Microbiota of pest insect Nezara viridula mediate detoxification and plant defense repression

Abstract The Southern green shield bug, Nezara viridula, is an invasive piercing and sucking pest insect that feeds on crop plants and poses a threat to global food production. Given that insects are known to live in a close relationship with microorganisms, our study provides insights into the community composition and function of the N. viridula-associated microbiota and its effect on host–plant interactions. We discovered that N. viridula hosts both vertically and horizontally transmitted microbiota throughout different developmental stages and their salivary glands harbor a thriving microbial community that is transmitted to the plant while feeding. The N. viridula microbiota was shown to aid its host with the detoxification of a plant metabolite, namely 3-nitropropionic acid, and repression of host plant defenses. Our results demonstrate that the N. viridula-associated microbiota plays an important role in interactions between insects and plants and could therefore be considered a valuable target for the development of sustainable pest control strategies.


Introduction
Pest insects are among the major threats to global food production and cause significant crop losses [1][2][3][4].Despite current pest management strategies, a third of the annually produced crops are lost due to pest insects and insect-transmitted plant diseases, and climate change is predicted to further increase these losses [5][6][7][8][9].In addition, there is an increasing need for reducing pesticide usage to relieve environmental pollution and negative effects on non-target organisms.Together with a rapidly increasing human population, serious problems for food security are expected.Therefore, expanding our knowledge of how insects cope with plant defenses is essential for the development of sustainable pest control strategies [10,11].
To ward off insects, plants make use of structural barriers (e.g.trichomes) and chemical defenses (e.g.toxic glycosides) that are constitutively present or induced upon recognition of potential threats.In the case of herbivorous insects, the induction of attacker-specific plant defenses is triggered within seconds after wounding, leading to the production and accumulation of phytohormones and toxic secondary plant metabolites [12].Plant defensive phytohormones salicylic acid (SA) and jasmonic acid (JA) participate in defenses against biotic threats, often acting antagonistically [13].Piercing and sucking insects, such as those from the order of Hemiptera, usually induce either one or both phytohormones resulting in the expression of downstream signaling genes such as pathogenesis-related 1 (PR-1) in the case of SA, and lipoxygenase 2 (LOX2) and myeloblastosis (MYB) transcription factors (e.g.MYB28) in the case of JA [14][15][16][17][18][19][20].Defense signaling eventually leads to the production of secondary metabolite chemical defenses of which thousands are known, covering alkaloids, glucosinolates, polyphenols, and terpenes [21][22][23][24].
The Southern green shield bug, N. viridula (Pentatomidae: Hemiptera), is an invasive piercing and sucking pest insect that feeds on plant sap of plants, including crop species from the cruciferous, solanaceous, and fabaceous plant families that represent major food crops (e.g.tomato, beans [35]).As in other shield bugs, N. viridula has crypts in the M4 section of the midgut that carry species-specific symbionts that are vertically transmitted via egg surface contamination to the offspring [36,37].Geerinck et al. [38] revealed that the egg-associated community of N. viridula consists of mainly Gammaproteobacteria and some Bacilli, specifically a Pantoea-like symbiont, Sodalis sp., Serratia sp., Niallia sp., Staphylococcus sp., and Bacillus sp.strains and a study on adult N. viridula showed the dominance of Pantoea sp., Yokenella sp., and Enterococcus sp. in the midgut [32].N. viridula microbiota has been hypothesized to facilitate detoxification and repression of plant defenses, because insect-associated microbes transmitted while feeding causes vein necrosis in soybean seedlings [32,39].Likewise, other insects have been described to transmit their microbiota and redirect plant defenses to their benefit [40].In the case of the Colorado potato beetle (Leptinotarsa decemlineata), the transmission of microbiota during feeding represses plant JA-defenses directed against chewing insects.This repression is caused by microbial SA induction that represses the plant's JA-defensive pathway via crosstalk [41][42][43].
To further explore the involvement of insect-associated microorganisms in insect-plant interactions, we studied N. viridula-associated microbiota.The objective of this study is to characterize N. viridula core microbiota across different developmental stages and assess whether it supports plant defense repression and detoxification to the benefit of its host.To this end, we performed 16S rRNA gene and metagenome sequencing, microscopy of insect tissues, bacterial isolation, detoxification assays, and plant infection experiments, revealing that the N. viridula microbiota has a profound impact on plantinsect interactions.

Insect dissection
To determine the core microbiota of N. viridula with 16S rRNA gene amplicon sequencing, eggs and whole first until fourth instars were used.From the fifth instar onward, gut and salivary glands were dissected from insects and used either combined (for fifth instars) or separated for further analysis (performed in biological triplicate n = 3; an overview of the type of specimen used can be found in Supporting Information Fig. S1, Table S1).Insects were dissected directly after submersion in 70% ethanol for 1 min after which movement stopped.Dissection was performed under nonsterile conditions using a stereomicroscope, scalpel, scissor, and forceps.Separation of the complete gut system from the insect body was performed in phosphate-buffered saline solution (PBS; 137-mM NaCl, 2.7-mM KCl, 10-mM Na 2 HPO 4 , 1.8-mM KH 2 PO 4 , pH 7.4) to prevent rupture of the tissue.Both salivary glands and complete gut systems were disrupted by vigorously pipetting or vortexing in lysis buffer for DNA isolation or PBS for culturing and isolation.
Saliva, frass, egg, and phyllosphere collection N. viridula saliva was collected via a modified method as previously described [44].Brief ly, an artificial feeding solution was put onto a surface-sterilized watch glass, covered with parafilm allowing adult insect to feed exclusively on it for 3 days.Afterward, the solution was plated on Luria-Bertani (LB) agar (0.5% peptone, 0.3% yeast extract, 0.5% NaCl, and 1.5% agar; Supporting Information Methods).Insect frass was collected from adult insects by pipetting deposited frass droplets directly into an Eppendorf tube (Supporting Information Fig. S1).Sterilized egg clusters (collected from black nightshade plants) were obtained with a 70% ethanol wash for 30 s and left to dry before DNA isolation.Phyllosphere was collected by 30-min PBS wash of plant material.For detailed instructions, see the Supporting Information Methods.

Pure culture isolation and culturing of insect-associated microorganisms
Gut and salivary gland suspensions, saliva (i.e.feeding solution) and frass samples were diluted (10, 100, 1000, and 10 000 times) in PBS for CFU determination and plated on either LB or mannitol agar (2.5% n-mannitol, 0.5% yeast extract, 0.3% peptone, 1.5% agar) plates (Supporting Information Fig. S2).Culturing was performed under oxic conditions because anoxic culturing led to comparable results.Pantoea sp. was isolated from N. viridula gut systems on mannitol and LB agar after 2 days of incubation.Serratia sp. was isolated from frass with an overnight culture on LB agar.Sodalis sp. was isolated on LB agar from salivary glands samples and after 18-day growth at room temperature.Yeast was isolated from salivary glands by culturing on LB agar for 5 days.All isolated species were transferred to new LB agar plates at least six times before they were considered axenic.Bacterial density was determined with microscopy, using a Bürker-Türk counting chamber, and photometrically using a Cary 60 UV-Vis (Agilent) at 600 nm.

DNA isolation
Isolation of DNA for 16S rRNA gene and metagenome sequencing was performed using a DNeasy PowerSoil kit (QIAGEN, the Netherlands) including the optional heating step of the samples at 70 • C for 10 min in the Powerbead tubes.Tissue was directly put into the lysis buffer in the Powerbead tubes and tubes were vortexed for 10 min at 50 Hz using a TissueLyser LT (QIAGEN).DNA was eluted in 50-μl autoclaved ultrapure demineralized water and yield was measured with a Nanodrop1000 Spectrophotometer and a Qubit dsDNA HS assay kit (Thermo Fisher Scientific Inc., Waltham, USA).For detailed instructions, see the Supporting Information Methods.
Identification of pure cultures by 16S rRNA gene amplicon Sanger sequencing was performed by Baseclear (Netherlands) using the Bac341F and Bac806R primers and additional 18S rRNA EUKA21F and EUKB1791R primers [47]; Supporting Information Table S2).Raw Sanger sequencing data were analyzed using the Chromas software with a subsequent sequence alignment of forward and reverse reads (excluding 18S rRNA reads that did not overlap) using EMBOSS Needle Alignment and an NCBI nucleotide megablast to determine the identity of the microbial isolates.
Template DNA extracted from guts and salivary glands was used for metagenome sequencing to obtain metagenomeassembled genomes (MAGs) of the N. viridula core microbiota.DNA libraries were prepared with a Nextera XT Library Preparation Kit (Illumina) according to the manufacturer's instructions.The libraries were checked for quality and size distribution using an Agilent 2100 Bioanalyzer and a High Sensitivity DNA kit (Agilent Technologies, Santa Clara, California, USA).Quantitation of the libraries was performed with a Qubit dsDNA HS assay kit.Paired-end sequencing (2× 300 bp) was performed using an MiSeq sequencer (Illumina) and a MiSeq Reagent Kit v3 (Illumina), according to the manufacturer's protocol.The quality of Illumina paired-end genomic sequencing data were assessed according to our analysis pipeline (Supplementary Methods).Out of 78 373 844 reads, 64 920 532 passed filtering and were used for downstream analyses.

16S rRNA gene amplicon sequencing analysis
16S rRNA gene amplicon data was analyzed with R using the "DADA2" pipeline and the "SILVA" database (v138) for assigning taxonomy [48].Raw forward reads were trimmed by removing the eight nucleotides from the 5'end, removing 40 nucleotides from the 3'end, an additional 3 end quality threshold of 20 and an average quality threshold of 20 using "FastqCleaner" [49], leaving 2 984 222 reads after removal of plant chlorophyll sequences.The total number of microbial reads that exceeded >10 000 reads in all insect-related samples (excluding plant samples) were considered the core microbiome, covering 89% (2 650 717) of all reads.Relative abundance within each three biological replicates was calculated as a percentage of the total number of reads (Supplemental Fig. S3 and Tables S2 and S3).A heatmap of the average abundance of three biological replicates was produced using the "R" "pheatmap" package.The Shannon index for all amplicon sequence variants (ASVs) and for the core set of microbiota was calculated accordingly [50].

Fluorescence in situ hybridization
Isolated gut and salivary glands were hybridized with a f luoresceinlabeled general bacterial probe (Eub-mix [67,68]) and a Cy5labeled Gammaproteobacterial probe (GAM42A [69]) along with a GAM42A competitor probe.An additional Cy3-labeled probe specific for Sodalis sp. was used for targeted detection (Sod1238R [70]).Samples were visualized using laser scanning microscopy.For a detailed protocol, see Supporting Information Methods.

Quantification of NPA, nitrite, and nitrate
High-performance liquid chromatography (HPLC) was used to determine NPA concentrations.Using an Agilent 1100 system equipped with a diode array detector and a Merck C-18 column (Lichrospher 100 RP-18 end-capped [5 μm] column, 250 mm × 4.6 mm), isocratic analysis was performed with 100% 0.1% ortho-phosphoric acid in water with a f low rate of 1.2 ml min −1 .Before the analysis, 200 μl of supernatant was acidified with 25-μl 1M sulfuric acid, and 100 μl of the sample was injected.NPA was measured at 210 nm with a retention time of 5.2 min.
To determine nitrite (NO

Arabidopsis inoculation and insect feeding assay
One leaf per mature 5-week-old Arabidopsis thaliana plant (Supporting Information Methods) was inoculated with a 5-μl droplet of 1 × 10 8 CFU/ml bacterial suspension in 10-mM MgSO 4 , followed by a puncture with a sterile 0.4-mm needle through the inoculum into the leaf.Per treatment, three to six plants were inoculated or exposed to insect feeding.After 24 and 72 h, plant leaves were harvested and directly put into liquid nitrogen after which they were stored at −70

RNA extraction and cDNA synthesis
Plant RNA extractions were performed using a RNeasy Plant Minikit (Qiagen) according to the manufacturer, with the alteration that frozen plant material was ground with a micropestle directly in the extraction buffer.∼150 ng/μl of RNA per sample was used for cDNA synthesis using a QuantiTect Reverse Transcription Kit (Qiagen) according to the manufacturer's protocol with the prolonged incubation.RNA and cDNA qualities were determined with a Nanodrop.

N. viridula absolute microbial abundance
Real-time quantitative PCR (RT-qPCR) was performed to measure the absolute abundance of bacteria within samples.To quantify all bacteria, 16S rRNA gene primer pair Bac341F and Bac806R, amplifying V3-V4 regions was used.To determine the abundance of Pantoea sp., Sodalis sp., Serratia sp., and Commensalibacter sp., specific primer sets targeting single-copy groL and rpoB marker genes were designed (PantF, PantR.SodF, SodR, SerF, SerR ComF, and ComR) based on the obtained MAGs from the metagenome analysis (Supporting Information Table S2).RT-qPCR was performed using a pipeline described in detail under the Supporting Information Methods.

Real-time quantitative PCR of plant cDNA
To determine the relative expression of both plant SA-and JAdefense pathways and the specific activity of the aliphatic glucosinolate secondary metabolite pathway, A. thaliana gene primers of PR-1 (At2g14610), LOX2 (At3g45140), and MYB28 (At5g61420) were used (Supporting Information Table S2; [14,16,17,73]), comparing gene expression with the 2 -CT method to the housekeeping gene PP2AA3 [74,75].For detailed RT-qPCR methods, see the Supporting Information Methods.

Experimental approach
To characterize the core microbiota of N. viridula and subsequently unravel its role in insect-plant interactions we compared host plant leaf microbiota with insect eggs, gut systems, salivary glands, saliva, and frass samples using 16S rRNA gene amplicon sequencing (Fig. 1, Supporting Information Fig. S3).Fluorescence in situ hybridization (FISH) was performed to visually localize N. viridula symbionts.Metagenome analysis was performed on adult insect gut systems and salivary glands to unravel the microbial metabolic potential.Nezara viridula-associated microbes were isolated and tested in vitro for degradation of plant toxins.Lastly, plant inoculations using isolated pure cultures were performed to determine the effect of N. viridula microbiota on host plant defenses.

N. viridula core microbiota consists of two main symbionts
The microbiota of N. viridula was determined by 16S rRNA gene profiling of all developmental stages.Single plant phyllosphere samples of black nightshade, black mustard, and crown vetch were taken along as a background for transient microbiota from ingested food.Of all N. viridula microbiota, the most abundant reads were assigned to the genus Pantoea followed by the genera Sodalis, Serratia, Klebsiella, Commensalibacter, Pseudomonas, and Cutibacterium (Fig. 2, Supporting Information Figs S3 and S4A, and Tables S3 and S4).Plant material contained minor numbers of bacterial reads belonging to the genera found highly abundant in N. viridula; the high abundance of Pantoea sp. and Sodalis sp. in surface-sterilized eggs is particularly suggestive of vertical transmission as the analyzed eggs were collected from black nightshade that had only 0.1% Pantoea sp. and no Sodalis sp. in their phyllosphere samples (Supplemental Fig. S4B).This indicates that transient microbiota ingested via plant material is not responsible for the high numbers of bacteria found in N. viridula.Alpha diversity throughout the different developmental stages showed a steep and significant increase from egg to third instar, indicating that young insects obtained diverse microorganisms from their environment.From the 4th instar onward, a significant alpha diversity decline is seen that can be explained by the fact that insect guts and salivary glands were dissected from these larger animals, whereas younger instars were used as a whole.From fourth and fifth instar to adult another significant decrease in alpha diversity is seen, that can partially be explained by the change in sample type, in which the adult gut and salivary glands were separated (Supplemental Fig. S3C, D, F, and G).
As 16S rRNA gene amplicon sequencing does not account for the variation in 16S rRNA gene copy numbers, absolute abundances of Pantoea sp., Sodalis sp., Serratia sp., and Commensalibacter sp. were quantified in individual N. viridula adults, using RT-qPCR comparing 16S rRNA gene abundances with single-copy groL and rpoB genes (Fig. 3A).Similar to our amplicon sequencing results, Pantoea sp. was the most abundant genus, followed by Sodalis sp., Serratia sp., and Commensalibacter sp.Variation in the absolute abundance of bacterial genera between different insects was minimal, except for Serratia sp., indicating the stability of the N. viridula microbiota.To confirm that aforementioned species (Pantoea sp., Sodalis sp., Serratia sp., Commensalibacter sp.) are part of N. viridula core microbiota, 16S rRNA gene profiles obtained from adult insects collected from three other locations in the Netherlands were investigated and found to be in accordance with the microbial profile of our greenhouse-reared N. viridula (Fig. 3B).The Bleiswijk sample shows the largest deviation from the other locations, which can be explained by extensive inbreeding of this particular research facility rearing population.Overall, all gut samples taken at different locations showed high abundance of Pantoea sp. as found earlier.
To determine N. viridula microbiota throughout the different developmental stages, we monitored the changes in the relative abundance of the most abundant bacterial genera.Over 25% of the amplicon reads of N. viridula egg clusters belonged to the genus Pantoea, whereas Sodalis sp.represented over 73% (Fig. 2, Supporting Information Table S3).These numbers suggest that both Pantoea sp. and possibly also Sodalis sp. are vertically transmitted to N. viridula and therefore most likely fulfill essential symbiotic functions to their host.In addition, during the development of N. viridula from egg to adult, a clear increase in the relative abundance of Pantoea sp. was observed.Although Sodalis sp.seemed to decrease during development, it could be found highly abundant in salivary glands, indicating that it serves a different role than Pantoea sp., which is predominantly found in the gut system (Figs 2 and 3b).All other genera showed a lower abundance throughout N. viridula's development.Our data demonstrate that Serratia sp. is obtained by young first instar animals from the environment, and because its abundance in N. viridula is higher than that on plants, it seems N. viridula also supports the growth  Heatmap showing the relative abundance (in percentage) of the core genera (exceeding >10 000 reads in all insect-related samples, covering 89% (2650717) of all reads) within each sample type consisting of three biological replicates for N. viridula samples and one replicate for host plant samples (39 samples in total; Supporting Information Table S1). of this microorganism.Eventually, Serratia sp.declined throughout the development of N. viridula, suggesting it is eliminated via an unknown mechanism or resides in organs that were not analyzed.Klebsiella sp. was absent from the first instar and found in low abundance on the second instar onward and is, therefore, most likely acquired from the environment.For Commensalibacter sp., a similar lack of developmental pattern is observed with a small increase in abundance during the third and fourth instar stage.Pseudomonas sp. and Cutibacterium sp. were detected in low abundance in both plants and N. viridula, suggesting that these genera are transferred via the environment or feeding.The low abundance of Pseudomonas sp. in the phyllosphere is probably due to the sampling of young plants grown under controlled greenhouse conditions.Our analysis also revealed that there is no clear difference in the microbiota profiles of adult males and females.Overall, our 16S rRNA gene sequencing results show that the N. viridula core microbiota is limited to two symbionts, namely, Pantoea sp. and Sodalis sp., and a limited consortium of Serratia sp., Klebsiella sp., and Commensalibacter sp. with yet-unclear relationships to their host.

N. viridula core microbiota is transmitted from insects to plants via frass and saliva
16S rRNA gene profiling of frass revealed that it contained ∼10% Pantoea sp., 33% Sodalis sp., 34% Serratia sp., and 6% Klebsiella sp. (Fig. 2 and Supporting Information Table S3).Frass plated on LB agar resulted in ∼1.8 × 10 8 CFU/ml within a 24-h incubation at room temperature (Supporting Information Figs S5 and S6).The genera Sodalis and Serratia were present in a relatively high abundance in frass, even though the gut system of the adult animals contained only low abundances of these bacteria.This could be explained by other organs that potentially harbor these bacteria such as Malpighian tubules, which are the extensions of the distal part of the gut system, that are connected to the hindgut.
In contrast to earlier findings, indicating that N. viridula saliva obtained through ice exposure-induced salivation is sterile [76], our collected feeding solution offered to N. viridula contained excreted saliva with 13% Pantoea sp., 16% Sodalis sp., 5% Serratia sp. and 42% Klebsiella sp. (Fig. 2 and Supporting Information Table S3).Control feeding solution systems that were inaccessible to N. viridula remained sterile, suggesting that N. viridula transmits its associated microbiota during feeding.Culturing of collected feeding solution resulted in ∼1.2 × 10 4 CFU/ml of a particularly slowgrowing microorganism on LB agar, which was later identified as yeast, highly similar to the genera Wickerhamia and Candida (Supporting Information Fig. S5 and Table S5).
Taken together, 16S rRNA gene amplicon sequencing and culturing revealed that the N. viridula core microbiota is limited to several microbial genera of which some are transmitted via feeding, which might have consequences on insect-plant interactions.

N. viridula salivary glands are colonized by Gammaproteobacteria and Sodalis sp.
To visually confirm the localization of N. viridula-associated microbiota, we performed FISH on the gut system and salivary glands.We targeted all bacteria with f luorescent probes for Gammaproteobacteria and Sodalis sp. and visualized them with confocal laser scanning microscopy.Microscopy images of the gut system confirmed that many rod-shaped Gammaproteobacteria Figure 3. Absolute and relative abundance of N. viridula microbiota.(A) A box plot showing the mean absolute number of 16S rRNA copies/ng (logarithmic scale) of 11 individual adult insects using RT-qPCR.16S rRNA from either Pantoea, Sodalis, Serratia, or Commensalibacter was used to determine the number of copies compared to genera-specific single-copy genes groL and rpoB.On the x-axis, the total number of 16S rRNA copies in the gut and salivary glands is shown with subsequent copy numbers for each genus.Error bars represent standard deviations and dots represent outliers.(B) Core bacterial community composition of the gut based on 16S rRNA gene amplicon sequencing.The samples were compared between N. viridula adult insects obtained from four geographical locations in the Netherlands.Ten guts and 10 pairs of salivary glands were dissected, pooled, and sequenced, and the taxonomy was displayed whenever possible at the genus level.Others represent the amplicon sequence variances (ASVs) that average below 1% of all reads.
Table 1.Overview of the recovered MAGs.The information for each of the MAGs includes bacterial ID, genome size, GC content, completeness, contamination, number of contigs, and the relative abundance assigned based on the percentage of mapped reads.The relative abundance of unmapped reads was assigned to either mitochondrial or chloroplast reads and therefore was left out of the were present in the cavities of the M4 midgut crypts of N. viridula ( Fig. 4A).Salivary glands of phytophagous shield bugs consist of a principal gland with anterior and posterior lobes, a hilum with a principal salivary duct, the duct of the accessory gland, and the accessory gland [77].Our analysis revealed that the entire principal gland tissue is colonized with Gammaproteobacteria and Sodalis sp. is the dominant member of the community (Fig. 4B and Supporting Information Fig. S7).

Metagenome analysis reveals a detoxifying potential of N. viridula microbiota
The role and microbial metabolic potential of N. viridulaassociated microorganisms was analyzed using metagenomics on N. viridula gut systems and salivary glands.Similar to the 16S rRNA gene amplicon data, the analysis revealed that Pantoea sp. and Sodalis sp. were the most abundant species in the gut and salivary glands, respectively (Fig. 5 and Supporting Information Fig. S8).Moreover, found that N. viridula gut and salivary glands harbored several strains of Pantoea sp. and Sodalis Further metagenomics analysis yielded one metagenome-assembled genome (MAG) for each of the core symbionts Pantoea sp., Sodalis sp., Serratia sp., and Commensalibacter sp.(Table 1).
Nezara viridula feeds on nutritious-poor and sugar-rich plant sap and relies on its symbiotic partners to biosynthesize deficient nutrients [37]; therefore, we analyzed the metabolic potential of N. viridula-associated microbes in terms of amino acid and vitamin metabolism and the degradation of carbohydrates (Supporting Information Tables S6 and S7).Pantoea sp., Sodalis sp., and Serratia sp. can biosynthesize most essential amino acids and all core microbes harbor partial vitamin B biosynthetic pathways, although none can biosynthesize thiamine or cobalamin.In terms of carbohydrate-degrading properties, Sodalis sp.harbors genes to degrade starch, and D-galacturonate (Supporting Information Table S7).A similar degradation potential was observed for Serratia sp., yet it lacked the genes to degrade D-galacturonate, a key constituent of pectin.Commensalibacter sp. has the metabolic potential to degrade starch and, unlike others, might convert arabinan to Larabinose and 1,4-beta-D-xylan to D-xylose, which are dominant in woody plants, plant seeds, and root components [78,79].Pantoea sp.lacked most carbohydrate-degrading enzymes and is therefore Panel "bacteria" shows FISH-probe "Eub-mix" in cyan (Fluos dye).Panel "γ -Proteobacteria" shows FISH-probe "GAM42A" in yellow (Cy5 dye).Panel "overlay" shows the overlay of both channels.Scale bar "overlay" = 50 μm.A dashed square in "overlay" marks the location of the magnified panel "detail".Closed arrowhead points at autof luorescent gut structure.Open arrowhead points at FISH-stained bacterium.Scale bar "detail" = 10 μm.(b) Confocal micrographs show an adult N. viridula salivary gland with the salivary duct.Panel "bacteria" shows FISH-probe "Eub-mix" in cyan (Fluos).Panel "γ -Proteobacteria" shows FISH-probe "GAM42A" in yellow (Cy5).Panel "Sodalis" shows FISH-probe "Sod1238R" in magenta (Cy3).Panel "overlay" shows the overlay of all three f luorescent channels merged with a transmitted-light brightfield channel.Open arrowhead points at salivary duct.Scale bar = 100 μm.

Figure 5.
Relative metagenome gut and salivary gland microbial abundance.A heatmap showing the relative metagenome microbial abundance (>1%) in adult N. viridula gut systems (gut) and salivary glands (Sg), calculated percentage of reads based on raw reads from the Blood&Tissue DNA isolation kit.When the species could not be assigned, the lowest taxonomic level was shown instead and was indicated with a taxonomic name between the parentheses.likely not involved in the degradation of carbohydrates.In addition, both Serratia sp. and Sodalis sp.contained chitin degradation genes, which could support insect resistance to fungal pathogens [ 80].
Nezara viridula microbiota could also provide its insect host with unique characteristics that may play a role in establishing symbiotic interactions with insects via cellular invasion, as well as the colonization of plants [81,82].We discovered that Sodalis sp.encodes a complete bacterial type III secretion system (T3SS, yscFCJRSTUVNQ) that is used by bacteria to secrete effector proteins, which can manipulate or even repress plant defenses [83].The latter can potentially reduce the production of anti-insect feeding compounds and toxic metabolites [84,85].
Given that insects used in this study were reared on native black nightshade and crown vetch plants that produce toxic αsolanine, α-chaconine, and NPA metabolites, detoxifying symbiosis was investigated.α-Solanine and α-chaconine are two structurally related glycoalkaloids that inhibit cholinesterase activity and disrupt eukaryotic cells and therefore are likely harmful to N. viridula [86].Serratia sp., Sodalis sp., and Commensalibacter sp. were confirmed to contain genes involved in α-solanine and αchaconine degradation via β-galactosidase, β-glucosidase, and α-rhamnose isomerase ( [87]; Supporting Information Table S7) although none of the isolated strains degraded α-solanine and α-chaconine in vitro under the chosen conditions.NPA is known to be toxic to eukaryotes by irreversibly inhibiting an enzyme from the tricarboxylic acid cycle [88].Detoxifying symbiosis by N. viridula symbiont Serratia sp. by nitronate monooxygenase (nmoA), 3-oxopropanoate dehydrogenase (bauC), nitrite reductase (nasD), and f lavohemoprotein (hmp) potentially supports its insect host with NPA degradation.Overall, genetic evidence points toward the possible involvement of N. viridula microbiota in the biosynthesis of nutrients, digestion of carbohydrates, suppressing plant defenses, and in detoxifying symbiosis.Figure 6.NPA degradation by N. viridula-associated Serratia.Serratia was grown in M9 medium until an OD 600 of 1.00 ± 0.05.Hereafter, 100 μM of NPA was added, and samples were immediately taken (0 h) and after 2, 4, 6, and 24 h (n = 2).NPA (blue) was measured with HPLC and nitrite (NO 2 − , red) and nitrate (NO 3 − , green) concentrations were determined using a Griess assay.On the x-axis, the time is shown in hours (h) and on the y-axis NPA concentration (μM).Shaded areas represent standard error.

Serratia sp. degrades 3-nitropropionic acid
Serratia sp., which harbored detoxifying genes, was cultured with NPA for 24 h to unveil whether N. viridula symbionts can degrade toxic plant metabolites.Rapid degradation of NPA was observed for the first 6 h of incubation, after which the rate of degradation decreased (Fig. 6).Simultaneously, nitrate and to a lesser extent nitrite by-product concentrations increased.No nitrate nor nitrite were present in control cultures, confirming that they are byproducts of NPA degradation (data not shown).These results demonstrate the ability of gut-associated Serratia sp. to degrade the toxic plant metabolite NPA.

N. viridula-associated microbiota repress plant defenses
To determine whether N. viridula's saliva and frass microbiota, namely Pantoea sp., Serratia sp., Sodalis sp. and yeast sp.could alter plant defenses, they were inoculated on mature Arabidopsis thaliana plants (Brassicaceae).The results showed that when N. viridula fed on Arabidopsis for 72 h, it significantly induced PR-1 expression and suppressed MYB28 expression (Fig. 7 and Supporting Information Fig. S9).We also observed that the four tested N. viridula-associated microorganisms were able to significantly repress artificial piercing-induced PR-1 expression at 24 h after inoculation, raising the possibility that they also repress N. viridula-induced PR-1 expression at 72 h of feeding.At 72 h of feeding, no LOX2 induction was observed.On the contrary, at 72 h, all tested N. viridula microbiota repressed artificial piercinginduced LOX2-expression.JA-associated defense responses have previously been shown to be induced by N. viridula 3 h after feeding [15].Taken together, the current data suggest that N. viridula microbiota repress both N. viridula-induced SA and JA pathways.Sodalis sp. was the only microorganism that could significantly repress MYB28 expression at 72 h after inoculation, thereby potentially directly suppressing the plant's aliphatic glucosinolate defense pathway.In conclusion, our hypothesis that N. viridula-associated microbiota support their host by counteracting insect-induced plant defenses was confirmed.

Discussion
N. viridula is a notorious pest insect threatening global food production [35].Different pest management strategies were shown ineffective, due to N. viridula's fast adaptive abilities, resulting in the proposition of an integrated approach in which a variety of preventive and therapeutic methods are used [89][90][91][92][93].As most insects rely on symbiotic microbiota for essential tasks, pest management strategies specifically targeting insect microbiota may provide an additional facet in integrated pest management for N. viridula.Strikingly, N. viridula microbiota is still poorly investigated, leaving untapped opportunities.To this end, we characterized N. viridula microbiota throughout their development as well as for their functional potential.
Our study revealed that the core microbiota of N. viridula predominantly consisted of Pantoea sp., Sodalis sp., Serratia sp., and Commensalibacter sp. and to a lesser extent other genera such as Klebsiella and accounting for more than 99% of all microbial members.We determined that Pantoea sp. and Sodalis sp. were possibly vertically transmitted via eggs as they were present from (surface-sterilized) eggs to adult and thus most likely function as obligate symbionts [38,94].Another explanation for their continued presence would be the reacquisition from the environment at different developmental stages, which seems unlikely due to their low relative abundance on the host plants.Pantoea sp. was previously described in Pentatomidae such as N. viridula and elimination resulted in severe fitness defects of the insect host, confirming its obligate symbiotic nature [95,96].Sodalis sp. was associated with Pentatomidae before [97], and we were able to pinpoint its predominant localization and highly abundant colonization of salivary glands.Moreover, with metagenomic analysis we shed light onto the role and host-specific adaptations and of N. viridula core microbiota, such as genome reduction of Pantoea sp. and Commensalibacter sp.[98][99][100], presence of T3SS in Sodalis sp., which could allow cellular invasion and modulation of plant defenses [81,82] and the ability of microbiota to biosynthesize amino acids and B vitamin and degrade plant carbohydrates and toxic secondary plant metabolites.Although, Pantoea sp. and Sodalis sp.seem to be most dominant N. viridula symbionts, lower abundant Serratia sp., Klebsiella sp., Commensalibacter sp., Pseudomonas sp., and Cutibacterium sp. could also benefit its insect host.Serratia sp. was previously described as a facultative symbiont of insects that could be obtained via plants, and in this study, we demonstrated its possible participation in detoxifying symbiosis via culturing it with toxic NPA [101].Even though some Serratia species are notorious plant disease-causing organisms that colonize the plant's phloem [102], no visual disease symptoms were observed in our plant inoculation experiments.Furthermore, Klebsiella sp. was detected in symbiotic relations with insects including Mediterranean fruit f lies (Ceratitis capitata) and showed competitive capacities to pathogenic host gut inhabitants [103,104].Commensalibacter sp., isolated from the midgut of Drosophila, has been demonstrated to protect against gut pathogens and our finding additionally implies its contribution to degradation of α-solanine and α-chaconine [105].Also, Pseudomonas sp.Nvir isolated from N. viridula gut could detoxify NPA and therefore possibly benefits its host [88].Besides bacterial symbionts, we also isolated a yeast species closely related to Wickerhamia sp. and Candida sp. from salivary glands and although insect-associated yeasts are still poorly studied, a previous study suggested their nutritional support to insects by biosynthesizing amino acids and vitamins [ 106].
Along with nutritious, protecting and detoxifying properties of N. viridula-associated microbes, we found that N. viridula transmits bacteria during feeding via saliva that altered plant defenses.Nezara viridula induced SA (i.e.PR-1) plant defenses, which is in line with previously published results [15], whereas all isolated microorganisms repressed SA defenses early after inoculation.Nezara viridula repressed the aliphatic glucosinolate pathway (i.e.MYB28) involved in defense against insects and Sodalis sp., predominantly located in the salivary glands, was the only microorganism significantly suppressing the latter pathway.This observation along with the metagenomic evidence of T3SS presence further strengthens the hypothesis that Sodalis sp. is a key player in repressing the biosynthesis of secondary plant metabolites.Altogether, N. viridula microorganisms seem to have a secret agenda in supporting their host with detoxifying and suppressing properties.However, to fully verify that N. viridula microbiota mediates detoxification and plant defense repression, eliminating N. viridula microbiome would provide the required evidence for microbial involvement in these processes.Nonetheless, to this day eliminating all microbiota remains challenging, because N. viridula requires its microbiota for viability and development [30,37,107].
In conclusion, our study reveals the developmental and organspecific dynamics of N. viridula core microbiota, revealing the important roles of Pantoea sp., and Sodalis sp. as obligate symbionts and Serratia sp. and Commensalibacter sp. as facultative symbionts.Nezara viridula symbionts were found to be involved in host plant defense repression, and Serratia sp.revealed in vitro detoxifying activities on toxic crown vetch metabolite NPA.Our results show the importance of studying tri-trophic interactions between insects, associated microbiota, and host plants to obtain fundamental knowledge on interactions between different organisms that could lead to the development of sustainable pest management strategies.

Figure 1 .
Figure 1.Experimental approach.An overview showing the experimental setup to identify N. viridula core microbiota throughout different developmental stages and subsequently unravel its role in insect-plant interactions.In the right panel, the study materials (Supporting Information Fig. S1), techniques used, and eventually obtained data are depicted.

Figure 2 .
Figure 2. Microbiota of N. viridula over different developmental stages.Heatmap showing the relative abundance (in percentage) of the core genera (exceeding >10 000 reads in all insect-related samples, covering 89% (2650717) of all reads) within each sample type consisting of three biological replicates for N. viridula samples and one replicate for host plant samples (39 samples in total; Supporting Information TableS1).

table .
a The abundance of the mapped reads within a sample.b Sample isolated using a Blood&Tissue kit.