Different induction patterns of glutamate metabolizing enzymes in ripening fruits of the tomato mutant green flesh

Different induction patterns of glutamate metabolizing enzymes in

ripening fruits of the tomato mutant green flesh

Santiago Bortolotti, Silvana B. Boggio, Luciana Delgado, Elena G. Orellano and Estela M. Valle*

Division Biologıa Molecular, Instituto de Biologıa Molecular y Celular de Rosario (IBR-CONICET), Facultad de Ciencias Bioquımicasy Farmaceuticas, UNR, Suipacha 531, S2002LRK Rosario, Argentina*Corresponding author, e-mail: [email protected]

Received 19 December 2002; revised 4 April 2003

The stay-green mutant of tomato green flesh (gf) is deficientin the chloroplast degradation machinery and displays analtered ripening profile. The ripening process of tomato

(Lycopersicon esculentum Mill.) fruit is characterized by

a substantial induction in glutamate dehydrogenase (GDH,

EC 1.4.1.3) and a decline in glutamine synthetase (GS, EC6.3.1.2) levels, associated with an increase in the relative

glutamate content. In this study the level of free amino acids

and enzymes associated with glutamate metabolism werefollowed in mature gf fruits at green and ripe stages. From

these fruits, total RNA, protein extracts and intact plastids

were isolated. Specific activities of GDH, GS and NADH- or

ferredoxin-dependent glutamate synthase (NADH-GOGAT,EC 1.4.1.14; Fd-GOGAT, EC 1.4.7.1) were measured and

immunoblot analyses were performed. The total free amino

acid contents of gf mature fruits decreased during ripening,

contrasting with that observed in several cultivated varietiesof tomatoes. The relative glutamate content, however,

increased markedly in ripe gf fruits, as in all the tomato

cultivars tested so far. Although GDH activity and polypep-

tides were found in green gf fruits, GDH transcripts were notdetected in these fruits. In ripe gf fruits GDH was clearly

present. NADH- and Fd-GOGAT activities, on the other

hand, were barely measurable in gf fruits whereasFd-GOGAT protein was detected in green gf fruits and, to

a lower extent in ripe gf fruits. Levels of GS protein and

transcripts were correlated in all fruits tested. Collectively,

the expression of the enzymes involved in the primary glutam-ate metabolism seems to be differently regulated during the

ripening process.

Introduction

The ripening process is a co-ordinated regulatory eventinvolving changes in colour and texture of fruits, whichalso intensify their taste and flavour. These quality attri-butes are generated by a diversity of ways, implyingsignificant differences in the activity of the enzymesinvolved. During the early phases of tomato ripening,chloroplasts differentiate into chromoplasts, and conse-quently, chlorophyll is metabolized, thylacoids aredegraded and carotenoids, including b-carotene andlycopene, accumulate (Marano et al. 1993, Fraser et al.1994). Simultaneously, reductions in the levels ofmRNAs and proteins associated with photosynthesisoccur (Piechulla et al. 1987), through a controlled inter-action of complex gene expression patterns. Neverthe-

less, regulation and synergy of the multiple processescontributing to the ripe phenomena remain largelyunknown (Giovannoni 2001).

Changes in carbon and nitrogen assimilation pathwaysand different expression patterns of the enzymesinvolved during the ripening process has been reported(Piechulla et al. 1987, Gallardo et al. 1993, Boggio et al.2000). Furthermore, during ripening there is a substan-tial increase in the content of several components offruit, such as sucrose (Yelle et al. 1991), hexoses (glucoseand fructose), citrate (Rolin et al. 2000) and glutamate(Boggio et al. 2000, Kisaka and Kida 2003). In the caseof sucrose, these changes were associated with the co-or-dinated expression of sucrose synthase and sucrose

PHYSIOLOGIA PLANTARUM119: 384–391. 2003 Copyright# Physiologia Plantarum 2003

Printed in Denmark – all rights reserved

Abbreviations – BSA, bovine serum albumin; Fd, ferredoxin; GABA, g-aminobutyric acid; GB, grinding buffer; GDH, glutamatedehydrogenase; GOGAT, glutamate synthase; gf, green flesh; GS, glutamine synthetase.

384 Physiol. Plant. 119, 2003

phosphate synthase, which participate in the control ofsucrose import capacity (Schaffer and Petreikov 1997,D’Aoust et al. 1999). The sources of glutamate in ripefruits, however, are not clear. In an earlier report, wehave shown an apparent induction of enzymes partici-pating in the glutamate metabolism in ripening tomatofruits, such as aspartate aminotransferase and GDH, anda marked decreased in the levels of glutamate decarbox-ylase and glutamine synthetase (GS) (Boggio et al. 2000).These changes in the enzyme patterns were associatedwith an increase in the relative content of glutamate inripe tomato fruits.

Glutamate is found in the intersection of several meta-bolic pathways (Valle 2003). In the chloroplast, gluta-mate is produced by the concerted action of glutaminesynthetase (GS2) and ferredoxin-dependent glutamatesynthase (Fd-GOGAT), known as the GS-GOGATcycle, which is responsible for the prevalent NH4

1 assim-ilation in plants (Lea and Ireland 1999). In tomato cv.Platense, the levels of GS activity were similar whenassayed in leaf or green fruit pericarp extracts, decliningto undetectable levels in ripening fruits (Boggio et al.2000). On the other hand, Fd-GOGAT could only beassayed in intact green fruit chloroplasts (Buker et al.1998). The decay of GS/Fd-GOGAT activities in ripetomato fruits may be associated with the degradationof chloroplasts, since the integrity of the plastids seemsto be a prerequisite for optimal GS2 activity (Cren andHirel 1999). Moreover, in tomato leaves, GS2 showed arapid decline by natural senescence (Perez-Rodrıguezand Valpuesta 1996). Another putative source of gluta-mate is the reversible reaction catalysed by the mitochon-drial NAD(H)-dependent glutamate dehydrogenase(GDH), which was markedly induced during the ripeningprocess (Boggio et al. 2000). The direction of the GDHenzymatic reaction would depend on N and C sources(Lancien et al. 2000). In tobacco, the contribution ofGDH to the assimilation of ammonium was increasedin the older leaves in comparison with young leaves,whereas GS2/Fd-GOGAT involvement in ammoniarecycling and assimilation was greatly reduced(Masclaux et al. 2000). Similarly, in avocado (Loulakakiset al. 1994) and ripe tomato fruits, in which the sinkstrength is reduced (Ho 1988), the induction of GDHwas paralleled with a decrease of GS (Boggio et al. 2000).Recently, a GDH-shunt to the GS-GOGAT cycle hasbeen proposed (Miflin and Habash 2002) to provide amechanism that could respond to the differing needs ofcells for carbon and nitrogen compounds.

Tomato mutants have been amply used to study themolecular basis of the ripening process (Gray et al.1994). In some of them, multiple aspects of fruit ripeningare modified pleiotropically, like the lack of pigment inthe pericarp tissue in the dominant colourless non-ripening mutant (Thompson et al. 1999), the impairmentof colour change and softening by insensitivity to ethylenein the dominant Never-ripe (Rick and Butler 1956,Wilkinson et al. 1995), or the severely altered fruit flavour,colour, softening and ethylene production of ripening

inhibitor mutant (Tigchelaar et al. 1978). Among thetomato mutants where fruits develop and ripen atabout the same rate as the standard tomato varietiesbut showing alteration in the ripening process, is themutant gf (Kerr 1956, Akhtar et al. 1999). This mutantexhibits an alteration in the normal chloroplast to chro-moplast transition occurring during ripening (Cheunget al. 1993) and as a result, the ripe fruits present arusty or dirty red colour. The ultrastructure of thecarotenoid-containing plastids in ripe gf fruit retainedsignificant chloroplast thylakoid stacks that were nolonger found in the chromoplasts of wild-type fruit(Cheung et al. 1993). During ripening of wild-type fruitthe level of some photosynthetic transcripts and proteinsare highly reduced (Piechulla et al. 1987, Cheung et al.1993).

To study the specific decline in GS and the inductionof GDH associated with the ripening process in tomatofruits, we have chosen the green flesh (gf) mutant as ourexperimental system. We examined the free amino acidcontent and the expression patterns of GS, Fd-GOGATand GDH in the pericarp of gf fruits during the ripeningtransition.

Materials and methods

Plant material

Tomato (Lycopersicon esculentum Mill.) plants of cv.Platense and gf mutant (seeds were obtained from DrC.M. Rick, University of California, Davis) were grownin a controlled environment cabinet under a lightintensity at the top of a fruit-containing plant of700mmolm�2 s�1. The temperature ranged from 23�Cduring the light period (14 h) to 18�C in the dark andthe relative humidity was 70%. Plants were grown in soil,maintained under optimal irrigation and supplied withhalf-strength Hoagland nutrient solution (Boggio et al.2000). Fully expanded fourth leaves of 30-day-old plantswere used for activity assays and immunodetection ofGDH, GS and Fd-GOGAT. Fruits were allowed toripen naturally on the plant. They were harvested atmature stages (when fruits stopped growing), eithergreen, breaker or red just when fruits began to soften(cv. Platense was red and gf was rusty-red). Pericarptissue of the harvested fruits was obtained by removingthe locule tissues and seeds and immediately processedor frozen in liquid nitrogen and stored at �80�C untilanalysis.

Crude protein extraction

Proteins were extracted from fresh tissues for enzymeassays or from frozen tissues for western blot analyses.The procedure used for protein extraction from pericarptissues was essentially as described (Boggio et al. 2000)with the addition of fresh phenylmethylsulphonyl fluor-ide to a final concentration of 1mM in all extracts.

Physiol. Plant. 119, 2003 385

Plastids isolation

Intact chloroplasts from green gf fruits and chromoplastfrom rusty red gf fruits were isolated and purified fromfresh pericarp tissue according to Hadjeb et al. (1988).Briefly, 24 g fresh weight fruit pericarp was homogenizedwith grinding buffer (GB) containing 50mM Hepes,pH8.0, 330mM sorbitol, 2mM EDTA, 1mM MgCl2,2mM MnCl2, 5mM L-ascorbic acid (sodium salt),bovine serum albumin (BSA) 0.1% (w/v) in a Polytron(Bachofer GmBH, Weilheim/Teck, Germany). Thehomogenate was filtered through four layers of cheese-cloth and plastids were recovered by centrifugation at800 g for 10min. The plastids were suspended in GB andfurther purified on a continuous Percoll gradient [10 to79% (v/v) for chloroplasts and 5 to 40% (v/v) for chro-moplasts] by centrifugation at 6000 g for 20min. Intactplastids were collected from the lower layer recoveredafter centrifugation. Broken plastids remained in thetop layer. They were washed with GB and pelleted at3000 g for 20min. The integrity of the organelles wasjudged by microscopy. The plastids were washed withGB without BSA and plastidic proteins were obtained byre-suspending the plastids in 200ml GB one-tenth dilutedand centrifuging at 27 000 g for 15min.

Determination of pigments content

Pigments were extracted from the fresh pericarp tissuewith a mixture of acetone : hexane (4 : 6), cleared by cen-trifugation and the optical density of the supernatantwas measured at 663, 645, 505 and 453 nm. From thesevalues the content of chlorophyll a, b, lycopene andb-carotene were calculated (Nagata and Saito 1992).

Determination of free amino acid content

Frozen pericarp tissues were extracted with chloro-form :methanol (Valle et al. 1998). The amino acid com-position in the methanolic phase was determined byderivatization with o-phthaldialdehyde and analysed ina HPLC system (Riens et al. 1991).

Protein determinations and enzyme assays

Protein concentrations were determined according to themethod of Bradford (Bradford 1976). GS, GDH andGOGAT activities were measured as previouslydescribed (Boggio et al. 2000).

PAGE and immunoblotting

Proteins were fractionated by native PAGE or SDS-PAGEon 8–10% (w/v) acrylamide gels, and electroblottedto nitrocellulose membranes overnight at 4�C. Whenindicated, the SDS gels were stained using CoomassieBrilliant Blue. Immunodetection was carried out accord-ing to the manufacturer’s procedure (ECL AmershamBiosciences, Argentina), using antisera raised in rabbits

against Zea mays GS2 and Fd-GOGAT (Sakakibaraet al. 1992), and Vitis vinifera NADH-GDH (Loulakakisand Roubelakis-Angelakis 1990).

RNA analysis by semi-quantitative reverse transcription-

polymerase chain reaction (RT-PCR)

Total RNA was extracted from leaves and fruit pericarpusing RNeasy Plant Mini kit (Qiagen Inc., USA) accord-ing to the manufacturer’s protocol. The first strand ofcDNA was synthesized from 2mg total RNA with theoligo (dT)22 adaptor as the primer using 30 U AMVreverse transcriptase (USB, Cleveland, OH, USA) in afinal volume of 45ml. After heat inactivation (15min at70�C), one-tenth of the reaction products served as thetemplate in the PCR using oligonucleotides correspond-ing to L. esculentum cDNAs of GDH (Accession no.U48695) (sense primer 50-TGGATTCAGGGTACAG-CACGAC-30 and antisense primer 50-GCAGCCCAG-GAACCAACATTAC-30) or GS (Accession no.AF200360) (sense primer 50-TCAAGCTCCTGGA-GAAGACA-30 and antisense primer 50-CTCATC-GACTTGGTGCTGTA-30) giving rise to PCR productsof 509 or 591 bp, respectively. As a control, a fragment of328 bp corresponding to the gene encoding 25S riboso-mal rRNA from tomato was amplified with two specificoligonuleotides (50-GTTAGTCGATCCTAAGGGTC-30

and 50-CCTGATGCGGTTATGAGTAC-30). The oligo-primers were commercially synthesized (Bio-Synthesis,Lewisville, TX, USA). Amplifications were conducted ina thermal minicycler (Eppendorf, Mastercycler personal;Hamburg, Germany) in 34 cycles for GS and GDH or 20cycles for 25S ribosomal rRNA. The number of PCRcycles was chosen within the linear range of amplifica-tion of each transcript tested. The gene amplificationreaction was conducted at 94�C for 90 s, 55�C for 180 sand 72�C for 180 s for the specified cycles. Same aliquotsof the reaction products were analysed on 1% (w/v) agarosegel, blotted on Hybond-N membranes (Amersham) andhybridized with the corresponding cDNA probes32P-labelled by random-priming using a Prime-a-GeneLabelling System kit (Promega, Madison, USA). Membraneblots were prehybridized and hybridized in 6�SSC,5�Denhardt’s and 0.5% (w/v) SDS (Maniatis et al. 1982)at 65�C. The blots were washed at 65�C, initially in 2� SSC,0.5% SDS, then in 0.2� SSC, 0.1% SDS, and finallyexposed to X-OMAT AR film (Kodak, Rochester/NewYork, USA) at �70�C using a single intensifying screen.Signal intensities were determined with a scanner and theGEL PRO ANALYSER 4.0 software (MediaCybernetics, Silver-Spring, MD, USA). Data were normalized to the 25Sribosomal rRNA amplifier.

Results

Pigment contents of mature fruits

The most obvious difference observed between fruits ofgf and the cv. Platense is the contrast in the colour


change during the ripening process. The gf fruits undergonormal development to the red-ripe stage but, unlike thoseof other varieties, they keep their chloroplasts, and, as aresult, a rusty-red fruit is observed (Akhtar et al. 1999).Table 1 shows the pigment content of tomato fruits fromcv. Platense, with the normal ripening process, and fromthe gf mutant at two ripening stages. The mutant failed toshow any reduction in the chlorophyll (a1 b) content thatnormally occurs during ripening of fruits, whereas lyco-pene accumulated to similar levels in ripe fruits of bothtomato types. These results are similar to those previouslypublished (Akhtar et al. 1999). When the b-carotene con-tent was determined, a much higher content was observedin gf fruits, mature green as well as rusty-red fruits, incomparison with the fruit from cv. Platense (Table 1).

Free amino acid composition of ripening gf fruits

The free amino acid content of gf fruit pericarp wasdetermined during the transition from mature green torusty-red (Fig. 1). The total free amino acid contents ofgf mature fruits decreased during ripening contrastingwith that observed in several cultivated varieties of toma-toes (Boggio et al. 2000). The relative molar contents ofseveral individual amino acids were different at eachripening stage analysed. Glutamine was the most abun-dant amino acid in mature fruits at all stages, showing a12% decrease during ripening. Gamma-aminobutyricacid (GABA) was the second most abundant aminoacid in the green and breaker fruits showing a substantialreduction upon ripening (5.6-fold from green and 6.8-fold from breaker to rusty-red stage). This markedreduction in GABA content of ripe fruits was notobserved at all in cv. Platense or it was less pronouncedin cv. Vollendung (2- or 1.4-fold from green or breakerto rusty-red stage, respectively) (Boggio et al. 2000).Asparagine, serine and glycine contents showed a slightdecrease during fruit ripening. Glutamate contentshowed higher levels in ripe fruits (3.9 mmol g�1 FW ofrusty-red fruit compared to 1.6mmol g�1 FW of greenfruit) becoming the second most abundant amino acidin the rusty-red stage after glutamine. This pattern ofincrease in the glutamate content of ripe fruits is inagreement with earlier results obtained with othertomato varieties (5.2-, 6.4-, 8.7- and 11-fold forcvs. Platense, Minitomato, Cherry and Vollendung,

Table 1. Pigment content of tomato fruits from cv. Platense andgreen flesh mutant. The cv. Platense and gf mutant fruits wereharvested at two ripening stages: mature green and ripe (red for cv.Platense and rusty-red for gf fruits), 5 days after breaker stage. Thepericarp was ground with a mixture of acetone : hexane for pigmentextraction and the pigment content was determined at differentwavelengths as described in Materials and Methods. Each valuerepresents the means– SE of three independent experiments.

Fruit pigments (mg g�1 fresh weight)

Fruit stage Chlorophyll (a1 b) Lycopene b-carotene

cv. Platense, green 44.1– 4.1 4.9 – 0.5 4.1 – 0.4cv. Platense, red 8.6– 0.9 67.10– 5.7 7.7 – 6.7gf mutant, green 34.0– 3.5 19.7– 2.1 34.1– 2.9gf mutant, rusty-red 33.2– 4.0 63.0– 5.9 42.4– 4.5

Fig. 1. Relative free amino acid content of mature green flesh fruits during the ripening process. Fruits were harvested at mature green, breakerand ripe (5 days after breaker stage) and the molar contents of free amino acids were assayed in the methanolic phase of pericarp extracts byderivatization with o-phthaldialdehyde as described in Materials and methods. The values obtained for each amino acid (mmol g�1 FW) werereferred to the total free amino acid content of the sample analysed and indicated as relative molar content (%). Results are means of fourreplicates with different fruits of approximately the same fresh weight and chlorophyll or lycopene content for each maturation stage. The means–SE were less than 15%. Total free amino acid contents were 36.05– 5.05; 22.73– 3.10 and 14.02– 1.02mmol g�1FW of green, breaker and rusty-red fruits, respectively. Empty, grey and black bars represent relative amino acid values of green, breaker and rusty-red gf pericarp fruits.


respectively) (Boggio et al. 2000, Kisaka and Kida 2003).Aspartate and alanine contents also increased during theripening events reaching levels of 10 and 7%, respect-ively, of the total free amino acids in rusty-red fruits.

Patterns of glutamate-metabolizing enzyme activity and

protein in mature gf fruits

GS, GOGAT and GDH were determined in mature gffruits at different ripening stages. Specific activity andWestern blot analysis were carried out in the proteinextracts from pericarp of mature green, breaker andrusty-red gf fruits.

Glutamine synthetaseThe GS activity was assayed in gels by activity stainingafter separation of leaf and fruit extracts by non-denaturing polyacrylamide gels. GS activity was observedin both organs although in the leaf was detected at higherlevels than in green or rusty-red fruits (Fig. 2A). Thisresult differs from cv. Platense, where the GS activity inthe leaf and green fruit extracts were similar and GS wasnot detected in ripe fruits (Boggio et al. 2000). Leafextract revealed the presence of two clear bands withGS activity, which became diffuse a few seconds afteraddition of the staining media, and an additional band ofhigher mobility appeared later (Fig. 2A). In the case ofgreen or rusty-red fruit only one band with GS activitywas observed, which co-migrated with one of the upperleaf band (Fig. 2A). The presence of GS activity bandswith different mobility in the gf leaf extracts suggeststhat more than one catalytically active GS oligomer ispresent. An immunoblotting experiment was performed

with an additional gel run in parallel under the same non-denaturing conditions and using antiserum against GS2 ofZea mays as described in Materials and methods (Fig. 2B).Leaf extract showed three immunoreactive bandsco-migrating with the detected GS activity bands (compareFig. 2A and B). In fruit extracts, one additional GS bandwith a faster mobility was observed (Fig. 2A and B), whichwas inactive and clearly visible in red fruits (Fig. 2B).

The total protein profiles in SDS-PAGE differed con-siderably between green and rusty-red fruits (Fig. 3A).To determine the polypeptide composition of GS in theleaf and fruits, western blots experiments after SDS-PAGE were made using antibodies against Zea maysGS2, which recognized both cytosolic GS1 and chloro-plastic GS2 (Sakakibara et al. 1992). This experimentrevealed the presence of two polypeptides of 48 and45 kDa in the leaf extracts with the larger form dominat-ing (Fig. 3B). However, in the green and rusty-red fruitsonly one immunoreactive band of 45 kDa was observed(Fig. 3B). In order to clarify whether this 45 kDa bandcould be a GS1 or GS2 subunit, we isolated plastids fromleaf and fruits of wild-type tomato and gf mutant andperformed a similar immunoblotting experiment. GS wasdetected in plastids of both tomato plants, except inchromoplasts of wild-type fruits (Fig. 3C). Two bandsof 48 and 45 kDa were observed in leaf chloroplastsand a unique 45 kDa GS subunit was detected in plastidsof both tomato fruits (Fig. 3C).

Fig. 2. Analysis of native GS in fruits and leaf of green flesh plants.Protein extracts (46mg) were separated in two non-denaturingpolyacrylamide gels (8%) by electrophoresis and analysed for GSactivity (A) and protein (B). After electrophoresis, one gel wasincubated with GS reaction mixture and subsequently stained asdescribed (Boggio et al. 2000) (A). The second gel was blotted on anitrocellulose membrane and immunoassayed using antiserum raisedagainst Zea maysGS2 (Sakakibara et al. 1992) and ECL (Amersham)as detection system (B). L, G and R mean leaf, mature green fruit andrusty-red fruit, respectively. The arrows indicate the GS bandsdetected by activity (panel A) or immunodetection (panel B) assays.

Fig. 3. Comparison of GS polypeptide patterns of leaf and totalfruit extracts with leaf and fruit plastid lysates of tomato cv.Platense and green fleshmutant. Leaf and total fruit protein extracts(46mg) (panels A and B) or chloroplast and chromoplast lysates(from 3 g leaf or 6 g fruit fresh weight) (panel C) (see Materials andmethods for the isolation procedure) were fractionated by SDS-PAGE(10%) and then, either stained with Coomassie Brilliant Blue (A), orblotted on nitrocellulose membrane and immunoassayed as describedin Fig. 2 (B and C). The migration of molecular mass markers areindicated on the left part of panel A and C. L, G andRmean as Fig. 2.WT means wild type (cv. Platense) and Gf means green flesh.


Glutamate synthaseActivity and polypeptides of NADH- or Fd-GOGATwere barely detected in tomato fruits of cv. Platense(Boggio et al. 2000). We assayed in leaf and fruit extractsof the gf mutant the presence of these enzymes by meas-uring specific activity or immunoblot analysis. NADH-or Fd-GOGAT-specific activities were undetectable in gfleaf or fruit extracts. However, a distinct and uniqueband corresponding to Fd-GOGAT in the leaf as wellas in green and rusty-red fruit extracts was visualized ingf fruit extracts by western blot analysis using antiserumraised against maize leaves Fd-GOGAT (Sakakibaraet al. 1992). As expected, Fd-GOGAT was present inthe leaf extract and in green gf fruits, although its con-tent was much lower in rusty-red fruit extracts (Fig. 4).In another cv. Microtom with normal ripening Fd-GOGAT protein or activity were undetectable in greenfruits (S. B. Boggio unpublished data).

Glutamate dehydrogenaseAnalysis of GDH activity was performed in extracts ofmature gf fruits and the results are shown in Table 2. Theaminating activity of GDH was higher in green fruitsthan in rusty-red fruits. In the wild-type fruits, however,the aminating activity (NADH-GDH) was 1.9 timeshigher in red than in green fruits. The GDH activityratio of aminating- versus de-aminating-reaction shiftedfrom 0.58 in green gf fruits to 1.55 in rusty-red fruits(Table 2). In the wild type, this GDH activity ratio was1.1 in red fruits whereas in green fruits it was muchlower. Figure 4 shows an immunoblot analysis per-formed with pericarp extracts of gf fruits at differentripening stages. GDH protein was present in the leafand fruits gf mutant showing two distinct bands cross-reacting with NAD (H)-GDH antibody. In the gf fruitssimilar levels of GDH were observed regardless of ripen-ing stage, although the upper band was more intensethan in leaf extracts (Fig. 4).

Expression of GS and GDH

In order to correlate the GS and GDH protein patternswith the transcripts level, both mRNAs were detected bysemi-quantitative RT-PCR in gf leaves and fruits. As con-trol 25S ribosomal rRNA was used. This technique waschosen because in gf fruits the GS and GDH transcriptswere too low to be detected by northern blot analysis.

To detect GS transcripts encoding both GS1 and GS2,a cDNA probe (Accession no. AF200360) was used,corresponding to the region between the highly con-served domains I and V of GS (M. L. Marro unpub-lished data). GS transcripts were observed at similarlevels at both maturation stages in gf fruits (Fig. 5) andquantitative analysis revealed that they were twice thelevel found in the leaves. In contrast, GDH transcriptswere absent in green gf fruits and in the leaves the level ofGDH transcripts were more than 3.5 times higher than inripe gf fruits. In the cultivated varieties the level of GDHtranscripts in the leaves were also higher than in thefruits (S. Bortolotti unpublished data).

Fig. 4. Western blot analysis of Fd-GOGAT and GDH of leaf andfruit extracts of green flesh plants. Protein extracts (46mg) from leaf,green and rusty-red fruits were fractionated by SDS-PAGE (8%)and then blotted on nitrocellulose membrane, immunoassayed usingantiserum raised against Zea mays Fd-GOGAT (Sakakibara et al.1992) or Vitis vinifera NADH-GDH (Loulakakis and Roubelakis-Angelakis 1990) and detected by ECL (Amersham) system. Arrowson the right indicate two migrating bands of GDH.

Table 2. GDH activity in ripening green flesh fruits. The NAD (H)-GDH specific activities were determined in freshly preparedpericarp extracts as previously described (Loulakakis andRoubelakis-Angelakis 1990). Each value represents the means– SE

of three independent experiments.

Aminating reaction Deaminating reactionStage of ripening (mmol h�1mg�1 protein) (mmol h�1mg�1 protein)

Green 2.88– 0.24 4.98– 0.42Rusty-red 1.86– 0.30 1.20– 1.12

Fig. 5. GS and GDH expression in mature green and ripe fruits ofgreen flesh plants. RT-PCR products of total RNA isolated fromgreen or rusty-red fruits of gf mutant were electrophoresed on 1%agarose gel, blotted on Hybond-N membranes and hybridized with a32P labelled cDNA encoding GS (accession no. AF200360) or GDH(accession no. U48695). As control 25S ribosomal rRNA was used.All steps were carried out as described in Materials and methods.


Discussion

In this study experiments were performed to analyse theregulation of glutamate metabolizing enzymes, localizedin different cell compartments, during the ripening pro-cess in the mutant gf with altered fruit-ripening charac-teristics.

The total free amino acid contents of cultivated fruitvarieties increased about two-fold during ripening(Boggio et al. 2000). However, in the gf fruits, itdecreased approximately 2.5-times, probably due to themaintenance of chloroplasts in ripe fruits. In the case ofgf leaves, a prominent increase in free amino acids duringsenescence was observed (Akhtar et al. 1999), indicatingthat protein degradation takes place as in the wild-typegenotype. Nevertheless, a diminished proteolytic activityin ripe gf fruits cannot be ruled out.

The gf fruits showed the typical increase in the gluta-mate content during the ripening transition as observedbefore in other tomato varieties with a normal ripeningprocess (Boggio et al. 2000, Kisaka and Kida 2003).These results clearly indicate that the accumulation offree glutamate in the pericarp of ripe tomato fruits is notassociated with the chloroplast–chromoplast transition.Interestingly, the GABA content of gf fruits showed asharp decline during ripening, similarly to that observedin the cv. Cherry (Boggio et al. 2000, Rolin et al. 2000).These opposite changes in glutamate and GABA levelsobserved in tomato fruits during ripening parallels theday–night oscillations of these amino acids in tobaccoleaves, where a 5–6 h delay was observed between theaccumulation of GABA preceding that of glutamate(Masclaux-Daubresse et al. 2002). These data reinforcethe hypothesis that GABA is buffering the glutamateproduction (Masclaux et al. 2000, Masclaux-Daubresseet al. 2002). The increase in alanine relative contents (7%of total free amino acids) during ripening in rusty-redfruits is distinctive of gf fruit, since alanine contents werealways less than 2% at any ripening stage in the tomatocultivated varieties analysed so far. The variations in thecontent of other free amino acids in mature gf fruits weresimilar to that previously published (Boggio et al. 2000).

Activities or polypeptides of GS, GOGAT and GDHhad been detected at discrete levels in mature gf fruits atboth green and rusty-red stages. GS was presumablypresent at different oligomeric arrangements in leaf andmature fruits, although only one oligomeric stage seemedto be active in the fruits. The GS band of higher mobilitybut of undetectable GS activity observed in green gffruits could correspond to an oligomer formed by anassociation of a smaller number of subunits (Mack1998). Catalytically active octameric and tetrameric GShad already been observed in Beta vulgaris (Mack 1998).These results also showed an important difference withthe GS activity pattern of other tomato varieties withnormal ripening: in gf fruits GS activity declines in thetransition from green to rusty-red but do not disappearcompletely by ripening as in red fruits of cv. Platense(Boggio et al. 2000). The level of GS polypeptide, how-

ever, was unchanged during the ripening transition asrevealed by SDS-PAGE followed by western blot analy-sis. In this experiments the migration of GS in fruitextracts was faster than in leaf extracts. To test if GS2was present or not in fruit plastids, intact chloroplastsand chromoplasts were isolated from wild-type and gfleaf and fruits and analysed for their GS subunit type.Leaf chloroplasts showed two GS subunits of 48 and45 kDa apparent molecular masses, whereas fruit plastidsrevealed a clear band comigrating with the 45 kDa sub-unit. Two bands of GS2 of different molecular masseshave already been reported in tomato cotyledons (Miggeet al. 1996). These results suggest that in wild-type and gffruits, plastidic GS and, if present, the cytosolic GS arecomposed of subunits with the same size. Mack (1998)obtained similar results. Moreover, GS was present inplastids of green as well as of rusty-red gf fruits, indicat-ing that its presence was associated to the occurrence ofchloroplasts. The GS transcripts, like the polypeptides,varied similarly during fruit ripening in the gf mutant,suggesting that GS expression could be transcriptionallyregulated.

GDH and Fd-GOGAT exhibited distinct patterns ofinduction in gf and cv. Platense fruits. The lack of GDHprotein in green cv. Platense fruits (Boggio et al. 2000)was similar to the lack of GDH activity in sink leaves oftobacco (Masclaux et al. 2000). In tobacco, GDH activ-ity does not contribute to the increase in ammoniumcontent in source leaves but rather participates in thereplenishment of glutamate as a consequence of aprogressive disappearance of Fd-GOGAT activity(Masclaux et al. 2000). Considering the sink/source tran-sition of tobacco leaves in parallel to the ripening processwhere the sink strength of green fruits is higher than inred fruits (Ho 1988) it could be proposed that GDHplays a similar role in both processes. Recently, a com-pensatory role for GDH in stabilizing the glutamic acidpool when Fd-GOGAT becomes limiting was proposedfor transgenic tobacco leaves deficient in Fd-GOGAT(Ferrario-Mery et al. 2002). RNA analysis of gf fruitsrevealed that transcripts of GDH were only observed inripe gf fruits. The lack of correlation between GDHtranscript and protein levels suggests that post-transcrip-tional and/or post-translational regulatory mechanismsmust be acting. The possibility that a different GDHgene could be expressed in green fruits cannot be ruledout.

In gf fruits, where the senescence and ripening isaltered, the GS, Fd-GOGAT and GDH polypeptidesco-exist, although to different levels, under the wholeripening process. On the other hand, the glutamate accu-mulation occurring during tomato fruit ripening was notimpaired in ripe gf fruit, indicating that this mechanismis independent of the chloroplast to chromoplast transi-tion. Altogether these data suggest that there may beseveral methods of regulation that operate by controllingthe amount of active glutamate metabolizing enzymes intomato fruits, involving the turnover of transcripts andproteins.


Acknowledgements – We thank Dr C. Rick for the gf tomato seeds.For the generous gift of the antibodies against Zea mays GS2 andFd-GOGAT and against Vitis vinifera NAD (H)-GDH we wish tothank Dr H. Sakakibara and Dr K.A. Roubelakis-Angelakis,respectively. This work was supported in part by a grant from theVolkswagen Foundation (Germany).

References

Akhtar MS, Goldschmidt EE, John I, Rodoni S, Matile P, Grierson D(1999) Altered patterns of senescence and ripening in gf, a staygreen mutant of tomato (Lycopersicon esculentum Mill.). J ExpBot 50: 1115–1122

Boggio SB, Palatnik JF, Heldt HW, Valle EM (2000) Changes inamino acid composition and nitrogen metabolizing enzymesin ripening fruits of Lycopersicon esculentum Mill. Plant Sci159: 125–133

Bradford MM (1976) A rapid and sensitive method for the quanti-tation of microgram quantities of protein utilizing the principleof protein-dye binding. Anal Biochem 72: 248–254

Buker M, Schunemann D, Borchert S (1998) Enzymic propertiesand capacities of developing tomato (Lycopersicon esculentumL.) fruit plastids. J Exp Bot 49: 681–691

Cheung AY, McNellis T, Piekos B (1993) Maintenance of chloro-plast components during chromoplast differentiation in thetomato mutant Green flesh. Plant Physiol 101: 1223–1229

Cren M, Hirel B (1999) Glutamine synthetase in higher plants:regulation of gene and protein expression from the organ tothe cell. Plant Cell Physiol 40: 1187–1193

D’Aoust MA, Yelle S, Nguyen-Quoc B (1999) Antisense inhibitionof tomato fruit sucrose synthase decreases fruit setting andthe sucrose unloading capacity of young fruit. Plant Cell 11:2407–2418

Ferrario-Mery S, Hodges M, Hirel B, Foyer CH (2002) Photore-spiration-dependent increases in phospho enolpyruvate carbox-ylase, isocitrate dehydrogenase and glutamate dehydrogenase intransformed tobacco plants deficient in ferredoxin-dependentglutamine–a-ketoglutarate aminotransferase. Planta 214:877–886

Fraser PD, Truesdale MR, Bird CR, Schuch W, Bramley PM(1994) Carotenoid biosynthesis during tomato fruit develop-ment: evidence for tissue-specific gene expression. Plant Physiol105: 405–413

Gallardo F, Canton FR, Garcia-Gutierrez A, Canovas FM (1993)Changes in photorespiratory enzymes and glutamate synthasesin ripening tomatoes. Plant Physiol Biochem 31: 189–196

Giovannoni J (2001) Molecular biology of fruit maturation andripening. Annu Rev Plant Physiol Plant Mol Biol 52: 725–749

Gray JE, Picton S, Giovannoni JJ, Grierson D (1994) The useof transgenic and naturally occurring mutants to understandand manipulate tomato fruit ripening. Plant Cell Environ 17:557–571

Hadjeb N, Gounaris I, Price CA (1988) Chromoplast-specific pro-teins in Capsicum annuum. Plant Physiol 88: 42–45.

Ho L (1988) Metabolism and compartmentation of imported sugarsin sink organs in relation to sink strength. Annu Rev PlantPhysiol Plant Mol Biol 39: 355–378

Kerr EA (1956) Green flesh, gf. Tomato Genet CooperativeReports 6: 17

Kisaka H, Kida T (2003) Transgenic tomato plant carrying a genefor NADP-dependent glutamate dehydrogenase (gdh A) fromAspergillus nidulans. Plant Sci 164: 35–42

Lancien M, Gadal P, Hodges M (2000) Enzyme redundancy and theimportance of 2-oxoglutarate in higher plant ammonium assim-ilation. Plant Physiol 123: 817–824

Lea PJ, Ireland RJ (1999) Nitrogen metabolism in higher plants. In:Singh BK (ed) Plant Amino Acids: Biochemistry and Biotech-nology. Marcel Dekker, New York, USA, pp. 1–47

Loulakakis KA, Roubelakis-Angelakis KA (1990) Immunocharac-terization of NADH-glutamate dehydrogenase from Vitisvinifera L. Plant Physiol 94: 109–113

Loulakakis KA, Roubelakis-Angelakis AK, Kanellis AK (1994)Regulation of glutamate dehydrogenase and glutamine synthe-tase in avocado fruit during development and ripening. PlantPhysiol 106: 217–222

Mack G (1998) Glutamine synthetase isoenzymes, oligomers andsubunits from hairy roots of Beta vulgaris L. var lutea. Planta205: 113–120

Maniatis T, Fritsch EF, Sambrook J (1982) Molecular Cloning. ALaboratory Manual. Cold Spring Harbor Press, Cold SpringHarbor, NY, USA.

Marano MR, Serra EC, Orellano EG, Carrillo N (1993) The path ofchromoplast development in fruits and flowers. Plant Sci 94: 1–17

Masclaux C, Valadier MH, Brugiere N, Morot-Gaudry JF, Hirel B(2000) Characterization of the sink/source transition in tobacco(Nicotiana tabacum L.) shoots in relation to nitrogen manage-ment and leaf senescence. Planta 211: 510–518

Masclaux-Daubresse C, Valadier MH, Carrayol E, Reisdorf-Cren M,Hirel B (2002) Diurnal changes in the expression of glutamatedehydrogenase and nitrate reductase are involved in theC/N balance of tobacco source leaves. Plant Cell Environ 25:1451–1462

Miflin BJ, Habash DZ (2002) The role of glutamine synthetase andglutamate dehydrogenase in nitrogen assimilation and possibi-lities for improvement in the nitrogen utilization of crops. J ExpBot 53: 979–987

Migge A, Meya G, Carrayol E, Hirel B, Becker TW (1996) Regula-tion of the subunit composition of tomato plastidic glutaminesynthetase by light and the nitrogen source. Planta 200: 213–220

Nagata M, Saito R (1992) Changes in free amino acid contents oftomato fruits during ripening, especially changes in glutamine.Nippon Shokuhin Kogyo Gakkaishi 39: 64–67

Perez-Rodrıguez J, Valpuesta V (1996) Expression of glutaminesynthetase genes during natural senescence of tomato leaves.Physiol Plant 97: 576–582

Piechulla B, Glick RE, Bahl H, Melisand A, Gruissem W (1987)Changes in photosynthetic capacity and photosynthetic proteinpattern during tomato fruit ripening. Plant Physiol 84: 911–917

Rick CM, Butler L (1956) Phytogenetics of the tomato. Adv Genet8: 267–382

Riens B, Lohaus G, Heineke D, Heldt HW (1991) Amino acid andsucrose content determined in the cytosolic, chloroplastic andvacuolar compartments and in the phloem sap of spinach leaves.Plant Physiol 97: 227–233

Rolin D, Baldet P, Just D, Chevalier C, Biran M, Raymond P(2000) NMR study of low subcellular pH during the develop-ment of cherry tomato fruit. Aust J Plant Physiol 27: 61–69

Sakakibara H, Kawabata S, Takahashi H, Hase T, Sugiyama T(1992) Molecular cloning of the family of glutamine synthetasegenes from maize: expression of genes for glutamine synthetaseand ferredoxin-dependent glutamate synthase in photosyntheticand non-photosynthetic tissues. Plant Cell Physiol 33: 49–58

Schaffer AA, Petreikov M (1997) Sucrose-to-starch metabolism intomato fruit undergoing transient starch accumulation. PlantPhysiol 113: 739–746

Thompson AJ, Tor M, Barry CS, Vrebalov J, Orfila C, Jarvis MC,Giovannoni JJ, Grierson D, Seymour GB (1999) Molecular andgenetic characterization of a novel pleiotropic tomato-ripeningmutant. Plant Physiol 120: 383–390

Tigchelaar EC, McGlasson WB, Buescher RW (1978) Genetic reg-ulation of tomato fruit ripening. Hortscience 13: 508–513

Valle EM(2003)Amino acid andproteinmetabolism. In:GoodmanRM(ed) Encyclopedia of Plant and Crop Science. Marcel Dekker,Inc., New York, in press.

Valle EM, Boggio SB, Heldt HW (1998) Free amino acids contentof phloem sap and fruits in Lycopersicon esculentum. Plant CellPhysiol 39: 458–461

Wilkinson JQ, Lanahan MB, Yen HC, Giovannoni JJ, Klee HJ(1995) An ethylene-inducible component of signal transductionencoded by never-ripe. Science 270: 1807–1809

Yelle S, Cheteat R, Dorais M, DeVerna J, Bennett A (1991) Sinkmetabolism in tomato fruit. IV: Genetic and biochemical analy-sis of sucrose accumulation. Plant Physiol 95: 1026–1035

Edited by P. Gardestrom


Different induction patterns of glutamate metabolizing enzymes in ripening fruits of the tomato mutant green flesh

Documents