Aquaporins and lipids molecular analysis
The expression analysis of integral membrane proteins such as AQPs (Figure 4 ) showed that B deficiency had the greatest effect on the expression of these proteins in broccoli leaves. The B (-) treated plants leaf presented generally lower levels of mRNA transcripts compared to control leaves, particularly in PIP2 subfamily isoforms, PIP2-1, PIP2-2/3, and PIP2-7. Similarly, B (+) leaves also resulted in lower expression levels of PIP2-1. Out of the other treatments, only salinity had an impact on the expression of PIP1-2, which showed an almost threefold increase in mRNA transcript levels compared to control.
The vesicles obtained from microsomal fraction and plasma membrane isolation were analyzed to characterize their size and shape (seeFigure 5 ), as well as their polydispersity index (PDI). Overall, no differences were observed among the different samples with respect to the treatments applied. However, differences were found between the types of samples. Specifically, the plasma membrane vesicles were more size-stable with a PDI ranging from 0.06 to 0.18 and a mean size around 240-280 nm, while the microsomal fraction vesicles were larger (830 nm) and more polydisperse (PDI of 0.7-0.9).
The immunoblotting analysis (Figure 6 ) displayed the quantification of PIP1, PIP2 subfamilies, and PIP2-7 isoforms in different samples of microsomal fraction and plasma membrane from broccoli leaves of various treatments. In the case of microsomal fractions, changes were only observed in the membrane extractions of the Comb (+) treated plant leaves for PIP1 abundance (Figure 6A ). However, the levels of PIP1 in plasma membrane increased in the NaCl and B (-) treatments compared to the control, but not in the combination of these two, Comb (-). There was a general increase in PIP2 subfamily abundance compared to control, but only the Comb (-) treatment had the same levels of PIP2 as the control leaves. The levels of PIP2-7 decreased in the plasma membranes of the two combinations, Comb (-) and Comb (+), while they remained stable in other treatments compared to the control leaves (Figure 6D-F ).
Finally, during lipids analysis, no changes were observed in fatty acid composition (data not shown), but changes were found in sterol content (Table 2 ). For instance, lower levels of campesterol were observed in all treatments except B (+) when compared to the control in the microsomal fraction samples. Also, in the same fraction, lower levels of stigmasterol and sitosterol were found in salinity and B (-) treated samples. When analyzed the total sterol content of MF, only the combination treatment presented differences, with lower concentration in total sterol content. In plasma membrane sterols, changes were mainly seen in campesterol abundance, with lower concentrations in B (-), Comb (-), and Comb (+) when compared to control leaves. All treatments, except salinity, showed a reduction in sitosterol content, although no changes were observed in stigmasterol. Overall, the total presence of sterols was reduced in all treatment but salinity when compared to control. However, no changes were observed in the stigmasterol/sitosterol ratio in either the plasma membrane or microsomal fraction sample leaves among all treatments and control.
Figure 1 . (A ) Graphical representation of dry weight (g) root and upper part with the scale on the left and relative water content of the plants with the scale on the right represented with diamonds. Each measure is represented as the mean ± SE (n =6). (B ) Leaf per mass area (LMA) (g m-2) of treated plants after 15 days of treatment application. Data are represented as box plot. Different letters show statistical differences, the data corresponding both dry weight measures resulted in non-parametric data and Kluscal wallis post hoc was selected, on the case of relative water content and LMA statistical differences were evaluated with one-way ANOVA using Duncan test as post hoc . Both analyses were conducted with p < 0.05.
Figure 2. Stomatal conductance (mmol m-2s-1) of leaves from each treatment measured each 3-4 days starting the day of treatment application. Each measure is represented as means ± SE (n = 4). Different letters in lower case represent the statistical differences between treatments each separate day, statistical differences were calculated with one-way ANOVA using aspost hoc Duncan test when the data were parametric and non-parametric data was evaluated via Kluscal wallis test aspost hoc . Legend of the right represent statistical differences in capital letters of each treatment using repeated measures ANOVA. All analyses were conducted with p <0.05.
Figure 3. Analysis of mineral nutrients of broccoli leaves, (A ) PCA analysis of micro and macronutrients, (B ) concentration of B, (C ) Na, (D ) and K in in treated plants leaves. Data is represented as boxes (25-75%), error bars represent range within 1.5 quartile, and median line (n = 4). Different letters mean statistical differences of one-way ANOVA with Duncan test as post hoc , p < 0.05.
Figure 4. Gene expression of different aquaporins in broccoli leaves expressed as fold change (F. C) respect the control of isoforms of PIP1 group (A-D ) and PIP2 group (E-I ). Data is represented as means ± SE (n = 4). Different letters show statistical differences one-way ANOVA and Duncan test as post hoc . (E ) PIP2-1 and (F ) PIP2-2/3 statistical analyses for non-parametric data, Kruskal–Wallis. p < 0.05 in all cases.