The tomato variety SUN7705 (Nunhems, Parma, ID 83660, USA) was used in all the experiments. Seeds were sown in soil (Hawita Flor, Vechta, Germany) in plastic plugs and in a growth chamber under the following conditions: 25°C temperature, 55% relative humidity, 80 μmol m^-2 s^-1 PAR (photosynthetically active radiation). From germination to transplanting, plants were watered once a week with a nutritive solution, whose composition was the same as that used in the pot cultivation (see later in this paragraph). About 40 days after germination, tomato plants were transplanted to 24 cm diameter plastic pots (volume = 8 dm³) containing a mixture of soil and pumice (70:30, by volume), and transferred to a glass greenhouse. Pumice was used in order to facilitate water drainage. The main characteristics of the soil were: clay 8.4%, silt 32.0%, sand 59.6%; C/N 8.5; organic matter 1.31%; and electrical conductivity 0.44 mS cm^-1. Throughout the trial, day/night temperatures ranged from 25 to 31°C, and from 15 to 21°C, respectively. The composition of the nutritive solution, supplied to plants for 1 min three times per day, was: (in mM) N-NO₃ 11; N-NH₄ 0.5; P 1.2; K 7; Ca 4; Mg 0.94; Na 10; Cl 9.5; S-SO₄ 2.16; and (in μM) Fe 45; B 23; Cu 1; Zn 5; Mn 10; Mo 1; EC 2.79 mS cm^-1, and pH between 5.7 and 6.0. The moderate sodium and chloride content was due to the use of slightly saline irrigation water.

For pest management, a foliar application of copper was performed before transplanting to prevent tomato blight. Confidor (Bayer, Germany) was applied as a foliar application against aphids and white flies, once a week from transplanting to flowering. In addition, a systemic fungicide (Ridomil Gold^® EC, Novartis, NY, USA) was applied to the soil every 10 days from transplant to harvest.

Different experimental trials were performed, as later described. In all the trials, iodine treatment administrations were carried out, supplying iodine to pots as KI or KIO₃, dissolved in a volume of 200 ml water per plant. KI or KIO₃ concentrations ranged from 0 to 10 mM, depending on the type of experiment. Treatment applications were carried out weekly, starting from the development of the first branch of fruits.

Tomato plants were grown in soil in a glass greenhouse, fertirrigated with a nutritive solution, as above described. Starting with the set of the first truss of fruits, plants were root-treated with KI or KIO₃ once a week. Eight iodine administrations were performed.

Following a preliminary trial, with a very wide iodine dose–response curve (KI and KIO₃ concentrations ranging from 0 to 60 mM), performed to find out the most suitable doses of iodine without phytotoxicity symptoms, KI was supplied in concentrations of 1, 2, and 5 mM, while KIO₃ in concentrations of 0.5, 1, and 2 mM. Ten replicates for each experimental condition were carried out.

After the first four iodine administrations (total effective iodine supplied per plant: 0, 50.76, 101.52, 203.04, 507.6, and 1,015.2 mg I, corresponding, respectively, to 0, 0.5, 1, 2, 5 and 10 mM I applied as KI or KIO₃), the iodine content was measured in fruits from both the first and the second trusses. Other four iodine treatments were then carried out and the iodine content was measured in fruits collected from the first truss and used for the qualitative analyses.

Plants were grown in pots in a glass greenhouse, fertirrigated with a nutritive solution, as above described, and divided into two groups, according to the organic matter content of the soil mixture used. Two different soil mixtures, characterized by a low and a high organic matter content, respectively, were used. The composition of the soil mixture with the low organic matter content (approximately 1% on a weight base) was the same described above (soil:pumice, 70:30 by volume). The soil mixture with the high organic matter content was obtained by mixing soil, commercial peat (Hawita Flor) and pumice (28:41:30, by volume), considering the main characteristics of the different substrates, that were, respectively: organic matter content: 1.31, 40, and 0% on a dry matter basis; apparent density (kg/L): 1.5, 0.5, and 0.85 on a dry matter basis; dry matter content: 90, 72, and 90%. In the mixture soil enriched with peat the final organic matter content was approximately 10% (determined on a weight base).

Starting from the development of the first branch of fruits, four weekly administrations of 10 mM KI or KIO₃ (total effective iodine supplied per plant: 1,015.2 mg I) were performed in both the two groups of plants. Control plants, untreated with iodine, were also grown in the two types of soils. Ten replicates for each experimental condition were carried out.

Fruits were collected from the first fruit cluster at the end of the iodine treatments and analyzed for the iodine content. At the end of the trial, some plant growth parameters (fruit and shoot dry weight, fruit yield) were measured.

Plants were grown in soil in a glass greenhouse, as above described, and divided into three different groups, according to the nitrate level of the nutritive solution. Three different nutritive solutions, containing, respectively, a low (2 mM), medium (10 mM), and high (20 mM) nitrate content, were used. The medium nitrate nutritive solution had the following composition: (in mM) N-NO₃ 10; P 1.2; K 8; Ca 6; Mg 1; Na 10; Cl 9.5; S-SO₄ 4.97; and (in μM) Fe 56; B 23; Cu 1; Zn 5; Mn 11; Mo 1; EC 3.32 mS cm^-1, and pH between 5.7 and 6.0. The low nitrate nutritive solution contained 2 mM N-NO₃, while the high nitrate one 20 mM N-NO₃. Furthermore, some adjustments were made to the low and high nitrate solutions to maintain comparable macronutrient levels. The low nitrate nutritive solution contained additional 7 mmol l^-1 chloride, while in the high nitrate solution the sulfate content was reduced to 0.8 mM.

Starting from the development of the first branch of fruits, four weekly administrations of 10 mM KI or KIO₃ (total effective iodine supplied per plant: 1,015.2 mg I) were performed in all the three groups of plants. Control plants, untreated with iodine, were also grown with each of the three different nutritive solutions. Ten replicates for each experimental condition were carried out.

Fruits were collected from the first fruit cluster at the end of the iodine treatments and analyzed for the iodine content. At the end of the trial, some plant growth parameters (fruit and shoot dry weight, fruit yield) were measured.

For this experiment, fruits collected from plants of the Experiment 1, treated for 4 weeks with 5 mM KI (total effective iodine supplied per plant: 507.6 mg I) were used. Both turning red and red fruits were chosen.

The shelf-life experiment was performed by storing the turning red fruits under light at room temperature without any treatment for the following 2 weeks after harvest. The analyses of the iodine content on the stored fruits were carried out 1 and 2 weeks after harvest.

The cooking experiment was performed by boiling red tomato fruits for 30 min in deionized water. Processed fruits were divided into two groups, and boiled, with or without the external peel, respectively.

Fruits collected from plants of the Experiment 1 treated for 8 weeks with KI or KIO₃ were used. Both quantitative measures (fruit yield) and qualitative analyses (color, sugar content, total antioxidant power) were carried out.

For the analysis of the iodine content, fruits were harvested waiting at least 1 week after the last iodine treatment. Iodine as I was analyzed by inductively coupled plasma mass Spectrometry (ICP-MS), as previously described (Landini et al., 2011).

For the evaluation of dry weight (DW), fruits and shoots were weighed separately immediately after harvest and then dried in a ventilated oven at 80°C. All the fruits collected from plants at the end of the experiments were weighed for the analysis of fruit yields. Experiments were not continued after the collection of fruits from the first two branches. The calculated yields therefore always refer to the fruits collected from these two trusses, already developed, and those still growing in the third truss.

In order to analyse sugar content, whole fresh fruits were homogenized in a blender. An aliquot of the homogenate was then centrifuged twice for 10 min at 5,000 rpm and some drops of the supernatant were used to determine total soluble solids with a Refractometer (RL3 PZO). The content of sugars was expressed as degrees Brix (°Brix).

The total antioxidant power of fruits was evaluated using the “ferric-reducing/antioxidant power” (FRAP) assay (Benzie and Strain, 1996). Immediately after harvest, each fruit was homogenized in a blender (0.5 g of the flesh extracted in 5 ml of pure methanol) and stored overnight at -20°C. Samples were then centrifuged for 8 min at 5.000 rpm and 100 μl of the supernatant were added to 900 μl of freshly prepared FRAP reagent [1 mM TPTZ + 2 mM FeCl₃] and 2 ml of acetate buffer. Absorbance readings at 593 nm were taken after a reaction time of 4 min. The reagents used were: acetate buffer (0.25 M sodium acetate, pH 3.6), TPTZ (2,4,6-tripyridyl-2-triazine 0.01 M in methanol) and FeCl₃ (0.01 M in sodium acetate). Standard solutions of known Fe²⁺ concentration (0–50-200-500-1000 μM) were prepared by adding (NH₄)2Fe(SO₄)₂·6 H₂O to the acetate buffer, and were used for the calibration.

The experimental design adopted in the Experiments 1, 2, and 3 was completely randomized. The treatments (iodine source and organic matter in the Experiment 2; iodine source and nitrate level in the Experiment 3, respectively) were in factorial combination. Data were subjected to one-way and two-way analysis of variance (ANOVA; Statgraphics Centurion XV program), as described in the Figure legends, and the means were separated using the F-test (95% confidence level).

As a starting point, tomato plants, grown with the experimental set-up previously described (Figure 1), were root-treated with KI or KIO₃ concentrations ranging from 0 to 60 mM. Although clear damage was never observed in the fruits, plants started to show phytotoxicity symptoms (leaf chlorosis, epinasty, visible wilting) at iodine salt concentrations higher than 10 mM. Moreover, in the presence of 40–60 mM KI or KIO₃, plant development and biomass accumulation were severely compromised, with undeniable consequences on the development of the fruits (data not shown).

These preliminary results prompted us to focus on a narrower and lower range of iodine concentrations, to limit any phytotoxicity symptoms on the plants. KI was thus supplied in concentrations of 1, 2, and 5 mM, while KIO₃ was used at lower concentrations, namely 0.5, 1, and 2 mM, since this salt showed greater phytotoxicity in the preliminary trial. In this experimental set-up, the plants were healthy at the end of the experiment (Figure 2A), with the exception of those treated with the highest KI concentration (5 mM) which showed some discoloration and necrotic areas, limited to the basal leaves (Figure 2B).

The trial was interrupted when the third truss of fruits was developing and the iodine content was measured in fruits from both the first and the second trusses. Figures 2C,D show the iodine content detected in fruits collected from the second branch at the mature green stage. A very regular trend in the increase in fruit iodine content with the increase in its soil administration can be observed. After four treatments with 1, 2, and 5 mM KI, fruits contained an average of 1.5, 4.7, and 10 mg I kg^-1 fresh weight (FW), respectively (Figure 2C). A similar trend can be observed in fruits from the plants treated with potassium iodate, with fruits accumulating 0.3, 0.7, and 1.2 mg I kg^-1 FW following four applications of 0.5, 1, and 2 mM KIO₃, respectively (Figure 2D). Comparable results were obtained in fruits collected from the first truss (data not shown). In the analyses performed, the untreated control fruits showed a small amount of iodine (approximately 0.06 mg I kg^-1 FW) due to the trace amounts of this element present in both the irrigation water (0.109 mg I l^-1) and the soil used (0.084 mg I kg^-1); (Figures 2C,D).

In this second experiment, plants, grown into low or high organic matter soils, were treated with 10 mM KI or KIO₃, a concentration of iodine higher than those used in the previous trial, chosen to better quantify the possible negative effects of the organic matter on the iodine uptake. Over the 4 weeks of treatments, mild phytotoxicity symptoms appeared on the plants, depending on the form of iodine administered as well as the soil organic matter content. The most affected plants were those grown in the lower organic matter soil and treated with KI. Leaves of this group of plants presented some discolorations and necrotic areas (Figure 3A). Similar phytotoxic effects, though less severe, were observed on plants treated with KIO₃ (Figure 3B). In the high organic matter content soil all the plants appeared healthier (Figures 3C,D).

Fruits were collected from the first fruit cluster and analyzed for the iodine content (Figure 3E). The results obtained show that the increase in the organic matter reduced the iodine accumulation in KIO₃- but not in KI-treated plants.

Some plant growth parameters were analyzed in order to better quantify the effects of the different types of soil in combination with the iodine treatments. Plants grown in organic matter-rich soils showed a strong increase in the dry-matter production of their vegetative organs (Figure 3F), which was, on average, 1.5-fold higher than that quantified in the low organic matter soil. This effect was observed irrespectively of the iodine treatment performed. No significant effects were detected in fruit dry weight (Figure 3G), while plant yield was positively affected by the organic matter, as, on average, plants grown in the organic matter-enriched soil showed a fruit production 1.5-fold higher than those grown in the organic matter poor soil, but, again, this was observed irrespectively of the iodine treatment performed (Figure 3H).

The possible interaction between iodine and nitrate contained in the nutritive solution in terms of iodine availability and uptake by tomato plants was examined. Three different nitrate doses (2, 10, and 20 mM) were used to fertilize plants, which were also treated with 10 mM KI or KIO₃. Strong phytotoxicity symptoms on plants treated with KI and grown at the minimal nitrate level (2 mM) were observed (Figure 4B). These plants were strongly reduced in size and biomass production in comparison with plants fertilized with 2 mM nitrate but not treated with KI (Figure 4A), and, at the end of the trial, their basal leaves were completely burnt (Figure 4B). Leaves of the upper branches still showed chlorosis, necrotic areas, curling of the edges and a reduction in size (Figures 4B,C), whereas fruit appearance did not seem to be affected (Figure 4D). Similar phytotoxic effects, though less severe, were observed in plants grown at 2 mM nitrate dose and treated with KIO₃ (data not shown). On the other hand, iodine-treated plants fertilized with 10 and 20 mM nitrate did not show significant alterations in their growth, apart from some chlorotic and necrotic areas on the basal leaves of plants treated with KI. Control plants, not treated with iodine, also showed a slight chlorosis when fertigated with 2 mM nitrate (Figure 4A). The final amount of iodine in fruits collected from plants treated with the same iodine salt and increasing doses of nitrate was comparable (Figure 4E). Only fruits from 10 mM KIO₃-treated plants fertilized with 20 mM nitrate showed a small but significant reduction in the iodine content (Figure 4E).

Plant dry weight and yield were measured. A significant reduction in shoot DW was observed only in plants treated with 2 mM nitrate and 10 mM KI (Figure 4F), as a likely consequence of the strong iodine phytotoxicity under these conditions (Figure 4B). As far as fruit DW is concerned, no significant differences were detected in iodine-treated plants, whereas in control plants, not treated with iodine, a small trend toward a slight increase can be observed comparing, respectively, the 10 and 20 mM nitrate levels (Figure 4G). Finally the level of nitrate fertilization did not significantly affect the fruit yield in control plants, whereas the KI treatment reduced fruit yield in all the plants and in particular in those grown at the lowest nitrate concentration (Figure 4H), probably due to the phytotoxic effects described above. On the contrary, fruit yield in KIO₃-treated plants slightly increased with the increase in the nitrate level (Figure 4H).

To evaluate the possible effect of storage on the level of iodine accumulated in tomatoes, fruits were collected from 5mM KI-treated plants at the breaker stage (Figure 5A). The shelf-life experiment was performed by storing the fruits under light at room temperature without any further treatment for the following 2 weeks, during which fruit ripening continued. The iodine content remained constant in the fruits over time (Figure 5C), showing that 2 weeks of storage did not alter their value of biofortified fruits.

To evaluate the possibility of transforming the iodine-enriched tomatoes into processed food, a cooking experiment was performed by boiling red ripened fruits (Figure 5B) for 30 min. Both raw and processed fruits were divided into two groups, maintaining or removing the external peel. Iodine was finally measured in intact and peeled fruits and also in the fruit skin. Boiling did not alter the amount of iodine present in fruits, and, irrespectively of the treatment, the content of iodine in fruits without peel was lower than that measured in the same intact fruits (Figure 5D). Indeed, the peel alone contained very high levels of iodine (Figure 5D).

For the qualitative analyses of fruits, tomatoes were harvested at the red stage of ripening. In this experiment, iodine treatments were thus prolonged, and fruits were collected after eight iodine applications. Iodine accumulated in these fruits (Figure 6A) with a trend similar to that previously observed after four administrations (Figures 2C,D). The values of the iodine content, with the exception of a few samples, were also comparable. Fruits from KI-treated plants accumulated approximately 0.3–4.5 mg I kg^-1 FW following treatments with 1–5 mM KI respectively, while fruits from 0.5 to 2 mM KIO₃-treated plants ranged from 0.2 to 1.9 mg I kg^-1 FW (Figure 6A). The DW of fruits was not significantly different (Figure 6B).

Fruit quality was evaluated in terms of sugar content and antioxidant power. Treatments with KI and KIO₃ mildly reduced the fruit sugar content, as the °Brix progressively decreased, slightly but significantly, with the increase in potassium iodide or iodate concentrations (Figure 6C). On the contrary, no significant differences were detected in the ferric-reducing ability of tomatoes, with the exception of the value measured in fruits from 5 mM KI-treated plants, which was significantly higher than the control (Figure 6D). These fruits were those that accumulated the highest level of iodine (Figure 6A).