Phenotype of P. alba seedlings under salt stress
In this study, within one day of salt stress treatment, no significant phenotypic changes were observed in P. alba seedlings (Fig. 1). After three days of salt stress treatment, the mature leaves of P. alba turned yellow, and the tips and edges of leaves became dry and curled (Fig. 1A). The tips and margins of the leaves at the top of the P. alba were dry and slightly curled after three days of salt stress treatment (Fig. 1B). After seven days of salt stress treatment, the leaves of the whole plant turned yellow and the mature leaves were partially shed. After 11 days of salt stress treatment, the leaves of the whole plant turned yellow and withered.
POD, CAT activities and MDA content in roots and leaves of P. alba seedlings under salt stress
Compared with normal growth seedlings, POD activities in seedling roots were significantly increased after salt stress for 1 h (independent sample t-test, P < 0.01). Compared with other salt stress treatment time points, POD activity in seedling roots was the lowest at 72 h of salt stress (independent sample t-test, P < 0.01), and was the highest at 264 h of salt stress (independent sample t-test, P < 0.01) (Fig. 2A).
We found that CAT activities in seedling roots were higher than that in leaves under normal growth conditions or salt stress (independent sample t-test, P < 0.01) (Fig. 2B). Compared with other salt stress time points, CAT activity in seedling roots was the highest at 12 h of salt stress (independent sample t-test, P < 0.01).
Compared with normal growth seedlings, MDA content in seedling leaves significantly increased after 2 h of salt stress treatment (independent sample t-test, P < 0.05) (Fig. 2C). Compared with other salt stress time points, MDA content in seedling leaves was the highest at 264 h of salt stress (independent sample t-test, P < 0.05).
Root viability of P. alba seedlings under salt stress
The root viability of P. alba was significantly decreased at 1 and 2 h after salt stress (independent sample t-test, P < 0.01) (Fig. 3). Compared with other salt stress time, the root viability of P. alba seedlings was the lowest at 2 h of salt stress (independent sample t-test, P < 0.01), and then gradually increased. Compared with other salt stress time, the root viability of P. alba seedlings was the highest at 168 h under salt stress.
Photosynthetic parameters of P. alba seedlings under salt stress
We determined the photosynthetic parameters of the third leaf in P. alba under normal growth conditions and salt stress. Compared with normal growth conditions, the significant changes of net photosynthetic rate (Pn), stomatal conductance (Cond) and transpiration rate (Tr) were not observed within 12 h of salt stress. After 12 h of salt stress, values of Pn, Cond and Tr decreased gradually, and decreased to zero at 264 h (Fig. 4). Compared with normal growth conditions, the intercellular CO2 concentration (Ci) of leaves of P. alba did not change significantly within 72 h of salt stress. However, after 72 h of salt stress, Ci value increased significantly. The maximum quantum efficiency of photosystem II (PSII) photochemistry (Fv/Fm) of leaves of P. alba gradually decreased after 24 h of salt stress, and then decreased to zero at 264 h (Fig. 4). Values of the ABS/RC (maximal energy fluxes for absorption per reaction center), DI0/RC (maximal energy fluxes for energy dissipation per reaction center), ET0/RC (maximal energy fluxes for electron transport rate per reaction center), and TR0/RC (maximal energy fluxes for trapping per reaction center) were significantly decreased after 168 h of salt stress, and were decreased to zero at 264 h. The contents of chlorophyll a, chlorophyll b and carotenoid in leaves of P. alba seedlings gradually decreased after 9 h of salt stress (independent sample t-test, P < 0.01) (Fig. 4).
Na + and K+ contents in roots and leaves of P. alba seedlings under salt stress
This study determined the concentration of Na+ and K+ and the Na+/K+ ratio in the roots and leaves of P. alba under normal growth and salt stress (Fig. 5). Compared with the normal growth conditions, the significant change of K+ concentration in the roots of P. alba was not observed under salt stress during 0 ~ 24 h, but after 24 h of salt stress, the K+ concentration in the roots was gradually decreased. After 264 h salt stress, the K+ concentration in P. alba roots decreased by 0.57 fold compared to normal conditions.
After 1 h of salt stress, the content of Na+ in the roots of P. alba was increased gradually. Under salt stress for 264 h, the Na+ concentration in the roots of P. alba increased by 6.00 folds compared with that under normal growth conditions. The ratio of Na+/K+ in the roots of P. alba was increased gradually after 1 h of salt stress. Under salt stress for 264 h, the ratio of Na+/K+ in the roots of P. alba increased by 15.22 folds compared with that under normal growth conditions.
The K+ content in the leaves of P. alba increased significantly during the 1 ~ 12 h of salt stress. The Na+ content in P. alba leaves gradually increased after 2 h of salt stress. After 264 h for salt stress, the Na+ content was increased by 489.22 folds compared with normal growth conditions. The ratio of Na+/K+ in P. alba leaves gradually increased after 2 h of salt stress.
The increase rate of Na+ content in leaves was much higher than that in roots, while the K+ content retained stable in leaves and decreased in roots. These lead to a sharper increase in the Na+/K+ ratio of leaves than that of roots. These physiological observations, including Na+ transport from root cells to xylem and higher levels of K+ retention, have been found in halophytes [9].
Hormone content in roots and leaves of P. alba seedlings under salt stress
In this study, we determined content changes of endogenous hormones in the roots and leaves of P. alba under salt stress (Fig. 6, Table S1 and Table S2). Compared with 0 h, contents of SL (strigolactone), IAA (indole acetic acid), GA1 (gibberellin A1), GA3 (gibberellin A3), GA4 (gibberellin A4), GA7 (gibberellin A7), mT (meta-Topolin), 6-BA (6-benzyladenine), tZ (trans-Zeatin), and DZ (dihydrozeatin) did not change significantly in the roots of P. alba under salt stress (independent sample t-test, P > 0.05). Compared with 0 h, contents of ABA (abscisic acid ), SA (salicylic acid), iP (N6-(Δ2-isopentenyl)-adenine), pT (para-Topolin), oT (ortho-Topolin), and cZ (cis-Zeatin) in the roots of P. alba were increased significantly at a certain time point of salt stress, but contents of JA (jasmonic acid), ETH (ethylene), and K (kinetin) were decreased significantly at a certain time point of salt stress (independent sample t-test, P < 0.05).
Compared with 0 h, contents of OPC-4 (3-oxo-2-(2-(z)-pentenyl)-cyclopentane-1-butyric acid) and OPDA (cis(+)-12-oxophytodienoic acid) in the roots of P. alba were decreased after 3 h of salt stress, while the contents of iPR (N6-isopentanyladenosine), KR (kinetin riboside), pTR (para-Topolin riboside), and cZR (cis-Zeatin riboside) were the highest at one of salt stress time points. Compared with 0 h, SAG (salicylic acid 2-o-β-glucoside) in the roots of P. alba was increased after 72 h of salt stress, while JA-Ile (jasmonoyl-L-isoleucine) was decreased after 3 h of salt stress (Table S1).
Contents of SL, JA, GA1, GA3, GA4, 6-BA and K in the leaves of P. alba did not show changes significantly compared with 0 h (independent sample t-test, P > 0.05), while, contents of ABA, SA, ETH, IAA, GA7, iP, pT, mT, oT, cZ, tZ and DZ in the leaves showed changes significantly after 72 h of salt stress (independent sample t-test, P < 0.05). Compared with 0 h, contents of salicylic acid derivative SAG, abscisic acid derivative ABA-GE (ABA-glucosyl ester), cytokinin derivative oT9G (ortho-Topolin-9-glucoside), cZROG (cis-Zeatin-O-glucoside), and tZOG (trans-Zeatin-o-glucoside) in the leaves of P. alba were increased after 72 h of salt stress, while the contents of cytokinin derivatives DHZROG (dihydrozeatin-O-glucoside) and mT9G (meta-Topolin-9-glucoside) were decreased after 72 h of salt stress (Table S2).
Transcriptome of P. alba seedlings under salt stress in different tissues and different time points
The time-series transcriptome of roots, stems, leaves, and apical buds of P. alba under salt stress were detected. Sequencing of all examined tissues at 11 time points under salt stress resulted in 8.08 billion raw reads (Table S3). After quality control, a total of 7.3 billion clean reads and 106.63 Gb clean bases were obtained (Table S3). The average proportion of these clean reads mapped to the reference genome of P. alba was 95.78% (Table S3). Among them, the average proportion of clean reads with uniquely mapped position in reference genome was 91.02% (Table S3). Subsequently, we used these clean reads with uniquely mapped position in reference genome to perform gene expression analysis. In addition, the transcriptomic data of three biological repeats within each group were highly correlated, indicating stable and reliable results (Fig. S1, R2 > 0.96).
Differential genes in different tissues of P. alba seedlings under salt stress
During the 1 ~ 24 h of salt stress, the number of up-regulated differentially expressed genes (DEGs) in the roots of P. alba was greater than down-regulated (Fig. 7). On the contrary, after 72 h of salt stress, the number of down-regulated DEGs in the roots of P. alba was greater than that of up-regulated. In the leaves and apical buds of P. alba under salt stress, the DEGs number was increased significantly after 72 h. In the stems of P. alba under salt stress, the DEGs number was increased significantly after 168 h.
We randomly selected three DEGs (Poalb16G007310, Poalb08G012240, and Poalb18G005570) for RT-qPCR validation (Fig. S2). The RT-qPCR results showed, compared with 0 h, Poalb18G005570, Poalb16G007310, and Poalb08G012240 were specifically expressed in stems, leaves, and apical buds during the salt stress, respectively, which was consistent with the transcriptome results.
Identification of key candidate genes responding to salt stress in roots of P. alba seedlings
Based on transcriptome-based gene expression data, and the physiological and biochemical data in P. alba roots under salt stress, we used WGCNA (Weighted Correlation Network Analysis) software to perform weighted gene co-expression network analysis (Fig. 8). A total of 61 modules related to phenotype data were identified. The modules whose absolute value of correlation coefficient between phenotype and module eigengene was greater than 0.8 were MEblue, MEmidnightblue, MEturquoise, and MEdarkslateblue. These four modules were mainly related to Na+, Na+/K+ and phytohormone contents (Fig. 8A).
Modules MEturquoise, MEblue, MEmidnightblue, and MEdarkslateblue contained 7506, 4784, 298, and 64 genes in roots of P. alba, respectively. GO enrichment showed that the genes in MEturquoise module were mainly related to histone modification, chromatin organization, and vesicle-mediated transport (Fig. S3). The genes in MEblue module were mainly related to cell wall biogenesis, cell division, cytokinesis, and organelle fission (Fig. S3).
Only six genes from module MEblue (Poalb04G009630, Poalb09G004310, Poalb02G001570, Poalb10G017000, Poalb16G000560, and Poalb18G012460) and five genes from module MEturquoise (Poalb01G003930, Poalb01G005590, Poalb13G004180, Poalb02G018040, and Poalb06G007330) exhibited sustained differential expression at all time points under salt stress. Among these 11 DEGs, except for Poalb06G007330, the other 10 genes were up-regulated expression under the salt stress in the roots of P. alba (Fig. 8B). Co-expression network analysis showed that the P. alba ANN1, bHLH112, CC1/2, KT2, PGDH1/3, SKOR/GORK, CCAOMT1, NAC019, and HKT1 showed co-expression patterns with the six sustained differential expressed genes in the MEblue module (Fig. 8C). In the MEturquoise module, the five sustained differential expressed genes showed co-expression patterns with bHLH112, SOS2, SOS3, SOS1, SAG29, OSCA1, FER, KT2, and AKT1 (Fig. 8D). Sequence analysis showed that these genes belonged to 11 gene families, which included RanBP2-type zinc finger family, calcium-binding EF-hand family, small heat-shock proteins (smHSPs) family, SIAMESE/SIAMESE-RELATED (SIM/SMR) family, QWRF motif-containing protein family, aldo-keto reductase (AKR) family, leucine-rich-repeat receptor-like kinase (LRR-RLK) family, cinnamoyl-CoA Reduce (CCR) family, sugars will eventually be exported transporters (SWEETs) family, thioredoxin reductase family, and TEOSINTE BRANCHED 1 and CYCLOIDEA, PCF1 (TCP) family (Table 1).
Table 1
Key candidate genes respond to salt stress in P. alba seedlings.
Gene ID | Family Name |
Poalb06G007330 | RanBP2-type Zinc Finger (ZnF) Family |
Poalb02G018040 | Calcium-binding EF-hand Family |
Poalb09G004310 | Small Heat-shock Proteins (smHSPs) Family |
Poalb10G017000 | SIAMESE/SIAMESE-RELATED (SIM/SMR) Family |
Poalb13G004180 | QWRF motif-containing Protein Family |
Poalb16G000560 | Aldo-keto Reductase (AKR) Family |
Poalb18G012460 | Leucine-rich-repeat Receptor-like Kinase (LRR-RLK) Family |
Poalb01G003930 | Cinnamoyl-CoA reductase (CCR) Family |
Poalb01G005590 | Sugars Will Eventually be Exported Transporters (SWEETs) Family |
Poalb02G001570 | Thioredoxin Reductase Family |
Poalb04G009630 | TEOSINTE BRANCHED 1, CYCLOIDEA, PCF1 (TCP) Family |
Poalb02G006720 | Unknown |
Poalb16G007310 | FISSION1 (FIS1) Family |
Poalb01G001990 | Plant P-type H+-ATPase (P-ATPase) Family |
Poalb09G006800 | Lateral Organ Boundaries Domain (LBD) Family |
Poalb12G010830 | Type 2C Protein Phosphatases (PP2Cs) Family |
Poalb18G005570 | Raffinose Synthase (RS) family |
Poalb01G036340 | Vascular-related Unknown Protein Family |
Poalb03G010320 | MAPK kinase kinase (MAP3K) Family |
Poalb02G006200 | Unknown |
Poalb04G005680 | Unknown |
Poalb06G010440 | HXXD-type Acyltransferase Family |
Poalb07G008970 | Unknown |
Poalb09G012950 | Cytochrome P450 Family |
Poalb08G012240 | Metallo-phosphoesterase Family |
Poalb11G009160 | Calcium-binding EF-hand Family |
Poalb14G005890 | Phospholipase D (PLD) Family |
Poalb01G007070 | Ethylene-responsive Factor (ERF) Family |
Poalb06G004370 | Ethylene-responsive Factor (ERF) Family |
Poalb18G011550 | Ethylene-responsive Factor (ERF) Family |
Poalb02G003950 | Cytochrome P450 Family |
Poalb11G014910 | NAM, ATAF1/2, CUC2 (NAC) Family |
Identification of key candidate genes responding to salt stress in leaves of P. alba seedlings
Based on transcriptome-based gene expression data, and the physiological and biochemical data in leaves of P. alba under salt stress, we used WGCNA software to perform weighted gene co-expression network analysis (Fig. 9). A total of 43 modules related to phenotype data were identified. The modules whose absolute value of correlation coefficient between phenotype and modules was greater than 0.8 were MEpurple, MEblack, MEsteelblue, MEdarkred, MElightgreen, MEpink, MEroyalblue, MEturquoise, MEgreen, and MEblue (Fig. 9A).
The genes in these 10 modules were all related to plant phytohormone. In addition, the MEturquoise module was also related to MDA, Chl a, Chl a/b, Ci, Pn, Cond, Tr, Na+, and Na+/K+ ratio. The MEgreen module was related to the phenotypic traits of Ci, Fv/Fm, TRo/RC, ETo/RC, etc. The MEblue module were related to Chl a/b, Pn, Cond, Tr, MDA, Na+, and Na+/K+ ratio.
The modules of MEturquoise, MEblue, MEblack, MEgreen, MEpink, MEroyalblue, MEpurple, MEsteelblue, MElightgreen, and MEdarkred contained 11795, 6853, 618, 834, 600, 167, 380, 62, 173 and 135 genes in leaves of P. alba, respectively. Because MEturquoise and MEblue contained more DEGs, it indicated that these two modules may be more important in response to salt stress in leaves of P. alba.
GO enrichment showed that the genes in the MEturquoise module were mainly related to photosynthesis, cytoplasmic translation, and ribosome biogenesis (Fig. S4). The genes in MEblue module were mainly related to cell wall biogenesis, organelle fission, and mitotic cell cycle process (Fig. S4).
The module of MEturquoise did not contain any genes, which were DEGs at any time point under salt stress. While, the module of MEblue contained two genes (Poalb02G006720 and Poalb16G007310), that were DEGs at different time points under salt stress. These two genes were highly connected with the key genes (such as SOS1, SOS2 and SOS3) in the known molecular regulation mechanism of salt stress (Fig. 9C). Sequence analysis showed that Poalb02G006720 was an unknown functional gene, and Poalb16G007310 belonged to FISSION1 (FIS1) family (Table 1). These two genes were up-regulated in the leaves of P. alba under salt stress (Fig. 9B).
Identification of key candidate genes responding to salt stress in stems and apical buds of P. alba seedlings
Because of the lack of phenotypic data of apical buds and stems, we used another method to identify the key candidate genes responding to salt stress. Firstly, according to the functional data in the public papers, we identified 531 genes related to salt stress from Arabidopsis thaliana. Using these 531 genes as templates, the genome of P. alba and P. trichocarpa was searched by TBLASTN software. All protein sequences of each gene family obtained from P. alba, P. trichocarpa, and A. thaliana were used to construct a phylogenetic tree. On the phylogenetic tree, the P. alba gene closest to the salt stress gene in A. thaliana was considered to be related to salt stress. Based on this principle, 1215 homologous genes to salt stress genes in A. thaliana were found in the genome of P. alba (Table S4). Secondly, we regarded the DEGs obtained in apical buds or stems at different stress time points as a set. K-means algorithm divided this set into 10 clustering subsets (Fig. S5). We calculated the total number of homologous genes related to salt stress in each clustering subset in P. alba. The clustering subsets with the first and second largest number of salt stress homologous genes were selected for further analysis. In these two subsets, the genes that exhibited sustained differential expression under salt stress were considered as key candidate genes.
Based on above principle, this study identified six key candidate genes (Poalb12G010830, Poalb18G005570, Poalb01G001990, Poalb01G036340, Poalb03G010320 and Poalb09G006800) responding to salt stress in P. alba stems (Fig. 10A). The expression of these six genes was up-regulated in P. alba stems under salt stress.
Thirteen key candidate genes (Poalb02G003950, Poalb09G012950, Poalb14G005890, Poalb08G012240, Poalb11G014910, Poalb01G007070, Poalb04G005680, Poalb11G009160, Poalb06G004370, Poalb18G011550, Poalb06G010440, Poalb02G006200, and Poalb07G008970) responding to salt stress were found in P. alba apical buds (Fig. 10B). The expressions of Poalb14G005890, Poalb08G012240, Poalb01G007070, Poalb04G005680, Poalb11G009160, Poalb06G004370, Poalb18G011550, Poalb06G010440, Poalb02G006200, and Poalb07G008970 were up-regulated in P. alba apical buds under salt stress, while Poalb02G003950 and Poalb09G012950 were down-regulated. The expression of Poalb11G014910 was down-regulated at 72 h and was up-regulated at other time points.
Functional verification of key candidate genes of P. alba responding to salt stress
Based on the above analysis, we identified 32 candidate genes from the roots, leaves, stems, and apical buds of P. alba responding to salt stress (Table 1). Among these 32 genes, Poalb13G004180 was a pseudogene. Thus, this study first cloned the remaining 31 candidate genes. Except for Poalb02G006720, 30 candidate genes were cloned successfully. For heterologous gene expression, we subcloned the coding regions of these 30 genes into yeast vector pYES2 and transferred them into competent cells of Saccharomyces cerevisiae BY4741.
Compared with the negative control, the yeasts that overexpressed Poalb01G003930, Poalb02G018040, Poalb10G017000, Poalb04G005680, Poalb01G007070 or Poalb18G011550 genes could not survive under the stress of 1.8 M sodium chloride, 2.0 M potassium chloride, or 4.0 M sorbitol, while the yeasts that overexpressed other 24 genes showed tolerance to these stresses (Fig. 11). The yeasts that overexpressed Poalb01G001990, Poalb09G006800, Poalb14G005890, Poalb11G009160, Poalb02G001570, Poalb07G008970 or Poalb02G003950 could only survive under the stress of sodium chloride, but could not under the sorbitol and potassium chloride stress. The yeasts that overexpressed Poalb06G007330, Poalb12G010830, Poalb04G009630, Poalb11G014910 or Poalb06G004370 could survive under the stress of sodium chloride and sorbitol. The yeasts that overexpressed Poalb18G012460, Poalb08G012240, Poalb02G006200, Poalb09G012950, Poalb03G010320 or Poalb18G005570 could survive under the stress of sodium chloride and potassium chloride. The yeasts that overexpressed Poalb01G005590, Poalb16G007310, Poalb01G036340, or Poalb06G010440 could survive under the stress of sodium chloride, potassium chloride and sorbitol.