Ectomycorrhizal Fungal Strains Facilitate Cd2+ Enrichment in a Woody Hyperaccumulator under Co-Existing Stress of Cadmium and Salt

Cadmium (Cd2+) pollution occurring in salt-affected soils has become an increasing environmental concern in the world. Fast-growing poplars have been widely utilized for phytoremediation of soil contaminating heavy metals (HMs). However, the woody Cd2+-hyperaccumulator, Populus × canescens, is relatively salt-sensitive and therefore cannot be directly used to remediate HMs from salt-affected soils. The aim of the present study was to testify whether colonization of P. × canescens with ectomycorrhizal (EM) fungi, a strategy known to enhance salt tolerance, provides an opportunity for affordable remediation of Cd2+-polluted saline soils. Ectomycorrhization with Paxillus involutus strains facilitated Cd2+ enrichment in P. × canescens upon CdCl2 exposures (50 μM, 30 min to 24 h). The fungus-stimulated Cd2+ in roots was significantly restricted by inhibitors of plasmalemma H+-ATPases and Ca2+-permeable channels (CaPCs), but stimulated by an activator of plasmalemma H+-ATPases. NaCl (100 mM) lowered the transient and steady-state Cd2+ influx in roots and fungal mycelia. Noteworthy, P. involutus colonization partly reverted the salt suppression of Cd2+ uptake in poplar roots. EM fungus colonization upregulated transcription of plasmalemma H+-ATPases (PcHA4, 8, 11) and annexins (PcANN1, 2, 4), which might mediate Cd2+ conductance through CaPCs. EM roots retained relatively highly expressed PcHAs and PcANNs, thus facilitating Cd2+ enrichment under co-occurring stress of cadmium and salinity. We conclude that ectomycorrhization of woody hyperaccumulator species such as poplar could improve phytoremediation of Cd2+ in salt-affected areas.

In this study, we examined the impact of ectomycorrhizal fungi on root Cd 2+ uptake under combined stress of salt and cadmium, aiming to elucidate the underlying mechanisms. We used two different P. involutus isolates, MAJ and NAU, for this study. Strain MAJ forms a complete ectomycorrhiza composed of a thick hyphal mantle ensheathing root tip and a typical Hartig net structure inside the roots for nutrient exchange, while strain NAU forms only the outer mantle [83]. We studied Cd 2+ uptake in the presence and absence of NaCl and analyzed gene expression of annexins because previous studies show that PeANN1 facilitates Cd 2+ enrichment through CaPCs [79]. P. involutus activates H + -pumps and hyperpolarizes membrane potential in EM roots [48,64,65]. Therefore, the PM H + -ATPases-promoted Cd 2+ flux was also verified in EM roots under salt stress. Our data reveal that P. involutus inoculation stimulates Cd 2+ influx under salt stress, resulting from the upregulated H + -ATPases and annexins in the ectomycorrhizal roots. Both MAJ and NAU conserved the Cd 2+ uptake capacities under co-occurring stresses of cadmium and salinity, regardless of the formation of Hartig net in the ectomycorrhizal symbioses.

Cd 2+ Concentrations in Roots and Shoots of Ectomycorrhizal Poplars under NaCl Stress
Cd 2+ concentrations were analyzed in roots, stems and leaves of NM and EM P. × canescens after 24 h exposure to CdCl 2 (50 µM) or combined stress of CdCl 2 and NaCl (100 mM). Under CdCl 2 stress, non-mycorrhizal (NM) roots displayed remarkably higher Cd 2+ concentrations than stem and leaves ( Figure 1). Compared to NM plants, Cd 2+ concentrations were 0.8-to 1.4-fold higher in roots and stems of poplars colonized with P. involutus isolates, MAJ and NAU ( Figure 1). However, the addition of NaCl (100 mM) significantly decreased Cd 2+ accumulation in roots and shoots of both NM-and EM-plants ( Figure 1). Of note, EM-plants retained significantly higher Cd 2+ concentrations in roots and stems than NM poplars under salt stress ( Figure 1). Therefore, EM fungi enhanced Cd 2+ enrichment in both root and aerial parts of P. × canescens under co-occurring stresses of cadmium and salinity. Figure 1. Cd 2+ concentrations in roots, stems and leaves of non-mycorrhizal (NM) and ectomycorrhizal (EM) Populus × canescens under cadmium and salt stress. Poplar plantlets inoculated with or without Paxillus involutus isolates (MAJ or NAU, 30 d), were hydroponically acclimated and subjected to 24 h of CdCl2 (0 or 50 μM) in combination with NaCl (0 or 100 mM). Mean values of Cd 2+ concentrations in control (−Cd), CdCl2 stress (+Cd), and combined stress of CdCl2 and NaCl (+Cd + NaCl) are shown. Each column is mean ± SD obtained from 3 individual plants. Statistically significant differences (p < 0.05) among treatments are indicated with different letters (a-d).

Steady-State Cd 2+ Influx in Ectomycorrhizal Poplar Roots and Fungal Mycelia under NaCl Stress
To determine whether the Cd 2+ enrichment in EM P. × canescens resulted from the P. involutus-stimulated uptake, Cd 2+ fluxes were examined in NM-, EM-roots and fungal mycelia under CdCl2 and NaCl stress. CdCl2 exposure (50 μM, 24 h) resulted in an apparent Cd 2+ uptake, 34.9 pmol cm −2 s −1 , along NM-roots of the hyperaccumulator, Populus × canescens (Figures 2 and S1). EM-roots exhibited 36% to 39% higher Cd 2+ fluxes than the NMroots ( Figure 2). The presence of NaCl (100 mM) significantly decreased the flux rates in both NM-and EM-roots but the EM-roots still exhibited 1.2-1.4-fold greater Cd 2+ uptake Figure 1. Cd 2+ concentrations in roots, stems and leaves of non-mycorrhizal (NM) and ectomycorrhizal (EM) Populus × canescens under cadmium and salt stress. Poplar plantlets inoculated with or without Paxillus involutus isolates (MAJ or NAU, 30 d), were hydroponically acclimated and subjected to 24 h of CdCl 2 (0 or 50 µM) in combination with NaCl (0 or 100 mM). Mean values of Cd 2+ concentrations in control (−Cd), CdCl 2 stress (+Cd), and combined stress of CdCl 2 and NaCl (+Cd + NaCl) are shown. Each column is mean ± SD obtained from 3 individual plants. Statistically significant differences (p < 0.05) among treatments are indicated with different letters (a-d).

Steady-State Cd 2+ Influx in Ectomycorrhizal Poplar Roots and Fungal Mycelia under NaCl Stress
To determine whether the Cd 2+ enrichment in EM P. × canescens resulted from the P. involutus-stimulated uptake, Cd 2+ fluxes were examined in NM-, EM-roots and fungal mycelia under CdCl 2 and NaCl stress. CdCl 2 exposure (50 µM, 24 h) resulted in an apparent Cd 2+ uptake, 34.9 pmol cm −2 s −1 , along NM-roots of the hyperaccumulator, Populus × canescens (Figures 2 and S1). EM-roots exhibited 36% to 39% higher Cd 2+ fluxes than the NM-roots ( Figure 2). The presence of NaCl (100 mM) significantly decreased the flux rates in both NM-and EM-roots but the EM-roots still exhibited 1.2-1.4-fold greater Cd 2+ uptake than the NM-roots (Figures 2 and S1). The effect of NaCl on root Cd 2+ fluxes resembles the trend of Cd 2+ accumulation in salinized NM-and EM-roots (Figures 1 and 2). than the NM-roots (Figures 2 and S1). The effect of NaCl on root Cd 2+ fluxes resembles the trend of Cd 2+ accumulation in salinized NM-and EM-roots (Figures 1 and 2).  Figure S1). Mean values of Cd 2+ fluxes in control (−Cd), CdCl2 stress (+Cd), and combined stress of CdCl2 and NaCl (+Cd + NaCl) are shown. Cd 2+ flux was not detectable in salt controls that were treated without CdCl2. Each column is mean ± SD obtained from 5 individual plants. Statistically significant differences (p < 0.05) among treatments are indicated with different letters (a-e).
Fungal hyphae of the two tested P. involutus isolates, MAJ and NAU, showed a drastic Cd 2+ influx, 28.9-30.1 pmol cm −2 s −1 , under CdCl2 treatment (50 μM, 24 h, Figure 3). NaCl reduced the Cd 2+ influx by 84-85% in the mycelia (Figure 3), which is similar to the reduction in EM-roots upon salinity stress ( Figure 2). . Net Cd 2+ fluxes in fungal hyphae of Paxillus involutus isolates (MAJ and NAU) under cadmium, salt, and inhibitor treatments. MAJ and NAU mycelia (the youngest and active hyphae) were hydroponically acclimated and subjected to 24 h of CdCl2 (0 or 50 μM) in combination with NaCl (0 or 100 mM). The short-term Cd-and Cd + NaCl-stressed fungal mycelia were treated with an inhibitor of plasmalemma H + -ATPase (sodium orthovanadate, 0 or 500 μM) or an inhibitor of Ca 2+ -permeable channels (LaCl3, 0 or 5 mM) for 30 min. Following 30 min equilibration in measuring solutions, Cd 2+ flux recordings were continued for 15 min on the surface of pelleted hyphae. Mean values of Cd 2+ fluxes in control (−Cd), CdCl2 stress (+Cd), and combined stress of CdCl2 and NaCl (+Cd + NaCl) in the presence and absence of inhibitors are shown. Cd 2+ flux was not detectable in salt controls that were treated without CdCl2. Each column is mean ± SD obtained  Figure S1). Mean values of Cd 2+ fluxes in control (−Cd), CdCl 2 stress (+Cd), and combined stress of CdCl 2 and NaCl (+Cd + NaCl) are shown. Cd 2+ flux was not detectable in salt controls that were treated without CdCl 2 . Each column is mean ± SD obtained from 5 individual plants. Statistically significant differences (p < 0.05) among treatments are indicated with different letters (a-e).
Fungal hyphae of the two tested P. involutus isolates, MAJ and NAU, showed a drastic Cd 2+ influx, 28.9-30.1 pmol cm −2 s −1 , under CdCl 2 treatment (50 µM, 24 h, Figure 3). NaCl reduced the Cd 2+ influx by 84-85% in the mycelia (Figure 3), which is similar to the reduction in EM-roots upon salinity stress ( Figure 2). than the NM-roots (Figures 2 and S1). The effect of NaCl on root Cd 2+ fluxes resembles the trend of Cd 2+ accumulation in salinized NM-and EM-roots (Figures 1 and 2).  Figure S1). Mean values of Cd 2+ fluxes in control (−Cd), CdCl2 stress (+Cd), and combined stress of CdCl2 and NaCl (+Cd + NaCl) are shown. Cd 2+ flux was not detectable in salt controls that were treated without CdCl2. Each column is mean ± SD obtained from 5 individual plants. Statistically significant differences (p < 0.05) among treatments are indicated with different letters (a-e).
Fungal hyphae of the two tested P. involutus isolates, MAJ and NAU, showed a drastic Cd 2+ influx, 28.9-30.1 pmol cm −2 s −1 , under CdCl2 treatment (50 μM, 24 h, Figure 3). NaCl reduced the Cd 2+ influx by 84-85% in the mycelia (Figure 3), which is similar to the reduction in EM-roots upon salinity stress ( Figure 2). . Net Cd 2+ fluxes in fungal hyphae of Paxillus involutus isolates (MAJ and NAU) under cadmium, salt, and inhibitor treatments. MAJ and NAU mycelia (the youngest and active hyphae) were hydroponically acclimated and subjected to 24 h of CdCl2 (0 or 50 μM) in combination with NaCl (0 or 100 mM). The short-term Cd-and Cd + NaCl-stressed fungal mycelia were treated with an inhibitor of plasmalemma H + -ATPase (sodium orthovanadate, 0 or 500 μM) or an inhibitor of Ca 2+ -permeable channels (LaCl3, 0 or 5 mM) for 30 min. Following 30 min equilibration in measuring solutions, Cd 2+ flux recordings were continued for 15 min on the surface of pelleted hyphae. Mean values of Cd 2+ fluxes in control (−Cd), CdCl2 stress (+Cd), and combined stress of CdCl2 and NaCl (+Cd + NaCl) in the presence and absence of inhibitors are shown. Cd 2+ flux was not detectable in salt controls that were treated without CdCl2. Each column is mean ± SD obtained . Net Cd 2+ fluxes in fungal hyphae of Paxillus involutus isolates (MAJ and NAU) under cadmium, salt, and inhibitor treatments. MAJ and NAU mycelia (the youngest and active hyphae) were hydroponically acclimated and subjected to 24 h of CdCl 2 (0 or 50 µM) in combination with NaCl (0 or 100 mM). The short-term Cd-and Cd + NaCl-stressed fungal mycelia were treated with an inhibitor of plasmalemma H + -ATPase (sodium orthovanadate, 0 or 500 µM) or an inhibitor of Ca 2+ -permeable channels (LaCl 3 , 0 or 5 mM) for 30 min. Following 30 min equilibration in measuring solutions, Cd 2+ flux recordings were continued for 15 min on the surface of pelleted hyphae. Mean values of Cd 2+ fluxes in control (−Cd), CdCl 2 stress (+Cd), and combined stress of CdCl 2 and NaCl (+Cd + NaCl) in the presence and absence of inhibitors are shown. Cd 2+ flux was not detectable in salt controls that were treated without CdCl 2 . Each column is mean ± SD obtained from 5 fungal cultures. Statistically significant differences (p < 0.05) among treatments are indicated with different letters (a-e).

Transient Cd 2+ Kinetics and Membrane Potential upon Salt Shock
CdCl 2 shock (50 µM) created a transient Cd 2+ influx in roots of NM P. × canescens, although the flux gradually decreased with prolonged exposure time ( Figure 4A). EM-roots exhibited a pattern similar to NM-roots but with typically higher influx rates ( Figure 4A). The Cd 2+ influxes in both NM-and EM-roots were markedly reduced upon the NaCl addition ( Figure 4A), similar to reduction found for the steady-state Cd 2+ influx in salinized roots ( Figure 2). Compared with the EM-roots, the restriction effect of NaCl was more pronounced in NM-roots ( Figure 4A).
Transient kinetics of membrane potential upon CdCl 2 (50 µM) and NaCl (100 mM) shocks were compared between roots of NM-and EM-poplars because the membrane potential indicates activity of PM H + -ATPase [66]. NMT recordings showed that the resting membrane potential ranged from −54.4 to −59.2 mV in NM-roots under control conditions ( Figure 4B). EM-roots had a more strongly hyperpolarized PM, with a membrane potential ranging from −71.7 to −80.8 mV ( Figure 4B). CdCl 2 shock exerted no significant effects on the membrane potential in NM-and EM-roots, although a marginal rise (5.0-6.1 mV) was observed after the onset of CdCl 2 addition, which returned to the pretreatment level 1-2 min after Cd 2+ addition ( Figure 4B). However, the addition of NaCl together with CdCl 2 caused an immediate and substantial depolarization of the membrane potential in NM-and EM-roots, although the PM tended to be rehyperpolarized during prolonged exposure to NaCl + CdCl 2 ( Figure 4B). In comparison, the membrane potential in EMroots was less depolarized (−22.2 to −41.4 mV) after the onset of CdCl 2 + NaCl shock as compared to NM-roots (−4.1 to −13.0 mV, Figure 4B).

Transient Cd 2+ Kinetics and Membrane Potential upon Salt Shock
CdCl2 shock (50 μM) created a transient Cd 2+ influx in roots of NM P. × canescens, although the flux gradually decreased with prolonged exposure time ( Figure 4A). EMroots exhibited a pattern similar to NM-roots but with typically higher influx rates ( Figure  4A). The Cd 2+ influxes in both NM-and EM-roots were markedly reduced upon the NaCl addition ( Figure 4A), similar to reduction found for the steady-state Cd 2+ influx in salinized roots ( Figure 2). Compared with the EM-roots, the restriction effect of NaCl was more pronounced in NM-roots ( Figure 4A).
Transient kinetics of membrane potential upon CdCl2 (50 μM) and NaCl (100 mM) shocks were compared between roots of NM-and EM-poplars because the membrane potential indicates activity of PM H + -ATPase [66]. NMT recordings showed that the resting membrane potential ranged from −54.4 to −59.2 mV in NM-roots under control conditions ( Figure 4B). EM-roots had a more strongly hyperpolarized PM, with a membrane potential ranging from −71.7 to −80.8 mV ( Figure 4B). CdCl2 shock exerted no significant effects on the membrane potential in NM-and EM-roots, although a marginal rise (5.0-6.1 mV) was observed after the onset of CdCl2 addition, which returned to the pretreatment level 1-2 min after Cd 2+ addition ( Figure 4B). However, the addition of NaCl together with CdCl2 caused an immediate and substantial depolarization of the membrane potential in NM-and EM-roots, although the PM tended to be rehyperpolarized during prolonged exposure to NaCl+CdCl2 ( Figure 4B). In comparison, the membrane potential in EM-roots was less depolarized (−22.2 to −41.4 mV) after the onset of CdCl2+NaCl shock as compared to NM-roots (−4.1 to −13.0 mV, Figure 4B). At the apical zones Cd 2+ fluxes and membrane potential were recorded before and after the addition of CdCl2 (100 μM) or a combined solution of CdCl2 (50 μM) and NaCl (100 mM). The recordings continued respectively for 5 and 30 min before and after the cadmium and salt shock. Each data point is mean ± SD obtained from 5 individual plants. CdCl 2 and NaCl shock-altered Cd 2+ kinetics and membrane potential in non-mycorrhizal (NM) Populus × canescens and ectomycorrhizal (EM) roots. (A) Cd 2+ flux kinetics. (B) Membrane potential. Poplar plantlets were inoculated with or without Paxillus involutus isolates (MAJ or NAU) for 30 d. Root tips were excised from EM-and NM-poplars and equilibrated for 30 min in Cd 2+ or H + measuring solution. At the apical zones Cd 2+ fluxes and membrane potential were recorded before and after the addition of CdCl 2 (100 µM) or a combined solution of CdCl 2 (50 µM) and NaCl (100 mM). The recordings continued respectively for 5 and 30 min before and after the cadmium and salt shock. Each data point is mean ± SD obtained from 5 individual plants.

Effects of PM H + -ATPase Inhibitor and Activator on Cd 2+ Uptake
Cd 2+ transport in poplar trees is accelerated by the PM H + -ATPase [38,48,52]. An H + -pump inhibitor, orthovanadate, was used to testify the crucial role of H + -pumps for Cd 2+ uptake in NM-, EM-roots, and fungal hyphae under CdCl 2 and salt stress. In NMroots, orthovanadate decreased the Cd 2+ influx approximately two-fold, while in EM-roots only 17-25% decreases were found (Figures 5 and S2). In mycelia, vanadate also caused moderately reduced Cd 2+ influx ( Figure 3). In the presence of NaCl, the inhibition of orthovanadate was evident in the fungus and roots, although the Cd 2+ influx had been significantly lowered by the salt treatment (Figures 3, 5 and S2).

Effects of PM H + -ATPase Inhibitor and Activator on Cd 2+ Uptake
Cd 2+ transport in poplar trees is accelerated by the PM H + -ATPase [38,48,52]. An H +pump inhibitor, orthovanadate, was used to testify the crucial role of H + -pumps for Cd 2+ uptake in NM-, EM-roots, and fungal hyphae under CdCl2 and salt stress. In NM-roots, orthovanadate decreased the Cd 2+ influx approximately two-fold, while in EM-roots only 17-25% decreases were found (Figures 5 and S2). In mycelia, vanadate also caused moderately reduced Cd 2+ influx ( Figure 3). In the presence of NaCl, the inhibition of orthovanadate was evident in the fungus and roots, although the Cd 2+ influx had been significantly lowered by the salt treatment (Figures 3, 5 and S2).  (Figures S2 and S3). Mean values of Cd 2+ fluxes in control (−Cd), CdCl2 stress (+Cd), and combined stress of CdCl2 and NaCl (+Cd + NaCl) in the presence and absence of inhibitors are shown. Cd 2+ flux was not detectable in salt controls that were treated without CdCl2. Each column is mean ± SD obtained from 5 individual plants. Statistically significant differences (p < 0.05) among treatments are indicated with different letters (a-f).
Furthermore, the activator of PM H + -ATPase, fusicoccin (FC), was used to test the effect of H + pumping on Cd 2+ uptake in short-term stressed roots. Following the CdCl2 treatment (50 μM, 24 h), roots of NM-and EM-poplars were subjected to FC activation. Immediately after the onset of FC addition, a stimulation of Cd 2+ influxes was observed at the surface of NM-and EM-roots ( Figure 6A). H + efflux was correspondingly increased in FC-treated NM-and EM-roots ( Figure 6B), indicating that H + pumps were transiently activated [84][85][86][87]. The observation that the increase in H + efflux corresponded to the Cd 2+ influx in P. × canescens roots suggests that the uptake of Cd 2+ was promoted by the H + -ATPases in the PM.  (Figures S2 and S3). Mean values of Cd 2+ fluxes in control (−Cd), CdCl 2 stress (+Cd), and combined stress of CdCl 2 and NaCl (+Cd + NaCl) in the presence and absence of inhibitors are shown. Cd 2+ flux was not detectable in salt controls that were treated without CdCl 2 . Each column is mean ± SD obtained from 5 individual plants. Statistically significant differences (p < 0.05) among treatments are indicated with different letters (a-f).
Furthermore, the activator of PM H + -ATPase, fusicoccin (FC), was used to test the effect of H + pumping on Cd 2+ uptake in short-term stressed roots. Following the CdCl 2 treatment (50 µM, 24 h), roots of NM-and EM-poplars were subjected to FC activation. Immediately after the onset of FC addition, a stimulation of Cd 2+ influxes was observed at the surface of NM-and EM-roots ( Figure 6A). H + efflux was correspondingly increased in FC-treated NM-and EM-roots ( Figure 6B), indicating that H + pumps were transiently activated [84][85][86][87]. The observation that the increase in H + efflux corresponded to the Cd 2+ influx in P. × canescens roots suggests that the uptake of

Transcriptional Activation of H + -ATPase in Ectomycorrhizal P. × canescens
Transcript levels of the PM H + -ATPase-encoding genes, PcHA4, PcHA8 and PcHA11, were examined in NM and EM roots since these three PcHAs were previously shown to be differently expressed under control and Na + stress conditions [66]. EM-roots showed significantly higher (0.5-4.2 fold) transcript levels of PcHA4, PcHA8 and PcHA11 than NM-roots ( Figure 7A). This observation agrees with Sa et al. (2019) [66]. Cadmium treatment (50 μM CdCl2, 24 h) resulted in upregulation of PcHA4, PcHA8, and PcHA11 in NMroots ( Figure 7A). In contrast, Cd 2+ caused a 14-45% decline of PcHAs in EM-roots, with the exception of PcHA4 in NAU roots ( Figure 7A). It is notable that the transcript levels, in particular those of PcHA8, and PcHA11, still remained higher in the EM-than in NMroots, despite the decline caused by Cd 2+ stress ( Figure 7A).
NaCl treatment (100 mM, 24 h) lowers the transcript levels of PcHAs (4,8,11) in NMroots [66]. Here, exposure to NaCl of the Cd 2+ -treated roots did not result in decreased PcHA4 and PcHA8 transcript levels and an increase of PcHA11 was observed ( Figure 7A). Similarly, NaCl did not significantly change PcHAs transcription in EM-roots in the presence of Cd 2+ ( Figure 7A). We noticed that EM-roots retained overall higher transcript levels of PcHAs than NM-roots under co-occurring stresses of cadmium and salinity. At the apical zones, Cd 2+ and H + fluxes were recorded before and after the addition of fusicoccin (10 µM). The recordings continued, respectively, for 5 and 35 min before and after fusicoccin shock. Each data point is mean ± SD obtained from 5 individual plants.

Transcriptional Activation of H + -ATPase in Ectomycorrhizal P. × canescens
Transcript levels of the PM H + -ATPase-encoding genes, PcHA4, PcHA8 and PcHA11, were examined in NM and EM roots since these three PcHAs were previously shown to be differently expressed under control and Na + stress conditions [66]. EM-roots showed significantly higher (0.5-4.2 fold) transcript levels of PcHA4, PcHA8 and PcHA11 than NMroots ( Figure 7A). This observation agrees with Sa et al. (2019) [66]. Cadmium treatment (50 µM CdCl 2 , 24 h) resulted in upregulation of PcHA4, PcHA8, and PcHA11 in NM-roots ( Figure 7A). In contrast, Cd 2+ caused a 14-45% decline of PcHAs in EM-roots, with the exception of PcHA4 in NAU roots ( Figure 7A). It is notable that the transcript levels, in particular those of PcHA8, and PcHA11, still remained higher in the EM-than in NM-roots, despite the decline caused by Cd 2+ stress ( Figure 7A).
NaCl treatment (100 mM, 24 h) lowers the transcript levels of PcHAs (4, 8, 11) in NM-roots [66]. Here, exposure to NaCl of the Cd 2+ -treated roots did not result in decreased PcHA4 and PcHA8 transcript levels and an increase of PcHA11 was observed ( Figure 7A). Similarly, NaCl did not significantly change PcHAs transcription in EM-roots in the presence of Cd 2+ ( Figure 7A). We noticed that EM-roots retained overall higher transcript levels of PcHAs than NM-roots under co-occurring stresses of cadmium and salinity. Roots were harvested from EM-and NM-poplars and used for total RNA isolation and RT-qPCR. 18S rRNA was used as a reference gene. Specific primers designed to target PcHA4, 8,11,PcANN1,2,4 and 18S rRNA are shown in Table S1. Mean values of PcHAs and PcANNs relative transcript levels in control (−Cd), CdCl2 stress (+Cd), and combined CdCl2 and NaCl stress (+Cd + NaCl) are shown. Each column is mean ± SD obtained from 3 independent experiments. Statistically significant differences (p < 0.05) among treatments are indicated with different letters (a-d).

Calcium Channel Inhibitor Blocks Cd 2+ Fluxes
Cadmium ions enter the plasma membrane through CaPCs in plant cells [48,79,88]. To determine whether CaPCs contributed to the mediation of Cd 2+ influx under combined CdCl2 and NaCl stress, LaCl3 was used to block Ca 2+ -channels in the roots of NM-and EMpoplars. The inhibitor significantly decreased root Cd 2+ uptake in the presence and absence of NaCl, although NaCl treatment reduced the apparent Cd 2+ influx under coexisting stress (Figures 5 and S3). Similarly, the LaCl3 significantly reduced Cd 2+ uptake in fungal hyphae regardless of the NaCl addition ( Figure 3).  Table S1. Mean values of PcHAs and PcANNs relative transcript levels in control (−Cd), CdCl 2 stress (+Cd), and combined CdCl 2 and NaCl stress (+Cd + NaCl) are shown. Each column is mean ± SD obtained from 3 independent experiments. Statistically significant differences (p < 0.05) among treatments are indicated with different letters (a-d).

Calcium Channel Inhibitor Blocks Cd 2+ Fluxes
Cadmium ions enter the plasma membrane through CaPCs in plant cells [48,79,88]. To determine whether CaPCs contributed to the mediation of Cd 2+ influx under combined CdCl 2 and NaCl stress, LaCl 3 was used to block Ca 2+ -channels in the roots of NM-and EMpoplars. The inhibitor significantly decreased root Cd 2+ uptake in the presence and absence of NaCl, although NaCl treatment reduced the apparent Cd 2+ influx under coexisting stress (Figures 5 and S3). Similarly, the LaCl 3 significantly reduced Cd 2+ uptake in fungal hyphae regardless of the NaCl addition ( Figure 3).

The P. Involutus-Activated PM H + -ATPase Contributes to Cd 2+ Enrichment in EM Roots
Our data show that the woody hyperaccumulator, P. × canescens, exhibited strong Cd 2+ uptake and accumulation in root and shoots, which is further enhanced by colonizing with EM-fungus P. involutus (Figure 1). These findings are similar to previous reports in longterm studies [29,33,48,52]. The root flux recordings confirmed that the enhanced Cd 2+ entry in P. × canescens roots was due to the colonization with MAJ and NAU isolates, which were characterized by a remarkable Cd 2+ enrichment in the hyphae (Figures 2, 3 and S1) [48,90]. However, we observed that salt stress caused by NaCl reduced the Cd 2+ influx in roots and fungus (Figures 2, 3 and S1). Similarly, NaCl reduced root cadmium uptake and translocation in the halophyte Carpobrotus rossii [7,8] and Atriplex halimus [91]. An important novel result was that the P. involutus could alleviate the salt suppression of Cd 2+ uptake in P. × canescens roots (Figures 2, 4 and S1). To obtain a mechanistic understanding of the underlying processes, we inhibited and stimulated the Cd 2+ fluxes with pharmacological agents. The entry of Cd 2+ in the roots and fungal hyphae declined when the plasmalemma H + -ATPase was inhibited by vanadate (Figures 3, 5 and S2) [48] and increased when the plasmalemma H + -ATPase was stimulated by FC ( Figure 6). These data suggest that Cd 2+ uptake required a proton gradient [48,52]. Moreover, P. involutus colonization resulted in a higher H + efflux and correspondingly a more negative membrane potential (Figure 6), indicating that the PM H + -ATPases were activated by the ectomycorrhiza [48,64,66]. This is similar to the enhanced proton-ATPase in arbuscular-mycorrhizal symbiosis [92,93]. The highly activated H + -pumps hyperpolarize the PM, thereby facilitating Cd 2+ influx via hyperpolarization-activated CaPCs [48,73]. In accordance with our flux analyses, transcript levels of the PM H + -ATPase-encoding genes, PcHA4, PcHA8, PcHA11, generally remained at higher levels in ectomycorrhizal roots under control and CdCl 2 stress compared to NM P. × canescens roots, although two or three of the tested PcHAs were down-regulated by CdCl 2 in MAJ and NAU roots (Figure 7). Of note, EM-roots maintained higher transcripts of PcHA4, 8, 11 than non-colonized roots under combined stress of CdCl 2 and NaCl (Figure 7). Similarly, Sa et al. showed that both MAJ and NAU roots retain higher transcript levels of PcHA4 and/or PcHA8 than NM-roots under control and NaCl stress conditions [66]. Increased abundances of PM H + -ATPase transcripts are expected to contribute to the activated H + -pumps because the plasmalemma H + -ATPases are transcriptionally regulated in poplars [85,86,94]. Thus, the retained H + -pumping activity resulted in less depolarization of membrane potential under NaCl stress (Figure 4) [66], thereby upkeeping Cd 2+ influx into the EM-roots. This result concurs with those of , who found that upregulation of HA2.1 and AHA10.1 leads to Cd 2+ uptake in EM poplar roots [52].

The Fungus-Elicited Annexins Mediated Cd 2+ Uptake in EM Roots
Since LaCl 3 inhibited Cd 2+ uptake into roots and fungal hyphae, our results support that Cd 2+ uptake involves CaPCs in the PM (Figures 3, 5 and S3) [41,48,68,73,88]. Plant annexins, in particular ANN1, ANN2 and ANN4, have been shown to function as CaPCs in Arabidopsis, maize and rice [70][71][72][73]75,76]. Zhang et al. suggested that PeANN1 facilitates the flow of cadmium ions through CaPCs [79]. CdCl 2 treatment upregulated transcripts of PcANN1, PcANN2 and PcANN4 in roots of NM P. × canescens (Figure 7), similar to the findings in crop species, such as maize, peanut and rice [73,80,81]. Accordingly, the cadmium-elicited annexins might mediate root Cd 2+ inflow through CaPCs in the poplar, contributing to its hyperaccumulator character [33]. Noteworthy, PcANN1, PcANN2 and PcANN4 showed remarkably higher transcripts in EM-roots than in the non-colonized under control conditions; CdCl 2 treatment caused a further increase in PcANN1 in EM-roots and PcANN2 was specifically increased in MAJ-colonized roots (Figure 7). The arbuscular mycorrhiza-stimulated transcription of GmAnn1a was observed in soybean roots [95]. In addition, annexin proteins also showed enhanced accumulation in arbuscular mycorrhizal roots of Medicago sativa and M. truncatula following cadmium application [96,97]. Therefore, the fungus-induced annexins might have collectively contributed to the CaPCsmediated Cd 2+ enrichment in root cells of the poplar [73,79]. In accordance with this notion, we have previously shown that Paxillus-colonized roots showed higher Ca 2+ and Cd 2+ influxes than NM-roots [48,64,65]. We noticed that the transcripts of PcANN1, 2, 4 in EM-roots exhibited lower levels than non-colonized roots under CdCl 2 stress (Figure 7). However, root Cd 2+ influx remained higher in EM than in NM (Figures 2 and 4). Thus, it can be inferred that the annexin-mediated uptake of Cd 2+ was mainly promoted by the electrochemical gradient across the PM that was established by H + -ATPases. NaCl decreased PcANN2 and PcANN4 in NM-roots, but the transcript levels of PcANN1, 2, 4 were less reduced by NaCl in EM-roots (Figure 7). It is worth noting that in these ectomycorrhizal roots, transcription of PcHAs was retained at high levels under cadmium and salinity stress (Figure 7). Taken together, this suggests that the fungus-stimulated transcription of annexins contributed to Cd 2+ enrichment in EM-roots under combined stresses of cadmium and salt.

Fungal Inoculation with Populus × canescens
The two isolates of EM fungus P. involutus (MAJ and NAU) from Büsgen-Institute: Forest Botany and Tree Physiology (Göttingen University, Büsgenweg 2, Göttingen, Germany) were cultured on modified Melin Norkrans medium [83]. P. × canescens plantlets were micropropagated and rooted in modified Murashige and Skoog (MMS) medium [98]. Uniform and healthy plantlets were inoculated with MAJ or NAU for 30 d using a Petri-dish culture system [99].

Cadmium and NaCl Treatment
The agar plugs with hyphae, and plants colonized with or without EM fungus, were hydroponically acclimated in MMS nutrient solution for 2-3 d [66]. Then fungal mycelia, NM-and EM-plants were treated with CdCl 2 (0 or 50 µM) in combination with NaCl (0 or 100 mM) in MMS solution. Following 24 h of CdCl 2 treatment and combined stress of CdCl 2 and NaCl, steady-state Cd 2+ fluxes were recorded in fungal mycelia, NM-and EM-roots. Transcript levels of genes encoding annexins (PcANN1, 2, 4) and PM H + -ATPases (PcHA4, 8,11) were examined in control and stressed roots.
After 24 h exposure to 50 µM CdCl 2 , roots from NM-and EM-poplars were subjected to an activator of PM H + -ATPase, Fusicoccin (FC). FC produced by Fusicoccum amygdali, has the function of activating H + -ATPase in the PM [102,103]. Cd 2+ and H + transient kinetics were continuously recorded for 35 min after FC (10 µM) were added to measuring solutions.

Flux Recordings of Cd 2+ and H + 4.5.1. Microelectrodes Preparation and Calibration
Cd 2+ and H + flux profiles were recorded using an NMT system (NMT-YG-100, Younger USA LLC, Amherst, MA, USA). The glass microelectrodes were prepared as previously described [42,43,48,84,104]. Prior to flux recordings, the calibration of Cd 2+ -and H + -selective microelectrodes were carried out in the following standards (concentrations in mM): After calibration, the microelectrodes that showed Nernstian slopes of 58 ± 6 mV/decade (H + ) and 29 ± 4 mV/decade (Cd 2+ ) were used in our NMT recordings.

Steady-State Cd 2+ Flux Recordings
After 24 h exposure to CdCl 2 (0 or 50 µM) in combination with NaCl (0 or 100 mM), sodium orthovanadate (0 or 500 µM), and LaCl 3 (0 or 5 mM), fungal mycelia and root tips excised from NM-and EM-poplars were subjected to 30 min equilibration in the following measuring solutions (concentrations in mM), respectively: Following equilibration, net fluxes of Cd 2+ along root axis (100 to 2300 µm) were monitored at an interval of 200-300 µm. The flux recording at each point was continued for 6-8 min [41,64,101,105]. For the fungal mycelia, Cd 2+ flux recording of pelleted hyphae was continued 15 min [48]. Cd 2+ fluxes were recorded from at least five individual plants or fungal cultures for each treatment. The flux oscillations in EM fungus and poplars are not so pronounced as that observed in crop seedlings [48,106].

Data Analysis
The calculations of flux rate and membrane potential were processed using JCal V3.2.1 program (Xuyue (Beijing) Sci and Tech Co. Ltd., Suzhou street 49, Haidian District, Beijing, China, Available online: http://www.xuyue.net/, accessed on 12 March 2021). All experimental data were subjected to SPSS version 19.0 (IBM Corporation, Armonk, NY, USA). Differences between means were considered significant at p < 0.05.

Conclusions
Our data provide further evidence that cadmium can be enriched in ectomycorrhizal poplars under co-existing stress conditions of Cd 2+ and NaCl. P. involutus stimulated Cd 2+ influx through CaPCs in ectomycorrhizal P. × canescens roots, depending on the plasmalemma H + -ATPase. NaCl lowered the uptake of Cd 2+ in poplar roots, which was alleviated by ectomycorrhization with P. involutus. Ectomycorrhizal fungus colonization upregulated transcription of PM H + -ATPases (PcHA4, 8, 11) and increased transcripts of annexins (PcANN1, 2, 4), which might mediate Cd 2+ conductance through PM CaPCs. NaCl-treated EM-roots retained relatively highly expressed PcHAs and PcANNs. We hypothesize that the sustained transcription of PcHAs resulted in H + pumping activity and PM hyperpolarization in the ectomycorrhiza, thus promoting Cd 2+ enrichment through the PcANNs-mediated Ca 2+ channels in EM-roots under co-occurring stresses of cadmium and salinity. Although the colonization of MAJ and NAU varies with regard to the formation of intraradical hyphae, i.e., the Hartig net, both strains conserved higher Cd 2+ uptake under salt stress than NM-roots. We propose that P. involutus strains, which have been repeatedly shown to improve salt tolerance, may be applied as beneficial microbes to improve plant phytoremediation for cadmium in salt-affected areas.