Mass spectral imaging showing the plant growth-promoting rhizobacteria's effect on the Brachypodium awn

Plant growth-promoting rhizobacteria (PGPR) are a group of heterogeneous bacteria that colonize in the rhizosphere and they are well-known to enhance plant growth.¹ Bacteria such as Pseudomonas can function as efficient biofertilizers to promote seed germination, seedling vigor, and yields.^2,3 Soil salinization inhibits plant growth because of physiological changes caused by salt stress that lead to reduced nutrient uptake and growth. Arthrobacter sp., as a PGPR, has shown salinity tolerance on wheat growth; thus, it helps plant growth and nutrient uptake.⁴ However, the detailed interactive mechanism between plants, Gram-positive, or Gram-negative PGPRs at the biological interfaces and their effects on plant growth still remains speculative.^5,6

Brachypodium distachyon (Brachypodium) is an established and representative C3 plant model that is often used in studies to advance plant biology.⁷ It is also an excellent model plant for biomaterial production technology and engineering research due to its susceptibility to microbial interactions, minimal growth requirement, and short seed-to-seed life cycle with a fully sequenced genome.^8–10 Brachypodium has become increasingly prominent in studying the interactions between plants and microbes¹¹ because it provides insight into the biological studies of other types of major C3 grains like wheat and barley.¹² Therefore, Brachypodium is selected as an ideal model plant in this study.

The needlelike extensions that attach to the distal ends of Brachypodium seeds are called awns.¹³ Previous studies have reported that awns facilitate seed dispersal and germination—both are key stages of the plant life cycle.^14,15 During seed dispersal, awned seeds are buried deeper into the soil than awnless seeds, which promotes initial seed growth.¹⁶ However, reports on the effect of awn on germination behavior are still lacking.¹⁶ 

Mass spectral imaging (MSI) is widely used in surface and interface characterization.¹⁷ It provides molecular-level information on spatial distribution and chemical composition. Thus, MSI is an attractive tool in plant sciences and biomaterial characterization.¹⁸ Recent works have used MSI to study plant-omics.^19–23 Several MSI techniques are used in plant research, including matrix-assisted laser desorption ionization (MALDI) mass spectrometry, secondary ion mass spectrometry (SIMS), desorption electrospray ionization (DESI) mass spectrometry, and laser ablation electrospray ionization (LAESI) mass spectrometry. These techniques differ in ion source types, ionization, and ion detection.^24,25 MALDI permits the detection of larger ions such as peptides and proteins, and it has been used to analyze the distribution of metabolites, such as amino acids and phosphorylates with a spatial resolution of several micrometers (μm).^26,27 DESI and LAESI are used to capture the molecular information in plant leaves, albeit with much coarser spatial resolution at approximately 10 μm.^28,29 MALDI is a popular choice in plant and biomaterial characterization than DESI and SIMS. Although SIMS is not the most widely used MSI instrument in plant and biomaterial characterization, it has the potential to detect single cell interactions due to its superb lateral resolution.

Time-of-flight secondary ion mass spectrometry (ToF-SIMS), among the existing SIMS platforms, offers diverse measurement modalities, sensitivity, and parallel data collection compared to the other IMS techniques.^25,30,31 Specifically, ToF-SIMS has multiple measurement modes, providing full spectral (i.e., elemental, molecular, and isotopic) information and two-dimensional (2D) and three-dimensional chemical maps.^32–36 Sample preparation of ToF-SIMS is simpler and it is less matrix dependent than MALDI.^19,37,38 Specifically, molecular imaging of ToF-SIMS is attractive in studying biomaterials.^18,39,40 ToF-SIMS has been applied to investigate the distribution of chemicals in wood tissue.⁴¹ For example, several techniques were developed for SIMS sample preparation and analysis of corn seed and wood.^42,43 ToF-SIMS was used to identify small plant metabolites, such as cellulose, flavonoids, and other organics of biological materials including biomass.^44,45 Thus, ToF-SIMS is deemed an appealing approach to study biomaterials and cell-to-cell interactions due to its high spatial resolution and sensitive surface mapping.

In this study, the interaction of the Brachypodium awn after treatment with two PGPRs, Gram-negative Pseudomonas fluorescens SW25 (P.) and Gram-positive Arthrobacter chlorophenolicus A6 (A.),^46,47 respectively, were characterized using ToF-SIMS in different modalities. Because the awns have a bristlelike irregular shape with a millimeter dimension, we used a custom-made holder to immobilize them.⁴⁸ The delayed extraction mode was used to obtain both good spatial resolution and high mass resolution in plant imaging.^34,35,49 Secondary ions were extracted from the sample surface with a specific time delay to compensate for the effects of long primary ion beams to permit better resolution compared to the fast imaging mode.^33,35,49 High mass-resolution spectral analysis was used for peak identification to improve fidelity.

Our recent results demonstrated that ToF-SIMS can provide insight into metabolites and biomarkers in plant–PGPR interactions.⁵⁰ In this study, ToF-SIMS spectral results were used to characterize important plant metabolites and biomarkers including flavonoids, phenolic acids, fatty acids, and indole-3-acetic acid (IAA) on the Brachypodium awn surface. Additionally, spectral principal component analysis (PCA) was used with selected peaks to better evaluate the contribution of these compounds to the seedling. The delayed extraction mode measurements were used to validate that the Brachypodium awn surface was exposed to the PGPRs and they also captured the morphological features of the plant–bacteria interface. The novelty of using ToF-SIMS lies in the mapping of molecules indicative of the PGPR effects on the plant surfaces. We used the awns from a well-established C3 plant Brachypodium to demonstrate how ToF-SIMS can be used to study the PGPRs and awn interactions and their impact on the growth potential at the biointerphase.

In this study, the Arthrobacter and Pseudomonas planktonic cells and biofilms were used as bacterial controls. Two kanamycin resistant types, Gram-positive Arthrobacter chlorophenolicus A6 (abbreviated as A.) and Gram-negative Pseudomonas fluorescens SBW25 (abbreviated as P.) strains were acquired from previous studies.^46,47 Both bacteria were green fluorescent protein-tagged. First, bacteria were cultured on Luria Broth (LB) agar plates inside of the 30 °C incubator for 24 h. Then, 1 ml LB kanamycin medium was injected into a 2 ml centrifuge tube. Afterward, an individual colony was selected from the agar plate and transferred into the centrifuge tube. A separate sterile syringe was used to extract the mixed liquid from the centrifuge tube and it was transferred to the serum bottles. The serum bottles were brought to the orbital shaker (INNOVA 2300, New Brunswick Co., Inc.) that was set at 150 rpm at 30 °C for growth. A bubble trap was prepared to assist culturing of the bacterial biofilm in the sterilized system for analysis at the liquid vacuum interface (SALVI) microfluidic microchannel as described in previous studies.^51–53 9 ml of LB kanamycin medium in a 10 ml syringe was connected to a polyetheretherketone fitting on the microfluidic system. The flow rate was set as 20 μl/min to allow the medium to flow across the microchannel overnight. The cultures were inoculated inside the SALVI microchannel when bacteria in the serum bottles grew to the log phase up to 24 h according to their growth curves shown in Fig. S1 in the supplementary material,⁸⁰ and the flow rate was set to 2 μl/min for biofilm culturing.⁵⁴ The visible biofilm growth was observed in the microchannel after culturing for 5–6 days.

For planktonic bacteria samples, the active P. and A. were inoculated into 50 ml triangular flasks with 5 ml LB medium and cultured in a 30 °C incubator. It would take 1–2 days for the cells to reach an optical density (OD₆₀₀) of approximately 0.6. Then, planktonic cells were harvested by centrifugation (5430, Eppendorf) for 2 min at 734 g. After centrifuging, the supernatant was discarded prior to adding 1 ml of sterile de-ionized (DI) water to resuspend the precipitated bacteria for sample desalination. This step was repeated three times and the final 20 μl suspension was extracted by resuspending the bacteria with 200 μl of sterile DI water. All samples were air dried in the fume hood and deposited on a clean silicon (Si) wafer (10 × 10 mm² diced, Ted Pella Inc.) before ToF-SIMS analysis.⁵⁵ 

Brachypodium seeds (psb00001, Riken BRC Experimental Plant Division) were used as received. The awns were gently removed with forceps. Awns washed with DI water and nitrogen dried without PGPR exposure are referred to as the control and abbreviated as the DI control awn below. For the PGPR-treated awns, P. treated awns were soaked in DI water in a beaker followed by the inoculation of 400 μl P. bacteria in a 40 ml medium solution. Similarly, A. treated awns were prepared by inoculating the soaked awns in the DI water with A. bacteria following the same procedure. All beakers were covered with Parafilm® and placed at 25 °C room temperature without direct sunlight for 24 h. The DI control and PGPR-treated awns were nitrogen gas dried and cut into segments before getting secured in the sample holder. Figure 1(a) illustrates ToF-SIMS analysis in the Brachypodium awn. Figure 1(b) shows a representative 2D image of the P. awn acquired using the delayed image extraction mode, and Fig. 1(c) presents a representative SIMS spectrum of the P. awn collected in the static SIMS spectral mode.

High mass-resolution static SIMS spectral mode and delayed extraction imaging analyses were performed using an IONTOF ToF-SIMS V spectrometer (IONTOF GmbH, Münster, Germany). The primary cluster ion beam was a 25 keV Bi₃⁺ with a current of 0.56 pA at 10 kHz pulse energy. During the static SIMS analysis, the pressure in the main chamber was maintained at ∼8 × 10⁻⁹ mbar. Spectra of awns were acquired by rastering the primary ion beam over an area of 100 × 100 μm² for 100 scans with 128 × 128-pixel resolution. The delayed extraction imaging mode also used Bi₃⁺ primary cluster ion beam with a spatial resolution of approximately 0.4 μm. The current was 0.36 pA (150 ns pulse width) at a 10 kHz frequency. The size of the scanning area was 150 × 150 μm² in the awn segment of seeds for 50 scans with a pixel resolution of 256 × 256 in the delayed extraction mode. The pressure of the main chamber was maintained at ∼2.5 × 10⁻⁸ mbar during imaging analysis.

ToF-SIMS data were analyzed using the IONTOF Surface Lab 6.3 software and four positive and negative data points for each sample were collected. Mass spectra were calibrated using O⁻ (m/z⁻ 16), C₁₁H₁₆O⁻ (m/z⁻ 164), and C₂₀H₂₉O₄⁻ (m/z⁻ 333) peaks in the negative mode and CH₂⁺ (m/z⁺ 14), C₁₀H₈O₃⁺ (m/z⁺ 176), and C₂₃H₂₁⁺ (m/z⁺ 297) peaks in the positive mode, respectively.⁵⁶ Calibrated SIMS data were exported to Origin Pro 2019b for spectral plotting. Figures S2 and S3 in the supplementary material illustrate spectral results of the DI control awn and the results show good analysis reproducibility.⁸⁰ 

Spectral PCA was performed using unit mass in Matlab (R2018b).⁵⁷ In the first round of PCA, we analyzed all peaks in both the negative and positive modes. Inorganic peaks in the lower mass range of m/z ≤ 100 presented matrix effects from the growth media. Thus, we focused on the important metabolites between m/z 100 and 400 to conduct the second round of selected peak spectral PCA. The prominent peaks revealed in all peak spectral PCA results were included in the selected peak lists. The mass calibrated data were pretreated by mean centering, normalization to the total ion intensity of selected peaks, and square root transformation before running spectral PCA.^58,59

Figure 2 presents the high mass-resolution normalized spectral result of the DI control and PGPR-treated Brachypodium awns in the range of m/z⁻ 100–400 in the negative mode. Additional SIMS spectral comparison in the range of m/z⁻ 0–100 is provided in Fig. S4 in the supplementary material.⁸⁰ Flavonoids are important components in plants: they function as eco-sensing signals and protectants in the early seedling process.⁶⁰ In our SIMS spectra, these chemicals are observed in both the DI control and PGPR-treated awns, which are marked with triangles (Fig. 2). The fragments of quercetin^{[1, 2B]–} (m/z⁻ 121, C₇H₅O₂⁻) and galangin (m/z⁻ 183, C₁₂H₇O₂⁻) are prominent among all samples. Besides, the galangin fragments have much higher intensities in the PGPR-treated awns than those in the DI awn, indicating that this flavonoid is active in the plant–PGPR interactions. Specifically, the kaempferol fragment (m/z⁻ 285, C₁₅H₉O₆⁻) and quercetin fragment (m/z⁻ 301, C₁₅H₉O₇⁻) exist only in the P. awn, which suggests that these flavonoids are more active against the pathogenic Gram-negative P. infection.⁶¹ 

Phenolic acids are another major type of metabolite in plants. Phenolic acids and their derivatives play a key role in the rhizospheric plant-microbe interactions and they emerge rapidly during seed germination.^62–64 ToF-SIMS captured these components, marked with asterisks, in the awns with different PGPR treatments in the mass range of m/z⁻ < 200 (Fig. 2). For example, p-hydroxybenzoic acid (m/z⁻ 137, C₇H₅O₃⁻), cinnamic acid (m/z⁻ 147, C₉H₇O₂⁻), vanillin (m/z⁻ 151, C₈H₇O₃⁻), protocatechuic acid (m/z⁻ 153, C₇H₅O₄⁻), p-coumaric acid (m/z⁻ 163, C₆H₁₁O₅⁻), and gallic acid (m/z⁻ 169, C₇H₅O₅⁻) are observed in the awns. The spectral results suggest that inoculating PGPRs to the C3 model plant awn alters the intensities of these plant metabolites.

Fatty acid peaks are dominant and marked with circles in the range of m/z⁻ 200–400 (Fig. 2). Fatty acids are fundamental chemicals in plant membrane lipids.⁶⁵ Fatty acid peaks have long been identified and well documented using ToF-SIMS.^66,67 Herein, some fatty acids are detected among all the awn samples, such as myristic acid (m/z⁻ 223, C₁₄H₂₃O₂⁻), palmitic acid (m/z⁻ 255, C₁₆H₃₁O₂⁻), arachidic acid (m/z⁻ 311, C₂₀H₃₉O₂⁻), and docosanoic acid (m/z⁻ 339, C₂₂H₄₃O₂⁻). Stearic acid (m/z⁻ 283, C₁₈H₃₅O₂⁻) and cerotic acid (m/z⁻ 395, C₂₆H₅₁O₂⁻) are observed in the A. awn and DI control awn, while heneicosanoic acid (m/z⁻ 325, C₂₁H₄₁O₂⁻) exists only in the PGPR-treated awns. Spectral comparison among the DI control and PGPR-treated awns in the mass range of m/z⁺ 100–400 in the positive ion mode is depicted in Fig. S5 in the supplementary material.⁸⁰ Looking into the positive mode results, palmitic acid (m/z⁺ 257, C₁₆H₃₃O₂⁺), stearic acid (m/z⁺ 285, C₁₈H₃₇O₂⁺), arachidic acid (m/z⁺ 313, C₂₀H₄₁O₂⁺), and docosanoic acid (m/z⁺ 341, C₂₂H₄₅O₂⁺) are conspicuously detected in the A. and DI control awn.

It has been reported that certain types of PGPRs promote seed germination by producing indole-3-acetic acid (IAA, C₁₀H₉NO₂⁺), a naturally occurring plant growth promotion hormone.^68,69 Although the intensity of m/z⁺ 175 is relatively low (Fig. S5 in the supplementary material),⁸⁰ the loading of this peak is prominent in the selected peak spectral PCA results in the positive mode (Fig. S11 in the supplementary material).⁸⁰ Our results show that IAA exists in both the PGPR-treated awns and that the intensity is relatively high in the P. awn. This finding gives evidence that IAA is generated to assist subsequent seedling growth after the awn is exposed to the PGPRs. The total ion intensity spectral comparisons of the PGPR bacterial controls, nontreated dry awn, DI control awn, and PGPR-treated awns are presented in Figs. S6 and S7 in the supplementary material.⁸⁰ Possible peak identification is summarized in Tables S1 and S2 in the supplementary material.⁸⁰ 

Spectral PCA analysis is conducted to evaluate the metabolites’ contribution to seedling potential in the awn section. In the first round, all peak spectral PCA scores and loadings result in both the negative and positive modes and are presented in Figs. S8 and S9 in the supplementary material,⁸⁰ respectively. Based on the peak selection criteria described in the experimental section, the second round of selected peak spectral PCA was performed to study the effects of the Gram-positive and Gram-negative PGPR on awns with different treatments and their corresponding biofilms and planktonic cells as bacteria controls. PCA results in the negative mode are depicted in Figs. 3 and S10 in the supplementary material.⁸⁰ Figure 3(a) shows the spectral PCA scores plot. PC1 explains 60.1% of all data and mainly separates the DI control awn and PGPR-treated awns from the PGPR bacterial controls. PC2 explains 17.5% of data and it shows overlaps between the DI control and A. treated awns. PC3 explains 8.4% of data and separates the P. awn, A. planktonic cells, and P. biofilm with other specimens. These samples are well separated because of the difference in their surfaces.

The PCA score results [Fig. 3(a)] suggest that both the Gram-negative and Gram-positive bacterial control affect the Brachypodium awns. In addition, P. planktonic cells and A. biofilms unfold a stronger affiliation to the awns. In the corresponding loading results [Figs. 3(b)–3(d)], the peaks are labeled with their center masses. The peaks corresponding to flavonoids, phenolic acids, and fatty acids are marked with triangles, asterisks, and circles, respectively. Fatty acid peaks are dominant in the PC1, PC2, and PC3 loadings. This is expected because fatty acids not only play an important role in the bacterial biofilms but also form plant membrane lipids.^65,70 Palmitic acid (m/z⁻ 237, C₁₆H₂₉O⁻ and m/z⁻ 255, C₁₆H₃₁O₂⁻), margaric acid (m/z⁻ 265, C₁₇H₂₉O₂⁻), oleic acid (m/z⁻ 281, C₁₈H₃₃O₂⁻), stearic acid (m/z⁻ 283, C₁₈H₃₅O₂⁻), arachidic acid (m/z⁻ 311, C₂₀H₃₉O₂⁻), docosanoic acid (m/z⁻ 339, C₂₂H₄₃O₂⁻), and cerotic acid (m/z⁻ 395, C₂₆H₅₁O₂⁻) are important contributors to the PC1 negative and PC2 positive loadings. Combing with the scores plots, the sources of these fatty acids can be traced back to the DI and PGPR-treated awns. Fatty acids and lipids exist in plant seeds as a significant energy storage substance. In addition, their accumulation is closely tied with seed development.⁷¹ An earlier study investigated the changes in the composition of fatty acids in the safflower cultivars and showed that the saturated and unsaturated fatty acid contents fluctuated with seed growth and development.⁷² Our results provide fatty acids characterization in the PGPR-treated awns that are relevant to the seed growth. The sensitivity of ToF-SIMS makes it possible to detect these key components such as fatty acids and lipids inherent in awns.

The quercetin^{[1, 2B]–} fragment (m/z⁻ 121, C₇H₅O₂⁻) and galangin fragment (m/z⁻ 183, C₁₂H₇O₂⁻) contribute to the PC1 negative loadings, which indicates that these two compounds exist in both the DI control and PGPR-treated awns, when combing results from the scores plots. It is important to note that the galangin fragment (m/z⁻ 183, C₁₂H₇O₂⁻) is also a major contributor in the PC3 positive loadings. Referring to the scores plots, galangin tends to accumulate in both the A. and P. awns. This finding is consistent with the spectral result (Fig. 2), that is, the galangin fragment (m/z⁻ 183, C₁₂H₇O₂⁻) is present in the PGPR-treated awns. Another quercetin fragment (m/z⁻ 301, C₁₅H₉O₇⁻) contributes to the PC2 negative loadings and it is related to the P. awn from the scores plots. This result is also in agreement with the spectral result (Fig. 2), which suggests that the quercetin fragment (m/z⁻ 301, C₁₅H₉O₇⁻) is notable in the awn after exposure to the P. strain. Flavonoids are known to have antimicrobial properties against both Gram-positive and Gram-negative bacteria.⁷³ Our new finding gives additional evidence that the flavonoid responds differentially between Gram-negative and Gram-positive PGPR. In addition, quercetin and galangin show inhibitory effects on seed germination;^74,75 our data reasonably imply that the inhibition might be dependent on the PGPR strains.

Peaks of some phenolic acids are outstanding in the PC3 negative loadings. Bar plots and 2D overlay images are used to distinguish the accumulations of some of the significant phenolic acids in the negative mode. Figure 4(a) presents the normalized peak intensities of the DI control, P.-, and A. awns, respectively. Most of the phenolic acid compounds, such as p-hydroxybenzoic acid (m/z⁻ 137, C₇H₅O₃⁻), cinnamic acid (m/z⁻ 147, C₉H₇O₂⁻), and gallic acid (m/z⁻ 169, C₇H₅O₅⁻), accumulate more in the DI control awn than in the PGPR-treated awns. Gallic acid (m/z⁻ 169, C₇H₅O₅⁻) decreases remarkably in the P. awn; while p-coumaric acid (m/z⁻ 163, C₆H₁₁O₅⁻) increases in the A. awn. The protocatechuic acid (m/z⁻ 153, C₇H₅O₄⁻) storage slightly fluctuates in different PGPR awns. These polyphenolic compounds are seedling growth inhibitors, and they are known to exist widely in plants and fruits.^75,76 The results in the comparison bar plot indicate that seed germination inhibitory activities of phenolic acids are sensitive to the PGPR treatments. The 2D overlay images of selected ions from the awns are depicted in Figs. 4(b)–4(d). The p-hydroxybenzoic acid (m/z⁻ 137, C₇H₅O₃⁻), p-coumaric acid (m/z⁻ 163, C₆H₁₁O₅⁻), and gallic acid (m/z⁻ 169, C₇H₅O₅⁻) are colored in blue, red, and green, respectively. The distributions of these inhibitors show that these three phenolic acids accumulate more in the DI control awn. This finding further suggests that the growth inhibitory effect of PGPR-treated awns is lower than that of the DI awn as a control.

Figure S11 in the supplementary material⁸⁰ shows selected peak spectral PCA results in the positive mode. PC1 captures 73.5% of the variance and distinguishes the DI control awn from the PGPR-treated awns and the bacteria controls. PC2 captures 16.5% of the variance and distinguishes the A. awn, A. planktonic cells, DI control awn, and P. biofilm from the other samples. The IAA peak (m/z⁺ 175, C₁₀H₉NO₂⁺) is labeled in pink and it contributes to the PC1 negative and PC2 positive loadings. As a vital phytohormone, IAA can be produced by root colonizing bacteria. IAA plays a vital role in seed germination regulation and plant host protection from environmental stresses.^77,78 The SIMS spectral and PCA results verify IAA occurrence in the awns exposed to the PGPRs.

It is challenging to characterize an awn—due to its tiny and uneven physical features—using the high lateral resolution imaging mode in SIMS because it requires flat and even samples.^18,79 Thus, the delayed extraction imaging mode in ToF-SIMS was used to achieve a balance between the mass and spatial resolutions, simultaneously.^35,49 The delayed extraction mode helps to capture awn molecular and morphology information. Figures 5(a)–5(d) illustrate the 2D total ion images of the PGPR-treated awns in both the negative and positive modes. The DI control awn images are depicted in Figs. S12(a) and S12(b) in the supplementary material.⁸⁰ These images were acquired by scanning an area of 150 × 150 μm². The color bars beside each plot represent the normalized ion counts and the dark blue to white color scale corresponds to the ion intensities that increase from low to high. The surfaces of the seeds of different treatments show differences. The edges of the P. awn are not as rough as the DI control awn, instead, they have a smoother looking surface. Fiberlike structures on the P. awn have high ion counts. In contrast, the A. awn does not seem to have the same surface roughness.

Figures 5(e), 5(f), and S12(c) in the supplementary material⁸⁰ exhibit the SIMS spectral results from the delayed extraction negative mode, and they provide similar observations compared to the spectral results acquired from the high mass accuracy spectral mode (Fig. 2). Taking the distribution of fatty acids as an example, myristic acid (m/z⁻ 223, C₁₄H₂₃O₂⁻), palmitic acid (m/z⁻ 255, C₁₆H₃₁O₂⁻), stearic acid (m/z⁻ 283, C₁₈H₃₅O₂⁻), docosanoic acid (m/z⁻ 339, C₂₂H₄₃O₂⁻), and cerotic acid (m/z⁻ 395, C₂₆H₅₁O₂⁻) are prominent in the A. awn and DI control awn, while arachidic acid (m/z⁻ 311, C₂₀H₃₉O₂⁻) and heneicosanoic acid (m/z⁻ 325, C₂₁H₄₁O₂⁻) have significant counts in the P. awn. Figures 5(g), 5(h), and S12(d) in the supplementary material⁸⁰ show correlative spectral results acquired from the delayed extraction positive mode. Fatty acid peaks, such as palmitic acid (m/z⁺ 257, C₁₆H₃₃O₂⁺), stearic acid (m/z⁺ 285, C₁₈H₃₇O₂⁺), and arachidic acid (m/z⁺ 313, C₂₀H₄₁O₂⁺), are prominent in the DI control awn. In contrast, these compounds do not have as high counts in the PGPR-treated specimens.

Figures S13–S15 in the supplementary material⁸⁰ illustrate the comparison of 2D SIMS total ion and selected ion images of the Brachypodium awns in the negative ion mode, showing ToF-SIMS capability in single ion detection. Even though the intensities of the selected representative flavonoid and fatty acid ions are relatively low, their distribution is still captured in the different awns. The instrument setting could be optimized to improve the image results in the future. To better evaluate the delayed extraction mode measurements in the plant imaging, a model C4 plant Seteria leaf was used, and leaf images were also collected from a high lateral resolution imaging mode for comparison. These results are shown in Figs. S16 and S17 in the supplementary material,⁸⁰ respectively. The 2D images acquired from the high lateral resolution imaging mode offer sharper images, however, with a loss of mass accuracy. In comparison, results from the delayed image mode offer a better choice for providing both reasonable morphological characterization and spectral information, simultaneously.

The SIMS 2D image and spectral results enhance the characterization of plant metabolites to understand their functions concerning the effect of awn on germination potential. Our results show that ToF-SIMS can be a useful tool to study the PGPRs-plant interactions in the tiny awn biointerphase.

The PGPR-Brachypodium plant model was studied using ToF-SIMS in this work. SIMS high mass accuracy spectral and PCA results show that P. and A. as representative PGPRs change the awn surface composition differently. In addition, the delayed image extraction mode effectively acquires both morphological and chemical speciation changes as a result of PGPR interactions at the awn biointerface. Fatty acids, flavonoids, phenolic acids, and IAA are observed in both the control and PGPR-treated awns. Molecular evidence acquired from ToF-SIMS suggests the potential of these vital secondary plant metabolites and their roles in protecting the seeds against pathogens, inhibiting, or promoting functions in seeds’ maturation, and, possibly, their relevance to the seedling potential.

Compared to other MSI techniques, ToF-SIMS offers the advantage of detailed elemental, molecular, and isotopic information of the biomaterial surfaces and interfaces with high lateral resolutions (<1 μm) and high sensitivity at the microbial–plant biointerface. The application of multiple measurement modes offers rich data sets and confidence for the future studies of plants and microbes using mass spectra and imaging. Our results show that ToF-SIMS, as a powerful MSI technique, can be used to explore the relationship between plant metabolites and seed germination. ToF-SIMS is a promising tool to advance plant biology science and engineering in the future.

X.-Y. Yu acknowledges the Laboratory Directed Research and Development Program of Oak Ridge National Laboratory (ORNL) for partial support of this work. R. Komorek and X.-Y. Yu acknowledge the Environmental and Biological Science Directorate (EBSD) Mission seed Laboratory Directed Research and Development for the initial support of culture and experimental effort. Part of the experiment was covered by the Pacific Northwest National Laboratory (PNNL), W. R. Wiley Environmental Molecular Sciences Laboratory (EMSL) Strategic Science Area of Intercellular Thrust. The authors thank Jiyoung Son and Wen Liu for technical assistance in the sample analysis. A portion of the research was performed using EMSL (grid.436923.9), a DOE Office of Science User Facility sponsored by the Office of Biological and Environmental Research under the general EMSL user Proposal Nos. 50093 and 50170. ORNL is managed by UT-Battelle, LLC, for the U. S. Department of Energy (DOE) under Contract No. DE-AC05-00OR22725. PNNL is operated by Battelle for the DOE under Contract No. DE-AC05-76RL01830. Y. Zhang acknowledges the Chinese Science Council and the PNNL Alternative Sponsored Fellowship (ASF) for the graduate student fellowship. This manuscript has been authored in part by UT-Battelle, LLC, under Contract No. DE-AC05-00OR22725 with the U.S. Department of Energy (DOE). The publisher acknowledges the U.S. government license to provide public access under the DOE Public Access Plan (http://energy.gov/downloads/doe-public-access-plan).

The authors have no conflicts to disclose.

Ethics approval is not required in this work.

Yuchen Zhang: Formal analysis (lead); Investigation (supporting); Writing – original draft (lead); Writing – review & editing (equal). Rachel Komorek: Data curation (lead); Formal analysis (equal); Methodology (lead); Writing – review & editing (equal). Zihua Zhu: Data curation (supporting); Writing – review & editing (equal). Qiaoyun Huang: Writing – review & editing (equal). Wenli Chen: Writing – review & editing (equal). Janet Jansson: Writing – review & editing (equal). Christer Jansson: Project administration (equal); Writing – review & editing (equal). Xiao-Ying Yu: Conceptualization (lead); Data curation (equal); Formal analysis (equal); Funding acquisition (equal); Investigation (lead); Methodology (lead); Resources (lead); Supervision (lead); Writing – original draft (equal); Writing – review & editing (equal).

The raw and processed data that support the findings of this study are available from the corresponding author upon reasonable request.