Populus simonii×P. nigra cuttings were purchased from Lindian County, Daqing City, Heilongjiang Province, China. The cuttings were propagated by the plant factory of Northwest A&F University to obtain clone seedlings.

Funneliformis mosseae (BGC XJ01A) was purchased from the Institute of Plant Nutrition, Resources and Environment, Beijing Academy of Agriculture and Forestry Sciences. Propagation of F. mosseae using tobacco plants as a host. The tested inoculants included spores (the spore density was approximately 81 per gram of inoculant), mycelia and root fragments.

Surface soil (5-20 cm) was obtained from the garden of Northwest Agriculture and Forestry University. The soil was passed through a 2 mm sieve and sterilized by moist heat at 121°C for 2 h twice. The river sand was passed through a 2 mm sieve, then washed, and sterilized by dry heat at 170°C for 12 h. The sterile soil and sterile sand were mixed uniformly at the 1:2 (V/V) ratio. The physicochemical properties of the soil substrate were pH 7.6 (soil and water ratio was 1:5), available N 25.77 mg·kg^-1, available P 12.30 mg·kg^-1, available K⁺ 112.67  mg·kg^-1, Ca²⁺ 78.51 mg·kg^-1, water soluble Na⁺ 203.54 mg·kg^-1.

The experiment consisted of a randomized complete block design with two factors: inoculation with F. mosseae (AM) and no inoculation with F. mosseae (NM); saline-alkali stress (S) and no saline-alkali stress (N). There were 18 replicates for each treatment with a total of 72 pots.

100 mmol·L^-1 Na⁺, pH 9.5 saline-alkali stress solution (the stress concentration was selected based on the data of the seven-year-old P. simonii×P. nigra rhizosphere soil sample and the data of the pre-test of potted plants in the greenhouse) was made by mixing 50 mmol·L^-1 Na₂CO₃ solution and 100 mmol·L^-1 NaHCO₃ solution according to the volume ratio of 4:5. Each pot (7 cm × 7 cm × 5 cm) was preloaded with 230 g growth substrate. The growth substrate in 36 pots (group S) was pretreated: each pot was filled with 35 mL of saline-alkali stress solution (the maximum water content of 250 g soil matrix was about 70 mL), then after drying at 25°C, 35 mL of saline-alkali stress solution was added again. The control group without saline-alkali stress (group N) was treated by the same amount of distilled water using the same method.

Subsequently, 18 pots of each of the two groups (S and N) were selected and added with 20 g F. mosseae inoculum. The control group was added with 20 g of F. mosseae inoculum after moist heat sterilization.

One seedling was planted in one pot. Plants were cultivated in the greenhouse of the Forestry College of Northwest A&F University. The greenhouse’ room temperature was 25-28°C, light intensity was 3000 lux, photoperiod was 16/8 h, and humidity was 50%-70%. Each pot was weighed and rehydrated daily to keep the soil well moist. Each Pot was watered weekly with 25 mL of Hogland’s nutrient solution (Johnson et al., 1957) containing 1/10 Pi. After 8 weeks of culture, relevant data were measured and samples were collected.

According to the method of Wu et al. (2016), the third fully expanded leaf was selected for 30 min shading treatment. Then PAM Fluorometer (Mini-imaging-PAM, Walz, Germany) was used to measure the initial fluorescence value (Fo) and maximum fluorescence value (Fm) of leaves under saturation pulse; the minimum fluorescence value (Fo’), maximum fluorescence value (Fm’), and steady-state fluorescence value (Fs) after light adaptation. The maximum quantum yield of PSII (Fv/Fm, Fv/Fm = (Fm – Fo)/Fm), photosynthetic electron transport rate (ETR, ETR = (Fm’ – Fs)/Fm’ × Fm), non-photochemical quenching (qN, qN = (Fm − Fm’)/Fm’), photochemical quenching (qP, qP = (Fm’ – Fs)/(Fm’ − Fo’)), non-photochemical quenching coefficient according to the following formulas (NPQ, NPQ = (Fm – Fm’)/(Fm’)) and PSII actual photon quantum efficiency (ФPSII, ФPSII = (Fm’ – Fs)/Fm’) were calculated by the instrument.

According to the method of Huang et al. (2010), during the period from 8:30 am to 11:30 am, the third fully expanded leaf of each plant was selected for measuring net photosynthetic rate (P_n ), stomatal conductance (g_s ), intercellular CO₂ concentration (C_i ), and transpiration rate (E) by using li-6400 portable open flow gas-exchange system (Li-Cor Inc., Lincoln, NE, United States). The measurement conditions: photo-synthetically active irradiation, 1000 µmol·m^-2·s^-1; temperature, 25°C; relative humidity, 60%; and CO₂ concentration, 400 µmol·mol^-1.

After 8 weeks of culturing in the greenhouse, the height of each plant was measured using a ruler (cm). Plant samples were collected in two groups, the aboveground part and the underground part. Each 6 plants were pooled into one replicate, with three replicates per treatment. The fresh weight of the aboveground and underground parts were weighed by using an electronic analytical balance (FA2004A, Shanghai Jingda). The aboveground samples were divided into two parts with one of which immediately frozen in liquid nitrogen, then ground into powder and stored in a -80°C refrigerator and the other of which dried at 75°C to a constant weight. The underground part was divided into three parts, two of which followed the same procedure as aboveground samples and the third part was deposited in the FAA fixative to determine F. mosseae infection.

Tissue water content (TWC) and relative water content (RWC) in aboveground parts were determined according to the method of Kovar et al. (2019). We obtained the same part of the aboveground part from each sample, and weighed the fresh weight (FW). The samples were immersed in sterile distilled water for 48 hours, and the saturated fresh weight (WT) was weighed. Then the samples were dried at 75°C until constant weight, and the dry weight (DW) was weighed. Finally, the TWC ((FW – DW)/FW × 100%) and RWC ((FW – DF)/(WT – DW) × 100%) were calculated.

Roots were stained with trypan blue (Phillips and Hayman, 1970), and then mycorrhizal colonization was determined using the gridline intersection method (Giovannetti and Mosse, 1980).

Referring to the method of Wang et al. (2021), weighed 10 g of air-dried soil samples that have passed through a 1 mm sieve, and measure the soil pH according to the water-soil ratio of 5:1 (W/V).

After drying, the aboveground and underground samples were ground into powder, dried at 80°C to constant weight, and passed through a 20 μm sieve (Wu et al., 2016). Weighed 0.05 g of plant samples, with reference to the method of Zhou (2022), determine the concentration of Na⁺ and K⁺ in plants. Weighed 0.05 g of plant sample, with reference to the method of Porcel et al. (2016), determine the concentration of Ca²⁺ in plants.

P. simonii×P. nigra genome data was obtained from http://www.wangsui.net.cn/resource/temp/Pxiaohei_DH_1_21_genome_v4.6.tar.gz (Zhou et al., 2019). The NHX protein sequences of Arabidopsis thaliana (https://phytozome-next.jgi.doe.gov/info/Athaliana_TAIR10) were used as the input sequences for the BLAST software (2.12.0+) to search for P. simonii×P. nigra NHX genes (PxNHXs). The protein sequences of candidate gene were input into HMMER software (3.3.2) and aligned with the PFAM database (https://ftp.ebi.ac.uk/pub/databases/Pfam/releases/Pfam35.0/Pfam-A.hmm.gz) to further verify their domains.

Protein physicochemical properties of PxNHXs were analyzed using ProtParam (https://web.expasy.org/protparam/). Subcellular localization predictions of PxNHXs were performed using Plant-mPLoc (http://www.csbio.sjtu.edu.cn/bioinf/plant-multi/#). Transmembrane domain predictions of PxNHXs were performed using TMHMM 2.0 (https://services.healthtech.dtu.dk/service.php?TMHMM-2.0). Signal peptide predictions for PxNHXs were performed using SignalP 5.0 (https://services.healthtech.dtu.dk/service.php?SignalP-5.0).

Phylogenetic tree of the NHX gene families of P. simonii×P. nigra, A. thaliana, Zea mays (Zorb et al., 2005), and P. trichocarpa (Tian et al., 2017) were constructed using MEGA X using the adjacency method (NJ) and 1000 bootstrap-up replicates. The phylogenetic tree was visualized by iTOL (https://itol.embl.de/).

Total RNA was extracted from these samples using the E.Z.N.A Plant RNA Kit R6827-01 (Omega Bio-Tek, Norcross, GA, USA) following the manufactory instructions strictly. RNA was reversely transcribed to cDNA using the FastQuant RT Kit KR106 (TIANGEN, Beijing, China) according to the manufactory instructions. By using NCBI Primer-BLAST (https://www.ncbi.nlm.nih.gov/tools/primer-blast/index.cgi?LINK_LOC=BlastHome), the PxNHXs’ quantitative primers was designed ( Supplementary Table 1 ). Two housekeeping genes (NCBI accession: BU875027, GQ253565.1) were served as the reference genes (Dong et al., 2021). The qRT-PCR reaction was conducted by the CF96X real-time PCR system (Bio-Rad, Hercules, CA, USA). Each reaction mixture was 10 µL containing 1 µL diluted cDNA template, 0.5 µL forward and reverse primers (10 µmol·L^-1), 5 µL ChamQ SYBR qPCR Master Mix (Vazyme, Nanjing, China), and 3 µL sterilized ddH₂O. The two-step qRT-PCR was run as follows: 30 s denaturation at 95°C, 40 cycles of denaturation at 95°C for 10 s, annealing at the annealing temperature for 10 s, extension at 72°C for 20 s, followed by heating from 65 to 95°C at a rate of 0.5°C every 5 s. All samples were amplified in duplicates from the same RNA preparation and the mean value was calculated. Meanwhile, we averaged three amplified samples from each RNA and the relative expression of each target gene was calculated according to the 2^−ΔΔCT protocol (Livak and Schmittgen, 2001).

All experiments were repeated at least three times. All results were expressed as the mean ± standard error (SE) in tables and figures. One-way and two-way analysis of variance (ANOVA) and Tukey’s tests using SPSS software (Version 26.0, SPSS Inc., Chicago, IL, USA) were used to evaluate significant differences across all parameters.

After 8 weeks, the roots of P. simonii×P. nigra established a symbiotic relationship with F. mosseae ( Figure 1A ). Under saline-alkali stress, the infection rate was significantly decreased by 56.56% ( Figure 1B ), which indicated that the infection activity of AM fungi was inhibited by saline-alkali stress.

As shown in Figure 2 , saline-alkali treatment significantly increased soil pH while F. mosseae inoculation decreased soil pH. The soil pH of NMS (i.e., 8.54) was the highest. While the soil pH of AMS was significantly reduced by 2.93%. Two-way ANOVA showed that soil pH was significantly affected by F. mosseae inoculation and saline-alkali stress. The results indicated that AM fungi played an important role in reducing the pH of the rhizosphere soil of P. simonii×P. nigra.

As shown in Figure 3 , saline-alkali stress significantly inhibited the growth of P. simonii×P. nigra. Compared with NMN, the plant height, fresh weight of aboveground part and fresh weight of underground part of NMS decreased by 75.25%, 53.69% and 53.69%, respectively, and AMS of which decreased by 61.16%, 49.80% and 58.52%, respectively compared with AMN. However, under the influence of F. mosseae inoculation, the plant height, fresh weight of aboveground part and fresh weight of underground part of AMS were 300.51%, 368.74% and 5.31% higher than those of NMS, respectively. Two-way ANOVA showed that the plant height and fresh weight of the aboveground part of P. simonii×P. nigra were significantly affected by AM fungal inoculation and saline-alkali treatment, indicating that AM fungi played an important role in promoting the growth of the aboveground part of P. simonii×P. nigra.

The fresh weight of plants was affected by water content, so the tissue water content (TWC) and relative water content (RWC) of the aboveground part was further investigated. As shown in Table 1 , under the influence of saline-alkali stress, the TWC of AMS and NMS was significantly increased, but RWC was not significantly affected, probably because P. simonii×P. nigra had adapted to the long-term stressed environment. Inoculation with F. mosseae could significantly increase the TWC and RWC of the aboveground parts. Compared with NMS, the TWC of AMS was significantly increased by 23.74%, and the RWC was increased by 11.70%, indicating that AM fungi play an important role in promoting plant water uptake.

Plants synthesize organic matter through photosynthesis for maintaining a symbiotic relationship with mycorrhizae and for their own growth. As shown in Figure 4 , F. mosseae inoculation increased the P_n (net photosynthetic rate, 140.57%), g_s (stomatal conductance, 183.33%), C_i (intercellular CO₂ concentration, 6.88%) and E (transpiration rate, 173.11%) of P. simonii×P. nigra without salt-alkali stress. The application of saline-alkali stress resulted in a decrease in the P_n of NMS (40.25%) and a significant increase in the C_i (15.97%). For AMS, although saline-alkali stress led to a significant decrease in P_n (29.28%), g_s (45.27%), C_i (11.05%) and E (29.69%) (compared with AMN), the P_n (70.13%), g_s (55.07%) and E (92.03%) were still higher than those of NMN. Two-way ANOVA analysis showed that g_s and C_i were affected between saline-alkali stress and AM fungi interaction, indicating that AM fungi may promote the growth of P. simonii×P. nigra by regulating these two pathways under saline-alkali stress.

As shown in Figure 5 , the effects of saline-alkali stress and AM fungal inoculation on the chlorophyll fluorescence parameters of P. simonii×P. nigra were different. Saline-alkali stress significantly reduced the Fv/Fm value regardless of whether F. mosseae was inoculated or not. However, the Fv/Fm value of AMS was still significantly higher than that of NMS (3.10%), which was similar to that of NMN. For NPQ, ФPSII, qN, qP and ETR, the inoculation with F. mosseae under saline-alkali stress had opposite effects on these parameters. Compared with NMN, NPQ and qN of NMS decreased by 19.90% and 2.39%, respectively while ФPSII, qP and ETR increased by 13.28%, 9.82% and 13.61%, respectively. Although inoculation with F. mosseae increased the values of ФPSII, qP and ETR (compared with NMN), under the influence of saline-alkali stress, compared with AMN, the ФPSII, qP and ETR values of AMS decreased by 16.63%, 2.61% and 15.50%, respectively, and at the same time, NPQ and qN increased by 23.67% and 13.64%, respectively. However, the ФPSII (11.08%), qP (0.84%) and ETR (10.64%) of AMS were still higher than those of NMS. Two-way ANOVA analysis showed that AM fungal inoculation and saline-alkali stress significantly affected NPQ, ФPSII, qN and ETR.

As shown in Table 2 , the saline-alkali treatment caused an increase in the Na⁺, K⁺ and Ca²⁺ concentrations of P. simonii×P. nigra, while the Na⁺, K⁺ and Ca²⁺ concentrations were further increased after inoculation with AM fungi. Compared with NMS, the Na⁺, K⁺ and Ca²⁺ concentrations in the aboveground part of AMS were increased by 444.92%, 63.14% and 62.31%, respectively while those of the underground part were increased by 106.28%, 8.53% and 0.93%, respectively. Two-way ANOVA showed that the Na⁺, K⁺ and Ca²⁺ concentrations in the aboveground and underground parts of P. simonii×P. nigra were significantly affected by AM fungal inoculation and saline-alkali treatment. It suggested that AM fungal inoculation in saline-alkali environment played an important role in promoting plant element uptake.

A total of 8 NHX gene family members were identified in the P. simonii×P. nigra genome ( Supplementary Table 2 ). The protein length of the eight PxNHXs ranged from 318 (Px_DH25939) to 1147 (Px_DH21638) amino acids; the molecular weight of the protein ranged from 35.13 kDa (Px_DH25939) to 127.02kDa (Px_DH21638); the isoelectric point ranged from 5.48 (Px_DH25989) to 8.85 (Px_DH28184); the number of transmembrane domains ranged from 7 (Px_DH25939) to 12 (Px_DH18325 and Px_DH21638). Subcellular localization showed that six PxNHXs were localized to vacuole and other two PxNHXs were localized to the cell membrane.

To further clarify the corresponding gene functions and evolutionary relationships of PxNHXs, a phylogenetic tree was constructed ( Figure 6 ). The eight PxNHXs were divided into three clades: Px_DH25939, Px_DH28184, Px_DH21008, Px_DH25989 and Px_DH11117 were divided into the same clades, which were similar to AtNHX2 in function; Px_DH30127 was divided into one branch, which was functionally similar to AtNHX6; Px_DH21638 and Px_DH18325 were divided into the same clade and were functionally similar to AtNHX7.

Through qPCR analysis of NHX gene family, the effects of AM fungus inoculation on saline-alkali tolerance of P. simonii×P. nigra were explored from the microscopic level. In the aboveground part ( Figure 7A ), saline-alkali stress led to the down-regulated expression of Px_DH21008, Px_DH25989, Px_DH30127, Px_DH21638, and Px_DH18325 in NMS to varying degrees (2.45%-25.36%). Under saline-alkali stress, inoculation with F. mosseae up-regulated the expression of PxNHXs except Px_DH25939. Among them, Px_DH21008, Px_DH25989, Px_DH11117 and Px_DH21638 were significantly affected by F. mosseae inoculation, and their expressions were up-regulated. Compared with NMS, their expression levels in AMS were increased by 182.95%, 57.24%, 27.34%, and 143.56%, respectively. In the underground part ( Figure 7B ), PxNHXs’ expressions were more active. Under the saline-alkali stress, these genes were up-regulated by 200.99% (Px_DH21008)-1562.91% (Px_DH21638) in NMS. Inoculation with F. mosseae also induced further up-regulation of these genes. In AMS, the expression of these genes increased by 254.16% (Px_DH25939)-34762.16% (Px_DH21638) compared with NMN, indicating that AM fungi endow plants with higher saline-alkali tolerance by inducing the up-regulated expression of these genes. One-way ANOVA of the underground part showed that saline-alkali stress mainly affected the expression of Px_DH28184 and Px_DH30127. While AM ​​fungi significantly affected the expression of PxNHXs except Px_DH25939. Two-way ANOVA indicated that the expression of Px_DH11117 and Px_DH30127 in the underground part were affected by the interaction of AM fungal inoculation and saline-alkali treatment.

In order to explore the potential relationship between PxNHXs induced by AM fungi and the growth status of P. simonii×P. nigra, Pearson’s correlation analysis was performed on the above parameters. In aboveground parts ( Figure 8 ), inoculation with AM fungi was significantly correlated with the expression of Px_DH21008, Px_DH25989, Px_DH11117, Px_DH25939, Px_DH21638, Px_DH18325. The expression of these genes significantly affected the physiological activities (TWC, RWC, plant weight and FW), photosynthesis (ETR, qN, qP, NPQ and ФPSII), gas exchange (P_n 、g_s and E) and ion concentration (Na⁺, K⁺ and Ca²⁺) of the aboveground parts, suggesting that AM fungi could affect the growth status of P. simonii×P. nigra by affecting the expression of PxNHXs in the aboveground parts.

In underground parts ( Figure 9 ), inoculation with AM fungi was significantly correlated with the PxNHXs excepted Px_DH25939. At the same time, these genes were also related to the Na⁺, K⁺, Ca²⁺ concentrations in the underground part of P. simonii×P. nigra. Although AM fungal inoculation had no direct and significant correlation with root Na⁺ concentration and rhizosphere soil pH, the Px_DH28184, Px_DH30127 were significantly correlated with underground Na⁺ concentration, while Na⁺ concentration was significantly correlated with underground FW and rhizosphere soil pH. This result implies that AM fungi can alleviate saline-alkali stress by regulating the expression of these two genes to regulate root growth and reduce soil pH.