TREHALOSE AND TREHALOSE-6-PHOSPHATE INDUCE STOMATAL MOVEMENTS AND INTERFERE WITH ABA-INDUCED STOMATAL CLOSURE IN GRAPEVINE

Methods and results : Epidermal peels were used to assess the effects of sugars. Low concentrations of trehalose and T6P (1 μM) induced an osmotic-independent reduction of the stomatal aperture in light conditions. Furthermore, ABA-induced stomatal closure was reduced by sugar application in association with lower accumulation of reactive oxygen species in guard cells. Similar effects, although weaker, were observed in response to the disaccharides sucrose and maltose, but not in response to the monosaccharides fructose and glucose.


INTRODUCTION
Stomata are natural pores bordered by two guard cellsandpresentintheepidermisofmostaerialparts ofhigherplants.Theyallowgasexchangesbetween theplantanditsenvironment,balancingCO 2 uptake foroptimalphotosynthesisandlimitingwaterlosses bytranspiration(Assmann1993).Openingorclosure of stomata is achieved by osmotic swelling or shrinking of guard cells driven by K + , Cl -, malate 2- (for review, see Outlaw, 2003), and probably also sucrose (Tallman and Zeiger, 1988 ;Lawson et al., 2014).Stomatal movements are finely regulated by endogenous and environmental factors such as phytohormones, light, CO 2 level, and atmospheric humidity (Outlaw,2003).Thephytohormoneabscisic acid (ABA) is a signal molecule for water stress which induces stomatal closure.This involves a complex cascade of signaling events, including the second messenger H 2 O 2 acting on ion channels and leadingtoguardcellplasmolysis (forreview,seeLi et al., 2006;Sirichandraet al., 2009).
The stomatal aperture was analyzed 2.5 h after incubationwiththefluoresceindiacetateprobe(FDA, 20 μg ml -1 ) (Allègre et al., 2007).Pictures of 30 stomataperepidermalpeelwererecordedusinga LeicaDMLBfluorescencemicroscopeequippedwith a digital camera and the stomatal apertures were measured using the NIS-Elements BR (Nikon Instruments) software.The results presented correspond to the mean values of three independent experiments, each performed with two epidermal peels.

Statistical analysis
Data were analyzed using a one-way analysis of variance(ANOVA)andthemeanswerecomparedby Fisher'sleastsignificantdifference(LSD)test.

Trehalose and T6P induce closure of grapevine stomata and inhibit their ABA-induced closure
Trehaloseappliedatlowconcentrations(0.1,1,and 10 µM) to grapevine epidermal peels induced a significant closure of stomata in a dose-dependent manner, with a maximum effect at 10 µM (Figure 1A).The differences in mean aperture between trehalose-treated stomata and the control were0.42,0.80,and1.24µmfor0.1,1,and10µMtrehalose, respectively.A higher trehalose concentration (100 mM) induced a stomatal closure of 1.14 µm compared to the control treatment, similar to the response to 10 µM trehalose.To exclude a possible osmotic effect, mannitol was applied but failed to induce closure below 0.5 M (datanotshown).
The effect of T6P, also reported as important in signaling,wascomparedtothatoftrehalose.T6Pat 1 µM induced a more pronounced stomatal closure thantrehalose,withdifferencesof1.16and0.59µmcompared to the control, respectively (Figure 2A).However, T6P affected the ABA-induced stomatal closure to a lesser extent than trehalose with differences in aperture of 1.13 and 1.70 µm for the ABA + T6P and ABA + trehalose co-treatments, compared to the ABA treatment, respectively (Figure 2A).

Monosaccharides and other disaccharides have none or weaker effects on the grapevine stomata aperture and ABA-induced closure compared to trehalose and T6P
Inordertoinvestigatethespecificityoftheeffectsof trehalose and T6P on stomatal movements, other disaccharides (sucrose, maltose) and monosaccharides (glucose, fructose) were applied to epidermal peels.Both disaccharides were applied at 1 µM, alone or combined with 50 µm ABA, on grapevine epidermal peels (Figure 2B).Sucrose treatment reduced the aperture of stomata in the light by 0.83 µm compared to the control, whereas maltose had a less pronounced effect with a difference of 0.4 µm.Both sugars reduced the ABA-induced stomatal closure by 0.68 and 0.36 µm, respectively.These results highlight a general effect of the disaccharides assessed on stomatal aperture, with trehaloseandT6Pidentifiedasthemostefficient.The monosaccharides glucose and fructose (1 µM) were appliedinthesameconditions,buthadnosignificant effectonthestomatalaperture(Figure 2C).Fructose induced a statistically significant reduction of the ABA-induced stomatal closure, whereas glucose showednosignificanteffect.

Trehalose and T6P inhibit the ABA-induced ROS production in grapevine stomata
Trehalose and T6P, being the most effective in interferingwiththeABA-inducedstomatalclosurein grapevine,wereanalyzedfortheireffectontheABAinduced ROS production in guard cells (Figure 3).ABAaloneinducedasignificantproductionofROS (247.6 % of the level of gray measured in the control),whereastrehalose(1µM)reducedtheABAinduced ROS production close to the background levelofthecontroltreatment.T6Palsodecreasedthe ABA-inducedROSproduction,buttoalesserextent.

DISCUSSION AND CONCLUSION
The effects of trehalose and T6P on grapevine stomatalmovementsinlightconditionswerestudied usingepidermalpeels.Twomono-anddisaccharides werealsoassessedforcomparison:sucroseasanonreducingsugar,maltoseasadisaccharideformedby twoglucosesubunits,andglucoseandfructoseasthe constituents of the assessed disaccharides.All the disaccharides used induced a reduction of the apertureofstomatainthelight,withT6Pshowingthe strongesteffect,whereasthemonosaccharideshadno effect.Altogether, our results suggest that disaccharides, but not monosaccharides, applied at low concentrations, interfere with the mechanisms regulating the "open state" of grapevine stomata.Dittrich and Mayer (1978) assessed the effects of a set of carbohydrates (used at 100 mM) on stomatal movements of Commelina communis epidermal peels.They reported an effect of hexoses and some disaccharides on stomatal aperture.In particular, trehalose inhibited the opening of stomata, but not sucrose nor maltose.However, as these results were obtained by studying the opening process of closed stomata, it makes it difficult to compare with our conditions.
An increasing number of studies investigates the function of trehalose and T6P in plants.Trehalose was first associated to desiccation tolerance of resurrection plants (Adams et al., 1990).Further studies showed that it is present in trace amounts in plants (Goddijn and Smeekens, 1998 ;Paul et al., 2008) and is accumulated to allow resistance to abiotic stresses, such as drought, heat, chilling, salinityorUV,andalsoresistancetoarangeofbiotic stresses (for review, see Fernandez et al., 2010).Transgenicplantsoverexpressingtrehalosephosphate synthasegeneshenceshowanimprovedtoleranceto abiotic stresses correlated with the accumulation of trehaloseand/orT6P (Romeroet al., 1997;Avonceet al., 2004).In grapevine, T6P accumulates in stems and leaves of Chardonnay in vitro plantlets in response to chilling and could contribute to the chilling resistance conferred by the plant growthpromoting rhizobacterium Burkholderia phytofirmans strainPsJN (Fernandezet al., 2012).
In downy mildew-infected grapevine leaves, we observed a stomatal dysfunction characterized by stomata abnormally opened in the dark and unresponsive to ABA.This deregulation occurs in theoilspotsymptomareatogetherwithasourceto sink transition characterized by higher levels of solublesugars,amongwhichtrehalose (Gammet al., 2011).Trehaloseand/oritsprecursorT6Pweregood candidates to explain the stomatal deregulation.As stomatahavenoplasmodesmata,theputativesoluble sugar signal was likely to occur at the apoplastic level.Interestingly, trehalose was only detected in apoplasticfluidsofinfectedleaves(0.98±0.04mM)(unpublished data).However, at this concentration trehalose induces a lower (27 %) stomatal opening than that obtained with apoplastic fluids extracted frominfectedleaves.TheroleoftrehaloseandT6P in downy mildew-infected leaves thus remains unclear.
Fewstudieshavereportedtheconcentrationsofboth trehalose and T6P in grapevine leaves.In our previous study (Gamm et al., 2011), only trehalose was quantified in leaves of cv.Marselan using HPAEC-PAD method.However, as reported by Fernandez et al. (2012), this method overestimates the actual concentration in some cases.The same authorsthereforeusedfluorescencespectrometryand HPLC methods allowing the determination of the actual concentrations of trehalose and T6P, respectively, in grapevine organs.In leaves of Chardonnay in vitro plantlets, they found higher levelsoftrehalosethanthoseofT6P(intherangeof 5 and 0.3 nmol g -1 FW, respectively).These levels increasesignificantlyinresponsetochillingtoreach valuesofabout14and2nmolg -1 FW,respectively.It isdifficulttocomparetrehaloseconcentrationswith those reported by Gamm et al. (2011) as they were obtained with different plant materials (in vitro plantlets/herbaceous cuttings) of different varieties (Chardonnay/Marselan) and expressed differently (freshweight/dryweight).Itwouldbeinterestingto quantify trehalose and T6P in grapevine leaves in response to abiotic and biotic stresses to compare theirrespectiveaccumulationprofileandtoprogress intheunderstandingoftheirrespectiverole.
Like some other sugars, trehalose and T6P are also considered as signaling molecules (Rolland et al., 2006 ;Paul et al., 2008)   Images correspond to representative guard cell fluorescence observed using microscopy.
ABA-inducedstomatalclosure,withtrehalosebeing the most effective.This effect appears to be less specificsincebothmono-anddisaccharides(except glucose) were effective to induce it.ABA-induced stomatal closure is known to be mediated by ROS productioninguardcells (Peiet al., 2000;Murataet al., 2001 ;Kwak et al., 2003).In this study, we showed that ROS production is affected in the presence of trehalose and T6P, suggesting that both sugars act upstream of ABA-signaling events associatedtostomatalclosure.Studieshavereported the impact of trehalose/T6P on ABA signaling in association with plant development mechanisms (Avonceet al., 2004;Gomezet al., 2010).However, only few papers have reported an impact of the modulationoftrehalose/T6PlevelsonABA-induced stomatal closure.In the Arabidopsis mutants Attre1 (affected in trehalase synthesis), Van Houtte et al.
(2013) reported a higher trehalose concentration correlated with an impaired ABA-induced stomatal closureanddroughtsusceptibility.Curiously,stomata of the Arabidopsis mutant tps1-12 (affected in T6P synthesis)aremoreclosedinthedarkthanthoseof the wild type and have also an impaired responsivenesstoABA (Gomezet al., 2010).These studies show the high complexity of ABA/trehalose signalingcrosstalk.
Usingasimplifiedstomatalmodelwethusobserved thattrehaloseandT6Pinduceareductionofstomatal opening in the light and partially inhibit the ABAinduced stomatal closure.At the leaf level, it is tempting to suggest that variations in their concentrationsintheguardcellapoplastcouldserve assignalforgrapevinestomatalmovementsandthe subsequent regulation of photosynthesis/ transpiration.

CONCLUSION
Altogether, these results highlight the role of exogenously applied carbohydrates as signaling molecules on grapevine stomata.The characterization and significance of the dynamics of trehalose and T6P at the scale of grapevine plants and their respective role in response to stress is one crucial questionthatwillrequirefurtherinvestigation.

Figure 1 .
Figure 1.Effect of trehalose on grapevine stomatal aperture in the light (A) and on ABA-induced stomatal closure of grapevine stomata (B).Epidermal peels were prepared and incubated with 0.1 µM, 1 µM, 10 µM, 100 µM trehalose (Tre) or 50 µM abscisic acid (ABA) alone or in combination.Data are the arithmetic means of 3 biological replicates with 60 stomata each, with bars representing the standard error of means.Values with different letters are statistically different (p < 0.05).
Figure 2. Effect of trehalose and T6P (A), sucrose and maltose (B), and fructose and glucose (C) on grapevine stomatal aperture in the light and on ABA-induced stomatal closure.Epidermal peels were prepared and incubated with 1 µM saccharides and 50 µM abscisic acid (ABA) alone or in combination for 2.5 h in the light.Data are the arithmetic means of 3 biological replicates with 60 stomata each, with bars representing the standard error of means.Values with different letters are statistically different (p < 0.05).
Figure 3. ROS production in grapevine epidermal peels treated with abscisic acid (ABA) alone or in combination with trehalose or T6P.Values correspond to the arithmetic mean of the mean gray value per pixel of 100 guard cells, with bars representing the standard error of means.Data represent the results of one representative experiment of three biological replicates using 50 µM 2'7'-dihydrofluorescein diacetate to assess the production of ROS in guard cells.Values with different letters are statistically different (p < 0.05).Images correspond to representative guard cell fluorescence observed using microscopy.