Optimization of Molecular Methods for Detecting Duckweed-Associated Bacteria

Abstract

The bacterial colonization dynamics of plants can differ between phylogenetically similar bacterial strains and in the context of complex bacterial communities. Quantitative methods that can resolve closely related bacteria within complex communities can lead to a better understanding of plant-microbe interactions. However, current methods often lack the specificity to differentiate phylogenetically similar bacterial strains. In this study, we describe molecular strategies to study duckweed-associated bacteria. We first systematically optimized a bead-beating protocol to co-isolate nucleic acids simultaneously from duckweed and bacteria. We then developed a generic fingerprinting assay to detect bacteria present in duckweed samples. To detect specific duckweed-bacterium associations, we developed a genomics-based computational pipeline to generate bacterial strain-specific primers. These strain-specific primers differentiated bacterial strains from the same genus and enabled the detection of specific duckweed-bacterium associations present in a community context. Moreover, we used these strain-specific primers to quantify the bacterial colonization of duckweed by normalization to a plant reference gene and revealed differences in colonization levels between strains from the same genus. Lastly, confocal microscopy of inoculated duckweed further supported our PCR results and showed bacterial colonization of the duckweed root-frond interface and root interior. The molecular methods introduced in this work should enable the tracking and quantification of specific plant-microbe associations within plant-microbial communities.

Keywords: Azospirillum brasilense Sp245; Azospirillum brasilense Sp7; RISA; bacterial colonization; bead-beating; duckweed; duckweed-associated bacteria; plant-bacteria associations; plant-microbe interactions; strain-specific primers.

Figure 1

Molecular detection of duckweed-associated bacteria. Axenic Lm5576 was inoculated with different bacteria in 0.5X SH media. After seven days, inoculated Lm5576 tissue was collected, washed with sterile water, and nucleic acids were isolated for analysis. (A) Representative gel electrophoresis results of end-point PCR using RISA, LmLFY, and strain-specific primers (File S2). RISA PCR fingerprints from inoculated Lm5576 samples were compared to the respective DNA controls from Lm5576 and bacteria alone. Samples: +G2-6 = Lm5576 inoculated with Bacillus simplex RUG2-6, +DAB 1A = Lm5576 inoculated with Microbacterium sp. RU370.1, +Sp7 = Lm5576 inoculated with Azospirillum brasilense Sp7, +Sp245 = Lm5576 inoculated with Azospirillum baldaniorum Sp245. Primers: RISA = PCR using 16S-1390f and 23S-e130r primers, LmLFY = PCR using LmLFY-F and LmLFY-R primers specific to Lm5576, DAB 1A = PCR using AmRU370.1-F and AmRU370.1-R primers specific to Microbacterium sp. RU370.1, G2-6 = PCR using BsRUG2.6-F and BsRUG2.6-R primers specific to Bacillus simplex RUG2-6, Sp7 = PCR using AbSp7-F and AbSp7-R primers specific to Azospirillum brasilense Sp7, Sp245 = PCR using AbSp245-F and AbSp245-R primers specific to Azospirillum baldaniorum Sp245. (B) Bacterial association with Lm5576 was quantified using real-time PCR. Bacterial DNA copy number was determined for each inoculated Lm5576 sample and normalized to Lm5576 DNA copy number. Different colors were used for the different bacterial genera. Each data point represents an experimental repeat except for +G2-6, for which each sample was measured twice. A significant difference was found in colonization levels between bacteria (Kruskal–Wallis, p-value = 4.58 × 10⁻⁶). Pairwise comparisons were performed using Dunn’s test, and results are displayed as compact letters (a–c). Bacteria with significantly different colonization levels from each other, according to Dunn’s test (FDR < 0.05), do not share any letters.

Figure 2

Two-dimensional confocal microscopy of inoculated duckweed samples. Confocal microscopy (40X/1.1 objective) was performed on inoculated Lm5576 in 0.5X SH media to characterize the spatial colonization dynamics of bacteria on duckweed. Calcofluor white was used to stain Lm5576 cellulose and visualized with the blue channel, whereas SYBR Gold was used to stain DNA and visualized with the green channel. Both bacterial and Lm5576 DNA were stained by SYBR Gold and pictured in green, but bacterial DNA (indicated by white arrows) is smaller in size than Lm5576 nuclei (indicated by red arrows), and these bacteria are often found in clustered colonies. For each main image, 1 unit of the grid scale is equal to 35.56 μm and zoomed-in images are pictured in the top-right corner. +G2-6 = Lm5576 inoculated with Bacillus simplex RUG2-6, +DAB 1A = Lm5576 inoculated with Microbacterium sp. RU370.1, +Sp7 = Lm5576 inoculated with Azospirillum brasilense Sp7, +Sp245 = Lm5576 inoculated with Azospirillum baldaniorum Sp245.

Figure 3

Molecular detection of duckweed-associated bacteria in a community context. Axenic dw9509 was inoculated with different bacteria in wastewater with (NS) or without (S) microbes. In addition, Sp245 was co-inoculated onto dw9509 with DAB isolates or non-sterile wastewater-containing microbes. For co-inoculated samples, bacteria were mixed at a 1:1 ratio based on OD₆₀₀. After five days, dw9509 tissue was collected, washed with sterile water, and nucleic acids were isolated for end-point PCR using RISA, SpLFY, and strain-specific primers (File S2). RISA PCR fingerprints from inoculated dw9509 samples were compared to DNA controls from dw9509 and bacteria alone. Wastewater: S = filter-sterilized wastewater not containing microbes, NS = non-sterile wastewater containing microbes; Bacteria: NB = axenic dw9509, +1A = dw9509 inoculated with Microbacterium sp. RU370.1, +3D = dw9509 inoculated with Bacillus sp. RU9509.4, +Sp245 = dw9509 inoculated with Azospirillum baldaniorum Sp245, +WW = dw9509 inoculated with non-sterile wastewater containing microbes; Primers: SpLFY = PCR using SpLFY-F and SpLFY-R primers specific to dw9509, RISA = PCR using 16S-1390f and 23S-e130r primers, DAB 1A = PCR using AmRU370.1-F and AmRU370.1-R primers specific to Microbacterium sp. RU370.1, DAB 3D = PCR using BsRU9509.4-F and BsRU9509.4-R primers specific to Bacillus sp. RU9509.4, Sp245 = PCR using AbSp245-F and AbSp245-R primers specific to Azospirillum baldaniorum Sp245.