Comprehensive analysis of AHL gene family and their expression under drought stress and ABA treatment in Populus trichocarpa

Identification of AHL genes in P. trichocarpa

We obtained the amino acid sequences of 29 AtAHL proteins (Zhao et al., 2014) from TAIR (The Arabidopsis Information Resource; https://www.arabidopsis.org/) database. These sequences were used as a query to blast in Phytozome (version 12.1; https://phytozome.jgi.doe.gov/pz/portal.html), with the following parameters: target type, proteome; program, BLASTP-protein query to protein db; and expect (E) threshold, −1. The redundant sequences were manually detected and deleted. In the candidate PtrAHL amino acid sequences, the existence of the conserved AT-hook motif and the PPC domain was confirmed using the SMART database (http://smart.embl-heidelberg.de/) (Letunic & Bork, 2018). The coding amino acid and gene sequences of each PtrAHL gene were directly downloaded from Phytozome. The subcellular localization was predicted using WoLF PSORT (https://www.genscript.com/wolf-psort.html) (Horton et al., 2007). The protein length, molecular weight, theoretical pI, aliphatic index and grand average of hydropathy (GRAVY) were identified by using the ProtParam tool on Expasy (https://web.expasy.org/protparam/) (Wilkins et al., 1999).

Phylogenetic analysis

The amino acid sequences from O. sativa were obtained from the NCBI database by standard Protein BLAST using the PtrAHL genes as query. The Clustal X (version 2.0) and Bioedit (version 7.2.5) were used to perform the multiple sequence alignment with a gap opening penalty and gap extension penalty of 10 and 0.1, respectively. MEGA (version 7.0) was used to analyze the molecular features and phylogenetic relationships of the AHL genes of P. trichocarpa, Arabidopsis and O. sativa using the maximum likelihood (ML) method with the following parameters: Poisson model; partial deletion (95%); and 500 bootstrap replications (He et al., 2013).

Gene structure and conserved motif analysis

The exon-intron organization of PtrAHL genes was generated using GSDS (Gene Structure Display Server version 2.0: http://gsds.cbi.pku.edu.cn/). MEME (Multiple Em for Motif Elicitation program 5.1.1; http://meme-suite.org/tools/meme) was used to identify the conserved motifs in PtrAHL genes using the default parameters and a conserved motif number of 15.

Calculation of Ka/Ks values

The orthologs of the PtrAHL genes in P. trichocarpa were identified in DnaSP (DNA Sequence Polymorphism, version 6.12.03). The paralogous genes’ coding sequences were first aligned using Clustal X, and the alignment results and selected protein-coding regions were loaded in DnaSP. Gene pairs originating from duplication events within genome of a single species was regarded as paralogous pairs (Altschul et al., 1997). Finally, the synonymous and nonsynonymous substitutions were selected to calculate the Ka/Ks values for all paralogous genes. The divergence time (T; million years ago, Mya) was calculated as follows: T = Ks/(2 × 6.1 × 10⁻⁹) Mya (Wei, Xu & Li, 2017).

Chromosomal location analysis

The chromosomal location of each PtrAHL gene in the P. trichocarpa genome was confirmed using the PopGenIE (Populus Genome Integrative Explorer, version 3 database (Sjodin et al., 2009). The schematic representation of the homologous chromosomes segments was generated using MapChart (version 2.3). Adobe Illustrator CS6 (Adobe System Incorporated) program was used to construct the physical map based on the information in PopGenIE. Genes separated by ≤5 gene loci in a range of 100 kb were considered as tandem duplicates (Hu et al., 2010).

Promoter cis-element analysis

The promoters of PtrAHL genes, 2,000 bp upstream of the translation start site, were obtained from Phytozome. The online database PlantCARE was used to search and locate the cis-elements in each promoter (Lescot et al., 2002).

Plant materials, drought stress and ABA Treatment

P. trichocarpa (genotype Nisqually-1) plants were cultured in vitro on woody plant medium (WPM) supplied with 20 g/L of sucrose and 5.5 g of agar and maintained in a growth chamber with 16 h/8 h light/dark cycles at 25 °C with a light intensity of 46 µmol photons m⁻² s⁻¹. For drought stress, three-week-old in vitro plants were transferred to WPM medium supplied with 7% PEG6000. For ABA treatment, three-week-old plants were exposed to the WPM medium containing 200 µM ABA. The plants were allowed to grow under these two stress conditions for 0, 3, 6, 12, and 24 h. The untreated plants served as controls. Samples were collected, immediately frozen in liquid nitrogen, and stored in a freezer (−80 °C) for further use. Three independent biological replicates were maintained for each treatment.

RNA isolation and real-time quantitative PCR (qRT-PCR)

Total RNA of P. trichocarpa was extracted using the CTAB (cetyltrimethylammonium bromide) method (Jaakola et al., 2001), and RNA samples were analyzed by agarose gel electrophoresis to ensure integrity. HiScript® II Q Select RT SuperMix for qPCR Kit (Vazyme) was used for reverse transcription, and gDNA Wiper Mix to remove the genome DNA. qRT-PCR was performed using the AceQ® Universal SYBR qPCR Master Mix Kit (Vazyme) following the manufacturer’s instructions. The qRT-PCR was performed on a qTOWER 3G Cycler (Analytik Jena, Germany), and the 2^−ΔΔCT method was used to analyze the relative gene expression level. Primer Premier 5 was used to design the primers, and NCBI primer designing tool (https://www.ncbi.nlm.nih.gov/tools/primer-blast/index.cgi?LINK_LOC=BlastHome) was employed to check their specificity. The P. trichocarpa actin gene (GenBank ID: XM_002298674) was used as the reference gene (Chu et al., 2014). All PtrAHL gene-specific primers used for qRT-PCR are listed in Table S1.

exHeatmap analysis

The tissue-specific expression values were directly downloaded from the published RNA-seq data (NCBI GEO number GSE6422) in PopGenIE using the accession numbers of each PtrAHL (Yang et al., 2008). Heml 1.0.3.7 software was used to visualize the data (Deng et al., 2014) The clustering method selected “Hierarchical” and “Average linkage (default)”. The expression values are listed in the Table S2.

Gene Ontology (GO) annotation

The functional annotation of PtrAHL genes was performed using Blast2GO (version 5.2) as follows: the full-length protein sequences were aligned and mapped using UniProtKB/Swiss-Prot (swissprot_v5). An annotation cutoff: 55, GO weight: 5, and E-value-hit-filte of 1.0E-6 was used for annotating into three GO categories (biological processes, molecular functions, and cellular components).

Protein interaction network analysis

The STRING (version 11.0; https://string-db.org/cgi/input.pl) database was used to predict the protein interaction network of PtrAHL proteins in P. trichocarpa (Szklarczyk et al., 2015). The sequence of each protein was used as query, and the “Organism” options P. trichocarpa was selected. Then basic settings included “meaning of network edges” set as “evidence” and “active interaction sources” set as “Text-mining, Experiments, Databases, Co-expression, Neighborhood, Gene Fusion and Co-occurrence”. The minimum required interaction score was customized to medium confidence of 0.4. Finally, the protein interaction network was exported and visualized using Cytoscape (version 3.7.2).

Statistical analysis

Data were analyzed using Student’s t test was performed using SPSS software (version 20, IBM, Chicago, USA) to assess the significant differences between the treatments and control, at p <0.05 (*) and p < 0.01 (**). The detailed statistical analysis was listed in Table S3.

Identification of AHL genes in P. trichocarpa

We identified 37 AHL genes from P. trichocarpa whole genome. According to the position on the chromosome, we named these AHL genes from PtrAHL1 to PtrAHL37. The protein length ranged from 248 aa to 413 aa, and the molecular weight ranged from 26.2 kDa to 44.2 kDa. Among the different proteins, PtrAHL22 was the largest and PtrAHL30 was the smallest. The PtrAHL proteins showed considerable variations in the theoretical pI values (5.05 to 10.11). PtrAHL25 had the lowest GRAVY score, while PtrAHL16 had the highest (Table 1). WoLF PSORT predicted nuclear localization for 23 proteins, cytoplasmic localization for 18.9% and chloroplast localization for four proteins. PtrAHL3/34 and PtrAHL8/9 were predicted in the plasma membrane and endoplasmic reticulum, respectively. The detailed information is provided in Table 1.

Phylogenetic analysis of PtrAHL genes

The full-length amino acid sequences of 37 AHLs from P. trichocarpa, 29 from Arabidopsis and 25 from O. sativa were used to construct an unrooted tree following the maximum likelihood (ML) method to analyze the evolutionary relationship. MEGA 7.0 program was used to visualize the results. Seven distinct clades were identified (Fig. 1; Bishop et al., 2020) named AHL Type I, AHL Type II a-c and AHL Type III a-c. Type-I possessed RGRPAGSKNKPKP and RGRPPGSKNKPKP conserved sequences and the PPC domain; Type II a-c contained two AT-hook motifs (RGRPRKY and RGRP conserved sequences) and the PPC domain; Type III a-c contained RGRPRKY sequence and the PPC domain. AHL Type-I had the largest number of members, with about 59.45% of PtrAHL genes, while AHL Type II a-c and Type III a-c had had 21.62% and 18.93%, respectively. Otherwise, PtrAHL1, 3, 4, 7, 8, 9, 10, 15, 16, 21, 22, 24, 31, 32, 33, 35 and 37 clustered with the Arabidopsis AHL protein. These foundings indicate that the PtrAHL proteins are more closely related to those of Arabidopsis than of O. sativa. The detailed phylogenetic distance is provided in Table S4.

Phylogenetic analysis, gene structure and conserved motifs

We constructed an phylogenetic tree to investigate the relationships among PtrAHLs using their corresponding protein sequences by MEGA 7.0. We identified 16 paralogous pairs with strong bootstrap support (90%; Fig. 2A). We analyzed the exon-intron organization using the full-length coding sequence of AHLs with their corresponding genomic DNA sequences in P. trichocarpa. Most PtrAHL genes of the same groups had similar exon-intron lengths and numbers, especially the paralogous pairs within the same group (Fig. 2B). PtrAHL1 and PtrAHL18 belonging to AHL Type-II a, possessed four introns of similar lengths. Both PtrAHL9 and PtrAHL22, of AHL Type-I subfamily, contained coding sequences and downstream, with no introns.

We further used MEME to confirm the conserved motifs in PtrAHL proteins. Fifteen conserved motifs, named motif 1 to motif 15, were found (Fig. 2C). All PtrAHLs contained motif 1, the AT-hook motif. The paralogous pairs of PtrAHLs possessed similar motifs. Interestingly, PtrAHL14 had one more motif (motif 15) than its homologous gene, PtrAHL5. PtrAHL32 had one more motif (motif 4) than its homolog PtrAHL37, however, PtrAHL37 possessed an additional motif 6 than PtrAHL32, Notably, PtrAHL31 had RGRPRKY and RGRPLESVKKQHN conserved motifs; however, it was classified into the AHL Type-I subfamily, indicating similarity with PtrAHL3/16 and PtrAHL11/34. These observations suggest functional diversity among the AHL proteins. Detailed information about motifs is listed in Table S5.

Chromosomal locations and gene duplication

We mapped on 19 linkage groups (LGs) to determine the distribution of PtrAHL genes. Six PtrAHL genes (PtrAHL7, 8, 9, 10, 11, and 12) were located on LG II, followed by another six (PtrAHL1, 2, 3, 4, 5, and 6) on LGI (Fig. 3). Five AHL genes were detected on LG III and four were located on LG V. PtrAHL34, 35, and 36 were located on LG XIV. Besides, LG IV, VIII, IX, X, and XIII each contained two PtrAHL genes. PtrAHL 31 and PtrAHL37 were located on LG XII and XIX, respectively. No genes were detected on the other LGs.

We further calculated the Ka/Ks values using DnaSP to better understand the evolutionary constraints associated with the PtrAHL genes. The Ka/Ks value was less than 1 for almost all PtrAHL genes, except for PtrAHL7 and PtrAHL24 (Table S5). indicating the evolution of the majority of paralogous pairs was under purifying selection. The Ka/Ks value of PtrAHL7/PtrAHL24 was 1.4901, suggesting the influence of positive selection. The divergent date of these 16 paralogous pairs was estimated to have occurred between 1.36 to 58.62 Mya (Table S6).

Promoter cis-element analysis

We analyzed the putative stress-related and phytohormone-related cis-elements in the promoter region of PtrAHL genes using PlantCARE to understand these genes’ possible regulatory mechanisms after phytohormone treatment and under abiotic stress.. Ten different phytohormone-related and stress-related cis-elements were identified (Fig. 4). All genes, except PtrAHL4, 16, 18, 24, 32, and 35 contained the ABA-responsive element (ABRE). PtrAHL29 possessed eight ABRE. Besides, PtrAHL21 and 22 contained methyl jasmonate (MeJA)-responsive element (CGTCA/TGACG), and 20 other PtrAHLs possessed salicylic acid-responsive TCA element. Ptr3, 4, 10, 12, 15, 21, and 35 contained one GARE-motif, which acted as gibberellin-responsive element. Further, 15 PtrAHL genes possessed TGA (auxin-responsive element), 9 members possessed P-box (gibberellin-responsiveness element), and 14 contained TC-rich repeats (defense- and stress- responsive element). Meanwhile, 15 PtrAHL genes possessed abiotic stress-related cis-elements, such as LTR (low-temperature- responsive element) and MBS (MYB binding site, drought-inducible element). Seven LTR cis-elements were found PtrAHL32. However, we failed to detect any phytohormone-related or stress-related cis-elements in PtrAHL24.

Gene Ontology (GO) annotation

The GO analysis of PtrAHL genes revealed their roles in multiple biological processes (Fig. 5). Cellular process and metabolic process had the largest number of PtrAHL genes, followed by regulation of biological process and biological regulation. Approximately 6% of the AHL genes were related to reproduction, multicellular organismal process, developmental process, and reproductive process. For molecular function, 14 PtrAHL genes were functionally enriched in heterocyclic compound binding and organic cyclic compound binding. Besides, 21% of PtrAHL genes were involved in transcription regulator activity. Cellular component prediction showed that 47% of PtrAHL genes were located in intracellular membrane-bound organelles and membrane-bounded organelles and two were integral components of membrane. The detailed information about GO annotation is listed in Table S7.

Expression profile of PtrAHL genes in different tissues or organs

To further investigate the function of PtrAHL genes in the development, we downloaded the available microarray datasets (accession number: GSE6422) (Yang et al., 2008) to generate a heatmap based on the tissue-specific expression in P. trichocarpa. We obtained the expression profiles of 36 PtrAHL genes. however, PtrAHL3 showed no expression in the selected dataset. The PtrAHL genes demonstrated different expression patterns across the various developmental stages of tissues or organs in P. trichocarpa (Fig. 6). Twenty-four PtrAHL genes showed high expression in roots; 64.86% exhibited high expression in young leaves, whereas most of the PtrAHL genes showed low expression in mature leaves. PtrAHL11 and PtrAHL14 showed high expression in both roots and young leaves, low expression in mature leaves and nodes, and no expression level in internodes. A similar expression profile was observed for PtrAHL24 and PtrAHL27 as well as PtrAHL33 and PtrAHL34. Notably, PtrAHL4, 26 and 29 showed high expression in root. Most of the PtrAHL genes showed low or no expression in internodes and nodes.

Expression pattern of PtrAHL genes under drought stress

Previous studies have shown that MBS cis-element helps the plant cope with drought stress (Xu et al., 2019). Our analysis of cis-elements predicted MBS cis-element in 15 PtrAHL genes (Fig. 4). We, therefore analyzed the relative expression levels of these 15 PtrAHL genes in both roots and leaves of P. trichocarpa under drought stress. In roots, 86.67% of PtrAHL genes were induced under drought stress (Fig. 7). PtrAHL27, 29, and 36 were significantly up-regulated at all time points (p < 0.01), while, PtrAHL12, 14, and PtrAHL17 were remarkably down-regulated PtrAHL34 was rapidly induced by drought stress at 6 h, with 15-fold expression and gradually declined to normal level at 24 h. PtrAHL31 showed no significant change in expression in roots compared with the untreated control. In leaves, all genes but PtrAHL14 and PtrAHL21, were induced by drought stress. PtrAHL14 was downregulated at 12 h, and PtrAHL21 was suppressed at all time points. PtrAHL16 and PtrAHL17 were only upregulated at 6 h and 12 h, respectively. PtrAHL31 showed no significantly early expression (3 h and 6 h) but was upregulated at 12 h and 24 h compared with the control, suggesting late response. Meanwhile, PtrAHL12 showed the opposite trend. PtrAHL20 and PtrAHL22 showed similar expression patterns in both roots and leaves; ohey were upregulated at 3 h and eaked at 6 h) and then declined at 12 h and 24 h (Fig. 7).

Expression pattern of PtrAHL genes under ABA treatment

To better understand PtrAHL genes’ role in response to ABA treatment, we analyzed the expression levels of 15 PtrAHL genes in both roots and leaves of P. trichocarpa under 200 µM ABA treatment at 3, 6, 12, and 24 h. In roots, PtrAHL8, 9, 16, 17, 20, 21, 22, 27, 29, 31, 34, and 36 were induced by ABA treatment; PtrAHL12 and PtrAHL23 were suppressed, PtrAHL14 showed no significant change in expression compared with the control (Fig. 8). PtrAHL8 and PtrAHL22 were significantly upregulated at the initial time points and peaked at 24 h (p < 0.01), with 30-fold and 20-fold expression levels compared with control, respectively. PtrAHL27 and 29 were remarkably upregulated at all time points. PtrAHL9, 12, and 16 were downregulated at 3 h. PtrAHL17 and PtrAHL31 were significantly upregulated at 6 h and 12 h (p < 0.05), respectively. PtrAHL20 was significantly upregulated during early response and elevated to a level comparable with that of the untreated control at 24 h. In leaves, 12/15 PtrAHL genes were induced, two were suppressed, and only one displayed no change. PtrAHL12, 17, 20, and 27 were dramatically upregulated at all time points (p < 0.01), PtrAHL12 and PtrAHL17 were gradually upregulated at 3, 6 and 12 h and peaked at 24 h. with 13-fold and 25-fold expression, respectively. The expression level of PtrAHL27 was about 60-fold at 24 h. PtrAHL21 was significantly downregulated (p < 0.01), whereas, PtrAHL9 and PtrAHL16 were remarkably upregulated at 6, 12, and 24 h (p < 0.01) (Fig. 8). PtrAHL16, 17, 20, 22, 27, 29, 31, and 34 were induced in both roots and leaves by ABA treatment (Fig. 8).

Analysis of the AHL protein interaction network in P. trichocarpa

We used the protein sequences of P. trichocarpa AHLs to predict the protein interaction network using STRING, We totally obtained 27 protein interaction networks, excluding PtrAHL6, 8, 14, 19, 23, 27, and 28 (Fig. 9), indicating the interaction of PtrAHL proteins interact with other proteins in response to phytohormone treatment and environmental stress as well as during growth and development in P. trichocarpa. Our analysis predicted an interaction between PtrAHL2 and POPTR_0018s08410, the homologous gene of AtCPK4,that regulates the calcium-mediated ABA signaling pathway (Zhu et al., 2007). We found that the homologous proteins, including PtrAHL1 and PtrAHL18, PtrAHL2 and PtrAHL17, PtrAHL3 and PtrAHL16, PtrAHL4 and PtrAHL15, PtrAHL7 and PtrAHL24, PtrAHL9 and PtrAHL22, PtrAHL10 and PtrAHL21, PtrAHL12 and PtrAHL35, PtrAHL25 and PtrAHL30,and PtrAHL26 and PtrAHL29, shared similar protein interaction networks, indicating a conserved protein interaction domain in each protein pair.

The AHL gene family in Populus

AHL family includes transcription factors with one or two AT-hook motifs and a PPC/DUF296 domain (Zhao et al., 2013). In this study, we identified 37 PtrAHL genes in P. trichocarpa using 29 Arabidopsis AHL protein sequences. A previous study identified 29 members in P. trichocarpa (Zhao et al., 2014). Here, we reveal another nine AHL genes (PtrAHL3, 5, 6, 14, 16, 23, 26, 28 and 34).

Our analysis of predicted localization of PtrAHL8 and its homologous gene, PtrAHL23 were in the endoplasmic reticulum and nucleus, respectively. Furthermore, the homologs, PtrAHL3/16, PtrAHL9/22, PtrAHL11/34, and PtrAHL26/29 did not have the same subcellular location, indicating that homologs may also differ in their functions and signal transduction. Studies have identified that AHL proteins localize to the nucleus, while our analysis predicted the localization of 14 PtrAHL proteins in the the plasma membrane, cytoplasm, endoplasmic reticulum, and chloroplast. Prior research has demonstrated that both the AT-hook motif and PPC/DUF296 domain contribute to nuclear localization (Sgarra et al., 2006; Street et al., 2008). Therefore, we speculate the that the presence of other sequences that influence the localization. AHL1 is mainly localized in the nucleoplasm, little in the nucleolus region (Fujimoto et al., 2004). Therefore, a lot of work is needed to verify this.

The promoter cis-elements of AHLs in Populus

Under biotic and abiotic stresses, promoter cis-elements such as ABRE, CGTCA, GARE-motif, TGA-element, P-box, LTR, TCA-element, TC-rich repeats, TGACG motif and MBS cis-element play pivotal roles in the transcriptional regulation in plants (Wang et al., 2020). We identified at least one phytohormone-related or stress-related cis-element in all PtrAHL family members, except for PtrAHL24, indicating their role in regulating the response.We also found the LTR and MBS cis-elements in many PtrAHL promoters, including PtrAHL9, 12, 16, 17, 22, 23, 29, and 36, indicating their role under low temperature and drought stress. These findings suggest that AHLs may play an important role in response to abiotic stress and phytohormone treatment in P. trichocarpa.

Transcript profiles of AHL genes under drought stress and ABA treatment in Populus

Plants require water to maintain normal physiological processes, including growth, development, and reproduction (Fang & Xiong, 2015). To overcome and survive drought, woody plants have developed several strategies, including reprogramming gene expression (Debnath, Pandey & Bisen, 2011). Besides, cis-elements, especially the MBS cis-element, play an important role in drought stress. We identified 15 PtrAHL genes with the MBS cis-element. In plants, roots are the first organs to perceive drought stress (Lynch, 1995), Moreover, drought influences various events in leaves, including stomatal closure to prevent water evaporation (Qin et al., 2020). Therefore, we analyzed the expression levels of these 15 PtrAHL genes in roots and leaves of P. trichocarpa under drought stress. PtrAHL8, 9, 16, 20, 22, 23, 27, 29, 34, and 36 were induced by drought stress in both roots and leaves, indicating their role in drought response. Interestingly, PtrAHL12 was induced by drought in leaves but significantly suppressed in roots at all time points. Meanwhile, the expression level of PtrAHL31 was higher than the control at 12 h and 24 h in leaves, with no change in expression in the roots, These results indicate the tissue-specific expression of the genes. OsAHL1 (LOC_Os11g05160), the homologous gene of PtrAHL34, enhanced drought tolerance by directly regulating its target genes (HSP101, OsCDPK7, OsRNS4, Rab16b and AP2-EREBP) (Zhou et al., 2016). OsAHL1 also participates in root development under drought (Zhou et al., 2016). In our experiments, PtrAHL34 was strongly induced by drought in both roots and leaves, suggesting that similar functions for the homologous genes in different species.

Studies have shown that ABA is the most critical abiotic stress-related hormones and is involved in various physiological processes, including growth, development, and reproduction (Dong, Park & Wang, 2015; Ullah et al., 2018). ABRE (CGTGG/TC) is generally found in the promoters of ABA-inducible genes (Nakashima & Yamaguchi-Shinozaki, 2013). Our analysis predicted the presence of the ABRE motif in 15 drought-related PtrAHL genes. The qRT-PCR showed a significant change in the expression levels of 13 PtrAHL genes, except for PtrAHL14 in roots and PtrAHL36 in leaves, compared with control, indicating their role in the ABA signaling pathway. Previous studies have demonstrate that ABA in the guard cells induces stomatal closure during drought (Lim et al., 2015). Moreover, signals initiated in the roots under drought stress are transmitted quickly to the shoot and leaves, which lead to stomatal closure (Malcheska et al., 2017). We found that many drought-related PtrAHL genes such as PtrAHL9, 20, 22, 27, 29, and 34, were induced by ABA treatment. These genes were significantly upregulated in roots and leaves under drought stress and ABA treatment, indicating their role in ABA-dependent signaling in response to drought stress. PtrAHL27 was remarkably up-regulated in both roots and leaves at all time points under drought stress and ABA treatment, which indicates its critical role in response to drought stress and ABA treatment in P. trichocarpa.

The predicted interaction network of PtrAHL proteins

Protein interaction network provides an understanding of orthologous proteins’ roles in biological processes (Hao et al., 2016). We used STRING to predict the protein interaction network of AHLs of P. trichocarpa. Our constructed networks revealed that, PtrAHL2/PtrAHL17 and PtrAHL12/PtrAHL35, which are homologs of the Type-II subfamily, shared a similar protein interaction network. These four proteins interact with HDA904, the Arabidopsis gene homologous to HDA9. Increasing evidence showed that HDA9 acts in association with an ABA-related transcription factor to repress gene expression through histone deacetylation during drought stress (Baek et al., 2020). Moreover, had9 mutant was insensitive to ABA and hypersensitive to drought stress by regulates the chromatin modification of genes responsible for regulation of both the ABA-signaling and ABA catabolosm pathways in response to ABA and drought stress (Khan et al., 2020). Research also has demonstrated that AtAHL22 interacts with HDAC, including HDA9, and influences the physiological processes such as flowering time (Yun et al., 2012). In our experiments, PtrAHL12 and PtrAHL17 were induced by ABA treatment and drought treatment, respectively. Therefore, we speculate that PtrAHL12 and 17 may interact with HDA904 and participate in drought stress response via a ABA-dependent pathway in P. trichocarpa.