Genome-Wide Analysis of the BBX Genes in Platanus × acerifolia and Their Relationship with Flowering and/or Dormancy

The B-BOX (BBX) gene family is widely distributed in animals and plants and is involved in the regulation of their growth and development. In plants, BBX genes play important roles in hormone signaling, biotic and abiotic stress, light-regulated photomorphogenesis, flowering, shade response, and pigment accumulation. However, there has been no systematic analysis of the BBX family in Platanus × acerifolia. In this study, we identified 39 BBX genes from the P. × acerifolia genome, and used TBtools, MEGA, MEME, NCBI CCD, PLANTCARE and other tools for gene collinearity analysis, phylogenetic analysis, gene structure, conserved domain analysis, and promoter cis-element analysis, and used the qRT-PCR and transcriptome data for analyzing expression pattern of the PaBBX genes. Collinearity analysis indicated segmental duplication was the main driver of the BBX family in P. × acerifolia, and phylogenetic analysis showed that the PaBBX family was divided into five subfamilies: I, II, III, IV and V. Gene structure analysis showed that some PaBBX genes contained super-long introns that may regulate their own expression. Moreover, the promoter of PaBBX genes contained a significant number of cis-acting elements that are associated with plant growth and development, as well as hormone and stress responses. The qRT-PCR results and transcriptome data indicated that certain PaBBX genes exhibited tissue-specific and stage-specific expression patterns, suggesting that these genes may have distinct regulatory roles in P. × acerifolia growth and development. In addition, some PaBBX genes were regularly expressed during the annual growth of P. × acerifolia, corresponding to different stages of flower transition, dormancy, and bud break, indicating that these genes may be involved in the regulation of flowering and/or dormancy of P. × acerifolia. This article provided new ideas for the study of dormancy regulation and annual growth patterns in perennial deciduous plants.


Introduction
The B-BOX (BBX) proteins exist in unicellular and multicellular eukaryotes, named for the B-BOX domain, and belong to the family of zinc finger proteins (ZFP) [1]. The BBX protein family is divided into five subfamilies based on structural features and conserved sequences. Among them, BBX I subfamily (BBX1-6) and BBX II subfamily (BBX7-BBX13) both contain two distinct B-BOX domains and a CONSTANS, CO-like, and TOC1 (CCT) domain; BBX III subfamily (BBX14-17) contains one B-BOX domain and a CCT domain, while BBX IV subfamily (BBX18-BBX25) contains two B-BOX domains without a CCT domain, and BBX V subfamily (BBX26-BBX32) contains only one B-BOX domain without a CCT domain [2].

Phylogenetic Analysis and Multiple Alignments of BBX Proteins
To elucidate the evolutionary and phylogenetic relationships between the PaBBX proteins and other BBX proteins, we created two unrooted phylogenetic trees using MEGA X software with the neighbor-joining (NJ) method. One phylogenetic tree was constructed using the conserved domain sequences of 39 PaBBX proteins together with 32 A. thaliana BBX proteins, 30 O. sativa BBX proteins, 18 Liriodendron chinense BBX proteins, 32 Nelumbo nucifera BBX proteins, and 25 Vitis vinifera BBX proteins ( Figure 2). Another phylogenetic tree was performed with a full-length sequence of 39 PaBBX proteins ( Figure S1). Both phylogenetic trees showed that PaBBX proteins were divided into five groups. BBX Ⅰ subfamily (PaBBX1-PaBBX5) and BBX Ⅴ subfamily (PaBBX28/29-PaBBX32) both contained six gene sequences. BBX Ⅱ subfamily (PaBBX7-PaBBX13) and BBX IV subfamily (PaBBX19-PaBBX25) both contained twelve gene sequences, and BBX III subfamily only contained three gene sequences.

Phylogenetic Analysis and Multiple Alignments of BBX Proteins
To elucidate the evolutionary and phylogenetic relationships between the PaBBX proteins and other BBX proteins, we created two unrooted phylogenetic trees using MEGA X software with the neighbor-joining (NJ) method. One phylogenetic tree was constructed using the conserved domain sequences of 39 PaBBX proteins together with 32 A. thaliana BBX proteins, 30 O. sativa BBX proteins, 18 Liriodendron chinense BBX proteins, 32 Nelumbo nucifera BBX proteins, and 25 Vitis vinifera BBX proteins ( Figure 2). Another phylogenetic tree was performed with a full-length sequence of 39 PaBBX proteins ( Figure S1). Both phylogenetic trees showed that PaBBX proteins were divided into five groups. BBX I subfamily (PaBBX1-PaBBX5) and BBX V subfamily (PaBBX28/29-PaBBX32) both contained six gene sequences. BBX II subfamily (PaBBX7-PaBBX13) and BBX IV subfamily (PaBBX19-PaBBX25) both contained twelve gene sequences, and BBX III subfamily only contained three gene sequences.  To analyze the conserved domain sequence of PaBBXs protein, we utilized the MEGA X software with Clustal W program to perform multiple sequence alignment of the PaBBXs protein ( Figure 3). The results revealed that these proteins contained the highly conserved B-BOX1 with C-D-X-C-X8-CX2-D-X-A-X-L-C-X2-C-D-X3-H-X2-N-X5-H-X-R-X2-L sequence ( Figure S2a) and a less conserved B-BOX2 with C-X2-C-X4-A-X2-C6-L-C-X2-C-D-X3-H-X6-H-X-R-X3 sequence ( Figure S2b). In addition, the CCT domain was highly conserved ( Figure S2c). It is worth noting that PaBBX11-3a/b belonged to BBX Ⅱ subfamily but lacked B-BOX1 and CCT domains, and PaBBX19-4 belonged to BBX IV subfamily but lacked the B-BOX1 domain. Moreover, there were some amino acid changes in the consensus sequence of B-BOX2. Cys-1 and His-33 of PaBBX11-3a/b were changed to Tyr and Glu, respectively, and Cys-13 of PaBBX9-2 was changed to Tyr ( Figure 3a). To analyze the conserved domain sequence of PaBBXs protein, we utilized the MEGA X software with Clustal W program to perform multiple sequence alignment of the PaBBXs protein ( Figure 3). The results revealed that these proteins contained the highly conserved B-BOX1 with C-D-X-C-X8-CX2-D-X-A-X-L-C-X2-C-D-X3-H-X2-N-X5-H-X-R-X2-L sequence ( Figure S2a) and a less conserved B-BOX2 with C-X2-C-X4-A-X2-C6-L-C-X2-C-D-X3-H-X6-H-X-R-X3 sequence ( Figure S2b). In addition, the CCT domain was highly conserved ( Figure S2c). It is worth noting that PaBBX11-3a/b belonged to BBX II subfamily but lacked B-BOX1 and CCT domains, and PaBBX19-4 belonged to BBX IV subfamily but lacked the B-BOX1 domain. Moreover, there were some amino acid changes in the consensus sequence of B-BOX2. Cys-1 and His-33 of PaBBX11-3a/b were changed to Tyr and Glu, respectively, and Cys-13 of PaBBX9-2 was changed to Tyr ( Figure 3a).
In addition to the BOX and CCT domains, the alignments of the conserved sequences of the NLS domains, VP domains, and M1-M7 domains in PaBBX proteins were also presented. The NLS domain was present in BBX I-III subfamilies except for PaBBX11-3a/b because the NLS domain partially overlaps with the CCT domain. Moreover, some BBX IV and V members also had NLS domains ( Figure S3a In addition to the BOX and CCT domains, the alignments of the conserved sequences of the NLS domains, VP domains, and M1-M7 domains in PaBBX proteins were also presented. The NLS domain was present in BBX Ⅰ-III subfamilies except for PaBBX11-3a/b because the NLS domain partially overlaps with the CCT domain. Moreover, some BBX Ⅳ and Ⅴ members also had NLS domains ( Figure S3a,b). The VP domain appeared only in the BBX Ⅰ family ( Figure S3c). M1-M7 conserved motifs were found in different subfam-

Gene Structure and Conserved Motif Analysis
To clarify the structural features of the PaBBX genes, the exon-intron distribution was analyzed ( Figure 4). Except for PaBBX30, all other PaBBX genes contained introns. Furthermore, the intron lengths varied widely, with the shortest intron being 49 bp of PaBBX32-1 and the longest being 23,518 bp of the first intron of PaBBX10. The intronexon structures of the BBX I and BBX III subfamilies genes were relatively similar, with one intron and two exons, while the structure and intron length of the BBX II and IV subfamilies genes varied widely. Notably, there were five genes with introns over 10,000 bp (Table S2). To verify whether these genes transcribed normally, PaBBX22-2 and PaBBX23 were selected as examples. The CDS of PaBBX22-2 and PaBBX23 were amplified and proved to be the same as those in the genome. Since introns are a reservoir of cis-elements and also have functional roles [32,33], we predicted the cis-elements on these super-long introns. The results showed that these introns contained a large number of light, temperature, and hormone response elements, tissue-specific expression cis-elements, and biotic and abiotic stress elements (Table S3). Moreover, PaBBX22-2 and PaBBX23 also contained AT-rich sequences involved in maximal elicitor-mediated activation (Table S3). This suggested that these super-long introns may have important functions in the regulation of gene expression.

Gene Structure and Conserved Motif Analysis
To clarify the structural features of the PaBBX genes, the exon-intron distribution was analyzed ( Figure 4). Except for PaBBX30, all other PaBBX genes contained introns. Furthermore, the intron lengths varied widely, with the shortest intron being 49 bp of PaBBX32-1 and the longest being 23,518 bp of the first intron of PaBBX10. The intron-exon structures of the BBX I and BBX III subfamilies genes were relatively similar, with one intron and two exons, while the structure and intron length of the BBX II and IV subfamilies genes varied widely. Notably, there were five genes with introns over 10,000 bp (Table  S2). To verify whether these genes transcribed normally, PaBBX22-2 and PaBBX23 were selected as examples. The CDS of PaBBX22-2 and PaBBX23 were amplified and proved to be the same as those in the genome. Since introns are a reservoir of cis-elements and also have functional roles [32,33], we predicted the cis-elements on these super-long introns. The results showed that these introns contained a large number of light, temperature, and hormone response elements, tissue-specific expression cis-elements, and biotic and abiotic stress elements (Table S3). Moreover, PaBBX22-2 and PaBBX23 also contained AT-rich sequences involved in maximal elicitor-mediated activation (Table S3). This suggested that these super-long introns may have important functions in the regulation of gene expression.  To better understand the functional divergence of PaBBX proteins, the online MEME tool was used to investigate the conserved motif compositions ( Figure 5). As shown in Figure 4, a total of 20 motifs were present in PaBBX proteins. Motifs 1, 3, 4, 6, 14, and 20 were identified as part of the B-BOX domain, while Motif 2 was identified as the CCT domain. In addition to these ubiquitous motifs, some motifs were only presented in a subfamily. For example, Motifs 12 and 18 were presented only in the BBX I subfamily, Motifs 7,8,9,15, and 16 were presented only in the BBX II subfamily, and Motifs 11, 17, and 20 were presented only in the BBX III subfamily. Similar motif compositions were observed in subfamily proteins, while significant divergences were found in different subfamilies, suggesting that PaBBX proteins may have redundancy of functions in the same subfamily and divergence of functions in different subfamilies.
tool was used to investigate the conserved motif compositions ( Figure 5). As shown in Figure 4, a total of 20 motifs were present in PaBBX proteins. Motifs 1, 3, 4, 6, 14, and 20 were identified as part of the B-BOX domain, while Motif 2 was identified as the CCT domain. In addition to these ubiquitous motifs, some motifs were only presented in a subfamily. For example, Motifs 12 and 18 were presented only in the BBX I subfamily, Motifs 7,8,9,15,and 16 were presented only in the BBX II subfamily, and Motifs 11,17,and 20 were presented only in the BBX III subfamily. Similar motif compositions were observed in subfamily proteins, while significant divergences were found in different subfamilies, suggesting that PaBBX proteins may have redundancy of functions in the same subfamily and divergence of functions in different subfamilies.

Expression Profiles of PaBBX Genes
To predict the potential roles of the PaBBX genes, we examined the expression profiles of PaBBX genes in various tissues by qRT-PCR, including roots, shoots, and leaves of juvenile and mature adult trees. The sequences of the CDS and untranslated region (UTR) of PaBBX11-3a and PaBBX11-3b are similar, differing by only a few bases, so it was not feasible to design primers to differentiate between them. Therefore, their expressions were jointly analyzed in this chapter and 2.5. As shown in Figure 6, the expression patterns of PaBBX genes could be classified into three groups. The first group was ubiquitously expressed in almost all organs tested, including PaBBX1-1, PaBBX4, PaBBX7, PaBBX8-1/2, PaBBX9-1/2, PaBBX10, PaBBX11-2, PaBBX11-3a/b, PaBBX19-1/2/3/4, PaBBX21-3, PaBBX22-1/2, PaBBX23, PaBBX28/29-1/2/3, PaBBX30, and PaBBX32-1/2 (Figure 6a). The second group was tissue-specific expression genes. Among these genes, PaBBX5-1, PaBBX5-2, PaBBX15-2, and PaBBX25 showed a low level of expression in both juvenile and adult roots. PaBBX11-1 was hardly detected in roots and juvenile stems, while PaBBX13 was mainly expressed in juvenile stems. The expression of PaBBX12 was low in leaves at both stages, and PaBBX21-1 showed very weak or no expression in juvenile leaves and adult stems, whereas PaBBX1-2, PaBBX1-3, and PaBBX15-3 showed significantly higher levels of expression in juvenile leaves. Moreover, PaBBX15-1 was highly expressed only in the stems and leaves of adult plants (Figure 6b). The third group consisted of genes with stage-specific expressions: PaBBX21-2 and PaBBX24. The expression of these two genes was higher in juveniles, with very weak or no expression in adults (Figure 6c). The different expression patterns 9 of 20 suggested that the PaBBX genes may participate in different processes of growth and development in P. × acerifolia. (c) The PaBBX genes were stage-specific expressed. J and M represented juvenile and maturely adult trees, respectively. The relative expression level was normalized to the PaTPI (P. × acerifolia triose phosphate isomerase) gene. Data were presented as the mean values ± SD (standard deviation) from three biological replicates.

Transcriptional Profiles of PaBBX Genes during Annual
CO promotes Arabidopsis flowering under long-day conditions, and CO-RNAi poplars limit growth early under short-day conditions [26,34]. To investigate whether PaBBX genes were involved in flowering and/or dormancy regulation, such as CO, we observed the expression patterns of PaBBX genes at the corresponding stages. The expression of PaBBX genes throughout the year is presented by two sets of data, and the date from mid-spring to late-summer was based on transcriptomic data, while the date from early autumn to next early spring was obtained from qRT-PCR results.

Identification of Cis-Elements in the Promoter of PaBBX Genes
To explore the potential regulatory mechanism of PaBBX genes, the promoter regions (2000 bp upstream sequence from the start codon of the genomic DNA sequence) of the PaBBX genes were submitted to the PlantCARE database [35]. Many cis-elements involved in abiotic and biotic stress responses, phytohormone responses, and plant growth and development were identified in addition to the core promoter elements TATA and CAAT boxes ( Figure 9 and Table S4). Among the abiotic and biotic stress response elements, anaerobic inducible elements (ARE and GC-motif), defense and stress response elements (TC-rich repeats), low-temperature responsive element (LTR), wound-responsive element (WUN-motif), and drought-inducible elements (MBS) were found in 32, 9, 24, 14, and 21 PaBBX genes, respectively. Many elements involved in hormone responsiveness were detected, especially the abscisic acid-inducible element (ABRE), which was abundantly found in the promoters of 36 PaBBX genes except PaBBX11-2, PaBBX19-2, and PaBBX24. Meanwhile, gibberellin response elements (GARE-motif, TATC-box, and P-box), auxin response elements (AuxRR-core and TGA-element), salicylic acid response elements (TCA-element), and MeJA response elements (TGACG-motif and CGTCA-motif) were also detected in 25, 18, 17, and 30 PaBBX genes, respectively. In addition, 24 different light response elements were detected in the PaBBX genes promoters, with the largest number and the widest range of G-box elements. Other elements involved in plant growth and development, such as Meristem expression elements (CAT-box and GCN4_motif), zinc metabolism elements (O2-site), circadian control elements (circadian), flavonoid biosynthetic genes regulation elements (MBSI), and palisade mesophyll cell differentiation elements (HD-Zip 1), were observed in 22, 18, 7, 3, and 6 PaBBX genes promoters, respectively. These elements indicate that PaBBX genes may be involved in plant growth and development, hormone and defense signaling, and adaptation to environmental changes.

Structural and Evolutionary Analysis of BBX Genes in P. × acerifolia
In this study, we identified 39 PaBBX genes from P. × acerifolia. Phylogenetic analysis revealed that the 39 PaBBX proteins could be classified into five subfamilies, BBXⅠ-Ⅴ subfamilies (Figures 2 and S1). The same subfamily shared similar exon-intron organization and exon length, except for the BBX Ⅳ subfamily ( Figure 6). Interestingly, some BBX II and IV genes were much longer than other species due to the super-long intron [36][37][38][39]. There are many cis-elements involved in expression regulation in the super-long introns of PaBBX genes (Table S3), suggesting that these introns may be involved in the transcriptional regulation of these genes.
The B-BOX and CCT regions of PaBBX proteins were highly conserved as in other species [36,37,40,41], indicating that their functional differentiation may depend on regions other than BOX and CCT, such as the VP domain, the M6 domain, the PF(V/L)FL motif, and other unknown motifs [4,10,42,43]. In addition, some BBX proteins have lost conserved domains during recent evolutionary events. For example, ZmBBX7 in Zea mays, BdBBX11 in Brachypodium distachyon, and VviBBX22a in V. vinifera all lack a B-BOX domain [4,44]. This loss also occurred in P. × acerifolia, where PaBBX11a/b lacked the B-BOX1 and CCT domains, and PaBBX19-4 lacked the B-BOX1 domain, although other common features were conserved ( Figure 5). B-BOX is crucial to the function of BBX genes [45], and the loss of the B-BOX domain may affect the function of these genes. Moreover, the conserved amino acids Cys and His of B-BOX2 had been replaced in PaBBX11-3a/b and PaBBX9-2 ( Figure 4). This variation is also found in the grapevine and may give rise to novel functions [44].

Structural and Evolutionary Analysis of BBX Genes in P. × acerifolia
In this study, we identified 39 PaBBX genes from P. × acerifolia. Phylogenetic analysis revealed that the 39 PaBBX proteins could be classified into five subfamilies, BBX I-V subfamilies (Figures 2 and S1). The same subfamily shared similar exon-intron organization and exon length, except for the BBX IV subfamily ( Figure 6). Interestingly, some BBX II and IV genes were much longer than other species due to the super-long intron [36][37][38][39]. There are many cis-elements involved in expression regulation in the super-long introns of PaBBX genes (Table S3), suggesting that these introns may be involved in the transcriptional regulation of these genes.
The B-BOX and CCT regions of PaBBX proteins were highly conserved as in other species [36,37,40,41], indicating that their functional differentiation may depend on regions other than BOX and CCT, such as the VP domain, the M6 domain, the PF(V/L)FL motif, and other unknown motifs [4,10,42,43]. In addition, some BBX proteins have lost conserved domains during recent evolutionary events. For example, ZmBBX7 in Zea mays, BdBBX11 in Brachypodium distachyon, and VviBBX22a in V. vinifera all lack a B-BOX domain [4,44]. This loss also occurred in P. × acerifolia, where PaBBX11a/b lacked the B-BOX1 and CCT domains, and PaBBX19-4 lacked the B-BOX1 domain, although other common features were conserved ( Figure 5). B-BOX is crucial to the function of BBX genes [45], and the loss of the B-BOX domain may affect the function of these genes. Moreover, the conserved amino acids Cys and His of B-BOX2 had been replaced in PaBBX11-3a/b and PaBBX9-2 ( Figure 4). This variation is also found in the grapevine and may give rise to novel functions [44].
To explore the evolution of the BBX family, a phylogenetic tree of BBX proteins from Magnoliales (L. chinense), monocots (O. sativa), basal eudicots (N. nucifera and P. × acerifolia), and core dicots (V. vinifera and A. thaliana) was constructed. Except for some rice proteins, homologous BBX proteins from different species clustered together (Figure 2), suggesting that BBX proteins shared a common ancestor and expanded individually after the divergence of magnolias, dicots, and monocots. Subsequently, the segmental and tandem duplication events were analyzed in order to elucidate the expansion of the BBX gene family in P. × acerifolia. Twelve pairs of duplicated segments, including 30 genes, were found (Figure 1), explaining that segmental duplication is the main driver of the BBX family expansion in P. × acerifolia, as in Pyrus bretschneideri [31] and Fagopyrum Tataricum [38]. Gene duplication events contribute to the generation of novel functions, and expression profile is one of the characteristics of functional differentiation [46]. PaBBX genes generated by segment duplication showed differential spatiotemporal expression (Figures 6-8), which may lead to neofunctionalization and subfunctionalization.

Expression Patterns and Potential Functions of PaBBX Genes
The functions of the BBX gene family are involved in many processes of plant growth and development, such as seedling photomorphogenesis, flowering, plant hormone signal transduction, pigment accumulation, and abiotic and biotic stresses [15]. However, there have been no studies on the function of BBX genes of P. × acerifolia. Expression level and function are closely related, and gene expression mainly depends on the cis-elements on the promoter [47]. Therefore, we predicted the promoter cis-elements of PaBBX genes and performed spatiotemporal expression analysis to anticipate their functions in P. × acerifolia.
BBXs are involved in the vegetative growth of plants. PtCO1/2 overexpressing poplars were smaller than control plants [48], overexpression of IbBBX29 resulted in increased biomass of Ipomoea batatas leaves [49], and BBX31 promoted primary root elongation under low-intensity white light in Arabidopsis [50]. In P. × acerifolia, most PaBBX genes were detected in all tissues/organs tested but with different expression patterns. PaBBX1-1/2, PaBBX13, PaBBX15-3, and PaBBX30 were strongly detected in juvenile leaves, implying that these genes may have a similar function in regulating leaf development in juvenile trees. The expressions of PaBBX5-1/2, PaBBX11-3a/b, and PaBBX13 were significantly higher in the juvenile stem, suggesting that they may contribute to the internode elongation of stems. PaBBX28/29-3 and PaBBX10 were highly expressed in roots at different stages, which may be involved in root growth. In addition, the expression level of PaBBX21-2 and PaBBX24 was different between childhood and adulthood and may have different functions in these two periods. The expression patterns of these genes suggested that they may play different regulatory roles in the development of P. × acerifolia. In addition, different copies of some genes showed almost identical expression patterns (Figure 6), such as PaBBX1-1 and PaBBX1-2, PaBBX5-1 and PaBBX5-2, and PaBBX8-1, and PaBBX8-2, possibly with redundant functions. However, some other PaBBX genes showed differential spatiotemporal expression between copies (Figure 6), suggesting that subfunctionalization and/or neofunctionalization events may also occur between different copies.
Flowering and dormancy have important biological significance for the reproduction and survival of woody plants. Hormones are an important internal factor in regulating flowering and dormancy. GA is the most important hormone in floral transformation, and it works together with ABA antagonism to ensure that plants survive in winter [51][52][53][54][55]. Previous research has shown that the contents of ABA and GAs were tightly linked to the dormancy process of P. × acerifolia (unpublished). Moreover, almost all BBX genes respond to exogenous treatment of phytohormones in different plant species, and some BBX genes are involved in the transduction of GA and ABA signaling pathways [25,[56][57][58]. Similarly, a large number of hormone response elements were present on the PaBBX genes promoter (Figure 9), suggesting that PaBBX genes may be involved in these hormonal pathways. In particular, a total of eight PaBBX genes contained more than six ABAresponse elements, which may be involved in ABA-mediated growth processes such as bud dormancy. Photoperiod and temperature are decisive external environmental factors that regulate flowering and dormancy. Previous evidence suggests that BBX genes play an important role in the light signaling pathway [15]. In P. × acerifolia, each PaBBX gene contains numerous light-responsive elements. Among them, G-BOX, the binding site of central light signal regulators such as HY5 and PIFs, is the most abundant and widely distributed [59]. The results suggested that PaBBX genes were likely involved in light signaling transcriptional cascades, such as photomorphogenesis and pigment accumulation mediated by HY5 and PIFs, flowering, and shade avoidance. It is noteworthy that 21 PaBBX genes have low-temperature responsive elements while chilling in the winter release dormancy [60], indicating that PaBBX genes may participate in the dormancy release.
BBX genes are known to regulate flowering in herbs [15,[61][62][63], and in perennial woody plants populus, PtCO2 has an effect on poplar growth cessation and bud set [26]. This suggested that the BBX genes may be involved in both flowering and dormancy regulation. In our study, a large number of PaBBX genes were highly expressed during mid-late spring (Figure 7), indicating that these PaBBX genes may be involved in the flowering transition and vegetative growth. In addition, some PaBBX genes were regularly expressed, corresponding to different stages of dormancy and bud break. Previous studies have shown that the expression of PaFT dramatically increased under low temperatures and short days from early autumn to late winter and was believed to serve to release buds from dormancy [27]. Expression patterns of PaBBX8-1, PaBBX8-2, PaBBX9-1, PaBBX28/29-1, and PaBBX32-2 were similar to these of PaFT, suggesting that these genes may act as an upstream regulator of PaFT to contribute to the break of dormancy. The expressions of PaBBX11-2 and PaBBX15-3, similar to EARLY BUD BREAK 1(PtEBB1), an important factor in poplar bud break, were not detectable during dormancy but increase rapidly before bud break [64]. Therefore, it is possible that the function of these two genes may be also involved in bud break, such as PtEBB1. Moreover, PaBBX11-3a/b, PaBBX15-1, PaBBX19-3, PaBBX19-2, and PaBBX30 displayed high expression levels at the endodormancy phase; these genes may play an important role in dormancy maintenance ( Figure 8). Overall, PaBBX genes may play vital roles in flowering and/or dormancy in P. × acerifolia and function at different stages of dormancy. However, the exact function of PaBBX genes needs further experimental confirmation.

Plant Materials
For the analysis of gene expression patterns, the adult tree materials were obtained from 40-year-old P. × acerifolia, field grown at Huazhong Agricultural University, Wuhan, China. The juvenile tree materials were harvested from one-year-old seedlings with 10 leaves. The seedlings were collected from adult plants and grown by sowing them in a growth chamber under LD conditions (16 h:8 h, light/dark, 24 • C).
It has been found that critical period for the flowering transition in P. × acerifolia occurs during mid to late spring (April to May). The endodormancy period is observed during autumn, which spans from September to November. Dormancy release takes place during early and mid-winter (December and next January). Additionally, the bud break period of P. × acerifolia occurs during late winter and the next early spring (February and March). For the annual expression pattern analysis, the transcriptome samples were the subpetiolar buds of adult trees from April to August 2021, and the subpetiolar buds of the above adult trees were collected from September 2021 to March 2022.

Identification of BBX Genes from P. × acerifolia
We aligned protein profiles to the verified BBX proteins and BBX domain (PF00643) of Arabidopsis via diamond and/or using hmmsearch in HMMER with an E-value of 10 −5 from P. × acerifolia genome databases (unpublished). Then, all obtained BBX sequences were confirmed to contain the B-BOX domain by NCBI Conserved Domain Database (CDD) website (https://www.ncbi.nlm.nih.gov/Structure/cdd/wrpsb.cgi accessed on 4 December 2022).
The molecular weight (MW) and isoelectric points (pI) of PaBBX proteins were calculated on the ExPASy website (https://web.expasy.org/protparam/ accessed on 6 December 2022).
The distribution and collinearity analysis of PaBBX genes on chromosomes were drawn using the Circos program of TBtools software [65]. The nonsynonymous (Ka) and synonymous (Ks) substitution rates of the gene duplication pairs were estimated via TBtools.

Phylogenetic Analysis and Protein Sequence Alignments
The BBX protein sequences of Arabidopsis were downloaded from the Arabidopsis Information Resource (TAIR) database (http://www.arabidopsis.org accessed on 6 December 2022), and the sequences of Liriodendron chinense, Oryza sativa, Nelumbo nucifera, Vitis vinifera were obtained from the Phytozome database (https://phytozome-next.jgi.doe.gov/ accessed on 6 December 2022) [66]. Protein sequence alignments of BBX proteins of P. × acerifolia, Arabidopsis, L. chinense, O. sativa, N. nucifera, and V. vinifera were conducted using the Clustal W program of MEGA X software. Phylogenetic tree of BBX in six species and only in P. × acerifolia based on the alignments and constructed by the neighborjoining (NJ) method using the Poisson model and 1000 replicates of bootstrap value. Then, the tree was uploaded to iTOL (https://itol.embl.de/ accessed on 14 December 2022) to beautify [67]. Sequence prediction of NLS, VP domain, and M1-M7 domains were performed according to previous articles [4,43]. Sequence logos for generating conserved domains were generated using the online WebLogo (http://weblogo.berkeley.edu/logo.cgi accessed on 8 December 2022).

Gene Structure and Conserved Motif Analysis
The exons, introns, and untranslational region (UTR) of the BBX genes were identified according to the genome annotation of P. × acerifolia. The conserved domains were predicted via the CDD database, and conserved motifs were identified via Multiple Expectation Maximization for Motif Elicitation (MEME) online database (http://memesuite. org/tools/meme accessed on 27 December 2022). The maximum motif number in MEME was set as 20, and the motif length was set as 6-50 amino acids. The gene structure and conserved motif of PaBBX genes were drawn using TBtools software.

Putative Cis-Element Analysis of PaBBX Genes Promoters and Intron
The 2000bp region upstream from the start codon of the PaBBX gene was extracted from the P. × acerifolia genome as a promoter, and introns of PaBBXs about 10 kbp were also extracted from the P. × acerifolia genome. Then, the sequences of promoters and introns were uploaded to the Plantcare website (http://bioinformatics.psb.ugent.be/webtools/ plantcare/html/ accessed on 9 December 2022) to predict putative cis-acting element [35].

Analysis Expression Pattern of PaBBX Genes
For the expression analysis of PaBBX genes, RNA from the plant material in 4.1 was extracted via CTAB according to a previously reported protocol [68], and complementary DNA (cDNA) was synthesized using PrimeScriptTM RT reagent Kit with gDNA Eraser (TaKaRa, Otsu, Japan). Quantitative Real-time PCR (qRT-PCR) was performed using cDNA as a template and using SYBR Premix Ex TaqTM (TaKaRa, Otsu, Japan) in an Applied Biosystems 7500 Fast Real-Time PCR System (Applied Biosystems, Waltham, MA, USA) and following the manufacturer's instructions. The qRT-PCR amplifications were performed in a 10 µL volume reaction containing 1.0 µL template, 5.0 µL 2 × SYBR Green Master Mix, 0.2 µL forward and reverse primers (10 µL mol/µL), and 3.6 µL water. The qRT-PCR reaction program is 95 • C for 30 s, 95 • C for 5 s, and 60 • C for 34 s (40 cycles). Primers for qRT-PCR were designed using the Primer3Plus website (https://www.primer3plus.com/ accessed on 13 December 2022). The PaTPI (P. × acerifolia triose phosphate isomerase) was used as reference genes to normalize all data. The expression levels of a relative gene were detected using the method of 2 −∆∆CT . Data were presented as the mean values ± SD (standard deviation) from three biological replicates. In addition, expression from April to August 2021 is based on unpublished transcriptome data. PaBBX19-3 and PaBBX30 did not obtain expression data because they did not annotate. The histogram was drawn using GraphPad Prism 9, and the hierarchical clustering heatmap was performed by TBtools software. The primers used are shown in Table S5.

Conclusions
In conclusion, like the BBX family in Arabidopsis, the BBX family in P. × acerifolia was divided into five subfamilies, BBX I-V subfamilies, with a total of 39 BBX genes. The same subfamily has similar gene structures and conserved motifs, but different subfamilies have different gene structures and conserved motifs. In addition, some genes contain super-long introns that may regulate their expression. The expansion of the P. × acerifolia BBX family was mainly caused by segmental duplication and had undergone purifying selection. The promoters of the PaBBX genes contained many elements related to plant growth and development as well as hormone and stress response. Expression pattern analysis showed that the expression of some BBX genes was closely related to the process of flowering and/or dormancy, which may be involved in the regulation of flowering and/or dormancy of P. × acerifolia. This study is the first comprehensive analysis of the BBX gene family in P. × acerifolia on a genome-wide scale, reveals the relationship of PaBBX genes with flowering and/or dormancy, provides new ideas for the regulation of flowering and/or dormancy in perennials, and enriches the genetic stock of P. × acerifolia for molecular breeding.