3.1 Plant biomass and macro and micronutrients contents
The low availability of nutrients in the soil has been considered a major limitation for plant growth and productivity. The lack of one or more nutrients causes physiological, morphological, and biochemical disturbs, reducing plant growth and inducing biomass reductions. To avoid the low availability of P, plants were provided with several strategies to maximize uptake by the roots through long-distance and local signaling, and improve P utilization in the several physiological processes (Shabala et al. 2015; Bulgarelli et al. 2019). The integration between endogenous signals and environmental factors is responsible for the regulation of plant growth and development. So plants are able to sense the internal and external P-status, triggering appropriate responses. External Pi concentration sensing involves local signaling and plant tissue Pi status involves systemic or long-distance signaling (Chiou and Lin 2011).
Here we used a split root system to study P uptake and allocation and the systemic response of the genes involved in P-use efficiency to P heterogeneity in the root environment. The split root system set-up is shown in Figure 1.
Fig.1 Eucalypt plants after the split-root experiments. Treatments (+P/+P, +P/-P, -P/-P) and the three repetitions of each of the treatments illustrate the biomass production of the studied species: (a) E. tereticornis, (b) E. globulus and (c) E. grandis
In our study, the presence or privation of P did not significantly influence foliar biomass production in E. globulus and E. tereticornis plants (Figure 2a). However, the lack of P on one or both sides of the split root system affected the leaf biomass of E. grandis. Considering the biomass of the whole root system, the lack of P decreased root biomass of -P/-P and +P/-P in E. globulus plants and -P/-P of E. grandis plants (Figure 2b). In E. globulus the roots of the -P side of the +P/-P treatment grew more than roots in the +P side. E. tereticornis plants showed a large variation of the root biomass production between the two sides of the +P/+P treatment. The +P side of the +P/-P treatments showed the highest root biomass accumulation (Figure 1b). In our experiment plants were maintained for six weeks at the split root system and irrigated with the -P and +P solutions and because P is mobile within the plant (Hinsinger et al. 2011) we were not expecting too many differences regarding biomass accumulation in +P/-P plants. In these plants, P might have been redistributed from the +P side to the -P side and supported growth. Also, the +P side might transfer P to the shoots, supporting growth. This is corroborated by the lower average root biomass in -P/-P plants of E. globulus and the clear trend of a lower leaf mass accumulation. The same may be applied to E. grandis, but in this species, leaf and root growth was significantly affected even when just one side was under -P conditions, suggesting that is more sensitive to low P availability than the other two species. A recent work reported that root biomass of E. grandis was higher in plants grown with sufficient P than with low P (Silva et al. 2022). It was observed by (Bulgarelli et al. 2019) that plants of E. tereticornis and E. globulus growing in soils with low and sufficient P did not differ regarding shoot biomass accumulation.
Fig.2 Leaf (a) and root (b) biomass of eucalypt species cultivated in a split-root system, with P-repleted (+P) and P-deprived (-P) conditions. Bars represent the means of five replicates and the lines are the standard error. The average value of the two root sides was calculated to compare the treatment in the same species and statistical differences are indicated by different small letters. Asterisk indicates a significant difference between the sides of each treatment in each species (Tukey's test, p < 0.05)
In plants of E. grandis and E. tereticornis, the P concentration in the leaves did not show significant differences among plants receiving P-repleted or P-depleted solutions, although a trend of a higher P concentration was observed in +P/+P and +P/-P plants if compared with -P/-P plants (Figure 3a). The average root P concentrations at both root sides were lower in E. globulus and E. grandis at the -P/-P treatment, but no difference was observed between the sides of +P/-P treatment. E. tereticornis showed higher P concentration in the leaves and roots than the other species. E. tereticornis had a large P concentration variation in roots but did not show a significant difference between -P/-P treatments and the other two treatments (Figure 3b). The total P content per plant was calculated (P concentration multiplied by biomass) to verify if the -P/-P treatment were effective in decreasing P accumulation. Because of the large variation of P content among plants, we did not observe significant differences for E. globulus and E. tereticornis, as observed in E. grandis, but the values found were lower than the other treatments (Figure 3c). This shows that P was supplied to the plants from the +P side of the +P/-P treatment.
Fig.3 Phosphorus concentration (g kg-1) in leaves (a) and roots (b) and total P content (mg plant-1) in plants (c) of eucalypt species cultivated under split root condition, with P-repleted (+P) and P-deprived (-P) conditions. Bars represent the means of five replicates, and the lines are the standard error. The average value of the two root sides was calculated to compare the treatment in the same species, and different small letters indicate statistical differences (Tukey's test, p < 0.05)
Supplementary Tables 2 and 3 show the concentrations of macro- and micronutrients in the leaves and roots, respectively. In general, plants grown under split root conditions with -P/-P and +P/+P showed varying concentrations of the nutrients, and the foliar concentration of most of the nutrients varied depending on the species. In leaves, P availability significantly affected foliar B and Cu concentrations. B was highest in leaves of +P/+P plants of E. tereticornis. In E. grandis, +P/-P plants had higher foliar B and Cu concentrations. No significant effect was observed for the concentration of other nutrients. In the roots, the concentration of nutrients had a significant effect that varied among species and treatments. Interestingly, although only P was varied in the nutrient solutions, other nutrients, like K, Mg, Cu, Fe, and B, showed significantly different concentrations, influenced by P presence or absence. Among the studied species, higher K concentrations were found in roots of E. tereticornis in +P/-P and -P/-P treatments and was highest in the -P side of +P/-P split root. For E. grandis, comparing the three treatments, K concentration was highest when one side of the split system was +P, either the +P/+P or +P/-P.
Similar to K, in E. grandis, the highest concentrations of Mg occurred when one side of the split root system was +P, either +P/+P or +P/-P. In E. tereticornis, the lowest concentrations of Mg were observed in the +P/+P treatment, either among or within the species. Regarding the Cu concentration, significant differences occurred only between species. Cu had the highest concentration when one side of the roots was -P, either in the -P/-P or the +P/-P treatments. For Fe, the highest concentration was observed in E. grandis in the +P/+P treatment. For B, the differences were only significant when comparing the sides of each split root, with the highest concentration when one side of the split system was -P, either the -P/-P or +P/-P.
Copper toxicity due to the deliberate use of fungicides induces nutritional imbalances and restrictions on plant growth, which is related to P (Feil et al. 2020). In eucalyptus plants, higher concentrations of Cu occurred in treatments that had P deficiency on at least one side of the root, either in leaf (E. grandis) or root (E. tereticornis). Interestingly, it was the root sides, or foliar treatment, with the lowest biomass (Figure 2). Hydroponically grown cucumber plants were used to investigate the influence of exposure to different Cu concentrations. High concentrations of Cu (above 25 μM) resulted in lower shoot and root growth, and the rate of P influx decreased. The data shown by the authors clearly show that high concentrations of Cu can disrupt the uptake of P by the root, probably through a direct action on gene transporters (Feil et al. 2020).
Responses related to Fe over accumulation and toxicity has been observed in plants under Pi starvation, as P may negatively and positively regulate genes associated with Fe metabolism (Gonçalves et al. 2020). Using available transcriptomic data, (Gonçalves et al. 2020) devised new functions for already known genes in Pi metabolism and identified responsive genes with unknown functions, although upregulated under Pi starvation. This includes the downregulation of genes involved in activating Fe starvation responses, such as Fe transporters, and genes involved in cellular Fe homeostasis. Although the regulation of these genes under Pi deprivation is likely a consequence of Fe over accumulation and the following responses to avoid Fe toxicity, some of these genes are likely also mediators of P-Fe interaction (Gonçalves et al. 2020), which possibly explains the fact that the Fe concentration was not higher in the roots under P-starvation conditions (-P/-P), or perhaps also because the +P/-P treatment allowed signaling between the sides of the roots, and the +P side has maintained a suitable P supply for the plant. Another hypothesis is that genes related to Fe transport and uptake were downregulated. In a split-root system with tomato plants, it was analyzed whether the sulfur (S) concentration could modify the plant's ability to absorb and accumulate Fe (Coppa et al. 2018). The authors concluded that the S supply, even if located only on one side of the root system, contributed to the Fe status in the shoot. In rice, expression levels of Fe-regulated marker genes on the non-Fe side of the root were increased, but not as much as when the two sides of the root did not contain Fe, indicating signaling processes between the two sides of the split roots (Chen et al. 2018).
P deficiency has been reported to influence K uptake in rice (Li and Rengel 2012). The interaction between these two nutrients can influence their absorption and utilization in tomatoes, and P and K deficiency (either individually or both) significantly reduced root biomass density, and the ratios of P, and K contents in shoots and roots were positively correlated with root volume (Naciri et al. 2022).
3.2 Principal component analysis
PCA was performed separately for variables related to the nutrient concentrations of leaves (Figure 4a) and roots (Figure 4b). For leaves, the first (PC1) and the second (PC2) components explained 48.7% and 28.8% of the total variation, respectively. For roots, PC1 and PC2 explained 58.1% and 18.3% of the total variation, respectively. In leaves, the treatments -P/-P and +P/+P of E. grandis and +P/+P of E. globulus were separated from all other variables. E. globulus (+P/-P and -P/-P) showed the closest correlation with Fe and K, Ca, and Mg. E. tereticornis (+P/+P and -P/-P) did not show a close correlation with any nutrient, and it was slightly correlated with B, while E. tereticornis and E. grandis (+P/-P) were closely correlated with Mn. The most correlated nutrients were Cu, P, and B. Fe, K, Ca, and Mg were positioned in another quadrant.
P separated from all other variables in the roots but close to E. tereticornis in the treatment +P/+P. The closest nutrients were Fe, Cu, and Mg, which correlated with E. globulus (+P in the +P/+P and +P/-P), E. grandis (+P in the +P/-P treatments), and E. tereticornis (-P in the treatments +P/-P and -P/-P treatments). B did not have a positive correlation with the other nutrients and was closest to E. globulus in the -P/-P treatment. The same occurred with K, which was more correlated with E. tereticornis (-P/-P treatment), and Ca, which was more correlated with E. tereticornis in the -P/-P treatment.
A negative correlation between Fe and P in the leaves was reported in Citrus grandis (Meng et al. 2021) and E. grandis (De Oliveira et al. 2022), species that accumulated more Fe in stems and roots under P deficiency. Opposite observations were obtained in the present study for E. grandis, which showed higher Fe levels in the roots of the +P/+P treatment compared to the low P treatment. Low Fe availability promotes P uptake, while P-starvation promotes Fe accumulation in tissues because of ROS activity, causing callose deposition and preventing plasmodesmal communication (Müller et al. 2015) by regulating genes involved in Fe homeostasis. In our study, interestingly, the lack of P did not promote the accumulation of Fe, with the highest concentration in the +P/+P treatment, which is possibly explained by the negative and positive regulation of genes involved in cellular Fe homeostasis and those related to the synthesis of ferritin, a protein that can bind over one thousands of Fe atoms, regulating the concentration of free Fe inside the cell (Stein et al. 2009).
Fig.4 Principal component analysis (PCA) of nutrient concentrations of leaves (a) and the split roots (b) of the eucalypt species. Different colors were used to indicate species and treatments in the leaves and on both sides of the split roots. The side of the split root is indicated using bold letters. The analysis was performed on the correlation matrix of least square means of averaged accessions. Numbers in parentheses indicate the variation percentage explained by PC1 (F1) and PC2 (F2)
3.3 PSR gene expression in leaves and roots
The split-root system is a relevant tool to address the long-distance signaling and remobilization effects of nutrient heterogeneity at the root environment, as it allows separating local and systemic responses to understand better their respective molecular mechanisms (Ferreira Torres et al. 2021). We observed a large variation in gene expression levels, but in general, the expression of the analyzed genes was highest in the leaves and roots of plants grown under -P. Genes from three functional groups were analyzed: P transport (PHT1), P uptake (PHO1), and sulfolipid synthesis (SQD).
3.3.1 Genes related to transport and uptake P (PHT1 and PHO1)
Most members of the PHT1 membrane protein family are involved in the direct uptake of Pi from the soil, its transport through root cortical cells, in arbuscular mycorrhizal symbiosis and in the transport and remobilization of this nutrient within the plant (Gu et al. 2016; Wang et al. 2017). The analyzed PHT1 genes responded differently among species and treatments. In the leaves of E. globulus, the expression level of the genes of the PHT1.1 and PHT1.6 transporters was higher in the -P/-P treated plants than in the +P/+P (Figures 5a, 5b). PHT1.6 expression was also higher in in leaves of E. tereticornis under -P/-P conditions. On the contrary, the expression of PHT1.5 and PHT1.6 transporters were higher in the leaves of E. grandis plants under +P/+P conditions (Figures 5b, 5c), which may be related to P transport and remobilization (Wang et al. 2022). In Medicago truncatula the PHT1 transporter MtPT5 was proved to play an important role in leaf growth and Pi accumulation (Wang et al. 2022).
Fig.5 Expression analysis of genes included in the "P transport (PHT1)" group in leaves of eucalypts species cultivated under split root condition, with P-repleted (+P) and P-deprived (-P) conditions. A= PHT1.1, B= PHT1.5, C= PHT1.6, D= PHT1.10, E= PHT1.11. The relative quantification of each transcript in each treatment was normalized against the constitutive Ef1 gene. Bars represent the means of three replicates, and the lines represent the standard error. Different letters indicate statistical significance among treatments in the same species (Tukey's test, p < 0.05)
In the roots PHT1.3 expression in the three species showed the highest levels in plants from the treatments +P/+P in E. globulus and -P/-P in E. tereticornis. The comparison of the root sides in each species showed statistical significance only in the -P side of the -P/-P treatment in E. grandis and E. tereticornis (-P/-P). The expression of the PHT1.3 in E. globulus roots clearly showed what might be expected regarding signaling between the two roots in the +P/-P treatment. On the -P side of the +P/-P treatment, the expression was slightly lower than on the -P/-P treatment, reflecting the better P status of the plant which probably triggered a signal to the -P side, indicating to the plant the non-total absence of P (Figure 6a).
The induction of the PHT1 transporters group may be related to the increase in P uptake. Studies with plants deficient in PHT1 expression showed that these multigene families are composed of transporters that may have high or low affinity for Pi, corroborating with kinetics studies of absorption performed with plants, which showed that the high-affinity system works in the range of 2,5 to 12 μM and the low-affinity system between 50 and 100 μM (Poirier and Jung 2015). The passage of Pi through the membrane through these transporters takes place via proton-Pi co-transport (Baker et al. 2015). It is not known what defines the greater or lesser affinity for Pi transport, but it was recently discovered that it is not related to the Pi binding site of the protein, suggesting that these proteins could have dual affinity and that post-translational changes would be involved in defining the high or low-affinity transporter function (Ceasar et al. 2016).
The PHO1 family genes mainly express in stellar cells of roots and hypocotyls (Zhang et al. 2023). In Arabidopsis, PHO1 is responsible for the efflux of Pi out of the cells towards the xylem vessel, that is, with an essential role in the transport of Pi from the roots to the leaves (Arpat et al. 2012). The PHO1 gene may be weakly upregulated under P-deprivation (Hamburger et al. 2002; Zhang et al. 2023). Here, PHO1.8 expression was species-dependent, it was induced only in E. grandis in -P (average root values in the +P/-P and -P/-P treatments) and in the -P side of the +P/-P treatment. This same side had higher expression in E. tereticornis (Figure 6b). Under low P, SPX and PHO1 were induced in several crop plants, including rice, chickpea, switchgrass, and alfalfa (Ding et al. 2021; Li et al. 2022).
Fig.6 Expression analysis of genes included in the “P transport”: PHT1.3 (a) and “P uptake”: PHO1.8 (b) group in roots of eucalypts species cultivated under split root condition, with P-repleted (+P) and P-deprived (-P) conditions. The relative quantification of each transcript in each treatment was normalized against the constitutive Ef1 gene. Bars represent the means of three replicates, and the lines represent the standard error. The average value of the two root sides was calculated to compare the treatment in the same species, and statistical differences are indicated by different small letters. Asterisk indicates a significant difference between the sides of each treatment in each species (Tukey's test, p < 0.05)
Plant cells generally have a higher P concentration, between 5 and 20 mM, than the soil solution, between1 and 10 μM (Hinsinger 2001). Thus, the uptake of P from the soil to the root is an energy-dependent process (Prathap et al. 2022) and that needs membrane transporters, including those codified by PHT1 and PHO1 genes. So far, 36 members of the PHT1 transporter family have been identified in wheat (TaPHT1;1-TaPHT1;36), 14 in soybean (GmPHT1;1-GmPHT1;14), 13 in corn (ZmPHT1;1-ZmPHT1;13), 13 the genome of rice (OsPT1;1-OsPT1;13), 11 in barley (HvPHT1;1-HvPHT1;11) and nine members in Arabidopsis (AtPHT1;1-AtPHT1;9) (Prathap et al. 2022). In E. grandis 22 transporters belonging to the PHT1 family have been identified (Che et al. 2022).
As a large family of transporters, PHT1 transporters show differences in expression sites and affinities for P (Nussaume et al. 2011; Teng et al. 2013). The function and location of these transporters are very broad, whether in Arabidopsis or rice, the two well-characterized species for this family of transporters. Studies carried out with PHT1 transporters in Arabidopsis showed variation in the expression of transporters in different plant organs. Mobilization of P from the root to the leaves was observed for PHT1;8 and PHT1;9, while mobilization from the leaves to the root by PHT1;5 (Liu et al. 2012). In Arabidopsis and rice, expression of PHT1;7, and PHT1;12 transporters are confined to the anthers (Mudge et al. 2002). AtPHT1.1 and AtPHT1.4 are responsible for P acquisition by roots in both high and low P sources (Catarecha et al. 2007), and have high activity at the root-soil interface, including the epidermis, root hair cells, and the root cap under low-P conditions (Mudge et al. 2002). AtPHT1.8 and AtPHT1.9 act during the translocation of Pi from root to shoot (Lapis-Gaza et al. 2014). In rice, transporters also have distinct roles; OsPHT1.6 is classified as a high-affinity transporter responsible for Pi uptake and translocation throughout the plant, with an active role in the epidermal and cortical cells of primary and young lateral roots. OsPHT1.2 is a low affinity transporter with a role in Pi translocation. It has a specific location in the stele of primary and lateral roots (Ai et al. 2009).
Some studies have been carried out with eucalypts and putative P transporter genes to assess expression in response to P deprivation. PHT1 and PHT2 were evaluated in E. grandis, and the expression of EgPHT1;4 and EgPHT2;1 was induced in P deficiency, mainly in roots (Pereira et al. 2019). Also under P deprivation, PHT1 was induced in C. citriodora and E. dunni, and in the latter, the high expression level was possibly related to the greater demand for P because of the high growth rate of this species (Niu et al. 2015). Most PHT1 transporters are found in various tissues and organs, often overlapping with other PHT1 members, evidencing their functional complexity. It is important to mention that this location is often influenced by environmental or development factors (Mudge et al. 2002). This would explain the different expression patterns obtained between the analyzed PHT1 transporters between roots and leaves. Most PHT1 transporters are preferentially expressed in roots, where their expression is activated by P-starvation-induced PHR1 transcription factor and directly mediates P acquisition from the soil (Bayle et al. 2011).
In the present study, the PHT1 transporters that showed the highest level of expression under +P condition may be involved in P uptake under conditions of greater P availability. Although the PHT1 genes expression is usually involved with P scarcity and partially repressed under high P concentrations to prevent its excessive accumulation, some members of the PHT1 family have basal expression levels that are important for maintaining P absorption even under adequate availability P (Zhang et al. 2021). Additionally, the repression of four PHT1 genes by P deficiency in wheat has been reported for transporters. Under low-P conditions, expression of TaPHT1.1, 1.2, 1.9 and 1.10 in roots was lower than under high-P conditions (Teng et al. 2013). This down regulation by P deficiency could be partially explained by the colonization of arbuscular mycorrhizal fungi, which are upregulated under P deficiency (not analyzed in this experiment). High concentrations of Cu may also disrupt P uptake by the root, probably through a direct action on gene transporters (Feil et al. 2020).
3.3.2 Sulfolipid synthesis (SQD)
The lipid composition of plants is responsive to environmental conditions, including changes in nutrient availability (Bolik et al. 2022). P-deficient plants synthesize the non-P glycolipid sulfoquinovosyldiacylglycerol (SQD). This sulfolipid may replace membrane phospholipids, releasing P for remobilization (Dissanayaka et al. 2018) and improving the efficiency of P utilization (Watanabe et al. 2020; Luo et al. 2022). Studies developed mainly on rice (Jeong et al. 2017; Gho et al. 2018; Gonçalves et al. 2020) show that OsSQD2 is highly induced under P starvation, as also shown in Arabidopsis (Yu et al. 2002). This mechanism seems to be widely exploited in leaves but much less in roots (Murakawa et al. 2019). The role of SQD in rice is poorly understood, which makes it interesting for further investigation, especially in tree species.
In our study, SQD was induced in the leaves of plants of E. globulus and E. tereticornis grown on -P/-P, with a similar trend in E. grandis, despite not statically significant (Figure 7a). In roots SQD expression was only higher for - E. globulus under the P/-P conditions (Figure 7c).
Fig.7 Expression analysis of SQDs genes in the leaves (a) and (b) roots of eucalypt species cultivated under split root condition, with P-repleted (+P) and P-deprived (-P) conditions. The relative quantification of each transcript in each treatment was normalized against the constitutive Ef1 gene. Bars represent the means of three replicates, and the lines represent the standard error. The average value of the two root sides was calculated to compare the treatment in the same species, and statistical differences are indicated by different small letters (Tukey's test, p < 0.05)
Lipid remodeling occurs more frequently in young photosynthetic tissues, as photosynthesis demands P (Pfaff et al. 2020). In leaves, E. tereticornis that showed an increase in biomass under P deprivation, had a greater expression of SQD under -P/-P conditions suggesting the presence of the strategy to replace P-lipids as an attempt to optimize P utilization. In addition, the low concentration of P in the leaves of -P/-P E. tereticornis may have contributed to membrane remodeling to overcome P-deficiency.
In a split root system experiment, transcriptomic data revealed that genes involved in Pi transport as well as Pi uptake and recycling are generally systemically regulated, and genes linked to stress- or hormone-related responses are locally regulated (Thibaud et al. 2010). The growth of shoots and roots is reciprocally dependent, reinforcing the importance of signaling pathway networks (Chiou and Lin 2011). A recent study showed sulfolipid and phospholipid remodeling in eucalyptus plants at low P (Silva et al. 2022).