Systematical Characterization of the AT-Hook Gene Family in Juglans regia L. and the Functional Analysis of the JrAHL2 in Flower Induction and Hypocotyl Elongation

AT-hook motif nuclear localization (AHL) proteins play essential roles in various plant biological processes. Yet, a comprehensive understanding of AHL transcription factors in walnut (Juglans regia L.) is missing. In this study, 37 AHL gene family members were first identified in the walnut genome. Based on the evolutionary analysis, JrAHL genes were grouped into two clades, and their expansion may occur due to segmental duplication. The stress-responsive nature and driving of developmental activities of JrAHL genes were revealed by cis-acting elements and transcriptomic data, respectively. Tissue-specific expression analysis showed that JrAHLs had a profound transcription in flower and shoot tip, JrAHL2 in particular. Subcellular localization showed that JrAHL2 is anchored to the nucleus. Overexpression of JrAHL2 in Arabidopsis adversely affected hypocotyl elongation and delayed flowering. Our study, for the first time, presented a detailed analysis of JrAHL genes in walnut and provided theoretical knowledge for future genetic breeding programs.


Introduction
The AT-hook motif nuclear localized (AHL) gene family encodes the conserved transcription factor that widely distributed in plants species [1]. The AHL proteins are generally distinguished by two characteristic domains: the AT-hook motif and the Plant and Prokaryote Conserved (PPC) domain. The AT-hook motif was first described as a high mobility group (HMG) non-histone chromosomal protein HMG-I/Y [2]. The AT-hook motif contains conserved core sequence 'Arg-Gly-Arg' that can bind to the minor groove of the AT-rich B form of DNA [3]. AT-hook motifs can be divided into two categories according to the conserved amino acid sequence: Type-I AT-hook motif contains 'Gly-Ser-Lys-Asn-Lys' consensus sequences at the end of the 'Arg-Gly-Arg' core, while the Type-II contains 'Arg-Lys-Tyr.' The PPC domain, also known as the Domain of Unknown Function #296 (DUF296), is also found in, archaea, prokaryotes, and higher plants. The PPC domain was involved in the physical interaction with other AHL (or themselves) or nuclear proteins [1,4]. According to sequence differences, the PPC can be divided into two categories: Type-A and Type-B. Type-B contains the conserved core 'Phe-Thr-Pro-His' in the upstream region, while the core sequence within the PPC domain in Type-A is more variable ('Leu-Arg-Ser-His' or 'Leu-Arg-Ala-His', etc.) [5]. The AHL genes in land plants have been grouped into Clades based on the phylogenetic relationship: Clades-A and Clades-B. The Clade-A AHLs are intron-free and encode proteins containing only one Type-I AT-hook motif and Type-A PPC domain. The Clade-B AHLs contain introns, and Clade-B AHL proteins can be further divided into two types: one type of AHLs contains a Type-II AT-hook motif and a Type-B PPC domain, and the other type of AHLs contains two AT-hooks (Type-I and Type-II) and a Type-B PPC domain [6]. 2 of 19 The AHL gene family plays vital roles in plant growth, development, and stimulus response. In Arabidopsis (Arabidopsis thaliana), the biological functions of several AHLs have been extensively studied. A number of AHL genes redundantly inhibit the elongation of hypocotyl, such as AHL22 [7], SOB3/AHL29 [6,8], and ESC/AHL27 [9]. AHL18, which is involved in the regulation of the length of the proliferation domain and the division in the root apical meristem, and inhibiting primary root growth and lateral root development [10]. In M. truncatula, the MtAHL1 and MtAHL2 were involved in the formation of nitrogenfixing nodules [11].
The AHL gene is related to regulating flowering time and reproductive development differently. The TEK/AHL16 directly binds to the floral suppressor gene FLC's repressive regulatory region, leading to its transcriptional silencing. Knockout of the TEK/AHL16 activates a variety of transposons, and up-regulate the expression of floral suppressor gene FLC through epigenetic regulation (histone acetylation and DNA methylation), thereby delaying flowering [12]. TEK/AHL16 also contributes to nexine formation and pollen wall development by directly binding to the nuclear matrix attachment region (MAR) and the promoter of AGP6, an arabinogalactan protein coding gene [13,14]. In addition, overexpressing AHL delays flowering by inhibiting FT expression [7,15]. AHL22 binds to the AT-rich sequence in the FT locus and then alters the structural design of the chromatin region by regulating both histone acetylation and methylation [15]. AHL18, which is in the same phylogenetic clade as AHL22, conferred a similar late-flowering phenotype when overexpression in Arabidopsis to that of AHL22 [7]. AHLs were also reported to be involved in hormone-related pathways. SOB3/AHL29 inhibited PIF4 protein accumulation, thereby reducing the binding of PIF4 to target loci involved in hormone transport, homeostasis, or response such as auxin, brassinosteroid, and ethylene [16]. SOB3/AHL29 also targeted downstream genes associated with auxin-related genes, such as YUCCA8 and SAUR19 [8]. AGF1/AHL25 regulated gibberellins homeostasis in Arabidopsis by targeting the GA 3-oxidase gene AtGA3ox1 [17]. In addition, AHLs participated in the biotic and abiotic stress responses. AHL13, AHL15, AHL19, AHL20, and AHL27 were reported to implicate in PAMP-triggered immunity [18][19][20]. Overexpression of AHL20 and its closely related family members (AHL15, AHL19, and AHL27) in protoplasts inhibited the innate immune response [19]. Especially in Arabidopsis, AHL13 regulated key factors in the jasmonic acid (JA) pathway and affected immunity toward Botrytis cinerea and Pseudomonas syringae pathogens [20]. AHL10 played a critical in response to low-water potential stress, and in the regulation of auxin-and jasmonic acid -related gene expression [21]. AtAHL4 was involved in the regulation of lipid catabolism during seedling establishment [22] and of root xylem development [4]. Interactions between AHL proteins are also crucial for their functions. In Arabidopsis roots, the AHL3 physically interacted with AHL4 and facilitated intercellular trafficking [23]. In maize (Zea mays), the AT-hook protein BAF1 is capable of forming homodimers and heterodimers with other members of the AT-hook family, controlling the formation of maize ears [24].
In recent years, the AHL gene family has been identified and analyzed in several species, such as Arabidopsis, soybean (Glycine max) [25], cotton (Gossypium raimondii) [26], Brassica napus [27], grape (Vitis vinifera) [28], carrot (Daucus carota subsp. sativus) [29], and so on. Walnut (Juglans regia L.) is an economically vital nut trees in China and other temperate parts of the world, but the long period before flowering and bearing fruiting after seed germination limits the progress of walnut breeding work. In addition, functional characterization of AHL genes in walnut is still missing. In this study, the AHL gene family was characterized in walnut based on the high-quality genome data [30]. The chromosomal locations, phylogenetic relationships, gene expansions and structures, conserved domains, putative protein interaction, and cis-elements in the promoter region were analyzed. Furthermore, the JrAHL2, a Clade-A member highly expressed in flower, was isolated, its protein subcellular localization was analyzed, and its functions involved in floral transition and hypocotyl elongation were dissected. The current study will provide a theoretical basis for further investigation of the functions of the JrAHL gene family, and an improved understanding of the molecular mechanisms underlying its role in regulating growth and development in walnut.

AHL Genes in Walnut
The HMMscan was executed against the protein databases using the PPC/DUF296 (PF03479) and AT-hook as queries. After manual checking, the 37 unique AHL genes were obtained from walnut (Juglans regia L.). The JrAHL genes were named according to their chromosomal locations (JrAHL1-JrAHL37) ( Table 1). The gene lengths of JrAHL genes ranged from 667 bp to 7791 bp, which encode polypeptides from 221 to 406 amino acids with predicted molecular weights ranging from 20.04 kDa to 42.63 kDa. The theoretical pI ranged from 5.08 to 10.61. Approximately 92% of JrAHL proteins (except for JrAHL10/11/33) were predicted to be located in the nucleus (Supplementary Table S2), consistent with the previous report [1].

Phylogenetic Analysis of JrAHL Genes
To infer the evolutionary relationship among the JrAHL proteins, phylogenetic analysis was performed on the full-length protein sequences. The result showed that the JrAHL proteins in walnut could be divided into two clades, as in Arabidopsis ( Figure 1)-Clade-A and Clade-B. JrAHL were further classified into Type-I, Type-II, and Type-III (Figure 2 left panel), based on the presence and combination of the two characteristic functional units-the PPC domain and the AT-hook motif (Figure 2 right panel). Clade-A contained 20 JrAHLs, and it was also classified into Type-I, harboring the Clade-A PPC domain and the Type-I AT-hook motifs in their putative protein sequence (Figure 2 right panel). The higher abundance of Type I members (54%) in walnut is also consistent with observations in other land plants [25,27,31]. There were 17 JrAHL members in Clade-B, which all harbored the Clade-B PPC domain. Clade-B was further divided into two types: 11 Type-II members and 6 Type-III members. Type-II members contained Type-I and Type-II AT-hook motifs, whereas the Type-III JrAHLs only contained the Type-II AT-hook motif (Figure 2

Phylogenetic Analysis of JrAHL Genes
To infer the evolutionary relationship among the JrAHL proteins, phylogenetic analysis was performed on the full-length protein sequences. The result showed that the JrAHL proteins in walnut could be divided into two clades, as in Arabidopsis ( Figure 1)-Clade-A and Clade-B. JrAHL were further classified into Type-I, Type-II, and Type-III ( Figure 2 left panel), based on the presence and combination of the two characteristic functional units-the PPC domain and the AT-hook motif (Figure 2 right panel). Clade-A contained 20 JrAHLs, and it was also classified into Type-I, harboring the Clade-A PPC domain and the Type-I AT-hook motifs in their putative protein sequence (Figure 2 right panel). The higher abundance of Type I members (54%) in walnut is also consistent with observations in other land plants [25,27,31]. There were 17 JrAHL members in Clade-B, which all harbored the Clade-B PPC domain. Clade-B was further divided into two types: 11 Type-II members and 6 Type-III members. Type-II members contained Type-I and Type-II AT-hook motifs, whereas the Type-III JrAHLs only contained the Type-II AThook motif (Figure 2    Neighbor-joining tree representing phylogenetic relationships among AHL genes from walnut and Arabidopsis. The evolutionary history was inferred using the Neighbor-Joining matrixbased method and are in the units of the number of amino acid substitutions per site. The rate variation among sites was modeled with a gamma distribution. Evolutionary analyses were conducted in MEGA7.

Gene Structure and Duplication of JrAHLs
The distribution of introns and exons was investigated to explore the diversity of gene structure. The structures of JrAHL genes could be divided into two types, one contained no or only one intron, and the other contained multiple introns ( Figure 3 left panel). All Type-II and Type-III JrAHLs contained four to five introns, while the JrAHLs in Type-I were intronless except for JrAHL21 and JrAHL26. Furthermore, the number and arrangement of introns are relatively conserved within the same clade, implicating the evolutionary similarity between these members. A total of seven conserved motifs in which two motifs (motif 1 and motif 2) have highly conserved 'R-G-R-P' amino acid sequences ( Gene duplication leads to the expansion of AHL family through segmental and tandem duplications [1]. We speculated the exact mechanism has driven the expansion of the AHL gene family in walnut. We analyzed gene duplication events in the JrAHL gene family. As shown in Figure 4, a total of 37 JrAHL genes were unevenly mapped onto 16 chromosomes of walnut. Chromosomes 00 and 14 did not contain duplicated genes, whereas Chromosome 12 had the highest number of duplications. More than twenty pairs of JrAHL genes, such as JrAHL1/6, JrAHL4/9, JrAHL24/34, and JrAHL31/25, were located in duplicated genomic regions. Duplicated pairs showed similar motifs (Figure 2), patterns of

Gene Structure and Duplication of JrAHLs
The distribution of introns and exons was investigated to explore the diversity of gene structure. The structures of JrAHL genes could be divided into two types, one contained no or only one intron, and the other contained multiple introns ( Figure 3 left panel). All Type-II and Type-III JrAHLs contained four to five introns, while the JrAHLs in Type-I were intronless except for JrAHL21 and JrAHL26. Furthermore, the number and arrangement of introns are relatively conserved within the same clade, implicating the evolutionary similarity between these members. A total of seven conserved motifs in which two motifs (motif 1 and motif 2) have highly conserved 'R-G-R-P' amino acid sequences ( Figure 3 right panel), which are the typical characteristic sequences of the AT-hook motif family.
Gene duplication leads to the expansion of AHL family through segmental and tandem duplications [1]. We speculated the exact mechanism has driven the expansion of the AHL gene family in walnut. We analyzed gene duplication events in the JrAHL gene family. As shown in Figure 4, a total of 37 JrAHL genes were unevenly mapped onto 16 chromosomes of walnut. Chromosomes 00 and 14 did not contain duplicated genes, whereas Chromosome 12 had the highest number of duplications. More than twenty pairs of JrAHL genes, such as JrAHL1/6, JrAHL4/9, JrAHL24/34, and JrAHL31/25, were located in duplicated genomic regions. Duplicated pairs showed similar motifs (  . The x-axis shows the inferred length of the different genes (5′ to 3′) and their respective CDS (red). Domain structure and motif were elucidated with the MEME platform (right).

Cis-Acting Element Analysis of JrAHLs Promoter
The promoter is critical for transcriptional regulation because of its responsiveness to specific hormone levels and environmental cues and plays a vital role in the biological regulation of gene expression under stress. We investigated the possible regulatory mechanisms by searching the promoter region of 1500 bp upstream of the JrAHL gene family.

Cis-Acting Element Analysis of JrAHLs Promoter
The promoter is critical for transcriptional regulation because of its responsiveness to specific hormone levels and environmental cues and plays a vital role in the biological regulation of gene expression under stress. We investigated the possible regulatory mechanisms by searching the promoter region of 1500 bp upstream of the JrAHL gene family. Four categories of cis-elements were identified related to hormone response, stress response, tissue-specific expression, and transcription factor binding ( Figure 5). Hormone response elements, such as gibberellin response element (GARE-motif), salicylic acid response element (TCA-element), abscisic acid-responsive element (ABRE), auxin-responsive (AUXRR-core), and MeJA-responsiveness (CGTCA-motif) were abundant in JrAHL promoter regions. Stress response elements, including defense responsive element (TC-rich repeat), lowtemperature response (LTR), and drought response element (MBS), were discovered in most of the promoter regions of JrAHL genes. As for cis-acting regulatory elements related to tissue-specific expression, endosperm-and meristem-expression elements were found in 4 and 14 JrAHL promoter regions, respectively ( Figure 5 and Figure S2). Four JrAHL genes contained the AT-rich element, which provided the binding site of ATBP-1, a class of AT-rich DNA binding proteins. These results imply that JrAHLs are involved in many biological processes.

Protein-Protein Interaction Analysis JrAHLs
Members of the AHL family are partially involved in various plant biological processes via protein-protein interaction [23]. We constructed the protein interaction networks between the JrAHL proteins. Consistent with the previous report [16,24], the JrAHLs are shown to interact with each other ( Figure 6). Additionally, JrAHL3/JrAHL8/JrAHL11/ JrAHL13/JrAHL27/JrAHL33 were predicted in the networks, which were clustered with the physically-interacting proteins AHL3/AHL4 from Arabidopsis [23]. In Poncirus trifoliata, both PtrAHL14 and PtrAHL17 could interact with other proteins to target the same downstream genes [32]. However, the interaction between JrAHL proteins must be further verified in vitro and in vivo.

Protein-Protein Interaction Analysis JrAHLs
Members of the AHL family are partially involved in various plant biological processes via protein-protein interaction [23]. We constructed the protein interaction networks between the JrAHL proteins. Consistent with the previous report [16,24], the JrAHLs are shown to interact with each other (Figure 6). Additionally, JrAHL3/JrAHL8/JrAHL11/JrAHL13/JrAHL27/JrAHL33 were predicted in the networks, which were clustered with the physically-interacting proteins AHL3/AHL4 from Arabidopsis [23]. In Poncirus trifoliata, both PtrAHL14 and PtrAHL17 could interact with other proteins to target the same downstream genes [32]. However, the interaction between JrAHL proteins must be further verified in vitro and in vivo.

Gene Expression Patterns of JrAHLs
The JrAHLs expression profiles of F26 (anthracnose-resistant) and F423 (anthracnosesusceptible) fruit bracts at five-time points after infection were analyzed based on the transcriptomic data. In the anthracnose-resistant F26, more than half of the JrAHLs gene members were up-regulated within 24 hours after inoculation. However, the up-regulation of these JrAHL genes was delayed in the susceptible F423 (Figure 7). This result suggested that JrAHLs are involved in the immune regulation of pathogens. Expression profiles of JrAHLs at different developmental stages of endopleura, one of the walnut's most commercially essential organs, were analyzed. More than half of JrAHL genes showed high transcription levels in the early endopleura development stage (Figure 8). The tissue-specific expression profile of some JrAHL genes from different clades was determined by realtime quantitative PCR (Figure 9). JrAHL1, JrAHL2, JrAHL7, JrAHL19, and JrAHL28 were highly expressed in the shoot tip. JrAHL1, JrAHL2, and JrAHL7 also showed high expression levels in pistillate flowers. Two genes were preferentially expressed in the vegetative organs, such as JrAHL10 in root and JrAHL33 in leaf. The JrAHL11 is highly expressed in fruit. In addition, JrAHL14 and JrAHL31 showed no tissue-specific expression.

Gene Expression Patterns of JrAHLs
The JrAHLs expression profiles of F26 (anthracnose-resistant) and F423 (anthracnosesusceptible) fruit bracts at five-time points after infection were analyzed based on the transcriptomic data. In the anthracnose-resistant F26, more than half of the JrAHLs gene members were up-regulated within 24 hours after inoculation. However, the up-regulation of these JrAHL genes was delayed in the susceptible F423 (Figure 7). This result suggested that JrAHLs are involved in the immune regulation of pathogens. Expression profiles of JrAHLs at different developmental stages of endopleura, one of the walnut's most commercially essential organs, were analyzed. More than half of JrAHL genes showed high transcription levels in the early endopleura development stage (Figure 8). The tissuespecific expression profile of some JrAHL genes from different clades was determined by real-time quantitative PCR (Figure 9). JrAHL1, JrAHL2, JrAHL7, JrAHL19, and JrAHL28 were highly expressed in the shoot tip. JrAHL1, JrAHL2, and JrAHL7 also showed high expression levels in pistillate flowers. Two genes were preferentially expressed in the vegetative organs, such as JrAHL10 in root and JrAHL33 in leaf. The JrAHL11 is highly expressed in fruit. In addition, JrAHL14 and JrAHL31 showed no tissue-specific expression.

Gene Cloning and Subcellular Localization of JrAHL2
JrAHL2 was cloned from walnut and found to contain a 927-bp CDS ( Figure 10A) that encodes a protein of 307 amino acids. Multiple sequence alignment analysis results showed that the JrAHL2 shared 59.75% sequence identity with AtAHL22. The amino acid sequence within the AT-hook motif and the large PPC/DUF296 domain are highly conserved ( Figure 10B). JrAHL2 protein is predicted to be localized in the nucleus (Supplementary Table S2). JrAHL2 was fused to a green fluorescent protein (EGFP) ( Figure 10C) and transiently expressed in tobacco leaf cells. The GFP signal was targeted to the nucleus ( Figure 10D), suggesting that JrAHL2 functions as a transcription factor.

Functional Analysis of JrAHL2 in Transgenic Arabidopsis
To dissect the function of JrAHL2, we generated JrAHL2-overexpressing (OE) transgenic Arabidopsis plants. The homozygous JrAHL2-OE plants flowered significantly later than the WT controls ( Figure 11A,B). Additionally, the flowering genes AtFT is expressed at a lower level in the JrAHL2-OE Arabidopsis plants ( Figure 11C). Overexpression of JrAHL2 also leads to a dominant-negative long hypocotyl phenotype under both long day and dark conditions ( Figure 12). Thus, JrAHL2 inhibits floral transition and hypocotyl elongation.

Gene Cloning and Subcellular Localization of JrAHL2
JrAHL2 was cloned from walnut and found to contain a 927-bp CDS ( Figure 10A) that encodes a protein of 307 amino acids. Multiple sequence alignment analysis results showed that the JrAHL2 shared 59.75% sequence identity with AtAHL22. The amino acid sequence within the AT-hook motif and the large PPC/DUF296 domain are highly conserved ( Figure 10B). JrAHL2 protein is predicted to be localized in the nucleus (Supplementary Table S2). JrAHL2 was fused to a green fluorescent protein (EGFP) ( Figure 10C) and transiently expressed in tobacco leaf cells. The GFP signal was targeted to the nucleus ( Figure 10D), suggesting that JrAHL2 functions as a transcription factor.

Gene Cloning and Subcellular Localization of JrAHL2
JrAHL2 was cloned from walnut and found to contain a 927-bp CDS ( Figure 10A) that encodes a protein of 307 amino acids. Multiple sequence alignment analysis results showed that the JrAHL2 shared 59.75% sequence identity with AtAHL22. The amino acid sequence within the AT-hook motif and the large PPC/DUF296 domain are highly conserved ( Figure 10B). JrAHL2 protein is predicted to be localized in the nucleus (Supplementary Table S2). JrAHL2 was fused to a green fluorescent protein (EGFP) ( Figure 10C) and transiently expressed in tobacco leaf cells. The GFP signal was targeted to the nucleus ( Figure 10D), suggesting that JrAHL2 functions as a transcription factor.  To dissect the function of JrAHL2, we generated JrAHL2-overexpressing (OE) trans-genic Arabidopsis plants. The homozygous JrAHL2-OE plants flowered significantly later than the WT controls ( Figure 11A,B). Additionally, the flowering genes AtFT is expressed at a lower level in the JrAHL2-OE Arabidopsis plants ( Figure 11C). Overexpression of JrAHL2 also leads to a dominant-negative long hypocotyl phenotype under both long day and dark conditions ( Figure 12). Thus, JrAHL2 inhibits floral transition and hypocotyl elongation.

Identification and Analysis of JrAHL Gene Family Members
The walnut (Juglans regia L.) is one of the most economically essential nut trees, and the research on walnut focused on flowering and molecular breeding in recent years. However, less attention was paid to walnut than other plants because of the problems in the walnut industry development, such as the long juvenile phase, difficulty in flowering,

Identification and Analysis of JrAHL Gene Family Members
The walnut (Juglans regia L.) is one of the most economically essential nut trees, and the research on walnut focused on flowering and molecular breeding in recent years. However, less attention was paid to walnut than other plants because of the problems in the walnut industry development, such as the long juvenile phase, difficulty in flowering, and hypocotyl elongation through the husk. The high-quality walnut genome has been published [30], providing a reference for systematic research on the genetic composition and potential function. To date, some gene families from walnut have been identified, such as the Basic leucine zipper (bZIP) [33] and General Regulatory Factor (GRF) [34] genes, which may be involved in the regulation of flowering. However, the role of the AT-Hook Motif Containing Nuclear Localized (AHL) gene family has not been explored in higher plants beyond herbaceous plants. In this work, 37 JrAHL members were identified in the walnut genome (Table 1). More member was identified in walnut compared to Arabidopsis (29 members) [6], probably because the walnut genome (540 Mb) is larger than that of Arabidopsis (125 Mb). The JrAHL members were divided into two Clades, based on the phylogenetic tree ( Figure 1) and PPC/DUF domain ( Figure 2) as reported in other species [1]. AHL proteins interact with each other through their PPC domain [35]. Thus, the variability increase in PPC domains may extend the range of biological functions. The Clade-B is further divided into two types based on the AT-hook motif, Type-II and Type-III. At the same time, the Clade-A is referred to as Type-I (Figure 2). Because the AT-hook motif is responsible for recognizing and binding a specific nucleotide sequence as well as changing the chromatin structure [36], it is, of course, possible that JrAHLs from the different branch can target different downstream genes. The sequences of conserved motifs (Figure 2), the gene structure (Figure 3), and the arrangement of distinct domains ( Figure 3) and are highly consistent within each evolutionary branch. High consistency may be caused by segmental duplications (Figure 4) or the whole-genome duplication process [27], which are the major driving forces of AHL family evolution. Similarly, domain-swapping analyses showed that the N-terminus of AHL4 is required for its DNA-binding activity [4]. All JrAHL genes in Clade-B contain multiple exons, whereas all the Clade-A members are intronless ( Figure 3). Introns are ubiquitous in eukaryotic genomes, which affect gene function by affecting gene expression regulation [37]. For example, the number of introns in soybeans certainly affects the expression of AHL genes [25]. In addition to regulating the rate of gene evolution [38], introns also provide specific functional elements for selective splicing and exon shuffling, and some functions of non-coding DNA [39], and in so doing allow the occurrence of diversity in JrAHLs protein accumulation, structures (Supplementary Figure S1), and physically interactions ( Figure 6). Because the eukaryote genomes gained introns during or after prokaryote-eukaryote divergence [13], an early divergence occurred between the Type-I JrAHLs and the other two types (II and III) of JrAHL genes, and the JrAHLs of Clade-B in the walnut genome have diverged from Clade-A. Gene duplication, one of the most important features of plant genome structure, contributed to the evolution of novel functions [40] (Panchy et al., 2016). The ratio of non-synonymous to synonymous mutations in genes (Ka/Ks) of duplicated gene pairs was < 1, suggesting that purifying selection was the main source [40] of evolution for the JrAHL gene family. It is thus plausible that introns acquisition may cause the structural and functional changes of JrAHLs during duplication (Figures 3 and 4).
In addition to introns, the promoter, another important non-coding region, plays a critical role in regulating gene transcription. The cis-acting elements of some duplicated gene pairs are not the same, such as JrAHL1/6 and JrAHL4/9. These differences would allow the regulation of different aspects of plant development and the responses to hormones and stresses. Studies of promoters, especially cis-acting elements (CREs) are crucial for improving our fundamental understanding of gene regulation and function towards different environmental cues and stress. The CREs that regulate plant development and confer resistance stresses are found in the JrAHL promoters (Figures 5 and S2). Stress-related hormone response elements, including ABA, MeJA, and SA, are widely distributed on multiple JrAHL promoters. Fifteen JrAHLs contained GARE-motif, especially JrAHL6, contains up to four of these elements, suggesting a cross-talk between JrAHLs and GA hormones, such as negative feedback regulation as previously reported [17]. Aux-responsive CREs are predicted in ten JrAHL promoters. Specifically, JrAHL28 is homologous to AHL10 in Arabidopsis, which regulates genes related to the auxin signaling pathway, suggesting a cross-talk between JrAHLs and auxin [8,21]. AHL also negatively regulates jasmonic acid (JA) levels and affects the gene expression in the JA pathway [20]. Notably, two AT-Hook proteins, PtrAHL14/17 in Poncirus trifoliata, were reported to modulate cold tolerance [32]. Adversity stress-related CREs, including low temperature, drought, and defense, suggest that JrAHLs are relevant to controlling these processes. Four JrAHL genes contained the AT-rich element, which provided the binding site of AT-rich DNA binding protein (ATBP-1) [41]. It is plausible that there is reciprocal transcriptional regulation among the JrAHLs family members, in addition to the predicted protein-protein interactions ( Figure 6). In addition, tissue-specific elements such as endospermand meristem-expression ( Figures 5 and S2) suggested that JrAHLs regulate embryogenesis and meristem maturation like previously reported [35,42]. Specifically, JrAHL17 contains both endosperm-and meristem-expression CREs and is homologous to AtAHL10 from Arabidopsis, which affects the expression of the transcription factor STM (Shootmeristemless) required for meristem maintenance [21]. qRT-PCR revealed that the JrAHL genes containing meristemexpression tissue-specific elements in the promoter were highly expressed in the shoot tip ( Figures 5 and 9). Transcriptome data facilitate a comprehensive understanding of expression levels of multiple genes. Multiple JrAHL genes showed differential expression by analyzing the transcriptome data of two walnut materials after inoculation with Colletotrichum gloeosporioides. In response to C. gloeosporioides infection, more than half of the JrAHL genes were rapidly activated in F26 at the initial stage of inoculation. At the same time, the activation of JrAHL in F234 was relatively delayed, suggesting that JrAHL is involved in the immune regulation of pathogens and that early expression may confer high resistance to anthracnose in F26 fruits (Figure 7). In fact, there is already established evidence that AHL is involved in plant immunity and positively regulates plant resistance to pathogen infection [20,43]. More than half of the JrAHL genes are highly expressed at the beginning of endocarp development (Figure 8), including JrAHL11, which showed a profound expression in walnut fruit (Figure 9). JrAHL11 was clustered into the same clade with the Arabidopsis AHL4, and the latter is involved in the regulation of lipid mobilization and fatty acid β-oxidation [22], suggesting that the JrAHL11 plays an important role in fruit development. The three detected Type I JrAHL genes, JrAHL1, JrAHL2, and JrAHL7, were highly expressed in pistillate flowers and shoot tip ( Figure 9). JrAHL1 was homologous to the AHL20, JrAHL2 and JrAHL7 were homologous to the AHL22 (Figure 1) in Arabidopsis, respectively. Both AHL20 and AHL22 were involved in controlling flowering time in Arabidopsis [7,15,44], suggesting that these three genes are involved in floral regulation.

JrAHL2 Delayed Flowering and Inhibited Hypocotyl Elongation
Hypocotyl elongation is an essential event in the photomorphogenesis of walnut seeds, and flowering is an important trait for economic forest species. The functions and mechanisms of AHL genes from Arabidopsis have been extensively studied, whereas the function dissection information about the specific JrAHL gene in walnut is still unavailable. The JrAHL2 were highly expressed in the pistillate flower and was homologous to the Arabidopsis AHL22, which stimulates us to investigate the function of JrAHL2 in controlling flowering and hypocotyl elongation. The transgenic Arabidopsis with constitutive expression of the JrAHL2 gene was prepared. Overexpression of JrAHL2 caused a significant inhibition of hypocotyl elongation, whereas the Arabidopsis ahl22 (homologous to JrAHL2) mutant is longer in the hypocotyl length than wide type ( Figure 10) under both long-day conditions and darkness. It reported that AHL22, SOB3/AHL29, ESC/AHL27 [6][7][8] are negative modulators of hypocotyl growth in Arabidopsis [9]. This result indicates that JrAHL2 plays a negative role in controlling hypocotyl elongation and functions in both photomorphogensis and skotomorphogenesis. In addition to repressing the hypocotyl growth, JrAHL2 also inhibits the transcription of FT and delays flowering (Figure 9). Nevertheless, no perturbations to flower phenotype and FT expression were recorded in the ahl22 mutant (Figure 9), possibly because of extensive functional redundancy between AtAHL22 and other AHL genes [15]. These results suggested that the JrAHL2 has a much stronger effect on hypocotyl elongation than its effect on flowering regulation (Xiao et al., 2009). Additionally, despite the report that AHL regulates the expression of target genes, including FT, by directly binding to a specific nuclear MAR containing A/T-rich sequences within their promoter [15,32,44], the ubiquity of A/T-rich sequences in the promoter region makes it difficult to discriminate the elements specifically bound by JrAHL2. Thus, a systems approach such as ChIP-seq/qPCR could be used further to dissect the network of JrAHL.

Identification of the AHL Gene Family
For the discovery of the AHL family members in walnut, the high-quality common walnut (Juglans regia L.) reference genome data was downloaded [30]. The Hidden Markov Model scan (HMMscan) was used on the complete walnut protein database using PPC/DUF296 (PF03479) as well as AT-hook (PF02178) Pfam IDs obtained from the Pfam database (https://pfam.xfam.org/ (accessed on 7 January 2022)) [45] as queries. All nonredundant putative protein sequences were manually checked with the NCBI Conserved Domains Database (https://www.ncbi.nlm.nih.gov/Structure/cdd/wrpsb.cgi (accessed on 12 January 2022)). The unique sequences containing the AT-hook motif and the PPC domain were selected as putative JrAHLs. The online ExPASy program (http://www.expasy.org/ tools/ (accessed on 12 February 2022)) was used to predict protein physicochemical characteristics. The online system Cell-PLoc (http://www.csbio.sjtu.edu.cn/bioinf/Cell-PLoc-2/ (accessed on 15 February 2022)) [46] was used to predict the subcellular localization of these JrAHL proteins.

Phylogenetic Analysis of AHL Proteins
A Neighbor-Joining tree was constructed to infer the phylogenetic relationship between the AHL proteins from Arabidopsis and walnut. Sequence alignment was carried out using ClustalW in MEGA7 [47] with default parameters. The 1000 Bootstrap replicates were applied to present the evolutionary history. Multiple alignments were also performed with DNAMAN software (LynnonBiosoft, San Ramon, CA, USA. version 8.0).

Collinearity Analysis
The OrthoMCL algorithm [48] was used to identify paralogous genes within the walnut genome, and then the MCScan [49] algorithm was applied to detect syntenic blocks containing walnut AHL. The Circos [50] was used to visualize the syntenic relationships between the walnut genomes. The Ka/Ks ratio was calculated to estimate the selection pressure among duplicated JrAHL genes as previously described [51]. All plugins are used in TB tool software (TBtools-II v1.108) [52].

Promoter Sequence Analysis
The 1.5-kb upstream region of the start codon of each JrAHL gene was retrieved. The online PlantCARE database (https://bioinformatics.psb.ugent.be/webtools/plantcare/ html/ (accessed on 16 July 2022)) was used to identify the potential cis-acting elements.

Interactive Protein Network Analysis
The STRING database predicted the interaction network of AHL proteins (https:// string-db.org/ (accessed on 25 July 2022)). All JrAHL proteins were used as the query to search for the walnut data in the STRING database. Both the functional and physical protein associations were included in the network. The minimum required interaction score was selected as medium confidence (0.400).

Expression Profiles of JrAHL Genes
Gene expression data from the anthracnose-resistant F26 and the anthracnose-susceptible F423 fruits in response to anthracnose (Colletotrichum gloeosporioides) infection were downloaded from the GEO database (https://www.ncbi.nlm.nih.gov/geo/ (accessed on 20 July 2022)) at accession GSE147083 [53]. Transcriptomic data of the walnut endopleura during its development was also downloaded from GEO database at accession GSE185230 [54]. The log2 transformed FPKM (Fragments Per Kilobase of exon model per Million mapped fragments) values were used for preparing the heatmaps. Root, stem, leaf, flowers, fruits, and shoot tip were taken, frozen immediately in liquid nitrogen, and stored at −80 • C until use.

Quantitative RT-PCR (qRT-PCR)
Total RNA was extracted from different tissues using the RNA extraction kit (OMEGA, Doraville, GA, USA). The RNA concentration was determined by spectrophotometer, and RNA integrity was verified by electrophoresis. First-strand cDNA was synthesized using the EasyScript ® One-Step cDNA Synthesis SuperMix (TransGen Biotech, China) following the manufacturer's instructions. Primer pairs for qRT-PCR were designed using Primer Premier 6.0 (Premier Biosoft, San Francisco, CA, USA) (Supplementary Table S1). qRT-PCR was performed on a Bio-Rad CFX96 thermal cycler (Bio-Rad Laboratories, Hercules, CA, USA). The 18S rRNA gene was used as a standard internal control. The relative expression levels were calculated by the comparative 2 −∆∆Ct method [55].

Gene Cloning, Vector Construction, and Subcellular Localization
The coding sequence (CDS) of the JrAHL2 was cloned from the walnut pistillate flower's cDNA and then inserted into the modified pCAMBIA2300-EGFP vector. Subcellular localization was carried out according to our previously published article [56]. The leaves of 5-week-old tobacco plants were infiltrated with the Agrobacterium strain GV3103 cells containing the recombinant vector. After 3 d, the green fluorescent protein signals were detected with an Olympus fluorescence microscope (Japan). All primers used for gene cloning and vector construction were listed in Supplementary Table S1.

Plant Transformation and Phenotype Analysis
Arabidopsis was transformed with the Agrobacterium strain GV3103 mentioned above, according to the A. tumefaciens mediated floral dip method [57]. The transformants were identified based on kanamycin resistance, and the T 3 generation seedlings were used for functional analyses. The flowering phenotype of different Arabidopsis lines, including JrAHL2overexpressing (JrAHL2-OE), wild type (Col-0), and AtAHL22 mutant (ahl22-1, Salk_018866), were identified according to the previous description [58]. To analyze the hypocotyl growth phenotype, Arabidopsis seeds were surface-sterilized and sown on Murashige & Skoog (MS) medium [56]. The plate was held at 4 • C for 3 days, exposed to white light for 4 h to stimulate germination, and then cultured for 6 days at 22 • C with a 16-h light/8-h dark photoperiod (light intensity of 150 µmol m −2 s −1 ) (long day) or darkness.

Conclusions
Thirty-seven AHL genes were identified in the walnut (Juglans regia L.) genome and found to be unequally distributed on 16 chromosomes. The phylogenetic tree divided JrAHLs into two clades. Segmental duplications were the major driving forces of AHL gene family expansion and evolution. Promoter and expression analysis indicated that JrAHLs were involved in multiple biological processes. Functional analysis showed that the nuclear-localized JrAHL2 plays a negative role in floral transition and hypocotyl elongation. The first systematic exploration of the AHL gene family in walnut will provide helpful information for future work in walnut breeding.