Tomato (Solanum lycopersicum L.) cv. Micro-Tom plants were used for all tests performed. MT was used as control and compared to the transgenic genotypes MT-pDR5::GUS; MT-pARR5::GUS; MT-pGA20x::GUS; MT-pRD29B::GUS and MT-pEBS::GUS (Table 1). Seeds were disinfected using 5% sodium hypochlorite (HClO) plus 2 drops of detergent and mixed for 1 min and after 10 min drained

using a sieve and abundantly washed with distilled water. This experiment aimed at finding the Al concentration which inhibits approximately 50% of root growth. The results were used to determine concentrations for subsequent experiments with MT.

The nutrient solution used for all tests was that described by [16] at 10% ionic strength. All Al-containing solutions were prepared with aluminum chloride (AlCl₃) with added Homopipes buffer [17] and 80 seeds were used per treatment, with 40 seeds per replicate. Five treatments were used with increasing Al concentrations: control without Al and with 5, 10, 20 and 40 µM of Al. The seeds were placed in 250 mL Erlenmeyers containing 50 mL of nutrient solution and placed under constant shaking and 12 h of light. Root images were obtained by scanning in an Epson Perfection V800 Scanner and ImageJ software [18] on the 3^rd and 5^th days of root growth.

For cell viability assays, 15 seeds were used per treatment, as follows: control without Al and treatments with 3, 9, 18 and 27 µM Al. Seeds were placed in trays and kept in the same conditions described for 7 days. After this period, root cell viability was evaluated by analyzing Evans’ blue absorption by damaged cells. Roots were submerged in 5 mL 0.25% (w/v) Evans’ blue for 10 min, washed in distilled water and taken to a Leica ICC50 HP microscope coupled to a digital camera and a Leica Measure module, as described in [19] . To quantify Evans’ blue absorption, 0.5 cm long root tips were sectioned from 15 roots from each treatment. Roots from each treatment were placed in 3 flacons containing 300 μM dimethylsulfoxide (DMSO) for 2 h, totaling 5 root tips per flacon. After this period, the DMSO solution was removed from the flacons and analyzed in a spectrophotometer at 600 μm wavelength. β-glucuronidase (GUS) histochemical analysis was conducted in MT and MT transgenic plant roots. Twenty seeds from each genotype were used, divided in 10 seeds for the control and the remaining 10 for the 18 μM Al treatment. 48 h after germination, seeds were removed from treatments and analyzed. Five seeds from each treatment were placed in 1.5 mL tubes containing 300 μM GUS solution, 100 mM phosphate buffer at pH 7.0; 10 mM EDTA at pH 8.0; 0.5 mM potassium ferricyanide, 0.05% Triton X-100 and 10 mg∙mL⁻¹ X-Gluc (5-bromo-4-chloro-3-indolyl-beta-D-glucuronic acid, cyclohexylammonium salt) and 500 μM DMSO for 5 min. Roots were infiltrated in a vacuum 5 times, for 5 min each. Afterwards, they were incubated in darkness at 37˚C during 12 h, washed in distilled water and kept in 70% ethanol for 15 min. Then, they were once again washed in distilled water and immersed in 0.24 N HCl with 20% methanol, in a heating pad, at 57˚C, for 15 min. Subsequently, the solution was substituted for 7% NaOH in 60% ethanol for 15 min at room temperature. Samples were washed in distilled water and placed in 40% ethanol for rehydration for 5 min, and then placed in 70% ethanol. Roots were finally taken to a microscope for image acquisition. The whole-root images were obtained using Leica Application Suite – LAS software and data was obtained by observing the intensity of the blue staining in cells and tissues resulting from GUS activity. Parametric analyses were tested using the D’Agostino and Pearson omnibus tests. Significant differences in evaluated parameters were determined by one-way ANOVA, and when between treatments, using the multiple-scale HSD Tukey test. Correlation analysis was conducted using the Pearson correlation coefficient.

After 3 days of Al exposure, the mean average root growth of control plants was significantly greater than other treatments (Figure 1). There was no statistical difference between three of the Al-treated groups, 5, 10 and 20 μM. However, seedlings subjected to 40 μM treatment differed significantly from all others, having the smallest mean root growth and this implies that Al inhibits root growth in a dose-dependent way (Figure 1). After these results, the Al concentration in subsequent experiments was updated to 0, 3, 9, 18 and 27 μM Al.

MT roots subjected to increasing Al concentrations exhibited increasing losses of cell viability. Evans’ blue uptake was significantly greater under increased Al concentration, being clearly enhanced under 27 µM of Al (Figures 2(a)-(e)). These results indicate cell damage caused by Al toxicity, especially in the root elongation zone and corroborated by other studies in pea [19] [20] . Evans blue staining results underscore the data obtained for root growth, which was similar after 3 days in all treatments except in the seedlings subjected to the highest amount of Al (40 μM) (Figure 1). Similarly, absorbance was significantly higher in plants treated with 27 μM Al (Figure 2(f), Figure 2(g)). It was observed that

the root tip began to exhibit tissue necrosis in seedlings subjected to 27 μM Al, demonstrating severe cell damage. High Al concentrations can block water and nutrient uptake mechanisms and can ultimately lead to cytological damage [20] and these effects can, consequently, cause root growth inhibition, which can be explained by elongation and cell division issues [21] .

MT plants were used as the negative control for GUS histochemical assays (Figure 3(a), Figure 3(b)). Transgenic plants expressing DR5::GUS, which contain a synthetic promoter responsive to AIA [22] , showed restricted expression in quiescent center cells of the root tip in the absence of Al (Figure 3(c)). GUS activity increased after Al exposure, encompassing the entire apical meristem region (Figure 3(d)). Root growth is also intimately related to the synthesis, transport, and distribution of auxin [7] . Aluminum may interact with the polar transport of AIA and therefore alter its accumulation in roots [23] . In turn, these changes in AIA distribution contribute to growth inhibition. Indeed, our results

show that, while AIA accumulates in the root apex, basipetal transport to the elongation zone is apparently inhibited. Therefore, Al exposure leads to an interference in normal AIA levels required for cell elongation specifically in the elongation zone. Al-treated transgenic plants expression of EBS::GUS, responsive to an ethylene pathway transcription factor (EIN3), was enhanced in the root tip and transition zone, albeit diffusely. Activity was also verified in the vascular tissue of the differentiation zone (Figure 3(e), Figure 3(f)). In Al-treated plants, the ET response is generally seen as an upstream regulator of AIA synthesis in the root [11] [15] . Al-treated plants exhibit root elongation inhibition as well as an increase in ACC-oxidase activity [7] , indicating that ET is directly involved in Al-induced root symptoms, including cell autophagy and apoptosis. Transgenic roots expressing RD29::GUS, responsive to ABA, demonstrated activity in the meristematic region in absence of Al (Figure 3(g)), but Al-treated roots showed diffuse expression in root tip cells, and to a lesser extent to the transition and elongation zones (Figure 3(h)). While ABA has generally been correlated with alleviation of Al toxicity, precise mechanisms by which this is achieved are not known [24] . However, a recent study correlates ABA with reactive oxygen species protection in Al-stressed plants [9] . Accordingly, ABA levels were highest in the root tip, which is most affected by Al toxicity, and would, then, have the highest levels of oxidative cell damage. Nonetheless, the protective mechanisms induced by ABA were not enough to prevent tissue necrosis and root growth inhibition (Figure 1 and Figure 2(a), Figure 2(e)). The expression of ARR5:GUS, which responds to CKs in the root tips, including both the meristem zone and transition zone, was highly induced under Al treatment (Figure 3(i)), even if less heterogeneously in the transition and differentiation zones than in the meristematic region (Figure 3(j)). In the transition zone, CKs seem to be the last step in root growth inhibition [25] . The authors show that CKs production is essential for diminishing root growth, since mutants deficient in CK biosynthesis exhibit less inhibition of root elongation in the presence of Al. In addition, GA accumulation was observed using plants expressing the GA2OX::GUS construct, which is responsive to GA2-oxidase activity. There was no GUS expression in either treatment, with (Figure 3(k)) and without Al (Figure 3(l)), after 4 days of Al exposure. GA2OX::GUS activity was also measured after 5 days of Al exposure, and the same negative results were obtained for both treatments without and with Al. There is relatively little information on the role of GAs in the Al response [24] , however, it appears that GAs play a significant role in other stress responses, such as salt stress [26] . It is possible that a change in GA levels may only occur shortly after Al exposure, to elicit further responses from the root. In Picea abies, a GA peak in Al-stressed plants occurred after a 5 h treatment, but GA levels then decreased to sub-normal levels along nearly 9 months of treatment [27] .