Prolonged root hypoxia induces ammonium accumulation and decreases the nutritional quality of tomato fruits

ARTICLE IN PRESS

Journal of Plant Physiology ] (]]]]) ]]]—]]]

0176-1617/$ - sdoi:10.1016/j.

Abbreviation�CorrespondE-mail addr

Please cite thquality of tom

www.elsevier.de/jplph

Prolonged root hypoxia induces ammoniumaccumulation and decreases the nutritionalquality of tomato fruits

Faouzi Horchania, Philippe Galluscib, Pierre Baldetb,Cecile Cabassonb, Mickael Maucourtb, Dominique Rolinb,Samira Aschi-Smitia, Philippe Raymondb,�

aUR d’Ecologie Vegetale, Departement des Sciences Biologiques, Faculte des Sciences de Tunis,Campus Universitaire, 1060 Tunis, TunisiabINRA, Universite Bordeaux 1, Universite Bordeaux 2, UMR 619 Biologie du Fruit,Centre de Recherche de Bordeaux, BP 81, F-33883 Villenave d’Ornon Cedex, France

Received 28 July 2007; received in revised form 24 October 2007; accepted 24 October 2007

KEYWORDSFlooding;Fruit quality;Lycopene;Metabolite profiling;Nitrate

ee front matter & 2007jplph.2007.10.016

s: NMR; nuclear magneing author. Tel.: +33 5 5ess: raymond@bordeau

is article as: Horchani Fato.... J Plant Physiol

SummaryHere we examined the effects of root hypoxia (1–2% oxygen) on the physiology of theplant and on the biochemical composition of fruits in tomato (Solanum lycopersicumcv. Micro-Tom) plants submitted to gradual root hypoxia at first flower anthesis. Roothypoxia enhanced nitrate absorption with a concomitant release of nitrite andammonium into the medium, a reduction of leaf photosynthetic activity andchlorophyll content, and an acceleration of fruit maturation, but did not affect finalfruit size. Quantitative metabolic profiling of mature pericarp extracts by 1H NMRshowed that levels of major metabolites including sugars, organic acids and aminoacids were not modified. However, ammonium concentration increased dramaticallyin fruit flesh, and ascorbate and lycopene concentrations decreased. Our dataindicate that the unfavorable effects of root hypoxia on fruit quality cannot beexplained by two of the well-known effects of root hypoxia on the plant, namely adecrease in photosynthesis or an excess in ethylene production, but may insteadresult from disturbances in the supply of either growth regulators or ammonium, bythe roots.& 2007 Elsevier GmbH. All rights reserved.

Elsevier GmbH. All rights reserved.

tic resonance.7 12 25 42; fax: +33 5 57 12 25 41.x.inra.fr (P. Raymond).

, et al. Prolonged root hypoxia induces ammonium accumulation and decreases the nutritional(2007), doi:10.1016/j.jplph.2007.10.016

ARTICLE IN PRESS

F. Horchani et al.2

Introduction

The formation of tomato (Solanum lycopersicum)fruits of good nutritional quality depends on theadequate supply of water, minerals and photo-synthates from the roots and leaves (Davies andHobson, 1981) and on their storage or transforma-tion by fruit metabolism during development.Development itself is controlled by the supply ofnutrients (Joubes et al., 2001) and by growthsubstances produced by the seeds or importedfrom other plant parts, including the roots (Srivas-tava and Handa, 2005). Carbohydrate partitioningtowards the fruits is influenced by the competitionwith the roots, young leaves or flowers, which mayresult in carbohydrate limitations (Ho, 1996). Intomato under carbon starvation, carbohydrates areessentially accumulated by the fruits that act ashigh-priority organs, thus strongly decreasing theavailability of sugars for the roots (Baldet et al.,2002; Devaux et al., 2003).

Root function may also be disturbed by soilflooding, which slows down oxygen transfer to theroots, limiting their aerobic respiration. Thisdramatically affects plant development (Drew,1997). In hypoxic tissues, metabolic activity de-creases in parallel with the decrease in respirationrate and, at low oxygen concentrations, fermenta-tion is induced (Geigenberger et al., 2000). Theaccumulation of lactic acid and the lack of ATP forH+ extrusion by plasmalemma ATPase result incytoplasmic acidification, a major cause of celldamage in anoxic plant tissues (Drew, 1997). Anoxiaand hypoxia also lead to increased formation ofreactive oxygen species, which cause oxidation ofphospholipids and proteins, particularly upon re-aeration (Szal et al., 2004). Root anoxia often leadsto root death within a few hours or days. In somestudies, nitrate has been found to improve plantsurvival under anoxia (Stoimenova et al., 2003;Allegre et al., 2004), although in several others, noeffect was noted (Saglio et al., 1988). In anoxicroots of tobacco or tomato in the presence ofnitrate, nitrate reductase is activated (Stoimenovaet al., 2003; Allegre et al., 2004), and nitrite isreleased into the medium (Morard et al., 2004).The favorable effect of nitrate was first explainedby NAD regeneration through nitrate reduction,which would sustain energy production while limit-ing the activity of fermentation. More recently, thiseffect was attributed to a decrease in metabolicrate through NO signaling (Stoimenova et al., 2003;Libourel et al., 2006). Hypoxia is certainly morecommon and has less dramatic effects than anoxia.It may lead to root acclimation to anoxia (Saglioet al., 1988; Drew, 1997), but also may have

Please cite this article as: Horchani F, et al. Prolonged root hypoxiaquality of tomato.... J Plant Physiol (2007), doi:10.1016/j.jplph.20

detrimental effects on fruit development becauseit limits the transport of water and minerals, causesleaf epinasty and reduces photosynthetic activity(Pezeshki, 2001). It may also affect the synthesis ortransport cytokinins that are required for thenormal development of the aerial part of the plant(Rahayu et al., 2005). Because root hypoxia mayoccur in field-grown tomatoes, depending on soilstructure and water supply, our purpose was toevaluate the effect of root hypoxia on themetabolite content of tomato fruits, which mayaffect both their taste and nutritional quality.

Materials and methods

Plant material and growth conditions

Tomato (cv. Micro-Tom) seeds were germinated for 10 din vermiculite, and then grown hydroponically in agrowth chamber (16 h light at 23 1C/8 h dark at 18 1Cwith an irradiance of 350 mmolm�2 s�1, and 75–80%relative humidity). Each seedling was placed in a 25mLvermiculite plug on a polystyrene tray floating on thenutrient solution, with 6 plants per 20 L tank. Thenutrient solution, pH 5.8, contained 15mM NO3

� but noNH4

+ and was renewed weekly. For control plants thesolution was continuously bubbled with air. Hypoxictreatment was applied at first flower anthesis by stoppingair bubbling. These plants are called ‘‘hypoxicallytreated (HT) plants’’ and are compared to ‘‘control’’plants. Flowers were tagged when fully opened(anthesis). Six fruits were allowed to develop per plant,usually with 3 fruits per truss. Secondary stems andadditional trusses were eliminated.

Changes in fruit diameter were measured every 5 d, inthe equatorial part, with an electronic caliper.

Tomato fruits were harvested at the red ripe stage.The equatorial pericarp was isolated, immediately frozenin liquid nitrogen, ground to a fine powder and stored at�80 1C until used.

Oxygen pressure measurement

Oxygen pressure in the medium was measured polar-ographically (Strathkelvin Instruments, Clark electrodemodel 1302, oxygen meter model 782) at 3 positions ineach tank.

Extraction and assay of soluble metabolites

Polar metabolites (sugars, amino acids, lactate) wereextracted and analyzed by 1H-NMR spectroscopy, asdescribed (Mounet et al., 2007). Lycopene was extractedas in Mounet et al. (2007) and measured after HPLCanalysis according to Telef et al. (2006).

Ascorbate, ethanol and ammonium in fruit pericarpwere assayed in extracts prepared by mixing frozenpowder (300mg) with 750 mL 6% TCA and centrifugation

induces ammonium accumulation and decreases the nutritional07.10.016

ARTICLE IN PRESS

0

5

10

15

20

25

0 16 24 32 40 48 56 64 72 80

Time (h)

Oxygen (

%)

8

Figure 1. Oxygen concentration change in the nutrientsolution of hypoxically treated and control plants. Thiscurve was obtained during the fifth week of hypoxictreatment. At time zero, 20 L of a new (aerated) nutrientsolution was placed in the tank, and covered with thepolystyrene plate that supported 6 tomato plants.Measurements were taken by placing the calibratedoxygen electrode 5–10 cm below the surface, andmaintaining slow movements of the electrode until astable indication was read. This curve is representative ofmeasurements obtained during 6 different cultures.

Root hypoxia affects the nutritional quality of tomato fruits 3

at 25,000g for 20min. Ascorbate and dehydroascorbatewere quantified by the 2,20-bipyridyl method accordingto Kampfenkel et al. (1995). Ammonium was assayed bythe phenol–hypochlorite method described in Baldetet al. (2002). Ethanol was assayed enzymatically asdescribed in Beutler (1984), by measuring the productionof NADH at 340 nm.

Nitrate, nitrite and ammonium present in the mediumwere assayed spectrophotometrically by the salicylicmethod (Cataldo et al., 1975), the Griess reagent method(Miranda et al., 2001) and the phenol–hypochloritemethod described above, respectively.

Photosynthesis

Leaf photosynthesis was analyzed using an infrared CO2

analyzer (LCA3 – Analytical Development Corporation)following the recommendation of the manufacturer. Sixmeasurements were carried out for each plant.

Leaf chlorophyll content

Chlorophyll measurement was performed as describedin Wintermans and Mots (1965). Approximately 300mg offresh leaves was homogenized with a mortar and pestle in1.4mL 95% ethanol. The mixture was then stronglyagitated and kept 10min on ice. After centrifugationfor 1min at 20,000g, 1mL of supernatant was measuredat 665 nm for chlorophyll a and at 649 nm for chlorophyllb. Total chlorophyll content was calculated as follows:

Chlorophyll ðaþ bÞ ¼ 6:1OD665 nm þ 20:04OD649 nm ðmg=mLÞ.

Results

Oxygen shortage and plant morphology inhypoxic conditions

Because the stock nutrient solution was incontact with air, the oxygen concentration atmedium renewal was close to 21%, the oxygenpartial pressure in air, in both the control andhypoxic tanks. In aerated solutions, the oxygenconcentration remained close to 21%. In non-aerated solutions, oxygen shortage appeared pro-gressively as the roots consumed the oxygenpresent in the medium. Oxygen levels decreasedcontinuously to reach 3% within 1 d and stabilizedat about 1% or 2%. The oxygen depletion kineticsshown in Figure 1 was measured during the fifthweek of hypoxic treatment, which indicates thatthe roots were still able to respire actively after 4weeks of transient hypoxia. The pre-existing rootsof HT plants stopped growing in the first week (notshown). The gradual development of oxygen defi-ciency likely allowed root acclimation to hypoxia(Saglio et al., 1988). In spite of this, main roots

Please cite this article as: Horchani F, et al. Prolonged root hypoxiaquality of tomato.... J Plant Physiol (2007), doi:10.1016/j.jplph.20

turned light brown in the following weeks andbecame flaccid by the end of the culture. However,adventitious roots were formed by HT plants, andmore roots developed in the vermiculite plug of HTthan control plants (not shown). Due to proximitywith the atmosphere, these roots probably hadbetter access to oxygen than those bathing in thenutrient solution. For these reasons, the treatmentis called hypoxia rather than anoxia. A morestringent hypoxic treatment (not shown) obtainedby pre-bubbling the new solutions with nitrogendown to about 2% O2, and thus limiting the weeklyreoxygenation of the roots, led to more dramaticeffects on the aerial part of the plant with most ofthe leaves being dry at the end of the culture.

Photosynthesis and leaf chlorophyll content

Leaf photosynthesis of hypoxic and control plantswas measured at both the beginning and the end ofthe light period (Table 1). HT plants showed a 30%reduction in photosynthetic activity compared tothe control only at the beginning of the lightperiod. The chlorophyll content of leaves (Table 1)was 34% lower in HT compared to control plants.

Nitrate uptake and nitrite and ammoniumrelease under root hypoxia

Changes in the concentrations of nitrate, ammo-nium and nitrite over the first 3 weeks of hypoxic

induces ammonium accumulation and decreases the nutritional07.10.016

ARTICLE IN PRESS

Table 1. Photosynthetic activity and chlorophyll con-tent of leaves of tomato plants with roots in aerated(control) and hypoxic nutrient solution

Control (pmolCO2/cm

2 s)Hypoxic roots(pmol CO2/cm2 s)

PhotosynthesisBeginning of day 31.474 a 20.875.3 bEnd of day 1973.7 b 21.274.2 b

Chlorophyll Control (mg/gDW)

Hypoxic roots(mg/g DW)

Chl a 6.2170.47 a 3.5770.67 cChl b 2.9670.41 a 1.7970.31 cChl tot 9.1770.74 a 5.3670.87 c

Measurements of photosynthesis were performed at the begin-ning and end of the light phase after 4 weeks of hypoxictreatment. Results are the mean7SD of 6 measurements ondifferent plants. Chlorophyll content is expressed as themean7SD of 6 determinations in extracts of the last matureleaf from 6 different plants. Leaves were taken at the end ofculture, after 9 weeks of hypoxic treatment. Since chlorophyllwas analyzed on extracts from fresh leaves, conversion to DWused the ratio DW� 100/FW ¼ 14.3570.8 and 16.2770.99determined after freeze drying for control and hypoxicallytreated plants, respectively.Different letters in each row indicate significant differencescaused by treatment (Student t test at the 0.05 (b) or 0.01 (c)levels).

A

0

3

6

9

12

15

B

0

3

6

9

12

15

C

0.0

0.4

0.8

1.2

1.6

0 6 12 15 18 21

Time (d)

Nitra

te (

mM

)A

mm

oniu

m (

mM

)N

itrite

(m

M)

93

Figure 2. Time course of (A) nitrate consumption (B)ammonium release and (C) nitrite release in the nutrientmedium of hypoxically treated tomato plants during thefirst 3 weeks of root hypoxia (closed symbols). Nitratelevels in aerated media are shown with open symbols(J). Medium was renewed weekly. Each point is themean7SD of measurement on at least 4 measurementson duplicate samples.

F. Horchani et al.4

treatment are shown in Figure 2. In aeratedconditions, nitrate concentration in the mediumdid not vary significantly, which is consistent withthe determinate phenotype of the Micro-Tomcultivar which stops growing at the fruiting stage(Marti et al., 2006). In contrast, in the non-aeratednutrient solution, nitrate concentration decreasedmarkedly to reach about 1.4mM at day 6. This is inagreement with the increase of nitrate uptake byanoxic tomato roots described previously (Allegreet al., 2004; Morard et al., 2004). Strikingly, in ourconditions, high concentrations of ammoniumwere produced, indicating that most of the nitratewas reduced to ammonium. In the first week,the ratio of ammonium released to nitrate taken upwas lower than 1, which may be explained bythe storage of part of the ammonium in theplant tissues during the first week (see below).During weeks 2 and 3, ammonium present in themedium was close to equimolar with nitrateconsumed.

Nitrite appeared in the nutrient solution 2 daysafter the beginning of the experiment (Figure 2)and increased in the following days but remainedlow, around 1/10th of the amounts of absorbednitrate.

Please cite this article as: Horchani F, et al. Prolonged root hypoxiaquality of tomato.... J Plant Physiol (2007), doi:10.1016/j.jplph.20

Absence of ethanol production in hypoxicroot media

No ethanol was detected in hypoxic root mediataken at different stages of 2 different cultures,with a detection limit of 0.4 nmol/mL (data notshown). This result is in agreement with nitratereduction re-oxidizing NADH that would otherwisefeed fermentation (Stoimenova et al., 2003). In theprevious study by Stoimenova et al., the excisedtobacco roots were placed under strict anoxia,which may explain that some ethanol was produced.

induces ammonium accumulation and decreases the nutritional07.10.016

ARTICLE IN PRESS

Table 2. Composition of ripe tomato fruit pericarps(mean7SD) from control or hypoxically treated plants

Compounds Control(mMol/gDW)

Hypoxic roots(mMol/g DW)

Fructose N 7257139 a 7697140 aGlucose N 6507125 a 6867119 aSucrose N 48711 a 55713 aGalactose N 1273 a 1072Malate N 2476 a 30713.8Citrate N 280768 a 279790Lactate N 4.472.0 a 3.670.9Alanine N 9.172.6 6.071.7Aspartate N 4079 a 3376Asparagine N 2175 a 1674Glutamate N 272763 a 206763 bGlutamine N 79722 a 58716 bGABA N 2676 a 2275.8 aPyroglutamate N 2275 a 1574 bValine N 1.370.3 a 1.170.3 aIsoleucine N 1.870.6 a 1.370.4 aLeucine N 4.171.5 a 3.371.6 aPhenylalanine N 7.872.0 a 6.671.4 aTyrosine N 1.570.4 a 1.170.2 aCholine N 8.672.6 a 8.372.1 aTrigonelline N 2.270.9 a 1.770.7 aNH4+ S 9876 a 78507800 cAscorbate (AsA) S 1473 a 5.671 cTotal ascorbate S 2172 a 1472 cAsA/total asc 0.7 0.4Lycopene LC 1.5370.16 a 1.0370.19 c

Values are given as mean7SD with the number of replicate fruitsper condition as follows: NMR analysis (notedN) n ¼ 8. Spectro-photometric assays (notedS) for ammonium and ascorbate:n ¼ 6. Since these compounds were analyzed on extracts fromfresh tissues, conversion to DW used the ratio DW� 100/FW ¼ 7.1670.2 determined after freeze drying the powderedpericarps. Liquid chromatography with diode array detection forLycopene, n ¼ 6 and n ¼ 7 fruit samples from control andhypoxically treated plants, respectively.Different letters in each row indicate significant differencescaused by treatment. (Student t test at the 0.05 (b) or 0.01 (c)

Root hypoxia affects the nutritional quality of tomato fruits 5

Effect of root hypoxia on fruit

Fruit growth and maturation

Fruit growth followed a sigmoid curve (Figure 3)and reached the red ripe stage between 40 and 45DPA irrespective of the growing conditions of theplants. However, a more rapid growth was observedbetween 10 and 35 DPA for fruits of HTcompared tocontrol plants. Hypoxia did not affect the final sizeof the fruits.

Metabolite content of fruits

Pericarps of mature fruit were analyzed for theirmetabolite contents (Table 2). Twenty-one of themajor water-soluble metabolites were identifiedand quantified by 1H-NMR spectroscopy. Theseincluded 4 soluble sugars, 3 organic acids, 10 aminoacids and the glutamate (Glu) derivatives GABA andpyroglutamate, choline and trigonelline, which allhave been previously identified in 1H-NMR spectra(Mounet et al., 2007). The levels of solublecarbohydrates were not significantly different infruit pericarp of control and HT plants. Togetherwith the similar fruit sizes, this result indicates thatthe supply of carbohydrates was not affected bythe treatment. The levels of organic acids, mainlycitrate and malate, and of most amino acids werealso not significantly different between control andtreated plants. This was confirmed by performing aprincipal component analysis of the data accordingto Mounet et al. (2007). The first and secondcomponents explained 82% and 12% of totalvariability, respectively, but did not discriminatebetween HT and control fruits (not shown). MeanGlu levels were found to be lower in fruits from HTplants (Table 2), but detailed examination of the

levels).

0

5

10

15

20

25

30

0 10 20 30 40 50

Time (DPA)

Dia

mete

r (m

m)

Figure 3. Effect of root hypoxia on tomato fruit growth.The diameter of tagged fruits growing on control (J) orhypoxically treated (’) plants was measured with anelectronic caliper. Each point represents the mean (7SD)of 8 fruits from 4 plants per condition.

Please cite this article as: Horchani F, et al. Prolonged root hypoxiaquality of tomato.... J Plant Physiol (2007), doi:10.1016/j.jplph.20

data revealed that Glu levels were not affected bytreatment in normally seeded fruits (Figure 4). Thesame was true for glutamine (not shown).

A major difference between control and HTplants was the 80-fold higher ammonium contentin the pericarp from HT plants (Table 2). Ammo-nium concentration was close to 7.5mM in controlfruits (Davies and Hobson, 1981; Baldet et al.,2002), but was equivalent to 600mM in HT plants.

Total ascorbate (AsA+dehydroascorbate) wasstrongly decreased and in a more oxidized state infruits from HT than control plants (Table 2).Lycopene, another major antioxidant of tomatofruits, was also decreased by 30% (Table 2). Thus,

induces ammonium accumulation and decreases the nutritional07.10.016

ARTICLE IN PRESS

0

50

100

150

200

250

300

350

0 20 40 60 80

seed number / fruit

glu

(µ

mol / g D

W p

ericarp

)

Figure 4. Relationship between glutamate concentrationin mature pericarp and seed number per fruit, in fruitsfrom control (J) or hypoxically treated (’) plants.

F. Horchani et al.6

major determinants of the nutritional quality oftomato fruits were severely reduced in fruits of HTplants.

Discussion

Flooding is a major abiotic stress that is known toaffect plant yield, and our aim was to examine towhat extent fruit quality might be affected. Thepresent study showed that root hypoxia, whileseverely disturbing roots and accelerating fruitmaturation, did not affect final fruit size or majorfruit metabolites. Because these results wereobtained using the determinate Micro-Tom variety(Marti et al., 2006) and hypoxia was applied at firstflower anthesis, possible effects of root hypoxia onflower formation that might occur in indeterminatevarieties could not be observed. The majorobservations presented here are that root hypoxialed to high accumulation of ammonium in the fruitand to decreased contents in ascorbate andlycopene, two nutritionally important componentsof the tomato fruit.

In roots of tobacco (Stoimenova et al., 2003) andtomato (Allegre et al., 2004), anoxia stimulates thereduction of nitrate to nitrite, with only smallamounts of ammonium observed in the root tissues(Allegre et al., 2004; Morard et al., 2004). Incontrast, in the present experiments, ammoniumformed was nearly equimolar with nitrate con-sumed. Since ammonium was formed within 1 day(Figure 2), when no contamination by algae orother microorganisms was visible, and the observa-tion was repeated during 3 different cultures, wefavor the hypothesis that ammonium was formed bythe tomato roots rather than by contaminating

Please cite this article as: Horchani F, et al. Prolonged root hypoxiaquality of tomato.... J Plant Physiol (2007), doi:10.1016/j.jplph.20

microorganisms. Similarly, Morard et al. (2004)observed equal uptake of nitrate with non-sterileor uncontaminated anoxic roots. We suggest thatthe conversion of nitrate to ammonium, instead ofonly nitrite, in the present conditions may beexplained by the mild hypoxic treatment used here.The hypoxic acclimation of the roots under hypoxiamay allow the activation, or induction, of therequired nitrite reductase in addition to theactivation of nitrate reductase, which is observedin anoxic roots (Stoimenova et al., 2003; Allegreet al., 2004).

With respect to the decreased levels of ascorbateand lycopene observed in the fruits of HT plants, itis important to discuss which of the differentphenomena induced by root hypoxia may beresponsible for this negative effect on fruit nutri-tional quality.

One major signal produced in flooded plants isethylene (Drew, 1997). The tomato fruit is climac-teric, i.e., it requires ethylene for ripening(Srivastava and Handa, 2005), but it is also sensitiveto ethylene in the early phases of its development.The treatment of tomato plants by ethrel, acompound that leads to ethylene release in solu-tions, induces prolonged phases of fruit cell divisionand expansion (Atta-Aly et al., 1999). The fruits aresmaller in the early phases of development butlarger when mature, and maturation is delayed.Ethylene was most likely produced by the HTplants, as indicated by leaf epinasty and theproduction of adventitious roots. However, ourobservations of (1) accelerated fruit developmentin treated plants and (2) identical final fruit size incontrol and HT plants is in contrast with the effectsof ethrel. This suggests that the ethylene producedas a result of hypoxia had little effect on the fruitsof the HT plants.

The taste and nutritional quality of the tomatofruit are a direct function of its metabolitecontent, which essentially results from the supplyof carbohydrates by leaves during fruit develop-ment (Carrara et al., 2001). In addition to playing acentral role in metabolism, soluble sugars regulatemany developmental and physiological processes inplants (Baldet et al. (2002), and citations therein).Since sugar starvation was found to disturb thedevelopment of young fruit (Baldet et al., 2002)and to limit the accumulation of lycopene andphytoene in pieces of tomato pericarp developingin vitro (Telef et al., 2006), the decreasedphotosynthesis caused by root hypoxia (Table 1)might be expected to affect fruit growth andcomposition. However, our results showed equalfinal sizes and similar levels of carbohydrate andother major metabolites in mature fruits of HT and

induces ammonium accumulation and decreases the nutritional07.10.016

ARTICLE IN PRESS

Root hypoxia affects the nutritional quality of tomato fruits 7

control plants (Table 2). This likely indicates similarrates of carbohydrate supply to fruits, despite thedecrease in leaf photosynthesis. The strong de-crease in root growth that occurred soon after thebeginning of the hypoxic treatment probablybalanced the decrease in photosynthetic activity,so that sufficient carbohydrates remained availablefor the fruits. It therefore seems that sugarstarvation cannot explain the decreases in lycopeneand ascorbate contents in fruits from HT plants.

The roots were the most strongly affected organin HT plants. The decreased availability of bothoxygen and photosynthates led to the arrest ofgrowth and progressive deterioration of the mainroots. In contrast, the good state of the aerial partof the plant indicates that 2 of the major functionsof the roots, the supply of water and minerals,remained sufficient, perhaps owing to the smallnumbers of roots in the vermiculite plug and thefew newly formed adventitious roots. Whetherthese roots correctly supplied the plants withsufficient hormones like cytokinins is not clear. Onthe one hand, the fact that additional bunches offlowers appeared in the same way on the HTand the control plants (not shown) may suggestthat nutrients and hormones were not limiting(Srivastava and Handa, 2005). On the other hand,the decreased level of Glu in fruits with a smallnumber of seeds points to a difference in fruitmetabolism that may be related with a modifiedhormonal status of the fruits of HT plants. Forexample, cytokinins are produced by the roots andalso, inside the fruits, by the developing seeds(Srivastava and Handa, 2005). In HT plants, thelatter might compensate a deficiency in rootcytokinins, depending on seed number. Alterna-tively, NO signaling resulting from the reduction ofnitrite in HT plants might play a similar role(Stoimenova et al., 2003, Libourel et al., 2006).Therefore, the decreased levels of ascorbate andlycopene in the fruits of HT plants might result froma modification of the hormonal status of the fruitscaused by disturbed root function.

The high concentration of ammonium accumu-lated in the medium by HT plants resulted in a veryhigh accumulation in the fruits (Table 2). Anincrease in ammonium has previously been de-scribed in fruits from dark-treated plants, and wasattributed to the catabolism of amino acids causedby sugar starvation (Baldet et al., 2002). However,the ammonium content found in that experimentwas about 30 times lower than that found in fruitsfrom HT plants. Therefore, the source of ammo-nium is probably the hypoxic medium, whichcontained up to 12mM of ammonium formed fromnitrate reduction (Figure 2). This high ammonium

Please cite this article as: Horchani F, et al. Prolonged root hypoxiaquality of tomato.... J Plant Physiol (2007), doi:10.1016/j.jplph.20

concentration in the fruit is consistent with themodel based on (1) free diffusion of NH3 betweenthe medium and the whole plant, and (2) accumu-lation of NH4

+ proportional to local pH. It may beconsidered that the average ammonium concentra-tion in the nutrient solution is 6mM (Figure 2B).The pH of the mineral solution was initially 5.8, andreached pH 6.2 at the end of the week. With 6mMammonium at pH 6.0 in the nutrient solution, the600mM concentration in the fruit would correspondto a fruit pH of 4.0, which is close to both the pH of3.77 effectively measured in the fruit juice (notshown), and the vacuolar pH value of 3.86measured by 31P NMR in ripe tomato fruits (Rolinet al., 2000). Whether high ammonium is respon-sible for the decreased levels of ascorbate andlycopene remains to be established. This would beconsistent with the observation that wateringtomatoes with mineral solution containing highammonium, compared to nitrate, leads to lowerlevels of lycopene, although this effect was alsodependent on the presence of other ions, likechloride, and on the assimilation of sulfur (Tooret al., 2006). However, in that case, higherascorbate levels were obtained, which is in contrastto our results. This suggests a specific role of roothypoxia that remains to be clarified.

Acknowledgments

This work was supported by the French INRA andby the Comite Mixte Franco-Tunisien pour laCooperation Universitaire (CMCU, Grant no.03G0209). We thank A. Roos and D. Just formanaging the phytotronic chambers.

References

Allegre A, Silvestre J, Morard P, Kallerhoff J, Pinelli E.Nitrate reductase regulation in tomato roots byexogenous nitrate: a possible role in tolerance tolong-term root anoxia. J Exp Bot 2004;55:2625–34.

Atta-Aly MA, Riad GS, Lacheene ZE, Elbeltagy AS. Earlyapplication of ethrel extends tomato fruit cell divisionand increases fruit size and yield with ripening delay. JPlant Growth Regul 1999;18:15–24.

Baldet P, Devaux C, Chevalier C, Brouquisse R, Just D,Raymond P. Contrasted responses to carbohydratelimitation in tomato fruit at two stages of develop-ment. Plant Cell Environ 2002;25:1639–49.

Beutler H-O. Ethanol. In: Bergmeyer J, Grassl M, editors.Metabolites 1: carbohydrates, vol. VI. Weinheim:Verlag Chemie; 1984. p. 598–606.

Carrara S, Pardossi A, Soldatini GF, Tognoni F, Guidi L.Photosynthetic activity of ripening tomato fruit.Photosynthetica 2001;39:75–8.

induces ammonium accumulation and decreases the nutritional07.10.016

ARTICLE IN PRESS

F. Horchani et al.8

Cataldo DA, Haroon M, Schrader LE, Youngs VL. Rapidcolorimetric determination of nitrate in plant tissueby nitration of salicylic acid. Commun Soil Sci PlantAnal 1975;6:71–80.

Davies J, Hobson G. The constituents of the tomato fruit.The influence of environment, nutrition, and geno-type. CRC Crit Rev Food Sci Nutr 1981;15:205–80.

Devaux C, Baldet P, Joubes J, Dieuaide-Noubhani M, JustD, Chevalier C, et al. Physiological, biochemical andmolecular analysis of sugar-starvation responses intomato roots. J Exp Bot 2003;54:1143–51.

Drew MC. Oxygen deficiency and root metabolism: injuryand acclimation under hypoxia and anoxia. Annu RevPlant Physiol Plant Mol Biol 1997;48:223–50.

Geigenberger P, Fernie AR, Gibon Y, Christ M, Stitt M.Metabolic activity decreases as an adaptive responseto low internal oxygen in growing potato tubers. BiolChem 2000;381:723–40.

Ho L. Tomato. In: Zamski E, Schaffer A, editors.Photoassimilate distribution in plants and crops.Source–sink relationships. New York: Marcel Dekker,Inc.; 1996. p. 185–208.

Joubes J, LemaireChamley M, Delmas F, Walter J,Hernould M, Mouras A, et al. A new C-type cyclin-dependent kinase from tomato expressed in dividingtissues does not interact with mitotic and G1 cyclins.Plant Physiol 2001;126:1403–15.

Kampfenkel K, Vanmontagu M, Inze D. Extraction anddetermination of ascorbate and dehydroascorbatefrom plant tissue. Anal Biochem 1995;225:165–7.

Libourel IGL, Bodegom PMv, Fricker MD, Ratcliffe RG.Nitrite reduces cytoplasmic acidosis under anoxia.Plant Physiol 2006;142:1710–7.

Marti E, Gisbert C, Bishop GJ, Dixon MS, Garcia-MartinezJL. Genetic and physiological characterization oftomato cv Micro-Tom. J Exp Bot 2006;57:2037–47.

Miranda KM, Espey MG, Wink DA. A rapid, simplespectrophotometric method for simultaneous detec-tion of nitrate and nitrite. Nitric Oxide 2001;5:62–71.

Morard P, Silvestre J, Lacoste L, Caumes E, Lamaze T.Nitrate uptake and nitrite release by tomato roots inresponse to anoxia. J Plant Physiol 2004;161:855–65.

Please cite this article as: Horchani F, et al. Prolonged root hypoxiaquality of tomato.... J Plant Physiol (2007), doi:10.1016/j.jplph.20

Mounet F, Lemaire-Chamley M, Maucourt M, Cabasson C,Giraudel J-L, Deborde C, et al. Quantitative metabolicprofiles of tomato flesh and seeds during fruitdevelopment: complementary analysis with ANN andPCA. Metabolomics 2007;3:273–88.

Pezeshki SR. Wetland plant responses to soil flooding.Environ Exp Bot 2001;46:299–312.

Rahayu YS, Walch-Liu P, Neumann G, Romheld V, WirenNv, Bangerth F. Root-derived cytokinins as long-distance signals for NO3-induced stimulation of leafgrowth. J Exp Bot 2005;56:1143–52.

Rolin D, Baldet P, Just D, Chevalier C, Biran M, RaymondP. NMR study of low subcellular pH during thedevelopment of cherry tomato fruit. Aust J PlantPhysiol 2000;27:61–9.

Saglio PH, Drew MC, Pradet A. Metabolic acclimatation toanoxia induced by low (2–4 kPa partial pressure)oxygen pretreatment (hypoxia) in root tips of Zeamays. Plant Physiol 1988;86:61–6.

Srivastava A, Handa AK. Hormonal regulation of tomatofruit development: a molecular perspective. J PlantGrowth Regul 2005;24:67–82.

Stoimenova M, Libourel IGL, Ratcliffe RG, Kaiser WM. Therole of nitrate reduction in the anoxic metabolism ofroots. II. Anoxic metabolism of tobacco roots with orwithout nitrate reductase activity. Plant Soil2003;253:155–67.

Szal B, Drozd M, Rychter AM. Factors affecting determi-nation of superoxide anion generated by mitochondriafrom barley roots after anaerobiosis. J Plant Physiol2004;161:1339–46.

Telef N, Stammitti-Bert L, Mortain-Bertrand A, MaucourtM, Carde JP, Rolin D, et al. Sucrose deficiency delayslycopene accumulation in tomato fruit pericarp discs.Plant Mol Biol 2006;62:453–69.

Toor RK, Savage GP, Heeb A. Influence of differenttypes of fertilisers on the major antioxidant com-ponents of tomatoes. J Food Compos Anal 2006;19:20–7.

Wintermans J, Mots A. Spectrophotometric characteristicof chlorophylls a and b and their pheophytins inethanol. Biochem Biophys Acta 1965;109:448–53.

induces ammonium accumulation and decreases the nutritional07.10.016