 23S ribosomal RNA (rRNA) genes (GenBank accession no. AE016822.1). Sequence corresponds to positions 1,35,389–1,35,598 of L. xyli subsp. xyli strain CTCB07. The sense sequence is denoted by the right arrow, and the complementary sequence is denoted by the left arrow.

LAMP reaction conditions

LAMP mixtures for fluorescent or colorimetric analyses were prepared according to the protocols described in the WarmStart® Multi-Purpose RT-LAMP 2X Master Mix and WarmStart® Colorimetric LAMP 2X Master Mix (New England Biolabs, USA) manuals with slight modifications. WarmStart RT-LAMP reactions were carried out in a 25 μL mixture containing 2.5 μL of 10X LAMP primer concentration (2 μM each LxxF3 and LxxB3, 16 μM each LxxFIP and LxxBIP, 4 μM each of LxxLF and LxxLB), 12.5 μL of 2X WarmStart Multi-purpose master mix, 0.5 μL SYBR Green fluorescent dye, 2 μL of DNA template and 8 μL of nuclease-free water. WarmStart colorimetric LAMP reactions were also carried out in a 25 μL mixture containing 2.5 μL of 10X LAMP primer concentration (2 μM each of LxxF3 and LxxB3, 16 μM each of LxxFIP and LxxBIP, 4 μM each of LxxLF and LxxLB), 12.5 μL of 2X WarmStart colorimetric LAMP master mix, 2 μL of DNA template, and 8 μL of nuclease-free water. The mixture was incubated at 65°C for 40 min in a CFX96 Touch Real-Time PCR Detection System (Bio-Rad Laboratories Pty Ltd, Australia). Products were kept at 4°C for further analysis. The 25 μL colorimetric LAMP product volumes were inspected visually in the 200 μL tubes. Samples that turned yellow were considered Lxx-positive, and those that remained pink were assumed to be Lxx-negative, as per the manufacturer’s instructions (New England Biolabs, USA). Variations in colour gradient were linked to pathogen loads, with intense yellow colour indicating higher concentrations of Lxx, and pinkish-yellow shades suggesting lower levels of Lxx present in the samples. The colour change achieved in the colorimetric LAMP reaction was captured with a mobile camera. Sterile distilled water (ddH₂O) and fresh ‘clean’ sap served as the no-target controls (NTC), and Xalb cells and purified DNA from Xalb cells were used as negative controls (NC). Each experiment was performed in triplicate with three repetitions.

qPCR assay and gel documentation

To validate the LAMP assay, qPCR experiments were devised to detect the targeted intergenic spacer (IGS) region using synthetic targets, bacterial cell samples, and sugarcane sap samples collected from the field. RSD detection from sugarcane xylem sap samples was done using two conventional qPCR methods: (i) Using directly expressed xylem sap samples in the PCR reactions as a template (Goodwin and Lee 1993; Ghai et al. 2014), or (ii) Using a commercial kit to extract Lxx DNA from the sugarcane sap samples before using as a template. The qPCR analyses were conducted following the 2X SensiFAST SYBER No-ROX Master Mix manual (New England Biolabs, USA). A CFX96 Touch Real-Time PCR Detection System (Bio-Rad Laboratories Pty Ltd, Australia) was used with 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; followed by heating at 72°C for 2 min to terminate the reaction and a hold of 4°C for 5 minutes. The quantification cycle (Cq) value of each dilution was analysed at the end of the reaction and Lxx was considered present if a positive result was observed in less than 40 cycles. Each assay was performed in triplicate for three repeated experiments. For gel documentation, an aliquot of 5 μL of each LAMP or PCR product was loaded on 1% agarose, electrophoresed in 1X Tris-acetate-EDTA (TAE) buffer at 90 V for 40 min, stained with SYBR safe and visualised using UV light under gel documentation system. A 100 bp GeneRuler (Thermo Fisher Scientific, Australia) was used to estimate the sizes of amplified products.

Statistical analysis

Serial dilutions of known L. xyli subsp xyli target sequence, Lxx cells, and purified genomic DNA concentrations were used to generate a standard curve for the absolute quantification of Lxx. This was then used to determine the sensitivity and detection threshold of the qPCR. The standard curve was represented as a semi-log regression line plot of Cq values versus the −log of the input DNA template amounts. The efficiencies (E) of the fluorescence LAMP and qPCR assays were calculated using E = (10^−1/slope) − 1. Validation of fluorescence LAMP and qPCR results with target-specific primers was achieved when E-values were between 0.9 and 1.1, with E-values closer to 1.0 indicating higher amplification efficiency (Wu et al. 2018). The determination of Lxx quantity relied on Cq values derived from the analysis of xylem sap samples. This was then correlated with established disease resistance ratings, as demonstrated in earlier studies (Davis et al. 1988; Ngo et al. 2023). Consequently, a linear mixed model was fitted to the Cq datasets using proc mixed in SAS version 9.4 (SAS Institute, Cary, NC, USA), where varieties were treated as fixed effects, and block (replication) and the error term (residual) were treated as random effects. For the appropriate significance factors, protected-mean comparisons of all possible pairwise differences of the Cq values were tested at alpha = 0.05, using Fisher’s protected l.s.d. test. PDMIX800 SAS Macro was then used to convert mean separation outputs to letter groupings (Saxton 1998). Spearman rank correlation was then conducted between the outputs of the LAMP and qPCR methods to determine the relatedness and accuracy of the newly developed method. Statistical analyses were performed using OriginPro 2022 v.9.9.0.225 (OriginLab, Northampton, Massachusetts, USA), and Microsoft Excel 365 (USA). Graphical representations of data were carried out using BioRender, SnapGene software (www.snapgene.com), and Microsoft Powerpoint 365 (USA). A mobile camera (Samsung Galaxy A52s) was used to capture the colour change during the colorimetric LAMP reaction. The mean value ± s.e. among all three replications were determined for each sample.

Validation of novel Lxx target sequences and robustness of LAMP assay

The newly designed LAMP primer sets successfully amplified the targeted IGS region (210 bp) in both colorimetric and fluorescent LAMP assays using 10 pM (i.e. 10⁷ copies/μL) synthetic target sequences, confirming their efficacy (Fig. S2). In colorimetric LAMP, a colour change from pink to yellow indicated positive responses where the Lxx targets were present, whereas NTC and NC reactions remained pink, indicating that the primers were specific and useful for naked-eye detection of Lxx (Fig. S2). Additionally, the positive colour change to yellow occurred in all eight serially diluted concentrations of synthetic target sequence (representing 10⁷−10⁰ copies/μL) with a threshold of detection of 1 copy/μL (1 ag/μL) (Fig. 2a). Accordingly, the fluorescent LAMP assay detected the synthetic targets in the infected sap only, and not in the NTC (Fig. S2). The LAMP amplification signal level was directly and reproducibly proportional to the target concentration (Fig. 2b and S3a), with a correlation coefficient (r) of 0.99. Again, the threshold of LAMP assay detection was 1 copy/μL (1 ag/μL) and subsequent gel electrophoresis confirmed amplification of the expected size (210 bp) only in positive reactions (Fig. S2 and S3b).

Fig. 2.

LAMP and qPCR primer pairs checking with different synthetic target concentrations. (a) Colorimetric LAMP detection for the designated concentrations; Tubes 1 to 8: 1:10 dilutions of the synthetic target (10⁷−10⁰ copies/μL, or 10 pg/μL-1 ag/μL); Tube 9: No Target Control (NTC). Obtained Cq value of (b) Fluorescence LAMP and (c) qPCR detection for the designated concentrations of synthetic targets (10⁷−10⁰ copies/μL, or 10 pg/μL-1 ag/μL). Error bars represent the standard deviation of three repetitive experiments (biological replications).

Successful qPCR amplification of the synthetic target (1 nM) at an annealing temperature of 52°C validated the LAMP assay results (Fig. S6a). Consequently, the qPCR standard curve with titrated synthetic target (ranging 10⁷−10⁰ copies/μL of Lxx; Fig. 2c) was established and had a detection threshold of 10 copies/μL (100 ag/μL), 10 times less sensitive than the LAMP assay (Fig. 2b). Meanwhile, a single amplicon of the correct size (210 bp) was visualised from only the positive samples (Figs S6b and S7c).

Validation of LAMP-based in-situ detection and quantification of Lxx

Using predetermined and titrated concentrations of Lxx cells (10⁷ to 10⁰ cells/μL), the detection threshold of the colorimetric LAMP assay was approximately ~1 cell/μL (Fig. 3a) and was the same for the fluorescence LAMP assay (s.d. = ≤5% (n = 3), and r = 0.99; Fig. 3b and S4a). Also, the expected LAMP amplicon (210 bp) was observed on agarose gel in only positive samples (Fig. S4b).

Fig. 3.

Sensitivity and specificity analysis. (a) Colorimetric LAMP detection for the designated samples; Tubes 1 to 8: 1:10 dilutions created by spiking cells into clean fresh sap (10⁷−10⁰ cells/μL); Tube 9: No Target Control (NTC); Tube 10: Known number of spiked Xalb cells (10⁷ cells/μL). (b) Obtained Cq of fluorescence LAMP and qPCR detection for the designated number of Lxx cells spiked into fresh xylem sap (10⁷−10⁰ cells/μL); error bars represent the standard deviation of three repetitive experiments.

To validate the LAMP assay with Lxx cell samples, the standard plot for qPCR-based absolute quantification was obtained by plotting Cq values against a log of the corresponding amount of purified DNA extracted from the titrated Lxx cells. Consequently, qPCR effectively amplified the target region from the DNA extracted from as little as 100 Lxx cells (Fig. 3c and S7b). Again, the expected amplicon (210 bp) was observed on agarose gel in only positive sample reactions (Fig. S7d). There was a strong correlation (r = 0.99) between the Cq value from the LAMP and qPCR reactions. Also, the LAMP assay was 100 times more sensitive than the qPCR for in-situ detection and quantification of Lxx bacteria (Fig. 3b versus Fig. 3c).

Correlation of the new LAMP diagnostic with RSD resistance rating

To investigate the applicability of the assay for RSD resistance screening, the yellow colour was detected in all RSD-infected xylem sap samples from the 10 sugarcane varieties of known RSD resistance rating. Additionally, the gradient of the yellow colour was correlated with the industry resistance rating of the cultivar. The most intense colours were visualised from sap samples collected from more susceptible cultivars Q253, SRA26, SRA20, Q232, and Q242 (Fig. 4a, Tube # 1, 2, 8, 9 and 10), indicating high bacterial concentrations. In contrast, sap samples from more resistant cultivars Ho06-537, CP72-2086, Q208, SRA22, and WSRA24 (Fig. 4a, Tube # 3, 4, 5, 6, and 7) had a pinkish-yellow colour intensity in the reaction, suggesting lower bacterial concentrations (Table 1; Fig. 4a).

Fig. 4.

Field application of assay. (a) Colorimetric detection for all the analysed xylem sap samples collected from SRA Woodford RSD screening trials. Tubes 1 to 10: RSD-infected xylem sap samples- Q253, SRA26, Ho06-537, CP72-2086, Q208, SRA22, WSRA24, SRA20, Q232, and Q242; Tube 11: No Target Control (NTC); obtained Cq values for these samples, (b) fluorescence LAMP detection; qPCR detection of (c) directly on expressed xylem sap sample; (d) commercial kit-based DNA extraction technique; each bar is the mean of three replications, and bar(s) associated with same letter(s) are not significantly different according to Fisher’s protected l.s.d. test (P = 0.05).

The fluorescent LAMP assay successfully amplified Lxx DNA from all collected samples at different cycle times, indicating different levels of pathogen loads (Fig. 4b and S5a). The determined amount of Lxx DNA was then significantly correlated to the cultivar resistance rating, where higher LAMP values corresponded to lower known RSD resistance ratings (r = −0.82, P < 0.001; Table 3). Sap from the most susceptible cultivars, Q253, Q242, and SRA26 contained the highest number of bacteria (approximately 10⁶−10⁷Lxx cells/μL of sap samples), as depicted by their lower Cq values (20.02, 20.56, and 20.33) (Fig. 3b, 4b). In this manner, the LAMP results were validated with the susceptible to intermediate susceptible rating of cultivars Q232, SRA20 and WSRA24, with Cq values 21.09, 21.47, and 22.34 respectively, and the presence of around 10⁵−10⁶ cells/μL (Fig. 3b, 4b). Similarly, the LAMP results of SRA22 (Cq 23.38) and Q208 (Cq 23.49) were correlated with their intermediate resistant rating, exhibiting approximately 10⁴−10⁵ cells/μL, and were significantly (P < 0.05) different from that of CP72-2086 (Cq 24.50) and Ho06-537 (Cq 25.29), which showed approximately 10³−10⁴ cells/μL of sap (Fig. 3b, 4b). The LAMP results of CP72-2086 and Ho06-537 were supported by the previous findings by Ngo et al. (2023) that they are moderately resistant to RSD. Significantly different amounts of Lxx DNA 210 bp target amplicons were detected among sap samples from the cultivars and did not amplify in the control reactions (Fig. S5b; P = 0.05; Fig. 4b).

The target qPCR region was amplified from all infected field sap samples and assessed at all cycle times. The amount of amplified product was also directly correlated with the cultivar resistance rating (Table 1; Figs 4c, d, Figs S8a and S8b), albeit with different amounts of Lxx detected between the two methods and hence slightly different disease rating predictions (Figs 4, S8d). Both qPCR samples (sap and DNA) showed strong negative correlations with the resistance ratings. However, qPCRsap had slightly weaker (r = −0.72, P < 0.05) and qPCR_DNA showed slightly stronger (r = −0.87, P < 0.001) correlations (Table 3). Again, a single amplicon was present on the gel from positive samples only (Fig. S8c and d).