UNIFLORA, a pivotal gene that regulates floral transition and meristem identity in tomato (Lycopersicon esculentum)


  • Vincent Dielen,

    1. Unité de Biologie végétale, Département de Biologie et Institut des Sciences de la Vie, Université catholique de Louvain, Croix du Sud 5, boîte 13, B−1348 Louvain-la-Neuve, Belgium;
    Search for more papers by this author
  • Muriel Quinet,

    1. Unité de Biologie végétale, Département de Biologie et Institut des Sciences de la Vie, Université catholique de Louvain, Croix du Sud 5, boîte 13, B−1348 Louvain-la-Neuve, Belgium;
    Search for more papers by this author
  • Jaime Chao,

    1. Unité de Biologie végétale, Département de Biologie et Institut des Sciences de la Vie, Université catholique de Louvain, Croix du Sud 5, boîte 13, B−1348 Louvain-la-Neuve, Belgium;
    Search for more papers by this author
  • Henri Batoko,

    1. Unité de Biologie végétale, Département de Biologie et Institut des Sciences de la Vie, Université catholique de Louvain, Croix du Sud 5, boîte 13, B−1348 Louvain-la-Neuve, Belgium;
    Search for more papers by this author
  • Andrée Havelange,

    1. Laboratoire de Physiologie végétale, Département des Sciences de la Vie, Université de Liège, Bât. B 22 – Botanique, Sart Tilman, B−4000 Liège, Belgium
    Search for more papers by this author
  • Jean-Marie Kinet

    Corresponding author
    1. Unité de Biologie végétale, Département de Biologie et Institut des Sciences de la Vie, Université catholique de Louvain, Croix du Sud 5, boîte 13, B−1348 Louvain-la-Neuve, Belgium;
    Search for more papers by this author

Author for correspondence: Jean-Marie Kinet Tel: +32 10 472050 Fax: +32 10 473435 Email: kinet@bota.ucl.ac.be


  • • Flowering of uniflora (uf), a tomato (Lycopersicon esculentum) mutant which consistently produces solitary flowers instead of inflorescences, is late and highly asynchronous in winter. This puzzling behaviour prompted us to further investigate flowering regulation in this mutant to improve our understanding of UNIFLORA gene function.
  • • Growing plants under different daylengths and light intensities revealed that flowering time in uf is dependent on daily light energy integral. Transferring plants from low to high light energy integrals at different times after sowing showed that the light-conditions effect was stage dependent, suggesting that interactions between light energy integrals and endogenous regulatory pathways affect meristem sensitivity to flowering signals.
  • • Carbohydrate analyses suggested that one of these signals could be sucrose, but other interacting factors are probably generated by the root system, as indicated by grafting experiments.
  • • The UNIFLORA gene thus appears to have a dual role in tomato: floral transition regulation and the maintenance of inflorescence meristem identity.


Flowering is under the control of numerous genes which interact to regulate the timing of this event and the morphogenesis of inflorescences and flowers. In Arabidopsis thaliana, there are various pathways that coordinate flowering time with environmental conditions and developmental stage of the plant (Levy & Dean, 1998; Reeves & Coupland, 2000; Périlleux & Bernier, 2002). These pathways converge to integration pathways that ultimately regulate genes controlling meristem identity (Périlleux & Bernier, 2002; Zhao et al., 2002).

In tomato, flowering time, as measured by the number of leaves formed before the conversion of the vegetative meristem into a reproductive structure, is rather stable under various environmental conditions (Kinet & Peet, 1997). The major effect upon the extent of the vegetative growth is attributable to the light energy integral to which plants are exposed within a 24-h cycle. When this parameter is kept constant under different daylengths, short days (SD) may slightly reduce leaf number preceding the first inflorescence compared with long days (LD) in some cultivars (Kinet, 1977). However, as typical growing conditions of tomatoes in temperate regions are long days, this species is classified as an autonomously flowering plant: its conversion from vegetative to reproductive development is normally regulated by a developmental programme rather than by environmental cues. Flowering transition in tomato is also dependent on the genetic background of the plant and variation among cultivars in leaf number before the first inflorescence has long been documented, as well as variation between mutants (for review see Kinet & Peet, 1997). Recently, the spontaneous mutation uniflora (uf), in the Ailsa Craig (AC) genetic background, was reported to produce, in addition to a single-flower phenotype in place of the classical inflorescence of tomato, a late-flowering phenotype with an increased number of leaves below the first reproductive structure (Dielen et al., 1998). Moreover, flowering time of uf plants was strongly season-dependent, being markedly delayed in winter.

Another major aspect of the behaviour of the uf mutant in winter was that its flowering was typically asynchronous, some plants flowering far later than others, and some failing to initiate a single flower before the experiments were discontinued, after the production of more than 40 leaves. This intriguing observation prompted us to investigate further the uf mutant aiming at improving our understanding of the role of the UNIFLORA gene in the control of flowering in tomato. Unfortunately, the UF gene has neither been cloned nor mapped, hindering detailed molecular studies. In the present work, we thus favoured the physiological approach to investigate how environmental cues may interact with the developmental program during floral transition in the uf mutant.

Materials and methods

Plant material

Seeds of the ‘Ailsa Craig’ tomato (Lycopersicon esculentum Mill.) cultivar and of its near-isogenic uniflora mutant were obtained from the Tomato Genetics Resource Centre (University of California, Davis, CA, USA). They were multiplied in a glasshouse to constitute a seed stock for our experiments. The uf mutant was first described in 1967 by Fehleisen (1967) and it produces single normal and fertile flowers instead of classical wild-type inflorescences (see Fig. 1) (Dielen et al., 1998).

Figure 1.

Comparison of the reproductive structures of the Ailsa Craig (AC) tomato (Lycopersicon esculentum) cultivar (a) and of the uniflora (uf) mutant (b).

Growth conditions

Seeds were germinated at about 25°C in a peat compost. When approximately 2 wk old, seedlings were transplanted to 7 cm × 7 cm pots filled with the same compost and, about 3 wk later, plants were transferred to 15 cm pots, and fertilized weekly with a nutrient solution made of 15 g l−1 of a 16 : 18 : 21 N–P–K fertilizer. All experiments were carried out either in a glasshouse in Louvain-la-Neuve at different seasons or in phytotronic growth chambers at the University of Liège, Belgium, where day and night temperature was maintained at 20–25°C and light was provided exclusively by means of white fluorescent lamps (Osram Cool white, 40 W). As indicated in the Results section, different light regimes were used and, depending on the experiment, plants were kept either continuously under the same light regime or were transferred at different ages from one regime to another.

Grafting experiment

The grafting experiment was performed in a glasshouse under natural light conditions of winter and spring in order to extend the vegetative phase of uf.

Reciprocal grafts were made with 25-d-old AC and uf plants using a technique similar to that described by Mapelli and Kinet (1992). Briefly, plants were topped below the first true leaf and the lower part, comprising the decapitated stump with cotyledonary leaves, was used as a stock while excised apical shoot served as a scion. The appropriate combinations of stocks and scions were established by inserting the scion into a 1–2 cm longitudinal cleft at the top of the stock's stump and wrapping the graft with a cohesive, extensible and air-permeable strip. Grafted plants and their container were kept, for 1 month, in a tightly close plastic bag, which was progressively removed later on. Intact AC and uf plants were cultivated alongside the grafts, at the same time in the glasshouse, as control sets.

Apical exudates collection and sugar purification

Apical exudates were collected using the ethylenediaminetetraacetic acid (EDTA) method described by Lejeune et al. (1993) with some adaptations to suit our material. Plants were topped as close as possible to the apex and 2 ml Eppendorf microcentrifuge tubes, containing 1.5 ml of 1% ultrapure agarose (Life Technologies, Gaithersburg, MD, USA) supplemented with 5.8 g l−1 EDTA (pH 7.0), were immediately placed on the cut stump. After a 16-h exudation period, the tubes were removed from the plants and stored at −20°C until use. The agarose containing-exudates were thawed on ice and extracted with an equal volume of methanol followed by centrifugation for 30 min at 7800 g in a refrigerated centrifuge. The supernatant was kept and the agarose pellet was re-extracted overnight at 4°C with an equal volume of 50% methanol. After centrifugation, supernatants from the two extractions were pooled and evaporated up to dryness in a SpeedVac (HetoVac Model VR-1; Heto, Allerød, Denmark), at room temperature. The residue was dissolved in sterile distilled water and allowed to pass through a column containing a slur of an anion resin (AG-2×8, 200–400 mesh, 1 ml ml−1 of exudates; Bio-Rad, Hercules, CA, USA) to get rid of EDTA. The resin was rinsed with distilled water (2 ml ml−1 of resin). Before purification, galactose (1 mg ml−1 of exudates) was added as an internal sugar standard to evaluate the recovery rate, which ranged between 70 and 85%.

Carbohydrate analysis

The purified exudates were filtered through a 0.2-µm cut-off membrane (Acrodisc PVDF; Pall Gelman Laboratory, Ann Arbor, MI, USA) and fractionated by High Pressure Liquid Chromatography (Bio-Rad HPLC system Model 2700). Carbohydrates separation was achieved using a Bio-Rad Aminex HPX-87C column (300 × 7.8 mm internal diameter) maintained at 80°C using degassed ultrapure water as mobile phase at a rate of 0.6 ml min−1. A Carbo-C guard column (30 × 4 mm internal diameter; Bio-Rad) was introduced between the injector and the analytical column. Detection was achieved via a Bio-Rad Refractive Index Monitor (Model, 1755). Each compound was identified by matching its retention time to that of commercially available reference sugar standards (Sigma Chemical, St Louis, MO, USA) and quantification was accordingly determined by the peak area recorded using valuechrom software (Bio-Rad).

Criteria of flowering response

Two criteria were used to evaluate the flowering response: the percentage of plants forming a reproductive structure and the flowering time. Flowering time was assessed by two different measurements. (1) The number of leaves produced below the first reproductive structure; uf plants that did not flower after initiating up to 40 leaves were classified as vegetative (Dielen et al., 1998). (2) The number of days from sowing to macroscopic appearance of the reproductive structure.

Statistical analysis

The analysis of variance (anova i) at the 5% level was performed on the number of leaves initiated below the flower, the number of days before macroscopic appearance of the reproductive structure and on sugar content in the exudates. Differences between means were scored for significance by using the Student–Newman–Keuls test. Comparison of flowering percentages was performed using χ2 test.


Effect of the daily light energy integral on flower initiation

To investigate which seasonal component may be affecting flower initiation in uf tomato mutant, we cultured wild-type and mutant plants, from sowing, under different light conditions consisting of a combination of two daylengths (8 h, SD, and 16 h, LD) with two light irradiances (90 or 180 µmol m−2 s−1 over the waveband 400–700 nm at the top of the canopy). These light conditions represented three different daily light energy integral regimes as LD at low irradiance and SD at high irradiance provided the same total light energy during a 24-h cycle. Several independent experiments, with 12–20 plants per treatment, were performed and each independent experiment was analysed separately (Table 1). The results for the two treatments giving the same daily light energy integral were roughly similar and therefore pooled in data presented in Table 1. The percentage of wild-type plants flowering was not affected by the relative amount of energy given, and was the maximum in each case. By contrast, the relative number of uf plants flowering was significantly affected by the daily light energy integral, decreasing with the amount of energy received, with no plant flowering at the lowest energy level tested. However, although this general tendency was reproducible, the percentages of mutant plants flowering in response to a given light treatment varied from experiment to experiment, suggesting a quite flexible behaviour, which was consistent with all our previous observations under various growth conditions for almost a decade, as mentioned in the Introduction section. In both AC and uf, macroscopic appearance of the reproductive structure was earlier the higher the daily light energy integral. Differences were statistically significant for AC but not in two out of three experiments for uf due, once again, to the heterogeneous behaviour of the plants. The number of leaves formed by the wild type before the first inflorescence was stable (around 10), irrespective of the level of daily light energy integral (Table 1). In comparison, for mutant plants provided with the same amount of light, those individuals that flowered did so after forming at least twice the number of leaves of the wild type. Interestingly, in plants that failed to flower, particularly in the intermediate daily light energy integral conditions, we observed the development of lateral shoots at nodes 10–12 (Fig. 2) (i.e. at a position where the wild type produces its first inflorescence; Fig. 2a, arrow). Figure 2b shows one of these nonflowering mutant plants, with developed axillaries.

Table 1.  Effect of daily light integral upon flowering response of the Ailsa Craig (AC) cultivar and of the uniflora (uf) tomato (Lycopersicon esculentum) mutant
Flowering parameterRepetition of the experimentDaily light energy integral
  1. 4×, long days (LD) under high irradiance; 2×, either LD at low irradiance or short days (SD) at high irradiance; 1×, SD at low irradiance. MA, macroscopic appearance of the inflorescence or of the flower; ND, no data; NA, not applicable; > 40, no flowering when the experiment was discontinued. Values are means ± SE. Values followed by a same letter in a row are not statistically different at the 5% level

AC cultivar
Flowering percentage1100100ND
Days before MA ± SE134.1 ± 3.5 a38.4 ± 3.5 bND
244.9 ± 0.4 a47.2 ± 0.3 b53.6 ± 1.1 c
Leaves below inflorescence ± SE110.4 ± 0.9 a10.3 ± 0.7 aND
29.85 ± 0.1 a10.2 ± 0.1 a10.4 ± 0.2 b
uf mutant
Flowering percentage1100 a70.8 b0 c
255 a35 bND
361.5 a6.1 b0 b
Days before MA ± SE146.8 ± 6.9 a60.5 ± 12.8 bNA
278.4 ± 16.6 a80.1 ± 4.6 aND
376.8 ± 2.5 a76.0 ± 5.3 aNA
Leaves below flower ± SE114.6 ± 1.8 a15.9 ± 2.5 a> 40
226.4 ± 7.3 a28.9 ± 4.1 aND
322.9 ± 1.0 a23.5 ± 0.5 a> 40
Figure 2.

(a) Size of axillary shoots of nonflowering uniflora (uf) tomato (Lycopersicon esculentum) plants under intermediate light energy integrals, as a function of their position on the stem; arrow indicates the position where the wild type produces its first inflorescence. Vertical bars, SE. (b) Axillary shoots (after removal of main stem leaves) developing at nodes 10–12 on a vegetative uf tomato plant grown under intermediate light energy integrals.

The uf growth stage and responsiveness to light conditions

Plants grown from sowing under 8 h SD, given at an irradiance of 180 µmol m−2 s−1, were transferred at different ages to 16-h LD with the same light irradiance, and kept under these conditions up to the end of the experiment. The first transfer occurred on day 14 and subsequent transfers were done every 3–4 d. Control plants were maintained either under continuous 16 h LD or 8 h SD at the same light irradiance of 180 µmol m−2 s−1. Two experiments were performed with comparable results. As shown in Fig. 3, under continuous LD, flowering percentages were significantly higher than under permanent SD, which is consistent with earlier results. Plants that were transferred from SD to LD showed flowering percentages similar to controls that were kept continuously under LD (Fig. 3a), independently of the time of the transfer which, in contrast, affected flowering time, as measured by the number of leaves produced before floral transition (Fig. 3b). Flowering occurred later in response to late transfers to high light conditions but, interestingly, this delay did not follow the progressive delay in the transfer. Indeed, the number of leaves below the flower remained roughly similar to that of plants continuously grown under high light in response to early transfers, before increasing abruptly to numbers similar to those of plants continuously grown under a low light regime after a late transfer. Flowering time thus increased in a step-like manner, suggesting the occurrence of phases with different responsiveness or sensitivity to environmental conditions.

Figure 3.

Flowering percentages (a) and number of leaves below the first flower (b) of uf tomato (Lycopersicon esculentum) plants as a function of time of transfer from short days (SD) to long days (LD). Horizontal bars represent flowering percentages (a) and number of leaves recorded below the first flower (b) of control plants kept continuously under either LD or SD. Vertical bars, SE. Values followed by the same letter are not statistically different at the 5% level.

Grafting experiments and carbohydrate levels in apical exudates

In an attempt to approach the nature of the flowering signals that could mediate the interactions between light energy integrals and the endogenous regulatory pathways that affect meristem sensitivity, graft experiments and analyses of sugars circulating in the phloem sap of both the AC wild type and the uf mutant were performed.

Effect of grafting on flowering

All AC plants used for grafting experiments had started their sympodial development, alternating production of leaves and inflorescences, owing to their precocious floral transition occurring at the seedling stage (Kinet & Peet, 1997). Therefore, investigating AC plants during the initial vegetative phase was unfeasible. By contrast, all uf plants were strictly vegetative when graft was performed. When AC was used as a scion, all resulting plants flowered early regardless of stock (Table 2). Surprisingly, when uf was the scion, flowering occurred in almost all plants after producing 23–25 leaves, regardless of the genotype of the stock. Thus, all uf scions grafted, even on uf stocks, flowered despite the fact that only 13% of ungrafted uf control plants were able to produce a flower before having initiated 40 leaves.

Table 2.  Flowering response of scion following reciprocal grafts between the Ailsa Craig (AC) cultivar and the uniflora (uf) mutant of tomato (Lycopersicon esculentum)
Graft type scion/stockNumber of plantsNumber of successful grafts (%)% of flowering plantsNumber of leaves below the first reproductive structure1
  • 1

    Values are means ± SE.

  • 2

    2 nr, not relevant as floral transition of scions occurred before grafting.

Intact AC control1210010.4 ± 0.4
Intact uf control31 1330.0 ± 1.1
AC/AC1211 (92)100nr2
uf/uf1210 (83)10025.1 ± 1.0
AC/uf12 8 (67)100nr2
uf/AC1210 (83) 9022.9 ± 0.7

Carbohydrate concentrations in apical exudates of AC and uf plants

We next looked at the amount of carbohydrate reaching the apical bud in wild-type and mutant plants. Four experiments were conducted; three used plants grown in the glasshouse under natural light, as affected by seasonal variations throughout the year, and for the last experiment, plants were allowed to grow in a phytotronic chamber, under a 16-h LD photoperiod and light irradiance of 180 µmol m−2 s−1 at the top of the canopy. Because the exudation technique required the use of plants with a sufficiently elongated and rigid stem to support the Eppendorf tube used to collect the apical exudates, the age of plants used varied between 5 wk and 10 wk, depending on experiment. The variable ages and growth conditions allowed the analysis of uf plants at different physiological stages (vegetative, floral or even mixed set of plants; Table 3) compared with the AC individuals which had already started their sympodial development. Sucrose was the major sugar present in the apical exudates. Reduced quantities of glucose and fructose were also detected, probably as a result of the hydrolysis of sucrose during exudates collection and/or purification (Corbesier et al., 1998) as these two hexoses are usually absent from the phloem sap (Crafts & Crisp, 1971). Consequently, the amounts of sucrose, glucose and fructose were summed and expressed in sucrose equivalent content (sucrose + (glucose + fructose)/2).

Table 3.  Carbohydrate content of apical exudates of Ailsa Craig (AC) and uniflora (uf) tomato (Lycopersicon esculentum) plants
Experiment number1Physiological stage of the plantsSucrose equivalent content (µmol)2
  1. Plants were grown either in a glasshouse (experiments 1–3) or in a phytotronic chamber (experiment 4) under long days at an irradiance of 180 µmol m2 s1. Values are means ± SE for three replicate measurements. Values followed by different letters in a row are statistically different at the 5% level. 1Dates in parentheses refer to the day of the year the exudates were collected; nr, not relevant (experiment performed in phytotronic growth rooms). 2Carbohydrate contents are expressed in µmol of sucrose equivalents (µmol sucrose + [µmol glucose + µmol fructose]/2) recorded per plant during a 16-h period.

1 (08/01)FloralVegetative5.9 ± 0.7 a3.6 ± 0.12 b
2 (11/06)FloralFloral6.1 ± 0.3 a3.0 ± 0.06 b
3 (12/06)FloralFloral5.1 ± 0.3 a3.8 ± 0.12 b
4 (nr)Floral55% floral7.8 ± 0.4 a3.3 ± 0.06 b

The sugar concentration in apical exudates was relatively stable and reproducible from experiments to experiments in both the AC cultivar and the mutant. However, it was strongly and significantly reduced in uf plants compared with AC plants (Table 3) under all experimental conditions tested and thus independent of the physiological state of the uf plants.


UNIFLORA and the regulation of flowering transition in tomato

Investigating flowering time in uf is not straightforward given the variability of response between experiments, although every effort was made to reproduce experimental conditions. This behaviour was more obvious when plants were grown under conditions that were neither fully favourable nor fully detrimental to flowering. This inconsistency of the absolute flowering response, which has been repeatedly observed from our early work on this mutant (Dielen et al., 1998, 2001), suggests that UNIFLORA is a pivotal gene regulating floral transition in tomato. When mutated, passing the step controlled by UF becomes dependent on environmental parameters, and subtle variations in culture conditions, which are difficult to control and, hence, to reproduce from one experiment to another, may be critical to convey the decision whether an individual will remain vegetative or not. That UNIFLORA may be playing a critical role during floral transition is supported by the observation that, in plants that remained vegetative, lateral branches were developing at node levels where flowering normally occur on wild-type plant (AC cultivar). Release from apical dominance is known to be an early and essential event, always associated with floral evocation (Bernier et al., 1981). Its occurrence in the uf mutant strongly suggests that vegetative plants underwent a partial evocation but were unable to complete the process. Afterwards, the shoot apical meristem returned to a vegetative functioning and apical dominance was re-established as axillary outgrowth was inhibited at upper nodes (Fig. 2). Thus, the shoot apical meristem of uf plants appears to have the capacity to swing between successive vegetative and partly florally evoked states.

Discontinued light experiments clearly revealed that, during an extended period of vegetative growth, uf plants underwent changes in their developmental programme although maintaining a monopodial mode of growth. These changes affected the competence of the meristem to respond to signals generated by external cues, such as the level of daily light energy integral. This observation is reminiscent of results obtained in works showing that flowering of many plants has distinct phases of sensitivity to the environment, especially to the photoperiod (Ellis et al., 1992; Adams et al., 2001). In many species, seedlings are incapable of responding to inductive conditions, exhibiting a so-called ‘juvenile’ phase. When this initial phase is completed, plants become capable of responding to inductive conditions and are considered to be ‘ripe to flower’, a condition that is generally stable (Thomas & Vince-Prue, 1984). Interestingly, the situation in uf is different as young plants are first sensitive to light conditions promoting flowering and enter thereafter into a phase of reduced sensitivity. According to Pnueli et al., 1998), the gene SELF PRUNING acts, in indeterminate tomatoes, as part of a system which prevents early flowering in each of the de novo developing sympodial shoot meristems. This system should be downregulated in a step-like manner, with each internode of every new sympodial segment, to allow transition to flowering after having initiated a limited number of leaves. The question thus arises as to whether SELF PRUNING is implicated in the alternation of phases of higher and lower sensitivity to light conditions, exhibited by the monopodial shoot of uf.

Measurement of sucrose content in the apical exudates indicated that carbohydrate supply to the apex was lower in uf than in AC plants. This reduced sugar availability appeared to be independent of the physiological stage of the uf plants, suggesting that their apical bud is permanently undersupplied compared with their wild-type counterpart. This finding, associated with the fact that flowering is strongly delayed in uf plants, could suggest that sucrose is one of the signals implicated in the regulation of floral transition in tomato and that daily light energy integral may interact with the endogenous developmental programme through the supply of adequate levels of sugars to the meristem. That sucrose is required for flowering in tomato was suggested by in vitro studies (Dielen et al., 2001) and by work with transgenic plants which have an increased capacity for sucrose synthesis (Micallef et al., 1995).

We showed also that floral transition in uf was effectively triggered by grafting as flowering occurred in all grafted plants despite the fact that light conditions were unfavourable, as revealed by the poor response of intact control uf plants. The mechanisms by which grafting stimulated flowering under unfavourable light conditions are far from clear. Grafting is considered to be a valuable tool to investigate long distance signalling pathways (Turnbull et al., 2002). Our results suggest that a root signal may be affecting the fate of the shoot meristem and that the efficiency of this signal is not dependent on the stock genotype. The stimulatory effect of graft per se, could thus be caused by the fact that it brings the shoot apical meristem closer to the root system, a situation at odds with what was reported for Nicotiana tabacum, where the roots sustained vegetative growth when in close proximity to the apical meristem (McDaniel, 1996). However, a positive effect of roots upon floral transition in tomato was also suggested by previous experiments on uf, using topping and cutting treatments (Dielen et al., 1998), and by the observation that plants that were growing vegetatively for a long period of time before being submitted to light conditions conducive to flowering, produced flowers later than young individuals (V. Dielen & J. M. Kinet, unpubl. data). It is tempting to envisage that the potential stimulatory role of the root system, as revealed by grafting experiments, could be mediated by cytokinins as cytokinins are required for in vitro flowering of tomato (Dielen et al., 2001).

UNIFLORA and the regulation of reproductive morphogenesis in tomato

Throughout the years we have submitted uf plants to various growing conditions and have not found a single treatment capable of modifying the single-flower phenotype of this mutant, which is thus remarkably stable. Furthermore, work in progress in our laboratory aiming at generating double mutants suggests that UNIFLORA may be acting during floral transition upstream of genes affecting the structure of the tomato inflorescence, such as COMPOUND INFLORESCENCE (S) or BLIND (BL). Earlier observations (Dielen et al., 1998) led us to postulate, in agreement with Allen and Sussex (1996), that the reproductive structure of tomato is a raceme-like inflorescence; consequently, the uf mutation may be affecting a gene that controls inflorescence meristem identity. TERMINAL FLOWER (TFL) and CENTRORADIALIS (CEN) are genes that control the identity of the inflorescence meristem in Arabidopsis thaliana and in Antirrhinum majus, respectively (Bradley et al., 1997). Their orthologue in tomato is SELF PRUNING (Pnueli et al., 1998), a gene that is distinct from UNIFLORA as revealed by the phenotype of the F1 progeny of a cross between self pruning and uf mutants (M. Quinet et al. unpubl. data). Thus, the precise function of UNIFLORA and the way it interacts with other genes await further studies.

UNIFLORA, a key gene regulating flowering in tomato

To date, all the information at our disposal coming from segregation analyses in F2 populations issued from crosses between different mutants and backcrosses to wild-type plants indicate that the uf phenotypes are under the control of a single recessive mutation (M. Quinet et al. unpubl. data). The UF gene thus appears to act in both the promotion of floral transition and the regulation of reproductive morphogenesis in tomato. It appears to be involved in the control of inflorescence meristem identity, operating upstream of various genes regulating inflorescence edification. The gene UNIFLORA has not yet been mapped in the tomato genome. Cloning and sequencing this pivotal gene will allow an in-depth investigation of its functions and improve our knowledge of the genetic and molecular control of flowering in tomato.


This work was supported by the Fonds National de la Recherche Scientifique (FNRS) of Belgium (Fonds de la Recherche Fondamentale et Collective, 2001–4) and by the Université catholique de Louvain (Fonds Spécial de Recherche, 1995–99). V. D and M. Q., are, respectively, grateful to the UCL and to the FNRS for the award of a research fellowship.