Single-Cell Analysis of the Expression of Pseudomonas syringae Genes within the Plant Tissue

Summary

The present protocol describes a method that allows single-cell gene expression analysis on Pseudomonas syringae populations grown within the plant apoplast.

Abstract

A plethora of pathogenic microorganisms constantly attack plants. The Pseudomonas syringae species complex encompasses Gram-negative plant-pathogenic bacteria of special relevance for a wide number of hosts. P. syringae enters the plant from the leaf surface and multiplies rapidly within the apoplast, forming microcolonies that occupy the intercellular space. The constitutive expression of fluorescent proteins by the bacteria allows for visualization of the microcolonies and monitoring of the development of the infection at the microscopic level. Recent advances in single-cell analysis have revealed the large complexity reached by clonal isogenic bacterial populations. This complexity, referred to as phenotypic heterogeneity, is the consequence of cell-to-cell differences in gene expression (not linked to genetic differences) among the bacterial community. To analyze the expression of individual loci at the single-cell level, transcriptional fusions to fluorescent proteins have been widely used. Under stress conditions, such as those occurring during colonization of the plant apoplast, P. syringae differentiates into distinct subpopulations based on the heterogeneous expression of key virulence genes (i.e., the Hrp type III secretion system). However, single-cell analysis of any given P. syringae population recovered from plant tissue is challenging due to the cellular debris released during the mechanical disruption intrinsic to the inoculation and bacterial extraction processes. The present report details a method developed to monitor the expression of P. syringae genes of interest at the single-cell level during the colonization of Arabidopsis and bean plants. The preparation of the plants and the bacterial suspensions used for inoculation using a vacuum chamber are described. The recovery of endophytic bacteria from infected leaves by apoplastic fluid extraction is also explained here. Both the bacterial inoculation and bacterial extraction methods are empirically optimized to minimize plant and bacterial cell damage, resulting in bacterial preparations optimal for microscopy and flow cytometry analysis.

Introduction

Pathogenic bacteria display differences in diverse phenotypes, giving rise to the formation of subpopulations within genetically identical populations. This phenomenon is known as phenotypic heterogeneity and has been proposed as an adaptation strategy during bacterial-host interactions¹. Recent advances in the optical resolution of confocal microscopes, flow cytometry, and microfluidics, combined with fluorescent proteins, have fostered single-cell analyses of bacterial populations².

The Gram-negative Pseudomonas syringae is an archetypal plant pathogenic bacteria due to both its academic and economic importance³. The life cycle of P. syringae is linked to the water cycle⁴. P. syringae enters the intercellular spaces between the mesophyll cells, the plant leaf apoplast, through natural apertures such as stomata or wounds⁵. Once within the apoplast, P. syringae relies on the type III secretion system (T3SS) and the type III-translocated effectors (T3E) to suppress plant immunity and manipulate plant cellular functions for the benefit of the pathogen⁶. The expression of T3SS and T3E depends on the master regulator HrpL, an alternative sigma factor that binds to the hrp-box motifs in the promoter region of the target genes⁷.

By generating chromosome-located transcriptional fusions to fluorescent protein genes downstream of the gene of interest, one can monitor gene expression based on the fluorescence levels emitted at the single-cell level⁸. Using this method, it has been established that the expression of hrpL is heterogeneous both within bacterial cultures grown in the laboratory and within bacterial populations recovered from the plant apoplast^8,9. Although gene expression analysis at the single-cell level is typically performed in bacterial cultures grown in laboratory media, such analyses can also be carried out on bacterial populations growing within the plant, thus providing valuable information on the formation of subpopulations in the natural context. A potential limitation for the analysis of bacterial populations extracted from the plant is that classic inoculation methods by syringe-pressure infiltration into the apoplast, followed by bacterial extraction by maceration of the leaf tissue, typically generate a large amount of cellular plant debris that interferes with downstream analysis¹⁰. Most cellular debris consists of autofluorescent fragments of chloroplasts that overlap with GFP fluorescence, resulting in misleading results.

The present protocol describes the process of analyzing single-cell gene expression heterogeneity in two model pathosystems: the one formed by the P. syringae pv. tomato strain DC3000 and Arabidopsis thaliana (Col-0), and the other by the P. syringae pv. phaseolicola strain 1448A and bean plants (Phaseolus vulgaris cultivar Canadian Wonder). An inoculation method is proposed based on vacuum infiltration using a vacuum chamber and a pump, resulting in a fast and damage-free method to infiltrate whole leaves. Furthermore, as an improvement on conventional protocols, a gentler method is used to extract the bacterial population from the apoplast that significantly reduces tissue disruption, based on the extraction of apoplastic fluid by applying cycles of positive and negative pressure using a small amount of volume within a syringe.

Protocol

1. Plant preparation

1. Prepare Arabidopsis Col-0 plants following the steps below.
   1. Fill a 10 cm diameter pot with a 1:3 vermiculite-plant substrate mix (see Table of Materials), previously watered, and cover the pot with a 15 cm x 15 cm metal mesh with 1.6 mm x 1.6 mm holes. Adjust the metal mesh to the wet soil using a rubber band (Figure 1A).
   2. With a wet toothpick, sow Arabidopsis seeds into the holes of the metal mesh. Place three to four seeds in distant positions within the pot (Figure 1B, C).
   3. Cover the pots with a plastic dome to maintain high relative humidity and incubate them for 72 h at 4 °C for stratification.
      NOTE: Stratification (incubation at high humidity and low temperature, as described) improves the germination rate and synchrony of the seeds.
   4. Transfer the pots to a plant growth chamber under short-day conditions (8 h light/16 h dark at 21 °C, light intensity: 100 µmol·m⁻²·s⁻¹, relative humidity: 70%).
   5. After seed germination (8-10 days), use the tweezers to remove most of the seedlings, keeping one seedling in each of the positions of the pot (six seedlings/pot) (Figure 1D). Remove the plastic dome to uncover the pots.
      NOTE: The plants will be ready to use 4-5 weeks after germination.
2. Prepare Phaseolus vulgaris bean (cultivar Canadian Wonder) plants.
   1. Cover the bottom of a Petri dish with a wet piece of towel paper, and place the bean seeds on top of it. Seal the Petri dish with surgical tape and incubate at 28 °C for 3-4 days (Figure 2A).
   2. Transfer the germinated seeds into a 10 cm diameter pot filled with wet 1:3 vermiculite-plant substrate mix.
   3. Incubate in a plant growth chamber under long-day settings (16 h light/8 h dark at 23 °C, light intensity: 100 µmol·m⁻²·s⁻¹, relative humidity: 70%).
      NOTE: The plants will be ready to use 10 days after germination (Figure 2B).

2. Inoculation of Arabidopsis and bean plants

NOTE: In this study, the strains P. syringae pv. tomato DC3000 and P. syringae pv. phaseolicola 1448A were used.

1. Prepare the P. syringae inoculum.
   1. Streak out the P. syringae strain of interest from a −80 °C glycerol stock onto an LB plate (10 g/L tryptone, 5 g/L NaCl, 5 g/L yeast extract, and 16 g/L bacteriological agar, see Table of Materials) supplemented with the appropriate antibiotics. Incubate at 28 °C for 40-48 h.
      NOTE: The use of antibiotics is recommended if the strain of interest carries a plasmid or a genomic resistance gene. The recommended antibiotic concentrations for P. syringae are as follows: kanamycin (15 µg/mL), gentamycin (10 µg/mL), ampicillin (300 µg/mL) (see Table of Materials).
   2. Scrape out the bacterial biomass and resuspend in 5 mL of 10 mM MgCl₂. Measure the OD₆₀₀ and adjust to 0.1 by adding 10 mM MgCl₂.
      NOTE: An OD₆₀₀ of 0.1 of a P. syringae culture corresponds to 5 x 10⁷ CFU·mL⁻¹.
   3. Perform serial dilutions into 10 mM MgCl₂ to reach a final inoculum concentration of 5 x 10⁵ CFU·mL⁻¹. Prepare 200 mL of inoculum for the Arabidopsis plants and 50 mL for the bean plants.
   4. Right before inoculation, add the surfactant Silwett L-77 (see Table of Materials) to a final concentration of 0.02% for bean inoculation and 0.01% for Arabidopsis. Note that Silwett is somewhat detrimental to the Arabidopsis tissue.
2. Perform vacuum infiltration.
   1. For Arabidopsis infiltration, place two wood sticks forming an X over the pot (Figure 1E), and place the pot facing down over a 14 cm diameter Petri dish containing the 200 mL inoculum (Figure 1F).
   2. For bean leaves inoculation, introduce the leaf into a 50 mL conical centrifuge tube containing the inoculum (Figure 2C).
   3. Insert the plants immersed in the inoculum solution into a vacuum chamber (Figure 1G and Figure 2D) and give a pulse of 500 mbar for 30 s to infiltrate the leaves. Repeat the vacuum pulse 2-3 times until the leaf is completely infiltrated (Figure 1H and Figure 2E).
   4. Drain the excess inoculum solution with a piece of paper and return the plants to their corresponding growth chamber.

3. Extraction of bacteria from the apoplast

1. Four days after inoculation, cut either the aerial part of the Arabidopsis plant or the inoculated leaf from the bean plant and place it into a 20 mL syringe without a needle (Figure 2G). For bean leaves, roll the leaf on itself, leaving the abaxial face outward, as displayed in Figure 2F.
2. Add enough volume of distilled water to cover the tissue (usually 10-15 mL).
3. Insert the plunger, and with the syringe in the vertical position with the tip pointing up, remove the excess air and air bubbles inside the syringe by gently tapping the barrel until all the air is located near the tip. Slide then the plunger to take the air out. Once there is as little air as possible inside the syringe, cover the tip of the syringe barrel with paraffin film.
4. Carefully press the plunger to generate positive pressure until the tissue turns darker (Figure 1I and Figure 2H). Then, pull the plunger to generate negative pressure (Figure 1J and Figure 2I). Repeat this step 3-5 times.
5. Remove the paraffin film and the plunger and collect the fluid containing the apoplast-extracted bacteria, as in Figure 2J.

4. Single-cell analysis of the apoplast-extracted bacteria

1. Visualize by confocal microscopy following the steps below.
   1. Prepare a 1.5% agarose solution in distilled water. Once melted, add enough volume to fill the space between two microscopy slides set side by side and place another slide on top on it (Figure 3). Let them dry for 15 min and carefully remove the slide placed on the top. Using a blade, cut the agarose pad into 5 mm x 5 mm pieces right before use.
   2. Parallelly, centrifuge 1 mL of the apoplast-extracted bacteria at 12,000 x g for 1 min at room temperature, carefully remove the supernatant using a pipette, and resuspend the pellet into 20 µL of water to concentrate the cells. Place a 2 µL drop of the concentrated cells onto a 0.17 mm coverslip and cover the drop with a 5 mm x 5 mm piece of the agarose pad previously obtained in step 1, as is represented in Figure 3.
   3. Visualize the bacterial preparation under the confocal microscope (see Table of Materials). To identify green-fluorescent bacteria, use the excitation laser at 488 nm and an emission filter ranging from 500 nm to 550 nm. To identify all the bacteria, use the bright field and merge both fields.
   4. Process the confocal images using Fiji (see Table of Materials). To do this, use the MicrobeJ plugin to identify the contour of the bacterial cell and measure the fluorescence intensity within.
      NOTE: Image acquisition from isolated bacteria (not clustered) is recommended for this analysis.
2. Perform analysis by flow cytometry.
   1. Take an aliquot of the apoplast-extracted bacteria suspensions to analyze using the flow-cytometer. Acquire 100,000 events.
   2. To discriminate between bacteria and plant debris, analyze the apoplast extracted from a non-inoculated plant and compare the dot plot showing its forward scatter (FSC) cell size versus side scatter (SSC) cell size with that of the apoplast-extracted bacterial suspension. To identify non-fluorescent bacteria, use the apoplasts extracted from the plants inoculated with non-fluorescent isogenic bacteria and compare their fluorescent emissions.

Representative Results

The expression of the type III secretion system is essential for bacterial growth within the plant. The timely expression of T3SS genes is achieved through intricate regulation, at the center of which is the extracytoplasmic function (ECF) sigma factor HrpL, the key activator of the expression of T3SS-related genes¹¹. An analysis of the expression of hrpL was previously carried out using a chromosome-located transcriptional fusion to a downstream promoterless gfp gene and by following the expression patterns by confocal microscopy and flow cytometry⁹. These fusions are generated through homologous recombination-mediated integration of plasmid-encoded constructs generated through PCR and traditional cloning and carry the last 500 base pairs of the ORF of the gene of interest (including the STOP codon) and 500 base pairs of the sequence immediately downstream, flanking the ribosome-binding site and the ORF of the fluorescence reporter gene to be used (in this case, GFP), followed by an antibiotic resistance cassette (in this case, the NPT2 gene conferring resistance to kanamycin). Thus, it was established that apoplast-extracted P. syringae populations expressing T3SS-related genes, including hrpL, are heterogeneous in planta⁹. On this precedent, the representative results of confocal fluorescence microscopy and flow cytometry analysis of hrpL expression are shown in individual bacterial cells of the model strains Pph 1448A (Pph) and Pto DC3000 (Pto), extracted from the leaves of either bean or Arabidopsis, respectively, after colonization of the leaf apoplast (Figure 4). In each strain, the P. syringae chromosomal hrpL gene carries a downstream transcriptional fusion to a promoterless gfp gene that allows for monitoring the hrpL native promoter activity by following the GFP expression levels^8,9.

Apoplast-extracted bacteria can be used for different analyses. Here, 300 µL of the above-mentioned reporter bacterial strains were extracted from the apoplast of infected Arabidopsis or bean plants and used for data acquisition using a flow cytometer system (see Table of Materials). Laser emission at 488 nm and the FITC filter was used for GFP detection. Flow cytometry analysis shows the fluorescence distribution in individual bacterial cells within the population. Heterogeneous expression of hrpL can be clearly observed by flow cytometry in Arabidopsis apoplast-extracted Pto and in bean apoplast-extracted Pph, including HrpL^OFF (not expressing GFP) bacteria, and these results are supported by microscopic examination of the corresponding samples (Figure 4). The dot plot graphs (left panels) show the distribution of fluorescence intensity of GFP versus cell size in the population (Figure 4A), while the histograms show the fluorescence intensity of GFP versus cell counts (Figure 4B). These two different graphical representations of cytometry data allow the researcher to establish subtle visual nuances and provide complementary information. As a control for bacterial autofluorescence, the corresponding non-GFP wild-type strain was used (upper panel in Figure 4A and gray histogram in Figure 4B). This allows for identifying the graph area where non-GFP bacteria within the population are displayed. Flow cytometry analyses also generate quantitative data. As an example, the percentages of HrpL^ON (fluorescent) cells are extracted in apoplast-extracted Pto and Pph populations. Figure 4C shows how the HrpL^ON percentages were higher than the corresponding percentages of HrpL^OFF (non-fluorescent) cells in these populations, particularly in the Pto model. These data can also be used to calculate the average or median fluorescence for the entire population. In this particular experiment, the average GFP fluorescence intensity was obtained, which was higher for the Pto population than for Pph, in keeping with its higher percentage of ON cells. The average levels constitute population-level data comparable to data obtained through non-single cell techniques such as RT-qPCR or RNAseq. Finally, the robust coefficient of variation (RCV)¹², calculated as the third quartile minus the first quartile divided by the median, is also shown. The RCV is often used in flow cytometry studies to estimate data dispersion (i.e., heterogeneity of the expression within the population)¹³. In the current study, the RCV was slightly higher for Pph than for Pto, although the difference was not sufficient to characterize the distribution of the expression within the populations of these two strains as significantly different. The heterogeneity of hrpL expression can be visually confirmed with the confocal microscopy images (Figure 4D). The use of agar pads for microscopy preparation results in easier visualization and higher quality images of the individual cells since it pushes the bacteria onto a single layer and prevents bacterial movement. The inoculation and bacterial extraction procedures described in this protocol minimize the amount of plant debris within the bacterial preparation, thus allowing the analysis of bacterial expression by these different techniques (Figure 4).

Figure 1: Bacterial inoculation and apoplast extraction using Arabidopsis plants. (A) Preparation of the pot with the metal mesh properly attached. (B) Sowing seeds using a toothpick to distribute them as indicated in (C). (D) Arabidopsis plants ready for inoculation. (E) Two wooden sticks facilitate the immersion of the plants into the bacterial solution without reaching the pot or the soil (F). (G) The ensemble can be put inside the vacuum chamber. (H) The infiltrated leaves become darker after releasing the vacuum from the chamber. (I) The detached leaves are placed into a 20 mL syringe and covered with water. (J) Cycles of positive and negative pressure result in the extraction of the apoplastic bacteria. Please click here to view a larger version of this figure.

Figure 2: Bacterial inoculation and apoplast extraction using bean plants. (A) Germination of bean seeds in Petri dishes lined with wet toilet paper. (B) Bean plant ready for inoculation. (C) Bean leaf immersed in the bacterial solution contained in a 50 mL tube. (D) The leaves are inoculated inside a vacuum chamber until the tissue becomes darker (E). (F) The bean leaf is rolled, put inside a 20 mL syringe, and covered with water (G). (H) Cycles of positive and negative pressure (I) result in the extraction of the apoplastic bacteria that can be recovered into a tube (J). Please click here to view a larger version of this figure.

Figure 3: Schematic representation of the setting and preparation of the agar pad. Please click here to view a larger version of this figure.

Figure 4: Analysis of the apoplast-extracted bacteria obtained from Arabidopsis or bean leaves 4 days after inoculation with P. syringae pv. tomato (Pto) or P. syringae pv. phaseolicola (Pph), respectively. Expression of hrpL::gfp is monitored as GFP fluorescence. (A) Flow cytometry analysis is represented as a dot plot (forward scatter [cell size] vs. GFP fluorescence intensity). Non-GFP bacteria indicate a wild-type strain of either Pto or Pph. The vertical line delimits 99% of the non-GFP population. (B) Flow cytometry analysis is represented as histograms (cell counts vs. GFP fluorescence intensity). The grey histogram represents the non-GFP strain. (C) Quantitative data were generated from the flow cytometry analysis. The percentage of ON and OFF cells is calculated based on the distribution shown in the dot plot. The average indicates the mean fluorescence obtained in the experiment. The robust coefficient of variation (RCV) is calculated as the third quartile minus the first quartile divided by the median. (D) Fluorescence microscopy images showing heterogeneous levels of GFP associated with the expression of hrpL. The white arrows highlight bacteria displaying low levels or no GFP fluorescence. Scale bar: 3 µm. Please click here to view a larger version of this figure.

Discussion

The method presented here describes a non-invasive procedure that allows the infiltration of bacteria into the plant foliar tissue, allowing the rapid inoculation of large volumes while minimizing tissue disruption. One of the characteristics of the P. syringae species complex is the ability to survive and proliferate inside the plant apoplast and on the plant surface as epiphyte¹⁴. Thus, the possibility that the bacteria extracted using the present protocol come only from the plant apoplast cannot be ruled out. However, it was previously demonstrated using confocal microscopy that the proportion of bacteria on the leaf surface is notably minor compared to the bacterial growth inside the plant leaf⁵. This can be checked by microscopic examination of the leaf surface prior to extraction. Furthermore, the P. syringae strains used as a model in the present protocol, Pto DC3000 and Pph 1448A, have been shown to perform as weak epiphites¹⁵, representing an irrelevant proportion of the extracted bacteria. Nonetheless, if the method is to be adapted to other pathosystems, a prior step of leaf surface decontamination may have to be added to the present protocol.

Traditional inoculation methods, such as leaf infiltration using a syringe¹⁶, cause larger, hard-to-estimate tissue damage at the inoculation site, resulting in tissue necrosis. Moreover, the amount of damage is very variable between different inoculation points. Also, different plant species vary in their tolerance to leaf infiltration. For example, Arabidopsis leaves are easy to infiltrate, while bean leaves are more recalcitrant. Growth conditions also affect the level of tolerance to infiltration of a given species. More gentle traditional inoculation methods, such as leaf dip or leaf spray, resulting in a more natural and damage-free bacterial entry and colonization of the tissue, however, inherently limit the size of the bacterial populations obtained, frequently below the threshold required for subsequent analysis of the apoplastic bacterial samples⁵. The proposed inoculation method significantly reduces the activation of necrosis in the inoculated areas resulting from syringe pressure while avoiding bacterial sample size limitation and provides an easy, reproducible, and quick way to inoculate considerable leaf areas.

Some techniques requiring the inoculation of large tissue areas can be very time- and labor-consuming, increasing the chances of inflicting tissue damage. By applying this method, we can inoculate five or more Arabidopsis plants, or a whole bean leaf, in just one step without compromising tissue integrity. This technique may be adapted to inoculate a larger number of plants or to other plant species of agronomic or academic interest, such as Nicotiana benthamiana, tomato, or soybean. The concentration of the surfactant, used to reduce surface water tension to facilitate the infiltration of the bacterial suspension, must be adjusted and tested in advance depending on the plant species, as higher concentrations of the surfactant can result in tissue necrosis in some cases. Also, the amount of pressure applied and the duration of the inoculation process must be adjusted to the resistance offered by the plant leaf in different species to ensure optimal results.

With regard to apoplast extraction, the current method also minimizes tissue disruption, thus reducing the amount of plant debris present in the extracted sample. This allows for cleaner samples, resulting in more efficient analysis through different techniques such as RNA-seq, flow cytometry, or microscopy, as shown in the example. Recovery of the pathogenic bacterial population from its apoplastic environment with minimal tissue and plant cell disruption is challenging. P. syringae colonizes the intercellular spaces of the apoplast, forming microcolonies. Using an extracellular pathogen, such as P. syringae, allows for methods like the one presented here, since tissue and cell disruption is not required for bacterial recovery. The rounds of positive and negative pressure do disrupt the microcolonies, dispersing the bacteria and allowing their easy and rapid extraction through natural openings (stomata), avoiding any change in gene expression that may accompany long incubation times, such as those required for more gentle methods of apoplastic recovery. Microscopic examination of the leftover plant tissue does support the removal of most of the bacterial population. Apoplastic fluid extraction protocols have been widely used to study the complexity of the apoplastic fluid¹⁷. These protocols contain a final step of centrifugation to further clean up the apoplastic fluid recovered. Here, gentle cycles of positive and negative pressure using the syringe plunger were used instead, thus reducing sample contamination with host-cellular debris while having very little impact on the efficiency of bacterial recovery. One classic gentle bacterial extraction method consists of the incubation of leaf disks or seedlings with a surfactant for 1-2 h¹⁸. This method is not as efficient as the one presented here to extract bacteria; however, its principal limitation for gene expression analysis is the effect of the required incubation time on the expression profile of the population. The method hereby presented allows for the quick recovery and immediate analysis of the apoplastic population, thus reducing the risk of bypassing gene expression results at the cellular level.

Phenotypic heterogeneity in pathogenic bacteria has been widely studied using homogeneous laboratory media. Studying bacterial populations grown in their natural niche is often limited due to technical barriers, such as the difficulty of recovering bacterial populations without excess contamination with host cell debris that interferes with downstream single-cell analyses. This combined method of inoculation and extraction altogether allows the generation of large apoplastic populations of bacterial pathogens for single-cell analyses.

Materials

0.17 mm coverslip			No special requirements
1.6 x 1.6 mm metal mesh	Buzifu		Fiberglass screen mesh
10 cm diameter pots			No special requirements
140 mm Petri dishes			No special requirements
20 mL syringe			No special requirements
50 mL conical tubes	Sarstedt
Agarose	Merk
Ampicillin sodium	GoldBio
Bacteriological agar	Roko
Confocal Microscope Stellaris	Leica Microsystems
FACSVerse cell analyzer	BD Biosciences
Fiji software
Gentamycin sulfate	Duchefa	G-0124
Kanamycin monosulfate	Phytotechnology	K378
MgCl2	Merk
NaCl	Merk
Parafilm	Pechiney Plastic Packaging
Plant substrate			No special requirements
Silwet L-77	Cromton Europe Ltd
Toothpicks			No special requirements
Tryptone	Merk
Tweezers			No special requirements
Vacuum chamber 25 cm diameter	Kartell	554
Vacuum pump	GAST	DOA-P504-BN
Vermiculite			No special requirements
Yeast Extract	Merk