Introduction

Plant transcription factors are DNA-binding proteins that regulate gene expression at the transcriptional level [1] and are essential for plant growth and development, participating in the regulation of several plants physiological and biochemical processes [2]. The SQUAMOSA-promoter binding protein-like (SPL/SBP) gene was first identified in Antirrhinum majus as a transcription factor regulating early flower development [3]. The gene family was named after a conserved DNA-binding domain of 78–80 bp amino acid residues and was characterized by two DNA-binding domains (zinc finger structures) and a C-terminal NLS nuclear localization signal that overlapped with the zinc finger structure [4]. The SPL gene family has been identified in several cultivated economic trees, such as litchi [5], kiwifruit [6], chestnut [7], citrus [8], sweet orange [9], apple [10], grape [11], jujube [12], walnut [13], pecan [14]. In these plants, SPL genes were involved in a wide range of plant growth processes, including vegetative growth, flower development, fruit development and ripening, and abiotic stresses. The major SPL gene family had important roles in plant flower development, including vegetative to reproductive growth [15,16,17,18], anther development [19], GA flower regulatory pathways [16], and flower morphogenesis [20]. In Arabidopsis, AtSPL3 and its homologs SPL4 and SPL5 were highly expressed in flowering shoots and involved in flowering [21, 22], and AtSPL3 was directly regulated by the SOC1 gene [23]. Transcript levels of AtSPL4 were decreased in the double loss-of-function mutant of SOC1 and FUL [15]. Moreover, AtSPL3, AtSPL4, and AtSPL5 are involved in the regulated floral transition, which requires synergy with the FT-FD module to promote floral meristem transformation under long sunlight [16].On the other hand, AtSPL9/AtSPL15 and AtSPL13 were involved in juvenile to adult growth phase transition [15, 16], with AtSPL9 being associated with leaf growth rate and final size [17]. The miRNA-SPL regulatory network was involved in floral transition and development [20, 24]. miR156 played a key role in vegetative phase change and shoot development. And overexpression of miR156 in Arabidopsis led to a prolonged juvenile to adult phase [20]. In Arabidopsis thaliana, under short-day conditions, the SPL gene targeting miR156 had an important role in floral transition by activating the MADS-box gene at the shoot apical [20]. Under long-day conditions, in addition to a similar role at the shoot apical, the interaction between DELLA and SPL interfered with SPL transcriptional activity, thereby delaying floral transformation under long daylight conditions by inactivating miR172 in leaves and the MADS-box gene at the shoot apex [20]. Meanwhile, AtSPL9 /AtSPL15 interacted with the GA signalling protein DELLA, which was involved in the GA-mediated flowering pathway [16]. In addition AtSPL2, AtSPL10, AtSPL11, and AtSPL13 were involved in the regulation of the vegetative-reproductive transition [18], and AtSPL8 played an important role during sporophyte development [19]. In addition to dicotyledons, SPL genes also played an important role in monocotyledons. In rice, OsSPL10 was preferentially expressed in young panicle and stem, and negatively controls salt tolerance but positively controls trichome formation [25], the SPL14 mutant increased inflorescence branch number and grain weight as well as increasing rod strength [26, 27], and OsSPL2, OsSPL17 and OsSPL18 regulated tiller number and plant height [28].

In other woody plants, SPL genes also play an important role in flower development. GA might indirectly influence the expression of its target genes CmSPL6/CmSPL9/CmSPL16 through miR156, thereby playing an important role in chestnut flower bud development [7]. Grape VpSBP11 worked as a transcription factor during the transition from nutritional to reproductive stages, and overexpression in Arabidopsis activated the expression of FUL, AP1, and LFY genes, thus showing early flowering [29]. Overexpression of CclSBP6 and CclSBP7 in citrus could alter flowering time in Arabidopsis [8], and similarly, overexpression of EjSPL3, EjSPL4, EjSPL5, and EjSPL9 in litchi could significantly promote flowering and up-regulated the expression of downstream flowering genes [30]. In addition, the function of SPL in floral induction had been reported in walnut (Juglans regia) [13] and pecan (Carya illinoinensis) [14]. Flower development is an important work in plant growth and development, in which SPL is involved as a plant-specific transcription factor, so it is important to study the SPL gene family and use its function to regulate the flower development process in plants.

Blueberry is a photoperiod-sensitive berry shrub with short days to promote reproductive growth [31, 32]. The flower bud development process of blueberries is closely related to the flowering and fruiting time, which affects the marketing time of blueberries [33, 34]. At present, greenhouse cultivation has become the important method of blueberry in northern China, which can ensure safe overwintering and earlier fruit ripening and market time [35]. However, compared with open-field blueberry cultivation, the characteristics of the annual growth cycle of blueberries in the greenhouse were significantly different, especially in flower bud differentiation. To better study the process of flower bud development in blueberries, we conducted a genome-wide analysis of the SPL gene family in blueberries using genomics [36], including identification of the tetraploid blueberry gene families, phylogeny, chromosomal localization, evolutionary history, and the temporal and spatial expression characteristics during flower development in blueberry. Our study establishes a basis for investigating the regulatory role of SPL genes in the blueberry floral induction and initiation processes, which will contribute to the understanding of blueberry plant growth and flower and fruit development.

Material and methods

Identification of VcSPLs gene family members

Blueberry genomic data were obtained from (GIGA) nDB (http://gigadb.org/dataset/100537/). HMM and Blast searches were used to predict blueberry SPL genes [37, 38]. A total of sixteen Arabidopsis thaliana SPL protein sequences [39] were downloaded from NCBI as queries for BLASTP searches (e-value ≤ 1 × 10−5). In addition, the SPL DNA-binding domain (PF03110) was obtained from the Pfam database [40], and the HMMER toolkit was adopted to find putative SPL genes. The target genes obtained by these two methods were subjected to identification of the SPL structural integrity using the SMART program (http://smart.embl-heidelberg.de/). And the target protein properties, including isoelectric point (P), molecular weight (MW), number of amino acids (aa), lipophilicity, and hydrophobicity, were analyzed by the online analysis tool ProtParam (https://web.expasy.org/protparam/). Subcellular localization analysis was performed by the online analysis tool MBC (http://cello.life.nctu.edu.tw/).

Phylogenetic analysis of VcSPLs

To classify the evolutionary relationships of the VcSPLs gene family, the SPL protein sequences of Arabidopsis thaliana [39], rice (Oryza sativa) [41], tobacco (Nicotiana tabacum L.) [42], grape (Vitis vinifera) [11] and diploid blueberry (Vaccinium darrowii) [43] were aligned using the CluastW algorithm and MEGA11 software [44], where diploid bilberry SPL family members were identified in the same way to tetraploid blueberry, with gene names and location information in the Supplementary Table 2. Phylogenetic analysis based on amino acid sequence alignment was performed using the neighbor-joining (NJ) method with Bootstrap tests on 1000 resamples, and for more reliable phylogenetic analysis, maximum likelihood (ML) was also considered [45]. The final phylogenetic relationships were based on the results obtained from the two methods and visualized by EvolView software [46].

Conserved motifs and gene structure analysis of VcSPLs

Multiple sequence alignment of VcSPLs protein sequences was performed using Multiple Sequence Comparison by Jalview Version2 [47]. The motifs of blueberry SPL proteins were predicted using an online toolkit of MEME (https://meme-suite.org/meme/tools/meme). The conserved structural domains of blueberry SPL proteins were analyzed by the online toolkit of NCBI CD-Search (https://www.ncbi.nlm.nih.gov/Struct-ure/cdd/wrpsb.cgi). Extraction of target gene sequences from blueberry and visual analysis was performed with TBtools software [48].

Chromosomal location and duplication events among VcSPL genes

All 56 VcSPLs genes were located on the corresponding chromosomes using the blueberry genome information [36]. Based on the genome annotation information, gene densities were extracted by the TBtools [48]. The potential gene duplication events were used as an input file for Multiple Collinearity Scan Toolkit X [49]. The synteny relationship of the orthologous SPL genes between bilberry and blueberry, rice, and Arabidopsis were identified using the Synteny visualization plugin embedded in TBtools [48].

Expression pattern of the VcSPLs genes and analysis of the cis-acting element of the VcSPLs promoter

The fragments per kilobase million (FPKM) values of Vaccinium corymbosum L. RNA sequencing data (Bioproject ID#PRJNA494180) were utilized for the gene expression analysis of various tissues [50].

The 2000 bp upstream of the coding regions of the VcSPL genes were extracted from the blueberry genome data using by TBtools software. And PlantCARE10 was used to search for putative cis-acting elements (http://bioinformatics.psb.ugent.be/webtools/plantcare/html/).

Plant material and quantitative RT-PCR analysis

The northern highbush blueberry (Vaccinium corymbosum L.) plant was used as material, which was collected in the Blueberry Cultivation Garden, Miyun District, Beijing, China (N40°23 “15.86”, N116°47 “2.36”). The four different tissues (shoots, leaves, flower buds, and leaf buds) of blueberries were collected at the time of the bud burst. Genes from each subgroup of the VcSPL gene family (derived from evolutionary branches) were selected for differential expression analysis in different tissues. Secondly, to further investigate the role of VcSPL in floral induction and initiation, we selected genes that were highly expressed in floral buds, flower at anthesis, and flower post-fertilization for RT-qPCR using samples collected at different stages. A total of five periods in the process of blueberry bud differentiation were selected and the characteristics of the different periods were described in detail (Fig. 1). According to the anatomical characteristics of bud differentiation, the five periods of blueberry bud differentiation were characterized as follows: the vegetative growth stage (VES) showed a narrow meristem. The floral induction stage (FID) appeared in the loosed bud scales and increased the hemispherical of the apical meristem. During the floral initiation stage (FIN), the inflorescence primordium occurred. At the floral organ primordium development stage (FOD), the primordia of the sepals, petals, stamens, and pistil floral have appeared gradually. At the late stage of floral organ development (LOFD), the locules of the ovary developed further, the protruding tips coalesced and grew into the style and stigma, the corolla always coalesced at the tip, the calyx was less than half the length of the corolla, and the complete flower morphology had been presented (Fig. 1).

Fig. 1
figure 1

A-B The morphology of blueberries at different stages of bud differentiation, stages of the vegetative stage (VES), floral induction stage (FID); floral initiation stage (FIN); floral organ primordium differentiation (FOD); Late floral organ development (LFOD)

Full size image

The total RNA isolation and cDNA synthesis were according to our previous report [51]. Oligo 7 software was used for the design-specific primers, and related primers were shown in Supplementary Table 7. The quantitative kit was Taq Pro Universal SYBR qPCR Master Mix (Vazyme, Nanjing, China). Actin and UBC28 were used as the internal reference gene [52], and three replicates were performed for each sample. The relative gene expression values of the VcSPLs gene family were calculated using the 2−∆∆Ct method.

Results

Identification of VcSPLs gene family members

To identify SPL-encoding genes in blueberry, we initially performed a local BLAST library build with 16 SPL genes from Arabidopsis thaliana for comparison, followed by a further screening in combination with PFAM scan analysis to remove redundant sequences and then renaming them according to their position on the chromosome, finally identified a total of 56 VcSPL genes (Supplementary Table 1). The previous study indicated that 32 blueberry SBP/SPL gene families had been identified and screened, including two alternative SPL alternative splicing [53]. However, the previous identification of SPL encoding genes was based on a de novo assembly of the transcriptome, which was supplemented by further identification at the level of the whole blueberry genome in this experiment, with 24 new SPL gene families identified, and the relationship between the two SPL identifications corresponds in Supplementary Table 3. Two alternative splicings were not identified in the whole genome data, and in the previous study, two pairs of genes (VcSBP14a and VcSBP14aAS, VcSBP14c, and VcSBP14cAS) could be derived from different RNA splicing of the same genes [53]. In this study, VcSBP14aAS and VcSBP14cAS were not identified, while VcSBP14a and VcSBP14c corresponded to VcSPL43 and VcSPL30, respectively, in the present study. This may be due to the different blueberry varieties and tissues sequenced.

The basic information of the VcSPL family members, such as amino acid length, molecular weight, theoretical isoelectric point, and subcellular localization, were analyzed by ProtParam. The results (Supplementary Table 1), showed that 56 gene sequences were located on 32 chromosomes, with gene lengths ranging from 1968 bp (VcSPL26) to 9387 bp (VcSPL11). The length distribution of the encoded amino acids ranged from 126aa (VcSPL14) to 1072aa (VcSPL43), and the molecular weight was from 14.04 kDa (VcSPL14) to 117.84 kDa (VcSPL43), and the aliphatic coefficients ranged from 31.90 (VcSPL38) to 84.12 (VcSPL15). The average hydrophilic coefficients were all less than zero, suggesting that most members of the VcSPL transcription factor family were rich in aliphatic amino acids. All members were localized in the nucleus, except for VcSPL36, which was localized in both the cytoplasm and the nucleus.

Phylogenetic analysis of VcSPLs

To investigate the evolutionary relationships of blueberries, we performed the identification of bilberry SPL family members based on recently published diploid blueberry(bilberry) data [43], identifying a total of 24, with gene names and positional information in Supplementary Table 2, and then constructed phylogenetic trees for the amino acid sequences of Arabidopsis, tobacco, rice, and bilberry. The 56 SPL structural domains were divided into six subclasses based on topological relationships (Fig. 2). Each group contained at least one blueberry VcSPL gene, and Group1 contained the fewest number of VcSPL, only five. From the root of the evolutionary tree, the members of the G1 subclass clustered alone among multiple species to form an evolutionary branch, suggesting that it may have a large degree of differentiation among species and that its functions may differ in different plants. In Group2, VcSPL29/30/43 and VcSPL10/19/25/27/28/34 can be subdivided into G2a and G2b based on evolutionary branches. The number of VcSPLs in Group5 was at most 14, which can be further divided into two subclasses based on gene structure and conserved motif characteristics (Fig. 3B), namely VcSPL8, VcSPL20, VcSPL51, and VcSPL53 belong to subclass G5a, while the other members belong to subclass G5b. The other subclasses that have been divided contain similar gene structures and conserved motif features, suggesting that they remained highly conserved during evolution and that the function of these genes was potentially extremely important for the plant body. At the same time, we could see that at the end of the branches, blueberry SPLs showed a high degree of clustering with that of bilberry, indicating a high level of homology between the two species, with later differentiation between subspecies, presumably due to the effects of domestication.

Fig. 2
figure 2

Phylogenetic tree of blueberry (Vaccinium corymbosum), Arabidopsis thaliana, tobacco (Nicotiana tabacum), rice (Oryza sativa), grape (Vitis vinifera), and bilberry (Vaccinium darrowii) SPL proteins. Clades with different colors represent diverse subgroups. They were divided into 6 groups according to the evolutionary tree and were represented by different colours. The different icons represent the SPL genes of different species, with red pentagrams representing blueberry, green circles representing Arabidopsis, dark blue circles representing tobacco, cyan circles representing rice, light blue triangles representing grape, and yellow squares representing bilberry

Full size image
Fig. 3
figure 3

Phylogenetic relationships and protein structure of blueberry VcSPL. A Neighbor-joining trees constructed for VcSPL. B Protein motif. Schematic diagram of the conserved motifs of SPL/SBP proteins in blueberry, which were elucidated using MEME. Each motif is represented by a number in the coloured box. C Conserved domains. Different domains were represented in different colours

Full size image

Conserved motifs and gene structure analysis of VcSPLs

Multiple sequence alignment (Fig. S1) of the 56 VcSPLs showed that all genes, except VcSPL14, contain essentially two zinc finger structural domains and a nuclear localization signal, and the presence of conserved structural domains suggested that SPL family genes might have similar functions that were functionally conserved between species. The phylogenetic tree was built for the VcSPL gene family (Fig. 3A), and the classification pattern was consistent with Fig. 2 except for VcSPL14, which may be caused by the short amino acid sequence of VcSPL14 and the lack of a complete zinc-finger structure. In addition, the motifs on each gene were scanned (Fig. 3B), and sequence information of each motif was provided in Fig. S2. The number of motifs ranged from 1–9. Motif1, motif2, and motif3 together formed the complete SPL/SBP conserved structural domain, with motif1 encoding the C2HC zinc finger structure and NLS nuclear localization signal, motif2 and motif3 together encoding the C3H structural domain. However, VcSPL14 was unable to form a complete C3H zinc finger structural domain because it contained only one motif1 and lacked motif2 and motif3, VcSPL44 and VcSPL54 failed to form a complete NLS structural domain with the lack of motif1, and VcSPL55 failed to form a complete C2HC structural domain without motif2. The others all contain motif1, motif2, and motif3 sequences to form the complete SPL/SBP domain, consistent with the multiple sequence alignment results. Meanwhile, the structural distribution indicated that members of Group1 and Group2 were rich in motifs, while members of Group3 had shorter amino acid sequences and fewer motifs. Some motifs were group-specific, such as motif 6 was only found in Group2, suggesting that they may have unique mechanisms and functions in blueberry.

Within the gene structure, the VcSPL gene family was highly variable (Supplementary Table 1 and Fig. S4). Each VcSPL gene contains between 2 and 11 exons, with a maximum of 11 exons (VcSPL1/10/19/24/25) and a minimum of 1 exon (VcSPL4/13/16/22/26/38/44). The complete SPL structural domain (Fig. 3C) was present in all members except for some groups of variants (VcSPL14, VcSPL44, and VcSPL54). And VcSPL27, VcSPL28, VcSPL34, and VcSPL43 contain an additional intact Ank_2 superfamily structural domain, while VcSPL15 and VcSPL36 also contain LRR superfamily and PLN03210 superfamily structural domains. The variety and number of motifs and conserved structural domains differed markedly between the subclasses, which also demonstrates the reliability of our evolutionary classification.

Chromosomal location and duplication events among VcSPL Genes

To characterize the expansion patterns of the VcSPL genes family, we investigated duplication events in the blueberry genome and compared them with the chromosomal locations of all VcSPL genes (Fig. 4). Gene density analysis showed that 56 genes were unevenly distributed on 32 chromosomes, with Chr1,5,10,11 containing a high number of VcSPL genes. A total of 11 pairs of segmental duplication genes were found, whereas only two tandem duplications were found in the VcSPL gene family (Supplementary Table 4), and all segmental duplications occurred on the first chromosome of VcSPL1, VcSPL2, VcSPL3 and VcSPL4, and other family members, indicating that segmental duplication played a very important role in VcSPL genes expansion and that these four genes also played an important role in the evolution of the SPL genes family.

Fig. 4
figure 4

Chromosome distribution and synteny analysis of VcSPL genes. Coloured lines indicate duplication of VcSPL genes. The outer blue-filled rectangle represents chromosome length, and the approximate distribution of VcSPLs were marked with a short black line on the circle. Gene densities were drawn based on the annotation information of Vaccinium corymbosum L. genomes. The two rectangles in the inner circle represent the line pattern and bar pattern of chromosome density respectively, and the legend represents chromosome density

Full size image

We constructed a comparative syntenic analysis between blueberry and bilberry genomes to gain more insight into the evolution and differences. The number of 36 homologous pairs has been identified (Fig. S3). We also constructed a comparative syntenic analysis of blueberry and Arabidopsis and rice and identified 16 and 8 pairs of collinear SPL genes in Arabidopsis and blueberry, and rice and blueberry (Fig. 5 and Supplementary Table 5), respectively, suggesting that divergence of the common ancestor of rice and dicotyledons occurred before the divergence of blueberry and Arabidopsis. And we found that the collinear genes belonged to the same evolutionary branch (Fig. 2), suggesting that these genes may have similar functions.

Fig. 5
figure 5

The homologous gene pairs between blueberry and Arabidopsis and blueberry and rice. Grey lines in the background indicate the collinear blocks within blueberry and Arabidopsis and blueberry and rice genomes. Blue lines indicate homologous gene pairs

Full size image

Analysis of the cis-acting element of the VcSPLs promoter

To better understand the response and regulatory mechanisms of VcSPL genes, 2000 bp upstream of the putative translation start codon of genes were selected as cis-acting element sequences. Based on the cis-acting element predictions (Fig. 6 and Supplementary Table 6), it was found that in addition to the basic TATA-box and CAAT-box, cis-acting elements related to phytohormones, light response, plant growth and development, and biotic and abiotic stresses were included. Among the hormone response elements, the abscisic acid responsiveness (ABRE), and the MeJA-responsiveness (TGACG-motif or CGTCA-motif) were the most common of these, suggesting that abscisic acid and MeJA regulated this family of VcSPL transcription factors during growth and development. In plant biotic and abiotic stresses, anaerobic induction (ARE) and low-temperature responsiveness (LTR) response elements were more prevalent. Among the plant growth and development-related elements, the meristem expression element (CAT-box) was also detected in large numbers. These results suggested that VcSPL genes play an important role in light response, abscisic acid, MeJA, anaerobic induction, low-temperature responsiveness, and meristem expression.

Fig. 6
figure 6

Cis-regulatory elements analysis of the promoter region of VcSPL genes. Numbers indicate the number of cis-regulatory elements in the promoter region of the VcSPL gene. Based on functional annotation, VcSPL cis-acting elements can be classified into four categories: phytohormones, light response, plant growth and development, and biotic and abiotic stresses

Full size image

Expression pattern of the VcSPLs genes

Studying the expression pattern of VcSPLs in different tissues of blueberry, we found that VcSPLs were expressed at higher levels in flower buds, shoots, and roots (Fig. 7A). Considering the materials in our laboratory, genes from each subgroup (VcSPL18/23/32/40/43/53/54) were selected for spatiotemporal expression analysis with different tissues (flower buds, leaf buds, shoots, and leaves) of northern highbush blueberry and found that VcSPLs were differentially expressed in different tissues (Fig. 7B), with VcSPL53 being significantly more expressed in flower buds than in other tissues and VcSPL18/23/40 was significantly higher in leaves than in other tissues, while VcSPL32/43 was significantly higher in leaf buds than in other tissues. However, their relative expression in different tissues was not consistent with the gene expression profile data, probably due to conditions such as variety, and climatic period.

Fig. 7
figure 7

A Expression profiles of VcSPLs genes in different tissues. The legend represents gene expression, with red representing high expression and blue representing low expression. Clustering from evolutionary branches, Group1-6 represents subgroups of evolutionary branches. LD, Leaf day.LN, leaf, night. FL, Flower at anthesis.PF, Flower post-fertilization.grnfrt, green fruit.pinkfrt, pink fruit.ripe, ripe fruit. B Relative expression levels of VcSPLs in different tissues of northern highbush blueberry.Lower case letters, when present, indicate significantly different means between different tissue over the same period (using ANOVA, Duncan's test, p ≤ 0.05)

Full size image

However, the different tissue expression profiles only provided changes from bud burst to fruit development. As VcSPLs were involved in flower development and VcSPLs were expressed at higher levels in buds than in fruits, to further investigate the role of VcSPL in floral induction and initiation, we selected genes with high expression in floral tissues for analysis of relative expression during blueberry bud differentiation. VcSPL plays different roles in the five periods of blueberry flower bud differentiation (Fig. 8), with VcSPL2/18/20/32/35/40/42/45/47/51/53 being differentially expressed at different times. According to evolutionary branching, VcSPL45 was homologous to AtSPL3, and this expression was down-regulated during vegetative growth to floral induction, but the difference was not significant. And VcSPL45 was significantly up-regulated during late floral organ development. AtSPL13 was part of Group5 in the evolutionary branch, as was VcSPL20/35/47/51/53. VcSPL2/18 was homologous to AtSPL7 and was significantly down-regulated during floral initiation. VcSPL20 and VcSPL35 showed consistent expression trends from vegetative growth to floral initiation, with differences up-regulated, except that VcSPL20 was significantly down-regulated during flower organ development and VcSPL35 remained at a higher level. VcSPL51 and VcSPL53 showed an overall decreasing trend during blueberry floral bud differentiation. Notably, VcSPL51 and VcSPL20 were highly homologous in evolutionary branches, but showed different trends during the blueberry floral transition, indicating that these two genes were not functionally redundant, but functioned separately. VcSPL32 and VcSPL42 were homologous to AtSPL2/10/11, but the two genes had completely different roles in floral bud differentiation, with VcSPL32 significantly up-regulated during floral induction but VcSPL42 significantly down-regulated, and their functional characteristics need to be further investigated, considering their structural domain characteristics and different cis-acting elements.

Fig. 8
figure 8

Relative expression of VcSPL genes in blueberry bud differentiation at different stages. Lowercase letters, when present, indicate significantly different means between periods for the VcSPL gene (using ANOVA, Duncan's test, P ≤ 0.05). The coloured ones represent genes that are significantly different at different times and belong to different evolutionary subgroups, yellow for Group1, flesh pink for Group3, rose for Group5, and blue-purple for Group6

Full size image

Discussion

SPL encodes plant-specific transcripts that play important roles in the juvenile-to-adult growth phase transition, flower and fruit development, gibberellin signalling, sporulation, and responses to copper and fungal toxins [54]. Currently, 16, 18, 15 and 18 SPL family members have been identified in Arabidopsis [39], rice [41], tobacco [42], and grape [11], respectively, and a total of 56 VcSPL gene family members have been identified in tetraploid blueberry in this study. Different plants contain different numbers of VcSPL family members, which seems to be caused by functional redundancy or divergence of SPL family members during evolution. Notably, 24 SPL genes have been previously identified in blueberry [53], indicating that the number of previously identified blueberry SPL genes was underestimated, and this experiment further complemented the blueberry VcSPL gene family. To investigate the role of VcSBL genes in function, we showed the functional diversity of the blueberry SPL family in terms of phylogeny, gene structure, chromosomal localization, evolutionary history, and spatiotemporal expression. Firstly, in evolutionary branching and gene structure, genes belonging to the same subclass had the same motif and structural domain characteristics and were clearly distinguished from other subclasses, for example, motif6 was only present in Group2, presumably, this subclass may play a unique mechanism and function in blueberry. The biological functions of genes were often related to their gene expression patterns and structures [53]. In the present study, most members contained SPL structural domains, while Ank_2 superfamily structural domains were found in VcSPL27, VcSPL28, VcSPL34, and VcSPL43, and they may be involved in protein interactions [55]. In addition, VcSPL15 and VcSPL36 also contain LRR superfamily and PLN03210 superfamily structural domains, which may bind large hydrophobic ligands [56]. Secondly, the SPL gene family was suggested to undergo gene duplication in many plants, such as Arabidopsis [39], rice [41], and citrus [9]. Phylogenetic trees indicated that blueberry SPL genes were more closely related to bilberry, grape, and Arabidopsis, and more distantly related to rice SPL genes. This was consistent with the fact that blueberry, bilberry, grape, and Arabidopsis were dicotyledons. This was also supported by the number of homologous pairs identified in the comparative genomes of blueberry with Arabidopsis and rice, and it was hypothesized that divergence of the common ancestor of rice and dicotyledonous plants occurred before the divergence of blueberry and Arabidopsis. In blueberry, a total of two pairs of tandem duplications were identified and 11 pairs of segmental duplication, presumably segmental duplication playing a very important role in VcSPL gene amplification, in agreement with previously [43, 53]. It is clear that blueberry and bilberry show a high degree of clustering at the end of the evolutionary branch of SPLs, and 36 homologous pairs were identified, indicating a high degree of convergence in their relationship, consistent with the previous study [43].

Gene expression patterns could provide critical information for studying gene function prediction [57]. In Prunus mume, SPL genes were expressed at higher levels in flower buds and young fruits [58], and in walnut, eight JrSBP genes were expressed at higher levels in female flower buds than in leaf buds during adult floral induction [13]. In the present study, blueberry VcSPLs were expressed at higher levels in flower buds, shoots, and roots and lower levels after fruit ripening. The VcSPLs genes showed tissue specificity, which implied that the VcSPL family may have different functions in blueberries. On the evolutionary branch, VcSPL18 belonged to Group1, with significantly higher expression in leaves than in other tissues and significantly down-regulated expression during floral initiation. VcSPL18 was homologous to AtSPL7, which in more studies has been involved with the transcription factor CITF1 in regulating copper uptake and delivery to anthers, thereby affecting fertility [59]. In monocots, down-regulation of SPL7 and SPL8 in switchgrass could delay flowering and lead to increased biomass production and sugar release [60]. There were fewer studies on the flower development of SPL7 in dicotyledons, and the function of VcSPL18/AtSPL7 needed to be further explored. VcSPL34/27/28/1/19/25 was in the same branch belonging to Group2 and homologous with VpSBP5 and AtSPL1/12/14. During blueberry bud differentiation, VcSPL25/34 was not significantly differentially expressed at different stages of blueberry bud differentiation (Fig. 8). Analysis of the cis-acting elements showed that VcSPL19/25/10 had methyl jasmonate response elements and VcSPL19/25/10/1 had salicylic acid response elements, and they both also contained low-temperature response elements except for VcSPL10 (Fig. 6). AtSPL1/12 has been shown to improve plant heat tolerance and fruit set in Arabidopsis [61], but has not been reported to be involved in regulating flower development. Meanwhile, the AtSPL1/12/14 genes were involved in the developmental regulation of fumonisin B1 in Arabidopsis [62], and the same evolutionary branch, VpSBP5, has been shown to prevent powdery mildew in grapes [63]. It has been shown that VcSPL19 and VcSPL25 expression was upregulated and VcSPL10 expression was downregulated 48 h after spraying methyl salicylate [36]. Combined with the results of this experiment, we speculated that Group2 SPL genes may be involved in the regulatory pathway of salicylic acid and its function needs to be further investigated. VcSPL40 belonged to Group3 on the evolutionary branch and was differentially expressed in different periods of blueberry bud differentiation, with expression significantly up-regulated from the vegetative growth phase to floral initiation, significantly down-regulated during floral organ development, and significantly more expressed in leaves than in other tissues (Figs. 7B and 8). In the evolutionary branch, we found that VcSPL40 had no closer homology to Arabidopsis, but was homologous to OsSPL7, a target of miR156f that regulated plant architecture. Overexpression of OsSPL7 reduced the number of tillers in plants, whereas OsSPL7 RNAi was shown to increase the number of tillers and reduce plant height [64]. Meanwhile, in the previous study [53], heterologous expression of VcSPL40 (VcSBP8b) was found to exhibit late flowering in Arabidopsis and mi156-targeting VcSPL40. SPL genes targeted by miR156/157 in Arabidopsis could directly and positively regulate downstream flowering genes, such as SOC1, LFY, AP1, and FUL [18, 19], and GA might also promote flowering by directly binding to SPL transcription factors and activating miR172 and MADS-box genes [20]. Considering that VcSPL40 showed a gibberellin response element, combined with the expression profile and heterologous expression, we speculated that VcSPL40 might play a positive role during the floral transition (vegetative growth to floral initiation) and a possible repressive role during floral organ development.

The AtSPL3 was widely studied for its involvement in floral transition and flowering [21, 23]. In Antirrhinum majus, silencing of AmSBP1 homologous to AtSPL3/4/5 failed to flower [65]. In the present study, according to the phylogenetic tree, VcSPL45/33/6 were homologous to AtSPL3. In the expression profile, VcSPL45/33/6 had higher expression during bud, FL, and PF (Fig. 7A), but only VcSPL45 in blueberry expression was differentially down-regulated during flower bud differentiation (Fig. 8), speculating that the VcSPL45 gene may play a critical function in the floral induction. The miR156-SPL interaction constitutes a cue for endogenous vegetative phase transition and flowering [16], and the vegetative phase change in Arabidopsis was regulated by increased expression of SPL3 as well SPL4 and SPL5, and this increase was the result of decreased miR156 levels [66]. The BlSPL1, homologous to AtSPL3, played an important role in the change in the vegetative phase [67]. Age-dependent reduction of miR156 led to an increase in SPL, which in turn activated the expression of miR172, MADS-box genes, and LFY genes [20]. Currently, miR156/157 has been shown to target VcSPL45 (VcSBP6a) [31], but the regulatory pathway of miR156-VcSPL45 with downstream flowering genes (LFY, MADX-box) in blueberry flower formation needs to be further explored.

AtSPL9/15 was involved in juvenile-to-adult vegetative transformation [15, 16] and might play a functionally redundant role. In Petunia, PhSPL9a and PHSPL9c showed high transcript levels in the inflorescence, but PhSPL9b was expressed in all tissues [57]. In Walnut, JrSBP19, JrSBP20, and JrSBP24 were homologous to AtSPL9/15, and during floral induction, JrSBP19 and JrSBP20 were highly expressed in leaf buds, but JrSBP24 was highly expressed in flower buds [13]. In this study, VcSPL39/46/49/17/23/12/3 were homologous to AtSPL9/15, where the VcSPL expression profile showed that VcSPL23 was highly expressed during PF and FL (Fig. 7A). RT-qPCR results showed that VcSPL23 was significantly more expressed in leaves but not significantly expressed during blueberry bud differentiation (Figs. 7B and 8). AtSPL9 was associated with leaf growth rate and final size [17], and we speculated that VcSPL23 may function mainly in flower bursts and leaves. Tomato SlSPL13 positively regulated the expression of the tomato inflorescence-related gene SINGLE FLOWER TRUSS (SFT) by binding directly to its promoter region, thereby controlling inflorescence development [68]. In Medicago sativa, delayed flowering was observed in SPL13-silenced plants [69]. AtSPL13 was part of Group5 in the evolutionary branch, as was VcSPL20/35/47/51/53. However, VcSPL35 has higher homology with OtSPL16 in the evolutionary branch, and it has been reported that OsSPL16 positively conditions cell proliferation, and OsSPL16 missense mutations generate new alleles that affect plant height and inflorescence development [70], specifically showing dwarfism and shortened inflorescence length [71]. The expression of VcSPL35 was differentially up-regulated during blueberry flower bud differentiation, and it was hypothesized that VcSPL35 played an active role in promoting blueberry flower bud differentiation. VcSPL53 was significantly up-regulated during flower initiation, and was significantly higher in blueberry flower buds than in other tissues (Figs. 7B and 8). Meanwhile, gene structural analysis showed that VcSPL53 had salicylic acid response elements. Under long daylight, salicylic acid (SA) can induce flowering in Wolffia microscopic of short daylight plant [72], salicylic acid treatment can promote more flowering in citrus [73], and salicylic acid (SA) accumulation can inhibit the development of stamen abscission in tung tree [74]. Flower bud differentiation and flowering are two distinct stages in most woody plants [75, 76], and it can also be hypothesized that VcSPL53 is involved in floral initiation and flowering in blueberry. In summary, the blueberry VcSPL gene family has a significant role in blueberry plant growth, leaf development, flowering regulation, and abiotic stress, especially in the process of floral initiation and induction.

Conclusion

In this study, we performed a genome-wide identification and analysis of VcSPL in the blueberry. A total of 56 VcSPL gene families were identified, grouped into six subfamilies on the phylogenetic tree, and the results were further supported by gene structural domain, and motif analysis. 56 VcSPL genes were localized to the nucleus except for VcSPL36, which was localized in both the cytoplasm and the nucleus. In addition, the analysis of evolutionary origins and cis-acting elements provided further insight into the function of the VcSPL gene family, and segmental duplications may play a key role in VcSPL gene amplification. Expression profiles and RT-qPCR results showed that VcSPL genes were not only spatially and temporally specific, but also could play important roles in blueberry plant growth, leaf development, flower development, and abiotic stresses, with some genes acting synergistically. Among them, VcSPL40, VcSPL35, VcSPL45, and VcSPL53 may have critical roles in the floral transition of blueberries. This study will provide a theoretical basis for studying the blueberry flower and fruit development and improving agronomic traits.

Main conclusion

We identified 56 SPL genes by bioinformatic analyses. VcSPL40, 35, 45, and 53 play a role in the blueberry floral transition phase.