Culture-Free Detection of Crop Pathogens at the Single-Cell Level by Micro-Raman Spectroscopy

Abstract

The rapid and sensitive identification of invasive plant pathogens has important applications in biotechnology, plant quarantine, and food security. Current methods are far too time-consuming and need a pre-enrichment period ranging from hours to days. Here, a micro-Raman spectroscopy-based bioassay for culture-free pathogen quarantine inspection at the single cell level within 40 min is presented. The application of this approach can readily and specifically detect plant pathogens Burkholderia gladioli pv. alliicola and Erwinia chrysanthemi that are closely related pathogenically. Furthermore, the single-bacterium detection was able to discriminate them from a reference Raman spectral library including multiple quarantine-relevant pathogens with broad host ranges and an array of pathogenic variants. To show the usefulness of this assay, Burkholderia gladioli pv. alliicola and Erwinia chrysanthemi are detected at single-bacterium level in plant tissue lesions without pre-enrichment. The results are confirmed by the plate-counting method and a genetic molecular approach, which display comparable recognition ratios to the Raman spectroscopy-based bioassay. The results represent a critical step toward the use of micro-Raman spectroscopy in rapid and culture-free discrimination of quarantine relevant plant pathogens.

Keywords: crop pathogens; culture‐free detection; micro‐Raman spectroscopy; single cells.

Figure 1

Microscopic images, pathogenic symptoms, and micro‐Raman spectra of B. gladioli pv. alliicola and E. chrysanthemi. a) Microscopic images of B. gladioli pv. alliicola and the pathogenic symptoms in onion tissue infected by B. gladioli pv. alliicola. b) Microscopic images of E. chrysanthemi and the pathogenic symptoms in onion tissue infected by E. chrysanthemi. c) Comparison of the B. gladioli pv. alliicola and E. chrysanthemi Raman spectra. d) Discrimination of the B. gladioli pv. alliicola (red squares) and E. chrysanthemi (black squares) Raman spectra. For each species, 30 bacterial spectra were used as training sets while ten bacterial spectra were used to validate the model. The open scatter symbols are for training spectra, and the filled symbols are for the testing spectra.

Figure 2

Discriminate B. gladioli pv. alliicola and E. chrysanthemi from typical quarantine‐relevant plant pathogens. a) Microscopic images and average Raman spectra of eight different pathogenic strains. The positions of the Raman bands at 1004, 1048, 1155, 1445, 1519, and 1655 cm⁻¹ are highlighted. Scale bar represents 2 µm. b) Discrimination of the Raman spectra of the eight typical plant pathogens. The pathogens are represented as follows: A. avenae subsp. cattleyae (green pentagons), B. gladioli pv. alliicola (black squares), C. michiganensis subsp. insidiosus (green squares), C. michiganensis subsp. sepedonicus (pink pentagons), E. chrysanthemi (pink squares), E. stewartii (red squares), P. syringae pv. pisi (black pentagons), and P. syringae pv. tomato (red pentagons). The open scatter symbols are for training spectra, and the filled symbols are for the testing spectra.

Figure 3

Culture‐free detection of E. chrysanthemi and B. gladioli pv. alliicola in onion tissues. a) Comparison of the identification ratios of micro‐Raman spectroscopy‐ and molecular marker (i.e., 16S rDNA)‐based diagnostic approaches for pathogen detection in onion infected by either E. chrysanthemi or B. gladioli pv. alliicola. b) Comparison of the identification ratios of micro‐Raman spectroscopy‐ or molecular marker (i.e., 16S rDNA)‐based diagnostic approaches for pathogen detection in onion infected by both E. chrysanthemi and B. gladioli pv. alliicola.