Arabidopsis Restricts Sugar Loss to a Colonizing Trichoderma harzianum Strain by Downregulating SWEET11 and -12 and Upregulation of SUC1 and SWEET2 in the Roots

Phosphate (Pi) availability has a strong influence on the symbiotic interaction between Arabidopsis and a recently described root-colonizing beneficial Trichoderma harzianum strain. When transferred to media with insoluble Ca3(PO4)2 as a sole Pi source, Arabidopsis seedlings died after 10 days. Trichoderma grew on the medium containing Ca3(PO4)2 and the fungus did colonize in roots, stems, and shoots of the host. The efficiency of the photosynthetic electron transport of the colonized seedlings grown on Ca3(PO4)2 medium was reduced and the seedlings died earlier, indicating that the fungus exerts an additional stress to the plant. Interestingly, the fungus initially alleviated the Pi starvation response and did not activate defense responses against the hyphal propagation. However, in colonized roots, the sucrose transporter genes SWEET11 and -12 were strongly down-regulated, restricting the unloading of sucrose from the phloem parenchyma cells to the apoplast. Simultaneously, up-regulation of SUC1 promoted sucrose uptake from the apoplast into the parenchyma cells and of SWEET2 sequestration of sucrose in the vacuole of the root cells. We propose that the fungus tries to escape from the Ca3(PO4)2 medium and colonizes the entire host. To prevent excessive sugar consumption by the propagating hyphae, the host restricts sugar availability in its apoplastic root space by downregulating sugar transporter genes for phloem unloading, and by upregulating transporter genes which maintain the sugar in the root cells.


Introduction
Beneficial interactions between fungi and host plant roots rely on the exchange of nutrients between the symbiotic partners. Well investigated examples are mycorrhizal symbioses, where the plant delivers photo-assimilates to the fungi and fungi inorganic ions, in particular phosphate (Pi), and water to their hosts [1]. Quite similar exchanges might occur in beneficial symbiotic endophyte/plant interactions, although the molecular, cellular, and physiological bases are less understood [2,3] and ref. therein. All these beneficial symbioses are fragile and can shift to neutral, saprophytic, or even pathogenic interactions when stress conditions impair the exchange balance, and the survival of one of the symbiotic partners is compromised [4].
Upon Pi limitation, roots respond with the upregulation of Pi transporter genes and some Pi starvation regulators. The high affinity Pi transporters Pht1;1 and Pht1;4 play a major role in Pi acquisition under low Pi conditions [5]. Pht1;1 and -1;4 are induced by Pi starvation, and the transporter activities are controlled by additional mechanisms, including protein trafficking, localization, and degradation [6]. Furthermore, PHOSPHATE1 (PHO1) is the main Pi transporter into the xylem and PHO1 is preferentially expressed in the root vascular system under low Pi [7]. Chen et al. [8] demonstrated that WRKY6 regulates PHO1 expression, whereas low Pi reduced WRKY6-binding to the PHO1 promoter, thereby stimulating the expression of the gene. In addition, expression of many Pi transporter genes is controlled by Pi starvation response (PHR) transcription factors [9][10][11][12][13]. Castrillo et al. [14] showed that PHR1 is a central regulator in balancing the Pi starvation response and plant immune regulation, which occurs mainly post-transcriptionally and is only weakly regulated at the transcriptional level.
Restriction of fungal growth due to nutrient limitations in the environment promotes colonization of the host plants by suppressing the host's immune system. On the other hand, the increased demand of the fungus for photosynthates forces the plant to control its sucrose distribution to prevent loss to the microbes. Under Pi limitation, the plants transport more sugar from the leaves to the roots through the phloem to support root growth [15]. However, this also has tremendous effects on the growth of the root-associated fungi. They obtain the sugar either through direct transport from host cell to microbial cell [16], from the host apoplast, which contains the secreted sugar after release from the phloem cells, or from the rhizosphere, which contains secreted sugar from different root cells. Several transporter families have been proposed to transport sugars to the symbiotic partners [16]. Loading and unloading of the root phloem parenchyma cells with sugar involves the SWEET transporters [17,18] and ref. therein. The unloaded sucrose in the root apoplast is mainly transported into the mesophyll or epidermal root cells by SUC1 [2] and further transported into the vacuole of these cells by SWEET2, where it is stored [19]. Most of the hyphae of beneficial root-colonizing endophytic fungi grow in the root apoplastic space, which makes it likely that they also use the sugar from this plant compartment.
We studied the symbiotic interaction of Arabidopsis with a recently characterized Trichoderma harzianum strain [20]. The fungus strongly promotes plant growth on soil and full culture medium [20]. However, under Pi limitation, the fungus colonizes Arabidopsis aggressively, shifting the beneficial relationship towards an interaction with no or fewer benefits for the host. By analyzing genes for sugar and Pi transporters, as well as for defense, we were able to dissect the influence of Pi limitation on the symbiosis and propose a model for the shift in this interaction.

Growth Conditions of Plants and Fungus
Arabidopsis thaliana (ecotype Columbia-0) was used for this study. The seeds were surface-sterilized with a solution containing sterile distilled water (dH 2 O), sodium lauroyl sarcosinate, and DanKlorix (GP GABA GmBH Hamburg, Germany; 64%, 4%, 32%; v/v/v) for eight minutes under constant shaking, followed by six rinses with dH 2 O. Surface-sterile seeds were sown on MS media supplemented with 0.3% gelrite [21]. After cold treatment at 4 • C for 48 h, plates were incubated for 10 days at 22 • C under continuous illumination (100 µmol m −2 sec −1 ). The Trichoderma strain was propagated on KM medium (pH 6.5) for a week at 25 • C in the dark, as described previously, and pH changes were monitored by adding 0.004% (w/v) bromocresol purple pH indicator [20]. All experiments with insoluble Pi were performed on National Botanical Research Institute's Pi growth medium (NBRIP) (glucose, 10 g/L: Ca 3 (PO 4 ) 2 , 2.5 g/L; MgCl 2 × 6 H 2 O, 5 g/L; MgSO 4 × 7 H 2 O, 0.25 g/L; KCl, 0.2 g/L; and (NH 4 ) 2 SO 4 , 0.1 g/L), either on plates with 0.3% gelrite or in liquid medium for fungal growth. The medium with soluble Pi, which was used as a control, contained the same molar Pi concentration of K 3 PO 4 instead of Ca 3 (PO 4 ) 2 .

Trichoderma-Arabidopsis Co-Cultivation
Cocultivation of A. thaliana and Trichoderma strain was performed under in vitro culture conditions. Ten-day-old A. thaliana seedlings of equal sizes were directly transferred from MS medium to NBRIP medium. The fungus was grown on NBRIP plates. The cocultivation of Arabidopsis plants and Trichoderma strain were performed on NBRIP medium. A 5 mm plug of the Trichoderma strain was placed at the center of the cocultivation's plate. Cocultivation was monitored over a period of 10 days at 22 • C under continuous illumination (100 µmol m −2 s −1 ) [20,22].

RNA Extraction and cDNA Synthesis
RNA was isolated from the roots of colonized and uncolonized Arabidopsis seedlings with an RNA isolation kit (RNeasy, Qiagen, Hilden, Germany) and used for quantitative RT-PCR. Reverse transcription of 1 µg of total RNA was performed with an oligo(dT) primer. First strand synthesis was performed with a kit from Qiagen (Omniscript RT Kit, Qiagen, Hilden, Germany).

Real-Time PCR
Real-time quantitative RT-PCR was performed using the iCycler iQ real-time PCR detection system and iCycler software version 2.2 (Bio-Rad, Munich, Germany). For the amplification of the PCR products, iQ SYBR Supermix from Bio-Rad was used according to the manufacturer's instructions in a final volume of 23 µL. The iCycler was programmed to 95 • C 3 min, 40 × (94 • C 30 s, 57 • C 30 s, 72 • C 40 s), 72 • C 10 min, followed by a melting curve program (50-85 • C in increasing steps of 0.5 • C). All reactions were repeated at least 3 times. The mRNA levels for each cDNA probe were normalized with respect to the Arabidopsis glyceraldehyde-3-phosphate dehydrogenase 2 (GAPC2) and/or ACTIN2 mRNA level. Fold induction values for the tef1 mRNA levels (Genbank accession number: MT591352) from the Trichoderma cDNA were calculated with the ∆Ct equation. Root colonization was determined relative to the plant GAPC2 cDNA levels. The primer pairs for the Arabidopsis genes are given in the Supplementary Table S1.

Fluorescence Microscopy
The entire Arabidopsis roots, root tips, hypocotyls, and leaves were imaged using an AXIO Imager.M2 (Zeiss Microscopy GmbH, Jena, Germany) equipped with a 10× objective (N-Achroplan 10×/0.3). The bright field and fluorescence images (EX 545/25 and EM 605/70) were recorded with a color camera (AXIOCAM 503 color Zeiss, Jena, Germany) by use of an EGFP (EM 525/50 nm) and DsRED filter (EM 605/70 nm). Digital images were processed with the ZEN software (Zeiss, Jena, Germany), treated with Adobe R PhotoShop to optimize brightness, contrast, and coloring, and to overlay the photomicrographs to confirm fluorescence information.

Measurement of Photosynthesis Parameters
The Arabidopsis seedlings were dark-adapted for 15 min and the chlorophyll fluorescence was measured using a FluorCam 700F (Photon System Instruments, Czech Republic). Program parameters of FluorCam were set according to Wagner et al. [23]. The maximum quantum yield of the photosystem II (Fv/Fm) was calculated according to Maxwell and Johnson [24].

Acid Phosphatase (AcP) Assay
Extraction for the AcPase activity assays was carried out using 50 mL of sterilized NBRIP broth in a 100 mL conical flask. The flasks were inoculated with 1 mL of Trichoderma spore solution. The inoculated flasks were incubated at 25 • C for three days. The samples were centrifuged at 10,000 rpm for 10 min at 4 • C. The cell-free supernatant was assayed for crude AcP activity according to the method outlined by Tabatabai and Bremner [25]. One mL of the supernatant was mixed with 4 mL of modified universal buffer (pH 6.5). Further to this, 0.025 mM disodium p-nitrophenyl Pi (tetrahydrate) was mixed with the culture supernatant and incubated at 37 • C for 1 h. After 1 h of incubation, the reaction was stopped by adding 4 mL of 0.5 M NaOH and 1 mL of 0.5 M CaCl 2 . The concentration of p-nitrophenol was measured in triplicate by measuring the absorbance at 420 nm using a UV-Vis spectrophotometer, and the amount was quantified based on a standard curve with serially diluted solutions of p-nitrophenol. One unit (U) of phosphatase activity was defined as the amount of enzyme required to release 1 mol of p-nitrophenol/mL/min from di-Na p-nitrophenyl Pi (tetrahydrate) under the assay condition.

Quantitative Estimation of Soluble Pi
Erlenmeyer flasks containing 100 mL of NBRIP broth without bromophenol were inoculated with 1 mL spore solution of 5-day-old fungal cultures. Non-inoculated medium served as a control. The flasks were incubated in the dark at 25 • C for three days. The pH of the culture medium was measured 72 h after inoculation. A volume of 5 mL of the fungal culture was collected and centrifuged at 10,000 rpm for 10 min. The supernatant used to estimate the released Pi was determined spectrophotometrically (880 nm) according to Murphy and Riley [26].

Pi Limitation Promotes Plant Colonisation
We have previously shown that cocultivation of Arabidopsis seedlings with the Trichoderma strain resulted in the colonization of the roots and the establishment of a beneficial symbiotic interaction. The fungus promoted growth of the host and defended it against pathogen attack [20]. Without stress, the fungal colonization was restricted to the roots, and hyphae were never detected in the aerial parts of the Arabidopsis seedlings [20]. To test how nutrient limitation influences the symbiotic interaction, we co-cultivated Trichoderma and Arabidopsis seedlings on NBRIP medium with insoluble Ca 3 (PO 4 ) 2 as sole Pi source.
As expected, when 10-day-old Arabidopsis seedlings were transferred from MS medium to the NBRIP medium, they started to die after an additional 10 days ( Figure 1A). If these seedlings were transferred to soil, none of them recovered (n = 3 with 100 seedlings each; data not shown). Compared to control media with soluble Pi, the Trichoderma strain could only slowly grow in liquid NBRIP medium and on solid NBRIP agar plates ( Figure 1B), since the fungus had AcP activity ( Figure 1C) and can solubilize Ca 3 (PO 4 ) 2 . The AcP activity, as well as solubilized Pi in the medium, is shown in .
When Arabidopsis seedlings and the fungus were cocultured on NBRIY plates, we observed a fast root colonization ( Figure 1D). After 2 days, hyphae can also be detected on the stem and, ultimately, leaves ( Figure 2). Figure 2B shows the presence of hyphae at the stem/leaf junction 5 days after exposure to the fungus on insoluble Pi medium. At the same time, on soluble Pi medium, hyphae can only be detected in the roots ( Figure 2B). The seedlings did not benefit from the interaction ( Figure 1E). This can be shown by a recovery assay after transfer to soil. If the colonized seedlings were transferred to soil 8 days after colonization, 4.2 ± 0.6% (n = 3) survived, while 29.0 ± 3.7% (n = 3) survived when they were not exposed to the fungus.  The (Thtef1RNA)/(AtGAPC2 RNA) ratio can be utilized to quantify fungal colonization. A full 24 h after cocultivation, Thtef1 transcripts can be detected in the roots, but not in the shoots. Between 24 h and 72 h, we observed a greater than sixfold increase in root colonization ( Figure 3). Between 72 and 96 h, leaf colonization increased strongly. This indicates that 72 h after the cocultivation, the root tissue is fully colonized, and additional colonization of the host requires propagation of the hyphae towards the aerial parts of the seedlings. Propagation of the hyphae to the aerial parts of the seedlings does not occur when the cocultivation was performed on media with soluble Pi (Figure 2B, cf. also [20]). The effect of the fungus on plant performance under Pi stress was assessed by measurements of the efficiency of the photosynthetic electron transport (Fv/Fm) of photosystem II reaction center. A total of 48 h after transfer to Ca 3 (PO 4 ) 2 -medium, the Fv/Fm value of the uncolonized seedlings was lower than that of the colonized seedlings. Following 96 h after transfer to Ca 3 (PO 4 ) 2 -medium, this was reversed (Figure 4). This indicates that the plant initially benefits from the presence of the fungus, whereas, at the end of the experiments, this is no longer the case. The fungus tries to escape from the medium with insoluble Pi and prefers to grow on the host plant, despite the fact that it can solubilize Ca 3 (PO 4 ) 2 .

Trichoderma Effects on the Pi Starvation Response in Roots
PHOSPHATE1 (PHO1) transfers Pi from root to shoot via Pi export into root xylem vessels. As expected, expression of this Pi starvation marker gene was upregulated in the roots after transfer of the seedlings to Ca 3 (PO 4 ) 2 -containing medium ( Figure 5). Interestingly, 48 h after the transfer, the PHO1 mRNA level was higher in the roots of the colonized seedlings when compared to the uncolonized controls, whereas this was reversed 96 h after the transfer. A similar regulation was observed for the transcript levels of the high affinity Pi transporter Pht1;1 and Pht1;4 ( Figure 5). This suggests that the fungus reduces the Pi starvation response during the early colonization phase, whereas the beneficial effect of the fungus is lost after 96 h. A possible explanation could be that the fungus initially provides Pi to the roots, either by direct transport from the fungal to the plant cell, or from the residual external soluble Pi. The mRNA level for PHR1 increases only marginally under Pi stress and root colonization ( Figure 5), even though the transcription factor is considered as a central regulator controlling PHT1 expression and Pi homeostasis in Arabidopsis ( [27] and refs. therein). In addition, the WRKY6 mRNA level does not change significantly in response to the applied Pi stress or fungal root colonization ( Figure 5). In conclusion, and consistent with the photosynthetic parameters (Figure 4), it appears that, during the first period after transfer to the NBRIP medium, the colonized roots suffer less under Pi starvation than the uncolonized controls, as evident from the lower expression levels of PHO1, Pht1;1, and Pht1;4.

Fungal Colonization Does Not Activate Defense Gene Expression
The colonization of the roots, as well as aerial parts of the seedlings, by Trichoderma under Pi stress might be associated with defense gene activation. We checked classical marker genes for the jasmonate-dependent (PDF1.2) and salicylic-acid-dependent (PR1) defense pathways, but none of the investigated genes were upregulated more than twofold in the presence of the fungus ( Figure 6). This suggests that the roots did not respond to the colonization by the activation of its defense machinery. Furthermore, abiotic stress often induces abscisic acid accumulation. However, the transcript level for the zeaxanthin epoxidase (ZEP) that functions in the first step of the biosynthesis of ABA was also not significantly regulated in response to Pi stress or root colonization ( Figure 6). We conclude that the higher stress level of the Arabidopsis seedlings that were exposed to the fungus is not caused by more investment of the plant into defense against the aggressive fungal growth.

Root Colonization Alters Expression of SUC1, SWEET2, -11 and -12 in the Roots
Roots secrete a significant portion of sugars into the rhizosphere and deliver it to rootassociated microbes. The increase in root colonization may influence the sugar distribution within the plant body, subsequently affecting plant survival. The sucrose transporter genes SUC1 and SUC2 are expressed in roots and their expression does not respond to Pi starvation in our experiment (Figure 7, Supplementary Figure S2). However, at 48 and 96 h of Pi stress, we observed a greater than fivefold increase in the SUC1 transcript level in the colonized roots, while the SUC2 transcript level did not respond to the fungus (Figure 7,  Supplementary Figure S2). The monosaccharide transporter genes STP1, STP13, ERD6, and ERD6-like6 are also expressed in roots, but they did not respond to the Pi stress or root colonization (Supplementary Figure S2). Among the SWEET sugar transporters, the SWEET2 transcript level increased more than eight times in colonized roots (Figure 7), while there was no upregulation in the uncolonized controls. Furthermore, the SWEET11 transcript level decreased more than tenfold, and that of SWEET12 about sixfold 48 and 96 h after exposure to the fungus, but not in the uncolonized controls. SWEET3 can be used as a control, since it is expressed in roots but neither responds to the fungus, nor to the Pi stress (Figure 7). Interestingly, the investigated SUC, STP, ERD6, and ERD6-like6 transporter genes are also expressed in leaves, though to different extents than in roots, but we did not observe any significant response to the Pi stress or colonization (data not shown). Moreover, SWEET11 and -12 are expressed in leaves, although at much lower levels than in roots, and no regulation could be observed either (data not shown). The SWEET2 transcript levels in the leaves are too low for meaningful interpretations.
These results show that SUC1, SWEET2, -11, and -12 are specific targets of the fungus in roots under stress conditions.

Discussion
We established an artificial interaction system between Arabidopsis seedlings and a Trichoderma strain, which forces the fungus to escape from the growth medium and to colonize the roots, stems, and shoots of the seedlings. Without soluble Pi in the medium, the performance of the Arabidopsis seedlings is impaired and they ultimately die, which probably promotes the fast propagation of the fungus in the entire weakened host. The fungus can solubilize Ca 3 (PO 4 ) 2 and slowly grows on the medium, but prefers to escape to the plant.
One reason could be that it is easier for the fungus to withdraw Pi from the plant than solubilizing it from the medium. However, this appears to be unlikely, because we do not observe a stronger Pi starvation response of the plants when they are colonized by the fungus. In particular, during the first 2 days after cocultivation, the starvation response was even weaker, suggesting that the struggle for Pi is not the main reason for the colonization of the host.
Pi starvation triggers the colonization of the plant and propagation of the mycelium to the aerial parts, presumably because it is less energy-consuming for the fungus to grow on the plant than on the medium, and because the plant is weakened and does not activate processes to restrict fungal colonization. Without Pi stress, the colonization is restricted to the roots. It would be interesting to test whether other stresses will lead to similar strategies of the fungus.
However, the stronger colonization of the entire plant may force it to restrain the supply of reduced carbon to the fungus (cf. below). Therefore, our artificial system might mimic naturally occurring stress situations, in which beneficial plant/fungi symbiotic interactions in which only the roots are colonized by the microbes shift to more aggressive interactions which benefit the fungi. Since we did not observe a strong defense response of the weakened plant against fungal propagation, the interaction has more saprophytic than pathogenic features. Towards the end of the experiment, the plant dies and might provide material to the fungus which is used for their growth, similar or identical to saprophytic interactions. Several studies describe similar changes in fungal lifestyles. The establishment of a symbiotic interaction between Arabidopsis and Colletotrichum tofieldiae, which transfers Pi to the host and improves its growth under Pi limitation conditions, depends on the Pi availability, and Karandashov et al. [28] showed that the Pi stress causes a more saprophytic interaction. Grelet et al. [29] found that different Mycena species operate along a saprophytic-symbiotic continuum in association with Ericaceae. Intermediate interaction forms between endophytic and saprophytic lifestyles have also been proposed for wood-decaying fungi with different host species [30][31][32]; however, it is not clear how exogenous factors affect the interactions. An example for the adaptation to a symbiotic lifestyle provides Archaeorhizomycetes [33,34], which colonize the roots in summer and are mainly absent during the colder seasons. Schadt et al. [33] proposed that Archaeorhizomycetes might adjust their symbiotic lifestyle dependent on root-derived C compounds. Hacquard et al. [35] found genomic signatures in the transcriptomes of the beneficial root endophyte Colletotrichum tofieldiae and its pathogenic relative C. incanum, which are associated with the transition from pathogenic to beneficial lifestyles, and this included a narrowed repertoire of secreted effector proteins. Since we did not observe a strong defense response against the propagating fungal hyphae, nor a struggle for the limiting Pi source under our experimental conditions, access to sugar might be the main reason for the intense colonization. Like ectomycorrhia fungi, endophytes mainly colonize the apoplastic space within the root and live on the root surface. Therefore, most of the sugar required for the hyphal growth should be taken up from the plant apoplast.

Trichoderma Manipulates Sugar Transport in Roots
The unloading of sucrose from the phloem in the roots occurs through a combination of symplastic and apoplastic pathways [36]. Cells in the root tip are connected to the phloem by plasmodesmata and form a symplastic domain [37,38]. In more mature areas of the root, cellular connections to the phloem are more limited, and transmembrane transport may be more important [39].
Apoplastic sugar unloading of the phloem in the roots is mainly mediated by SWEET11 and -12 in Arabidopsis roots ( [18,40] and refs. therein). The higher demand for sugar due to the propagation of the hyphae results in the stimulation of photo-assimilate translocation from source to sink. In order to restrict sugar loss from the apoplast after unloading of the phloem, the plant downregulates SWEET11 and -12 ( Figure 8). The involvement of SWEET transporters in beneficial and pathogenic plant/microbe interactions in roots is well-documented. Several SWEET genes in Brasica crops were significantly upregulated upon P. brassicae infection, the causal agent of clubroot [41]. Clubroot disease stimulates early steps of phloem differentiation and recruits SWEETs within developing galls [42]. Furthermore, MtSWEET11 is a nodule-specific sucrose transporter in Medicago truncatula [17,43,44]. Figure 8. A model describing the regulation of sugar transporter genes in stress-exposed Arabidopsis roots in response to Trichoderma colonization. Upon increasing stress, the initial beneficial symbiotic interaction of Trichoderma with Arabidopsis roots shifts to an interaction with a saprophytic lifestyle of the fungus and a higher demand for sugar from the host. The endophytic fungus takes the sucrose mostly from the root apoplast, or from the rhizophere (not shown) after secretion of the roots. The host responds to it by downregulating SWEET11 and -12, which results in less unloading of sucrose from the phloem into the apoplastic space, and thus restriction of hyphal growth due to sugar shortage. Restricted sucrose availability in the apoplast stimulates uptake into the mesophyll root cells and upregulation of SUC1 expression to promote sucrose uptake from the apoplast into the mesophyll root cells. Furthermore, the shortage of sucrose in the root cells stimulates SWEET2 expression. SWEET2 sequesters sucrose in the vacuole of the root cells, to further reduce sugar loss to the microbe. SUC1 from Arabidopsis was one of the first H + -coupled sucrose transporters identified in plants [45,46], and, together with SUC2, is the most important sugar uptake carrier from the apoplast into the root cells [47,48]. Consistent with this, in Arabidopsis seedlings, SUC1 expression is found in the elongation zone of the root and less in the root cap or zone of cell division, and SUC1 expression in roots is induced by exogenous sucrose [48]. The most likely function for plasma-membrane-localized SUC1 [48] is in the uptake of sucrose released into the apoplast by SWEET sucrose efflux transporters as part of the phloem unloading mechanism [2]. SUC1 expression is upregulated by Trichoderma, but not by Pi stress, suggesting that the roots compete with the fungus for the apoplastic sucrose.
Removal of the sugar from the apoplast by stimulating SUC1-mediated uptake into the root cells restricts fungal growth ( Figure 8). Interestingly, the response to the fungus is mainly observed for SUC1, but not for SUC2 (Figure 7 and Supplementary Figure S2). Recently, Lasin et al. [49] demonstrated that SUC1 introns act as strong enhancers of expression. The authors showed that a SUC1 whole-gene GUS construct expressing a nonfunctional SUC1 mutant, that is transport inactive, is defective in sucrose-induced SUC1 expression when expressed in an suc1 knock-out mutant. The results indicate that sucrose uptake via SUC1 is required for sucrose-induced SUC1 expression, and that the site for sucrose detection is intracellular [49]. Therefore, it is likely that the root cell recognizes sucrose shortage due to the propagating hyphae in its apoplast, and responds to it by stimulating the uptake and upregulation of SUC1 (Figure 8).
Besides SUC1, SWEET2 is strongly upregulated in the stress-exposed colonized roots. SWEET2 is located in the tonoplast of the mesophyll and epidermal root cells, and Chen et al. [19] proposed that the transporter modulates sugar secretion, possibly by reducing the availability of glucose sequestered in the vacuole, thereby limiting carbon loss to the rhizosphere. The reduced availability of sugar in the rhizosphere contributes to resistance to Pythium. Consistent with their interpretation, SWEET2 might be involved in reducing the available sucrose by transporting it into the vacuole of the mesophyll cells (Figure 8). Several studies show that root infections by pathogenic microbes alter the expression of plant sugar transporters to restrict sugar loss to the microbes; for instance, SWEET genes are upregulated in pathogenic interactions with bacteria and fungi [50][51][52][53]. SWEET11 is involved in disease resistance against the obligate biotrophic protist P. brassicae, the causal agent of clubroot [41]. Chen et al. [19] showed that the root-expressed vacuolar SWEET2 regulates sugar secretion, specifically from epidermal cells of the root apex, which influences growth of pathogenic microbes.
Vargas et al. [53] investigated T. virens-colonized roots and showed that sucrose exuded by plants also has a tremendous influence on the sucrose-dependent network in the fungal cells. The authors investigated fungal sucrose transporters and showed that they are also important for the symbiotic association. This study clearly highlights the necessity for further investigations on the role of sucrose in establishing and maintaining symbiotic plant/fungus interactions.
In summary, it appears 4 out of the 56 sugar transporters in Arabidopsis play an important role in the regulatory circuit. The proposed model (Figure 8) can now be tested with other endophytes and host/plant combinations, different stress conditions, and genetic tools manipulating the sugar transporter genes discussed in our model.