Contrasted responses to carbohydrate limitation in tomato fruit at two stages of development


  • P. BALDET,

    Corresponding author
    1. Unité Mixte de Recherche en Physiologie et Biotechnologie Végétales, Institut de Biologie Végétale Moléculaire et Institut National de la Recherche Agronomique, Centre de Recherche de Bordeaux, BP 81, 33883 Villenave d’Ornon Cedex, France
    Search for more papers by this author
  • C. DEVAUX,

    1. Unité Mixte de Recherche en Physiologie et Biotechnologie Végétales, Institut de Biologie Végétale Moléculaire et Institut National de la Recherche Agronomique, Centre de Recherche de Bordeaux, BP 81, 33883 Villenave d’Ornon Cedex, France
    Search for more papers by this author

    1. Unité Mixte de Recherche en Physiologie et Biotechnologie Végétales, Institut de Biologie Végétale Moléculaire et Institut National de la Recherche Agronomique, Centre de Recherche de Bordeaux, BP 81, 33883 Villenave d’Ornon Cedex, France
    Search for more papers by this author

    1. Unité Mixte de Recherche en Physiologie et Biotechnologie Végétales, Institut de Biologie Végétale Moléculaire et Institut National de la Recherche Agronomique, Centre de Recherche de Bordeaux, BP 81, 33883 Villenave d’Ornon Cedex, France
    Search for more papers by this author
  • D. JUST,

    1. Unité Mixte de Recherche en Physiologie et Biotechnologie Végétales, Institut de Biologie Végétale Moléculaire et Institut National de la Recherche Agronomique, Centre de Recherche de Bordeaux, BP 81, 33883 Villenave d’Ornon Cedex, France
    Search for more papers by this author

    1. Unité Mixte de Recherche en Physiologie et Biotechnologie Végétales, Institut de Biologie Végétale Moléculaire et Institut National de la Recherche Agronomique, Centre de Recherche de Bordeaux, BP 81, 33883 Villenave d’Ornon Cedex, France
    Search for more papers by this author

Pierre Baldet. Fax: + 33 5 57 12 25 41; e-mail:


The metabolic consequences of long-term carbohydrate depletion have been well documented in many sink organs but not extensively in fruit. Therefore, in the present study the response to sugar limitation in tomato fruit (Lycopersicon esculentum Mill.) was investigated at two developmental stages; during the cell division and cell expansion phases. First, the response in excised fruit cultured in vitro was characterized. Sugar depletion caused an arrest of growth and an exhaustion of carbon reserves. The proteins that were degraded and the nitrogen released was transiently stored as asparagine and glutamine in both developmental stages and also as γ -aminobutyric acid (GABA) in expanding fruit. Fruit at the cell division stage appeared to be more sensitive to sugar limitation. The response to sugar depletion was then characterized in fruit from plants submitted to extended darkness. In planta, the effects of sugar-limitation were similar to those described in vitro but much more attenuated, especially in expanding fruit, which still accumulated dry matter. The expression of cell cycle genes, sugar- and nitrogen-related genes was reduced by darkness. Only asparagine synthetase gene expression was induced in both dark-treated fruit. Together the present data revealed that the effects of the carbon limitation are more pronounced in the youngest fruits as it is probably controlled by the relative sink strength of the fruit.


days post-anthesis


reverse transcriptase-polymerase chain reaction.


Plant growth and development depend largely upon the partitioning of assimilated carbon between photosynthetic sources such as mature leaves, and photosynthetically less active or inactive sink tissues such as flowers, roots and fruit (Farrar & Williams 1991). The activities of source and sink organs are regulated in different ways, so that crop growth and yield can be either source or sink limited (Ho 1988). The controls that actually regulate the partitioning of assimilates between sink organs are still poorly understood. Although this is still a matter of debate, it is now thought that the partitioning of assimilates between sink organs is not a property of the sink alone but a combination of co-ordinated mechanisms involving the whole system (Farrar 1996; Marcelis 1996; Minchin & Thorpe 1996). In this respect, sink strength can be defined as the competitive ability of an organ to import photo-assimilates. Among sink organs, fruit are defined as high priority sinks in the context of competition for assimilates between alternative sinks (Wardlaw 1990).

Plants with an indeterminate type of growth such as cherry tomato produce fruit continuously, thus constituting a collection of sinks competing for carbohydrates. The early development of tomato fruit can be divided into two distinct phases (Gillaspy, Ben-David & Gruissem 1993). In cherry tomato, the first phase lasts up to 8–10 d after fertilization and fruit set, and is characterized by a very active period of cell division inside the ovary. During the second phase, which proceeds for a further 13–16 d, fruit growth is mostly due to cell expansion, thus leading to a fruit that exhibits its final size and is able to ripen. The final fruit size depends on the number and size of cells that originate from the cell division and expansion phenomena, respectively (Gillaspy et al. 1993). Sink activity of developing fruit is comprised of three important physiological features: (i) the phloem unloading of assimilates; (ii) post-phloem transport and retrieval by sink cells; and (iii) the consumption and storage of imported carbohydrates (Ho 1988). Hence, the fruit sink strength results from its size and its metabolic activity.

Carbohydrate limitation is a common occurrence for most higher plants. Because plants manufacture their own carbohydrates, any exogenous factors such as abiotic or biotic stress, which alter the resource availability, may influence carbohydrate partitioning and lead to carbohydrate limitation. In addition, the plant life cycle is accompanied by changes with respect to the sink strength of individual organs and the number of sink organs competing for a common pool of carbohydrates. Such a competition based upon sink strength, if naturally or artificially enhanced, may lead to a state of carbohydrate limitation and influence fruit development (Bohner & Bangerth 1988; Bertin 1995).

The metabolic consequences of carbohydrate depletion have been studied in different sink organs in many plant models whereas such data are not available for fruit in the literature. Koch (1996) revealed that plant cells possess their own genetic potential to adapt to adverse growth conditions. Further, Brouquisse et al. (1991) observed that the response to sugar deprivation in maize root tips may be divided into three phases: an acclimation phase, a survival phase and a cell-disorganization phase leading to cell death. The acclimation phase entails a decrease in respiratory and growth capacities and a remobilization of the carbon reserves. When the conditions deteriorate plant cells increase their resistance to sugar depletion and survive by replacement with alternative carbon sources from cellular constituents such as lipids, proteins and other cellular materials. At the whole plant level, few data are available dealing with the changes of physiological and biochemical parameters in sink organs submitted to carbon limitation. In these studies, the sink response depends on how the photo-assimilate supply is reduced (e.g. shading, darkening, leaf pruning, excision or girdling). Brouquisse, Gaudillère & Raymond (1998) showed that in the main sink organs of maize plants, the consequences of carbon depletion induced by a period of extended darkness were identical to those established for the root tip, especially the decrease in sugar content and protein degradation.

As darkness affects fruit growth and its aptitude to accumulate carbohydrates in tomato plants (Pearce, Grange & Hardwick 1993; Grange & Andrews 1994), the aim of this work was to determine whether, as priority organ for photo-assimilates, the fruit is susceptible to sugar starvation and whether the sensitivity to carbohydrate depletion relies on its developmental stage. To investigate the effects of carbohydrate deprivation on the metabolism of young tomato fruit, we: (i) characterized the response to sugar limitation in excised tomato fruit maintained on sugar-free medium; (ii) validated the biochemical analysis of the starvation symptoms relative to fruit growth and gene expression analysis in tomato plants submitted to extended darkness; and (iii) compared the response between fruit at the cell division phase and fruit at the cell expansion phase. The effects of the carbohydrate limitation on fruit at these two distinct developmental stages are also discussed relative to sink strength.

Materials and methods

Plant culture

Cherry tomato (Lycopersicon esculentum var. cerasiformae Mill., cv. West Virginia 106) seeds were provided by Dr J. Philouze (INRA Montfavet, France). The seeds were germinated in vermiculite soaked in Algospeed solution (1 g L−1; Algochimie, Château Renault, France) fertilizer containing N : P : K: Mg in the ratio 13 : 13 : 24 : 3. The plants were raised under greenhouse conditions with temperature control within the ranges of 24–26 °C and 18–20 °C during day and night, respectively. About 10 d after sowing, seedlings at the two-leaf stage were transferred to a growth chamber, under a 15 h photoperiod with a photosynthetic photon flux density of 400 µmol photons m−2 s−1. The day/night temperature was 25/20 °C and the relative hygrometry was maintained close to 70%. Plants were grown in a hydroponic culture system. The culture medium consisting of Algospeed solution (1 g L−1) was constantly air-bubbled. The pH of the medium was checked and adjusted daily to 6.

Carbon limitation in excised tomato fruit

During cultivation, fruit was harvested at the end of the night period at 5 and 15 d post-anthesis (DPA), corresponding to the middle of the cell-division and -expansion phases, respectively. Immediately after harvest, the fruit were decontaminated by immersion for 1 min in a 2% (w/v) calcium hypochlorite solution, rinsed in a solution containing 0·1 m HCl and 1% (v/v) Triton X100, and several times in sterile water. Under sterile conditions, the fruit were set on a grid allowing the pedicel to be immersed in liquid MS medium (Murashige & Skoog 1962), and incubated at 25 °C in darkness. Excised fruit were removed at different times, promptly frozen and ground in liquid nitrogen prior to lyophilization and storage at −20 °C.

Prolonged darkness experiment

After 2 months of hydroponic culture, the tomato plants were submitted to prolonged darkness or maintained in day/night cycles as a control. Darkness was applied for 9 d to plants. Fruit at 5 and 15 DPA, at the beginning of darkness application, were harvested daily during the course of the treatment at a time corresponding to the end of the night period in the control plants. For all the harvested fruit the equatorial diameter and fresh weight were measured. Fruit of each treatment, stage and time were pooled. Each batch was treated just as excised fruit and after lyophilization dry weights were measured.

Sugars, amino acids, proteins and NH4+ extraction and assays

Sugars and amino acids were extracted from 10 to 15 mg dried powder using the alcoholic extraction method described by Brouquisse et al. (1991). Glucose, fructose and sucrose were assayed as described by Kunst, Draeger & Ziegenhorn (1984) adapted to a micro-assay using the MR 5000 reader (Dynatech, St Cloud, France). Starch was extracted and assayed as described by Moing, Escobar-Gutierrez & Gaudillère (1994). Free amino acids were analysed by reverse-phase high-performance liquid chromatography according to the AccQ-Tag method of Cohen & De Antonis (1994). Amino acid quantification was achieved using the Millenium 2·15 software (Waters, St Quentin en Yvelines, France). Total soluble proteins were extracted in 500 mm Tris-HCl (pH 7·5) and measured according to Bradford (1976) using bovine serum albumin as a standard. NH4+ was extracted with 0·1 m HCl and measured according to the phenol-hypochlorite method, as described in King et al. (1990).

Extraction of total RNA and estimation of relative transcript levels by reverse transcriptase-polymerase chain reaction

Total RNA from fruit was extracted according to Chevalier et al. (1996). For reverse transcriptase (RT)-polymerase chain reaction (PCR), total RNA samples were treated with RNAse-free DNAse RQ1 (Promega, Lyon, France) for 15 min at 37 °C, phenol/chloroform extracted, ethanol precipitated and dissolved in DEPC-treated water. In order to determine specifically the relative transcript level of each cDNA, RT-PCR assays were performed as previously described (Joubès et al. 2000), using a downstream primer corresponding to the 3′ UTR sequence, combined with an upstream primer corresponding to a sequence present at the C-terminus of the protein (Table 1). Typically 2 µg of total RNA was used in the RT reaction in a final volume of 20 µL. The RT was diluted 10-fold prior to PCR. For each RT-PCR product, positive control amplification was performed using a plasmid harbouring the cDNA of interest. The RT-PCR products were separated on a 1·2% agarose gel, blotted onto Hybond N+ membrane (Amersham, Saclay, France) and hybridized at 65 °C with the appropriate cDNA probes labelled with [α-32P]dCTP by random priming according to standard method (Sambrook, Fritsch & Maniatis 1989).

Table 1.  Sets of PCR primers used to amplify gene-specific regions and corresponding size of the amplified product
cDNAPrimer sequence (5′ → 3′)LocationSize (bp)Accession no.
Lyces;actinSenseTGG CAT CAT ACT TTC TAC AAT G 323–344615U60480
AntisenseCTA ATA TCC ACG TCA CAT TTC AT 915–938  
Lyces;H4SenseGTC TGG TCG TGG AAA GGG AGG CAA GGG  41–67309AW429173
Lyces;CDKA;1SenseGCT TAT TGT CAT TCT CAT AGA GTT CTT 479–505559Y17225
AntisenseCTG GAT GAA GGG GCA GAC AAT CAC GG1013–1038  
Lyces;CDKB2;1SenseGGA GGC TGC TGA AAA TGC TG  46–65521AJ297917
AntisenseGTA TAA GCT CTG CCA AGT CC 548–567  
Lyces;CDKC;1SenseGCA CCC ACA GCA ACA TTC GCG CCT CCC TCC1178–1207618AJ294903
Lyces;CycA2;1SenseTAT GAA GAA ATT TGT GCA CCT CGT G1208–1235554AJ243452
Lyces;CycB2;1SenseGAG GAG GTT TGT GCT CCA CTG GTG G1192–1216541AJ243455
AntisenseAAG TTA GTA ATA ACT GTA CAT GCC C1709–1733  
Lyces;CycD3;1SenseTTA TCT TTC ATT GAT CAT ATT ATG AGG 772–798525AJ243415
Lyces;Exp18SenseGCC ATC CCG GAA GTC CTT CCA 294–315554AJ270960
AntisenseCCC AAA ATA CCC ATT AAC AAC C 827–848  
Lyces;SuSy2SenseGGA ATG GTG AGC TCT ACC GAT1952 −1973562AI486064
AntisenseGGG GGG GGA TAC TAA ATG GAA2493–2514  
Lyces;SPSSenseGGG CCG CTT CTA TCA CTG ATA2644–2665590AB051216
AntisenseGCT GCC ATT CTT TTC ATG GCA3213–3234  
Lyces;L-AGPSenseGGG AGG ATA TTG GAA CAG TGA1037–1058566U85496
AntisenseCCA GCC TAC TTT ATT TCC CAA1582–1603  
Lyces;TivSenseTAT CAA TAC AAT CCA GAT TCA 403–424616Z12025
AntisenseGTA GAT ACC GGG TAA AAG TCC1088–1019  
Lyces;Sut1SenseCTT GAT GTT GCT AAT AAC ATG 499–520461X82275
AntisenseAAA GGG AAA CCA CGC AAT CCA 949–960  
Lyces;Sut4SenseACG TGG TCA TGT CTC TCT GCT1188–1209406AF176950
AntisenseTGC AAA GAT CTT GGG TCT CTC1573–1594  
Lyces;Ht3SenseTGG AAG TGA CAA GGA GCT ATC A 291–313414AJ132225
AntisenseAGC ATG GTT GAG TCC ATT GGT 684–705  
Lyces;Gs1SenseGGA CCT TCT GTT GGC ATC TCA 607–628524U14754
AntisenseCCC AAA TTG TGC AAA CCA GAA1110–1131  
Lyces;Gs2SenseGGA AAA GTC ATT TGG CAG AGA 393–414644U15059
AntisenseCCC CTC ATA TAC TCT AAA CCT1016–1037  
Lyces;As1SenseCCG TAT CTT CCA AAG CAT ATT 970–991489AI897557
AntisenseGAG TCA TGG ACA GCA AGA GAC1438–1459  
Lyces;Gdc1SenseCTT ATT CGT TTG GGT TAT GAG 994–1016636AI487384
AntisenseGCC TTT AAA ATG ATG GTC CAA T1608–1630  


Effects of sugar deprivation on excised tomato fruit

Excised tomato fruit at 5 and 15 DPA, referred to as dividing and expanding fruit, were used to analyse the effect of sugar depletion. Biochemical parameters such as soluble sugars, starch, proteins, amino acids and NH4+ content were analysed during a 17–34 d period (Fig. 1). When submitted to sugar starvation, the growth of fruit was strongly affected. This was not the result of the excision stress because the growth of excised fruit cultured with 0·2 m sucrose was maintained although at a lower rate (data not shown), as described by Guan & Janes (1991).

Figure 1.

Effects of carbohydrate depletion on the metabolite contents in excised tomato fruit at 5 and 15 DPA. Changes in sugar content (total soluble sugars including sucrose + glucose + fructose, □; starch, ▪), protein content (▴), amino acid content (asparagine, □; glutamine, ▵ GABA ○), NH4+ content (•) were measured in excised tomato fruit maintained in vitro on sucrose-free liquid MS medium for a period of 17 and 34 d. The data of the metabolite changes can be analysed on a dry weight basis, providing that during the course of starvation the mean dry weight of the fruit at 5 and 15 DPA corresponds to 9·1 ± 2 and 130 ± 7 mg per fruit, respectively. Each point represents the mean (± SD) of four independent experiments. In the main figures, insets describe the metabolite changes for longer time scale.

As total sugars can account for up to 35% of the dry weight in expanding fruit (Rolin et al. 2000), a carbon limitation will inevitably influence the dry weight value. Therefore, in Fig. 1 the metabolite concentrations were expressed per fruit rather than dry weight unit. The total soluble sugar and starch content exhibited significant variation during the incubation period. In dividing fruit, the levels of soluble sugar and starch dropped sharply within 4 d to reach 10% of the initial value. The decline in starch content was more attenuated in expanding fruit because 8–10 d were needed to reach the same level. In expanding fruit, the soluble sugar content decreased continuously to reach 10% of its initial value after 30 d. The analysis of the protein content revealed that it varied according to the fruit developmental stage (Fig. 1). In dividing fruit a decrease of 50% in the protein content was observed within 7 d, whereas only a 20% decrease in protein content was seen in expanding fruit over the same period. For both types of fruit the total amino acid content initially increased and then declined more-or-less rapidly (data not shown). This pattern is similar to that for the glutamine content alone (Fig. 1). Because asparagine, glutamine and γ -aminobutyric acid (GABA) accounted for 25–50% of the total amino acid content, only these three amino acids were represented in Fig. 1. In dividing fruit the glutamine content reached its maximum level after 4 d and then declined, whereas asparagine still accumulated up to 10 d. In expanding fruit both the asparagine and glutamine contents increased to a five-fold level within 7 d. During the same period of time, the GABA content doubled and then slowly declined.

The change in NH4+ content was the most spectacular in dividing fruit as it increased in value to a 10-fold level after 9–10 d (Fig. 1). In expanding fruit, this increase was initially much slower but accelerated sharply after 20 d. Table 2 summarizes the changes in the distribution of nitrogen between the main N-compounds of the fruit. At the beginning of the treatment, 90 and 70% of nitrogen was located in the proteins of dividing and expanding fruit, respectively. During starvation, a redistribution of nitrogen occurred from proteins to amino acids and ammonium, typically due to protein degradation and amino acid catabolism (Brouquisse et al. 1998), as revealed by the accumulation of glutamine, asparagine, GABA and NH4+ in carbohydrate-limited fruit. It should be noted that the total nitrogen content did not change significantly during the starvation period. This means that nitrogen is stored in starved fruit, but not released as observed in source- or weak sink-starved tissues (Brouquisse et al. 1998; Devaux et al. personal comm.)

Table 2.  Intracellular distribution of nitrogen between proteins, amino acids, asparagine, glutamine and ammonium in sugar-depleted tomato fruit (in vitro and in planta)
 StagesTime (days)Nitrogen (µmoles/fruit)Nitrogen distribution
ProteinsAmino acidsNH4+ΣNProt/[Aa + NH4+]
  1. In vitro experiment: excised tomato fruit were maintained on sucrose-free liquid MS medium for a period of 17 d for 5 DPA fruit and 34 d for 15 DPA fruit. The data correspond to the beginning (T = 0), the middle (T = 9 or 15) and the end (T = 17 or 34) of the sugar depletion.

  2. In planta experiment: fruit at the age of 5 and 15 DPA at the beginning of darkness application were harvested in the course of the treatment as well as from the control plants cultured in normal conditions. The data correspond to the beginning of the treatment (0), and 9 d of darkness (9D) and control conditions (9C).

  3. The amounts of proteins were calculated assuming that the nitrogen content is 16·7% (w/w); ΣN represents the sum of (proteins + amino acids + NH4+) nitrogen compounds present in the two types of fruit. On average, the standard deviation of the mean values varied between 7 and 19%.

in vitro5 DPA 0 14 0·1 0·7  1·4 0·3 175·7
 9  6·5 3·1 2·4  1·8 0·8 150·7
17  5·9 2·7 0·8  1·5 4 150·6
15 DPA 0110 2·418 27 91671·9
15 511317 44231480·5
34 25 711 39651470·2
in planta5 DPA 0 (Control) 13 0·2 0·8  1 0·5 166·7
 9 (Control) 76 519 34 8·21421·3
 9 (Dark)  6 4 3  3 0·9 170·6
15 DPA 0 (Control) 981029 39131891·3
 9 (Control)2872766119435421·4
 9 (Dark)1226669 86984410·6

Effects of sugar limitation in whole tomato plants submitted to extended darkness

Changes in fruit growth rate

The tomato plants were submitted to an extended period of darkness (9 d) to investigate the response to carbohydrate limitation in fruit still attached to the plant. Fruit harvested during the time course of darkness were analysed at the physiological, biochemical and molecular levels. Control fruit were collected on plants cultured under a day/night regime.

Figure 2 represents two parameters of the developing fruit from control and dark-treated plants, the fruit growth rate defined as the increase of the diameter and the change in dry weight. Under control conditions the growth rate of cherry tomato fruit exhibited a classical sigmoidal growth rate curve (Ho 1988; Fig. 2). Under darkness, fruit growth and consequently dry weight were significantly reduced, with the youngest fruit the most sensitive to darkness (Fig. 2). The dry weight of dividing fruit remained constant and the growth rate (inset of Fig. 2) was reduced 10-fold in comparison with control fruit. In expanding fruit the dry weight still increased but reached only half the value of the control fruit at the end of the experiment (24 DPA). Moreover, the growth rate was reduced two-fold relative to the control fruit. Under carbon deprivation conditions the growth rate of expanding fruit was maintained thus reflecting a process of assimilate importation.

Figure 2.

Effects of extended darkness on tomato fruit growth rate and dry matter accumulation. The changes in the growth rate were analysed by measuring the diameter (inset) of fruit harvested from control plants (day/night cycles) (○), of 5 DPA- (□) and 15 DPA-developing fruit (▵) harvested from dark-treated plants. In the course of development, the dry weight was measured from fruit of control plants (○) (from anthesis to 24 DPA) as well as from 5 DPA- (□) and 15 DPA-developing fruit (▵) from dark-treated plants. The arrows indicate the beginning of the darkness period for 5 and 15 DPA-developing fruit. The dry weight curves can be computed on a fresh weight basis as the ratio DW/FW for the fruits at 5 and 15 DPA are 10·6 ± 0·5 and 9·6 ± 0·4, respectively. Each point represents the mean (± SD) of four plants grown in day/night cycles and darkness from two independent experiments.

Changes in fruit metabolite contents

Under control conditions of culture the metabolite content of cherry tomato fruit was characterized by an increase in the sugar, protein, amino acid and NH4+ levels as expected for sink organs during development (Fig. 3). The accumulation process for starch, proteins, amino acids and NH4+ lasted up to the end of the expansion phase (∼ 23 DPA) and then slowed down, whereas the soluble sugars continued to accumulate up to the maturity of the fruit (Rolin et al. 2000).

Figure 3.

Effects of extended darkness on metabolite contents in tomato fruit. Changes in sugar content (total soluble sugars including sucrose + glucose + fructose (○,•); starch (□, ▪), protein content (○,•), amino acid content (asparagine, □; glutamine, ▵ GABA ○), NH4+ content (•) were measured in 5 and 15 DPA-developing fruit from control plants (day/night cycles) (open symbols, dashed line) and dark-treated plants (closed symbols, solid line). Each point represents the mean (± SD) of four plants grown in day/night cycles and darkness from two independent experiments. In the figures for fruit at 5 DPA, insets describe the changes in sugar and amino acid contents in order to magnify metabolite changes when the curve evolved within a range of low values. Moreover, the metabolite changes for each type of fruit and experimental condition can be computed on a dry weight and fresh weight basis by using the data of dry weight per fruit obtained in Fig. 2.

In dividing fruit the sugar content was affected by darkness, as displayed by a transient increase for 1 d (inset of Fig. 3) and then a rapid decline. Even though the growth of dividing fruit was impaired by darkness treatment, the protein content remained stable up to 7 d and thereafter decreased. Within the same period of time the glutamine, GABA and NH4+ contents increased slightly. Although the levels of GABA and NH4+ declined after 5 d of treatment, the glutamine content only began to decrease after 7 d. The accumulation pattern for asparagine in dark-treated fruit was strikingly similar to that of the control fruit up to 5 d; indeed, no difference could be observed (inset, Fig. 3). Whereas the accumulation of asparagine increased linearly to the end of the experiment in control fruit, this increase in dark-treated fruit was reduced after 5 d. Nevertheless, there was a correlation between the accumulation of asparagine and the drop in the protein content as observed in excised dividing fruit (Fig. 1).

In expanding fruit darkness had no effect on the soluble sugar content, which remained constant up to the seventh day and then declined. Starch accumulation persisted for 1 d, sharply dropped after 2 d and thereafter declined slowly up to the end of the experiment. In contrast with dividing fruit no sign of protein degradation was observed in expanding fruit, even though protein synthesis was highly reduced under darkness. With the exception of glutamine, whose accumulation pattern was similar to that of the control fruit, the increase of GABA and asparagine were also affected by darkness. Hence, the GABA content still increased but reached only half the value of the control fruit at the end of the experiment. Asparagine accumulated much more rapidly in the dark-grown fruit as the content was increased by three fold at 23 DPA compared with the control fruit. In the same way, the NH4+ content increased by at least two-fold in the dark compared with the control. The analysis of the distribution of nitrogen between the main N-compounds in fruit (Table 2) revealed that at the end of the darkness period the proportion of nitrogen in amino acids and NH4+ had increased, just as observed for excised fruit. However, if in excised fruit or in dividing fruit in planta the nitrogen redistribution may be obviously attributed to proteolysis, in expanding fruit in planta the increased proportion of nitrogen in the amino acid and NH4+ pool results clearly from a net nitrogen uptake by the fruit, as indicated by the stability of proteins and the increase in total nitrogen content (Table 2). These data show that in carbon-deprived plants the supply of nitrogen compounds from the other parts of the plant into the expanding fruit via phloem/xylem import is maintained.

Changes in fruit gene expression

The expression of genes involved in the growth and metabolic responses were analysed in fruit collected from dark-treated plants and compared to fruit from plants grown under normal conditions (Figs 4 & 5). The cell cycle Cyclin-Dependent Kinases (CDK) CDKA;1, CDKB2;1, CDKC;1 and the cyclins CycA2;1, CycB2;1, CycD3;1, histone H4 and expansin 18 (EXP18) were used as proliferation- and growth-associated gene markers (Joubès et al. 2000, 2001; Caderas et al. 2000). Marker genes encoding proteins associated with sugar metabolism were also used, including sucrose synthase (SUSY2), sucrose phosphate synthase (SPS), and vacuolar acid invertase (TIV) for sucrose metabolism, the large subunit of ADP-glucose pyrophosphorylase (L-AGPase) for starch metabolism, and sucrose (SUT1, SUT4) and hexose transporters (HT3) (Koch 1996 and references therein; Delrot, Atanassova & Maurousset 2000). Genes encoding cytosolic (GS1) and plastid (GS2) glutamine synthetases, asparagine synthetase (AS1) and glutamate decarboxylase (GDC1) were used as gene markers for amino acid metabolism (Lam et al. 1996). The expression of all these genes were analysed by semi-quantitative RT-PCR using a combination of primers (Table 1) in dividing and expanding fruit from plants grown for 9 d under darkness and compared with plants grown under control (day/night) conditions (Figs 4 & 5). Specific primers for actin cDNA were used as an internal control of the RT-PCR.

Figure 4.

Effects of extended darkness on cell cycle gene expression in tomato fruit. Semi-quantitative RT-PCR of cell cycle genes were performed using total RNA extracted from 5 and 15 DPA-developing fruit from control plants (day/night cycles) and dark-treated plants. The specific amplification of cDNA fragments was detected after gel electrophoresis, Southern blotting, and hybridization to the corresponding 32P-labelled probes. The positive control of PCR corresponds to the amplification of the purified PCR product used as a template and the negative control corresponds to the absence of DNA template in the PCR reaction.

Figure 5.

Effects of extended darkness on carbohydrate- and nitrogen-related gene expression in tomato fruit. Semi-quantitative RT-PCR of carbohydrate- and nitrogen-related genes were performed using total RNA extracted from 5 and 15 DPA-developing fruit from control plants (day/night cycles) and dark-treated plants, according to the legend of Fig. 4.

In good agreement with Joubès et al. (2000), the expression of proliferation- and growth-associated genes, especially cell cycle genes, was much higher in dividing fruit than in expanding fruit under normal growth and developmental conditions. Under darkness most of these genes were globally repressed (Fig. 4), among them Lyces;Actin, the expression of which has been shown to be affected by carbon limitation (Sheu et al. 1994; Joubès et al. 2001) and Lyces;Exp18, as reported for tomato hypocotyl growth (Caderas et al. 2000). However, slight differences were observed depending on the fruit developmental stage. For example, in dividing fruit the expression of the genes for CDKA;1 and CycD3;1 was unchanged up to the fifth day in either normal or dark conditions. Similarly, in expanding fruit the level of Lyces;CDKC;1 transcripts did not change in either normal or dark conditions.

In normal growth conditions the genes involved in sugar and nitrogen metabolism, such as genes for SUSY2, L-AGPase, GS1, GS2 and GDC1, were more highly expressed in the dividing fruit, whereas the genes for SPS, TIV, SUT1, HT3 and AS1 displayed maximum expression later on in expanding fruit (Fig. 5). Our expression data revealed that most genes were rapidly repressed when plants were submitted to darkness. As observed for Lyces;CDKA;1 and Lyces;CycD3;1, the genes for TIV in dividing fruit, and SUT1 in both fruit types, were expressed up to the fifth day (Fig. 5). Contrary to the other genes, the level of Lyces;As1 transcripts was strongly increased in both dividing and expanding fruit of plants grown in dark conditions (Fig. 5). This is in accordance with the studies of Lam et al. (1996) in A. thaliana, and Chevalier et al. (1996) in maize root tips, who showed the induction of asparagine synthetase gene expression in plants grown under dark conditions as well as in conditions where the cell carbohydrate supply is blocked.

In summary, the changes in transcript levels in fruit grown under dark conditions, namely the transcript level of genes associated with cell-proliferation and growth, and sugar and nitrogen metabolism decrease, whereas that of the gene involved in nitrogen storage (Lyces;As1) increases is in good agreement with biochemical and growth data (Figs 2 & 3). Such changes may be related to the ‘acclimation phase’ of starvation (Brouquisse et al. 1992).


In this study, we initially investigated the biochemical and molecular processes occurring in developing tomato fruit submitted to sugar depletion. Globally, sugar limitation induces symptoms and an acclimation response similar to that described for other sink organs, including: (1) the interruption of sugar accumulation related to the slackening of carbon metabolism; (2) the lessening of fruit growth and cell proliferation; and (3) the remobilization of alternative carbon sources, i.e. proteins and amino acids, associated with the initiation of the nitrogen storage processes. However, the extent and the schedule of the acclimation response differs depending on the fruit treatment (excised fruit or darkened plants), and the developmental stage of the fruit (dividing or expanding fruit). The metabolic consequences of carbohydrate depletion in excised tomato fruit are reproduced in fruit of plants grown under dark conditions. This validates the use of excised organs as a model to study the effects of carbon starvation in plants. However, as already observed in starved maize root tips, either excised or attached to the plant (Brouquisse et al. 1991, 1998), the response is somehow delayed and more attenuated in fruit still attached to the plant and submitted to extended darkness, compared with that in excised organs. Despite the fact that photosynthesis is arrested under darkness treatment, the fruit are still supplied with assimilates from starch reserves, which are considerable in leaf and stem tissue of tomato plants (Ammerlaan, Joosten & Grange 1986). Compared to other sink organs fruit possess a great ability to adapt and acclimate to carbohydrate starvation (Geiger, Koch & Shieh 1996). Thus, for quite a long period of darkness fruit are able to import and metabolize assimilates and maintain their potential for cell division.

The plant response to sugar depletion is usually a rapid consumption of the carbohydrate reserves and/or an arrest of the processes of carbon storage. However, in tomato fruit the sugar consumption is slowest when compared with other sink organs such as tomato young root (Devaux et al. personal comm.) or maize root tips (Brouquisse et al. 1991, 1998) in which total non-structural sugars were exhausted within 24 h of sugar depletion. Two explanations can account for this observation. First, it may be ascribed to the priority rank among sinks in the partitioning of assimilates (Wardlaw 1990). Under long-term carbohydrate depletion fruit appear to be the priority sink relative to other developing sink organs and thus, the last to undergo starvation (Ammerlaan et al. 1986). A second potential explanation is, that among tomato sink organs, those which contain the greatest carbohydrate reserve as a starch pool, sustain the starvation period for much longer. This is much more obvious and accentuated in expanding fruit (Figs 1 & 3).

In plants, many factors including light, nutritional and hormonal balance regulate the expression of carbohydrate-related genes (Koch 1996; Mustilli & Bowler 1997). In dark-treated expanding tomato fruit Grange & Andrews (1994) described a decrease in starch content, and related it to the down-regulation of L-AGPase activity (Guan & Janes 1991). We observed an arrest of the accumulation of soluble sugars and starch that was correlated with the deep reduction in the transcript levels of genes for SUSY2, SPS and L-AGPase (Figs 3 & 5). Developing fruit undergo a transient accumulation of starch that represents a carbohydrate reservoir contributing to the soluble hexose level in the mature fruit (Dinar & Stevens 1981). In the case of carbohydrate depletion, this starch pool is probably used for the survival of the fruit. The slow response in terms of starch remobilization (Figs 1 & 3) may be explained by the adaptation of the fruit, and generally speaking of plant organs, to environmental signals that often induce slow adjustments in order to establish a coarse control system in response to major alterations (Geiger et al. 1996). Thus, key enzymes of the starch synthesis pathway such as SUSY, L-AGPase, and soluble and insoluble starch synthases exhibit their maximum activities during the expansion phase (Schaffer & Petreikov 1997). Consequently, once submitted to sucrose limitation, the fruit reverse totally the metabolism from synthesis to degradation of starch and this adjustment, in terms of gene expression patterns, has to be slow. As the soluble carbohydrate content remains constant up to 2 d in dividing fruit and later in expanding fruit, light could be the main regulatory factor (Mustilli & Bowler 1997) while within the same period of time, the above-mentioned genes were deeply repressed.

The effect on fruit growth of sugar depletion under darkness (Fig. 2) and consequently on cell proliferation, was analysed at the molecular level through the investigation of cell-cycle gene expression (Fig. 4). Upon mitogenic stimulation, quiescent cells enter into the G1 phase of the cell cycle, especially through the induction of D-type cyclins by hormones and sugars (Riou-Khamlichi et al. 2000). Surprisingly the expression of Lyces;CycD3;1 was insensitive to darkness in tomato fruit. Similarly, some discrepancies were observed between the unaffected expression of Lyces;CDKC;1 in sugar-depleted tissues (Joubès et al. 2001) and the reduced expression in dark-treated fruit. It is too early to give any reliable hypothesis to explain such distinct responses to darkness in the regulation of these particular cell cycle genes in tomato fruit. However, we cannot rule out the possibility that the regulation of their gene expression is far more complex in fruit still attached to the plant than in isolated cells or organs, and especially at the level of the influence of the hormonal balance. Nevertheless we have previously demonstrated that in tomato cell suspension cultures and excised roots mitosis-specific genes such as Lyces;CDKB2;1 and Lyces;CycA2;1 and Lyces;CycB2;1 are repressed in response to sugar depletion (Joubès et al. 2000, 2001). We did confirm in fruit that the expression of these three mitosis-associated genes was also repressed in response to darkness (Fig. 4). In the developing tomato fruit the mitotic activity is related to the differential expression of CDK and cyclins (Joubès et al. 2000, 2001). Therefore it is tempting to associate the slackening of fruit growth with a strong reduction in mitotic activities inside fruit tissues, as a consequence of the repression of mitosis-specific CDK and cyclin genes exerted by sugar depletion. Interestingly the expression of Lyces;CDKA;1, the key cell cycle regulator which is characteristic of the competence of the cell to divide (Mironov et al. 1999), remained constant within the first 5 d of darkness treatment in dividing fruit. This suggests that fruit cells may maintain their aptitude to divide even though mitotic activity has stopped.

In the remobilization processes for alternative carbon sources, nitrogen released by protein degradation is differentially distributed into amino acids and then ammonium according to the fruit developmental stage (Figs 1 & 3). Asparagine is a well-known detoxification compound when high levels of ammonium result from very active proteolysis (Mazelis 1980). When the carbon depletion takes effect asparagine is consumed to supply the tricarboxylic acid cycle with carbon skeletons (Brouquisse et al. 1992) and this results in the increase in NH4+ content. In tomato fruit glutamine and GABA also appear as transient forms of storage for nitrogen in addition to asparagine. GABA is a non-protein amino acid whose functions are not well established (Satya Narayan & Nair 1990). In cherry tomato fruit GABA represents up to 61% of the total amino acids (per fruit) and has been proposed firstly to act as a temporary store for nitrogen, and secondly to play a role in the transport of nitrogen and the regulation of cytoplasmic pH (Rolin et al. 2000). Moreover, the accumulation of GABA has been observed under a variety of environmental stress conditions, thus suggesting that as for asparagine the nitrogen released by protein catabolism in expanding fruit could firstly be transiently stored as GABA and then released as NH4+ when GABA is used as a substrate to sustain respiration via the GABA shunt (Satya Narayan & Nair 1990).

The study of carbon starvation in plant tissue has been well documented in maize roots. In excised maize root tips submitted to sugar depletion Chevalier et al. (1996) described the induction of AS gene transcription and Brouquisse et al. (1992) observed the rise of the asparagine synthetase activity associated with the increase of the asparagine content. Furthermore, asparagine is the favoured compound for nitrogen storage and transport in dark-grown plants (Joy & Ireland 1990; Lam et al. 1996). Under darkness, in expanding fruit the changes in asparagine, glutamine and GABA contents could not result from proteolysis because no net protein degradation occurs (Table 2, Fig. 3). Hence, we hypothesize that the increase in these amino acids as well as NH4+ results from a net influx of N-compounds into the fruit from senescing organs (stem, mature leaves, roots) of less sink priority than fruit (Brouquisse et al. 1998). Moreover in the expanding fruit we observed a strong increase in asparagine content associated with a strong induction of the AS1 gene transcription. Hence, we cannot exclude that the massive importation of NH4+ in expanding fruit is partly detoxified through the asparagine synthetase activity resulting from the induction of AS gene transcription (Fig. 5). In contrast, the change in glutamine and GABA contents could not be explained by the expression patterns of the corresponding genes.

Dividing fruit are much more sensitive to carbon depletion than expanding fruit, as their growth and metabolism are rapidly impaired (Figs 2 & 3). This suggests the existence of a critical point related to the physiological capacity of cell proliferation. Thus, in expanding fruit the effects of sugar depletion are less pronounced, as if expanding fruit had overcome this critical point of cell division by being engaged in the process of metabolite storage. The fact that expanding fruit appear as priority organs is in accordance with the ‘priority rank order’ supporting the concept of competition for assimilates between alternative sinks (Wardlaw 1990), and thus the concept of sink strength (Marcelis 1996). Sink size may explain why expanding fruit attract assimilates more efficiently at the expense of dividing fruit. According to the definition of sink activity, SUSY, SPS, acid invertase and sucrose/hexose transporters participate in the control of sugar import into tomato fruit (D’Aoust, Yelle & Nguyen-Quoc 1999 and ref therein; Delrot et al. 2000). Under darkness, most of these sink-related genes were repressed. This could hamper the capacity of dividing fruit to compete for assimilates. However, the strong repression of the SUSY gene alone cannot explain the weak competitiveness for assimilates by dividing fruit. Indeed, D’Aoust et al. (1999) have demonstrated that SUSY does not participate in the control of the sucrose import capacity in dividing tomato fruit when the photo-assimilate supply is limited, as is the case for darkened plants. The aptitude of the tomato fruit to attract assimilates under unfavourable conditions of sugar supply depends also on the mechanisms controlling long-distance transport of reserve assimilates (Farrar 1996). As a consequence the position of the fruit within a truss is an important factor. Indeed within a truss, expanding fruit are at proximal settings and thus in a more favourable position for assimilate supply than the youngest fruit at distal settings (Bertin 1995). Under darkness treatment, dividing fruit experience a shortage in assimilates that could lead to fruit set failure, an incident which occurs in nature or when plants grow under artificially low carbon assimilation (Bertin 1995).


We thank L. Badulescu, who helped us with biochemical analysis of the numerous samples obtained in the darkness experiment. We thank A. Roos and J. P. Desbiens for managing the phytotronic chambers and the cultures. We wish to thank Dr C. Gary for critically reading the manuscript and Dr P. Scott for language corrections. This work was partly funded by a grant from Région Aquitaine, and AIP Agraf-INRA (‘Elaboration de la qualité des fruit’).

Received 15 April 2002;received inrevised form 2 July 2002;accepted for publication 3 July 2002