Inositolphosphate glycans and a fucosylated xyloglucan oligosaccharide are accumulated upon Arabidopsis thaliana/ Botrytis cinerea infection

In mammals, insulin is involved in controlling blood glucose levels and its role in modulating immunity is being more and more documented. This hormone promotes the release of inositolphosphate glycans (IPG) which act as mediators. In plants, one IG has already been identified in plant culture cells (Smith and Fry, 1999; Smith et al., 1999) but, to our knowledge, no IPG have been yet identified. Here, we discovered 7 IPG that are accumulated upon Arabidopsis thaliana-Botrytis cinerea interaction, concomitantly with oligogalacturonides and a fucosylated xyloglucan oligosaccharide. Further structural characterization showed that they come from the hydrolysis of polar heads of Serie A to H glycosylinositol phosphorylceramides presumably via a phospholipase C activity. Taken together with the emerging role of insulin as immune regulator, these results question the role of IPG as damage associated molecular pattern both in animal and plant kingdoms.


Introduction
Oligosaccharins are biological active oligosaccharide 1 . Their biological activities include the stimulation of skotomorphogenesis 2 , growth promotion [3][4][5][6] and anti-auxin activity [7][8] . We also know for decades that oligogalacturonides (OGs), oligomers of •-1,4-linked galacturonic acid trigger plant defense response when they are exogenously applied [9][10][11] . Until very recently 12 , almost none of the oligosaccharins studied have been detected in planta, thus questioning their role and their identity in planta. Here we identified unknown oligosaccharides accumulated during Arabidopsis thaliana-Botrytis cinerea interaction. While we rather expected to characterize cell wall derived oligosaccharides, we discovered that several inositol phosphate glycans (IPG) deriving from plant glycosyl inositol phosphoryl ceramides (GIPCs) are highly accumulated upon infection.
First, using high-performance size-exclusion chromatography (HP-SEC) coupled with high resolution mass spectrometry (HRMS)-based method in negative mode, we have set up a strategy aimed at identifying putative oligosaccharides accumulated over time upon Botrytis cinerea infection. We selected the main ions detected upon infection and increasing continuously between 12 and 18 hours post-infection with a mass-to-charge ratio (m/z) superior to 263 that corresponds to a dipentose (Fig. 1A). This leads to the identification of 36 candidates included 16 OGs, 4 being oxydized (Fig. 1B). Amongst the 18 non-OG candidates, we discarded 5 members for which we did not obtain confident elemental compositions. Finally, this workflow resulted in 13 candidates. 9 are phosphorylated and 3 do not present any elemental composition compatible with an oligosaccharide structure (Fig. 1C). We observed that 8 of them were also detected in cell wall water extract of non-infected plants.
We  13 suggest that m/z 597 ion corresponds to an inositolphosphate glycan (IPGs) composed of a phosphorylated inositol, an uronic acid and a hexose (Fig. 2B). We next fragmented the candidate at m/z 517 (C 18 Fig. 3A-D), these molecules contain one additional hexose and three and four additional pentoses respectively, likely linked to a Hex-UA-Ino/Hex-P core. The cross-ring fragments at m/z 353, 485, 617 and 749 (B 4• + C 2 H 4 O 2 ) might be produced by the cleavage of an inositol or a hexose residue. We favor the first hypothesis since no fragment corresponding to three or four pentoses plus two hexoses has been detected. Last, the cross-ring fragment at m/z 221 (C 6 H 10 O 5 + C 2 H 4 O 2 ) provides evidence that the Pent 4 Hex chain and Pent 3 Hex are branched on the inositol moiety via the hexose residue. Following the same reasoning, according to the fragmentation pattern obtained from candidates at m/z 716 and 724 ( Fig. 4A and 4B), we attributed the respective fragments at m/z 895 and 911 to the presence of side chains composed of one hexose linked to the Hex-UA-Ino-P core and four pentoses and one additional deoxyhexose (m/z 716) or hexose (m/z 724). We inally deduced that the candidates at m/z 782 and 790 are characterized by one additional pentose ( Fig. 4C and 4D).
Therefore, we identified 7 inositolphosphate glycans (IPG) and one inositol glycan (IG) over-accumulated upon A. thaliana/B. cinerea interaction and present in cell wall water extract of healthy leaves.
In order to explore if these I(P)Gs could derive from glycosyl inositol phosphoryl ceramides (GIPCs), we analyzed the cell wall water extract of gmt.1.3 (GIPC Mannosyl Transferase) mutant which is affected in GIPC glycan biosynthesis 14 . 14-day-old light-grown wild-type and gmt1.3 plants that still displayed similar heights were analysed. In WT cell wall water extract, we detected six of the IPGs and IG accumulated upon infection ( Fig. 5A and B). By contrast, none of the oligosaccharides described above were detected in the gmt1 mutant. We found ions at m/z 435 (C 12 H 20 O 12 P) and 355 (C 12 H 20 O 12 ) absent in WT ( Fig. 5A and B), that could correspond to GlcA-Ino-P and GlcA-Ino, respectively. The m/z 435 fragmentation gave rise to ions at 259 m/z, likely formed by the loss of an uronic acid, and at m/z 97 and 79 diagnostic of phosphorylated groups (Fig.  5C). The 355 fragmentation indicates the presence of a C 6 H 10 O 6 moiety and an uronic acid absent in the WT sample that we attributed to GlcA-Ino (Fig. 5D). It is worth to note that we did not detect any ion corresponding to the GlcA-Ino core substituted by side chains questioning the GIPC glycan biosynthetic machinery. We concluded that the 7 I(P)Gs accumulated upon A. thaliana/B. cinerea interaction derive from Serie A and Series E to H GIPCs and would be released by GIPC phospholipase(s) C. were next assigned to a 0,2 A n-1 cross-ring cleavage diagnostic of C 4 -and C 6 -substituted hexose. Last, the prominent and smaller single charged fragments at m/z 643 and 455 are accompanied by ions at m/z 541 and 353, respectively, which corresponds to a 102-Da neutral loss. This MS 2 pattern is diagnostics of xyloglucan oligosaccharide fragmentations which are characterized by prominent D-type fragment ions. They result from double cleavage and correspond to entire inner xyloglucan side chains 15 . They produce in turns 0,4 A i cross-ring cleavage ions, diagnostic for the presence of (1,6)-linkage, via a 102-Da loss from anhydro-glycosyl unit. We therefore attributed the D-type ion at m/z 455 and its 0,4 A • corresponding ion at m/z 353 to the presence of Galactose-Xylose-Glucose block (represented by the letter L) and m/z 643 (D), 541 ( 0,4 A i• ) and 499 ( 0,4 A i• -Acetyl) to the presence of an acetylated Fucose-Galactose-Xylose-Glucose block (represented by the letter F). According to these results and to the well-known structure of xyloglucan, we concluded that the candidate at m/z 809 is an acetylated XLFG oligosaccharide.
In conclusion our study shows that structurally diverse oligosaccharides are produced in planta. Upon infection, B. cinerea secretes a xyloglucanase 16 , it is therefore consistent to detect xyloglucan fragment produced concomitantly with OGs. By contrast, the presence of I(P)G is far more surprising. The study of gmt1.3 mutant affected in GIPC mannosylation showed that these I(P)Gs originate from GIPCs. The existence of an IG, originating from serie A GIPC, has already been described in extracellular space of culture cells during the period of rapid cell growth [17][18] . However, although in mammals IPGs were described over 30 years ago 19 , IPGs have not yet been identified in plants.
Here, we demonstrate that IPGs from serie A to H, present in cell wall extract of healthy plant are preferentially accumulated upon infection. This suggests that a GIPC phospholipase C activity which cleaves phospholipids before the phosphate group, is specifically promoted in that condition. It is worth to note that the non-specific phospholipase C4 (NPC4) has been shown to be involved in GIPC hydrolysis in response to phosphate deficiency 20 and consequently, IPGs could act more broadly as stress-related molecules. The accumulation of serie A IG (m/z 517) also suggests that a GIPC phospholipase D, which cleaves after the phosphate moiety, could be activated. A phospholipase D activity has been detected in cabbage and in A. thaliana 21 . This GIPC-PLD prefers GIPC containing two sugars 21 which is consistent the fact that Man-GlcA-Ino is the most highly produced in our conditions. However, we cannot rule out another scenario that would imply a phospholipase C activity and a subsequent phosphorylase. Finally, the I(P)G diversity observed questions the existence of diverse phospholipases C and D with different substrate preference. The identification of these enzymes, their impact in membrane properties as well as the role of such IPG and IG diversities need to be investigated.
The absence of series E to H IPGs lacking mannose residue in the gmt1.3 mutant suggests that mannose adding is indispensable prior to the transfer of other sugars. More interestingly, this also questions the role of complex GIPCs in the gmt1.3 phenotype. Indeed, this mutant gmt1.3 displays a severe phenotype and a constitutive hypersensitive response 14 . The role of IPGs as signaling molecules needs therefore to be investigated in this mutant and in plants in general. In mammals, IPGs are extracellularly-produced by insulin-sensitive cells in response to insulin treatment from glycosylphosphatidylinositol cleaving by phospholipase C 23 and promote lipogenesis, glycogenesis 22 and interestingly, are also important component of T-cell growth 23 . Insulin is involved in controlling blood glucose levels and its role in shaping immunity through modulating T cell metabolism is being more and more documented 24 . Taken together with the fact that I(P)Gs are accumulated upon infection in plants, we do believe that considering I(P)Gs as damage-associated molecular pattern and studying roles of I(P)Gs in defense will undoubtedly open new avenues of research.

Alcohol insoluble residue and cell wall water extract
Leaves and plantlets were grinded in five volume of ethanol 96 % and the supernatant was removed after centrifugation at 10,000g for 5 min. The pellet was washed with 70% ethanol with subsequent centrifugation until being uncolored and dried in a speed vacuum concentrator at room temperature. Two volumes of distilled water were added to the alcohol insoluble residue obtained and leave one hour at room temperature. The supernatant was collected after centrifugation at 10,000g for 5 min and dried in a speed vacuum concentrator at room temperature.

Oligosaccharides accumulated upon A. thaliana-B.cinerea interaction
Oligosaccharides were produced and analyzed according to Voxeur et al., 2019. Briefly, The wild-type B. cinerea B05.10 strain was grown on potato dextrose agar at 23 °C under continuous light. After 10 days, the spores were washed from the surface of the plate using Gamborg's B5 basal medium, 2% (w/v) fructose and 10 mM phosphate buffer. Fungal hyphae were removed by filtering. The concentration of spores was determined using a Malassez cell and adjusted to a final concentration of 3.10 5 conidia/mL. Isolated A. thaliana leaves of 5-week-old plants were immersed in a B. cinerea suspension (6 leaves for 10 ml of suspension at 3 x 10 5 spores/ml) and incubated on a rotary shaker at 100 rpm at 23 °C during 12, 15 and 18 h. The liquid medium was next collected and an equal volume of 96% ethanol was added. After centrifugation at 5000 g during 10 min, the supernatant was collected and dried in a speed vacuum concentrator at room temperature. The obtained pellet was then diluted. The equivalent of the digestate of 3 leaves of 5-week-old A. thaliana plants was dried and diluted in 200 μl. 10 μl were injected for MS analysis.

Oligosaccharide analysis
Samples were diluted at 1 mg/ml in ammonium formate 50 mM, formic acid 0.1%. Chromatographic separation on high-performance size-exclusion chromatography was performed on an ACQUITY UPLC Protein BEH SEC Column (125Å, 1.7 µm, 4.6 mm X 300 mm, Waters Corporation, Milford, MA, USA). Elution was performed in 50 mM ammonium formate, formic acid 0.1% at a flow rate of 400 μl/min and a column oven temperature of 40 °C. The injection volume was set to 10 μl. MS-detection was performed in negative mode with the end plate offset set voltage to 500 V, capillary voltage to 4000 V, Nebulizer 40 psi, dry gas 8 l/min and dry temperature 180 °C.
Major peaks were annotated following accurate mass annotation, isotopic pattern and MS/MS analysis. The MS fragmentation pattern is indicated according to the nomenclature of Domon and Costello 30 . For the targeted analysis, the theoretical exact masses were used with 4 significant figures with a scan width of 5 ppm. The resulting extracted ion chromatograms were integrated.

Data processing
The .d data files (Bruker Daltonics, Bremen, Germany) were converted to .mzXML format using the MSConvert software (ProteoWizard package 3.0 31 ) mzXML data processing, mass detection, chromatogram building, deconvolution, sample alignment, and data export were performed using MZmine 2.52 software (http://mzmine.github.io/). Mass list were built using a retention time window of 6.0-8.0 min and next, we used the ADAP chromatogram builder 32 with group size of scan 5, peak detection threshold of 800, a minimum highest intensity of 1500 and m/z tolerance of 0.01 m/z. We deconvoluted the data with the ADAP wavelets algorithm using the following setting: S/N threshold 8, peak duration range = 0.01-0.2 min RT wavelet range 0.02-0.1 min. Finally, we selected candidates with a surface area superior to 18000 at 18 hpi and results were deisotoped manually.