Background

Various transcription factors have revealed their critical roles in organism-specific function by activating or suppressing the expression of target genes [1]. The SQUAMOSA promoter binding (like) proteins (SBPs/SPLs) represent a major family of plant-specific transcription factors. SBPs/SPLs proteins share a highly conserved 76 amino acids in length DNA binding domain, also known as SBP binding domain [2]. The first SBP/SPL protein was identified in Antirrhinum majus, and this protein could interact with the promoter sequence of the floral meristem gene SQUAMOSA [3]. As a multigene family, SBP/SPL genes have been characterized from different species ranging from single-cell green algae to multicellular angiosperm [4, 5]. There are 16 SBP/SPL genes identified in Arabidopsis [6], 19 in rice [7], and 41 in soybean [8]. SBP transcription factors play central roles in various aspects of plant development including [2, 9, 10], flower development [11], leaf development [12], plant hormone signaling transduction [13], vegetative to reproductive phase transition [14, 15]. For example, AtSPL3 participates in regulating flowering under long photoperiod, and constitutively expressed SPL3 shows early flowering [6]. AtSPL8 is a central regulator involved in the regulation of microsporogenesis and megasporogenesis. spl8 mutant shows pollen sac development defects, and overexpression SPL8 affects plant fertility by GA-dependent signaling pathway [16]. Moreover, SPL8 and other SPL genes influence gynoecium patterning through mediating auxin homeostasis [17]. In monocot plants, such as rice and maize, SBP genes are also reported to modulate essential developmental processes. Overexpression of OsSPL14 during the reproductive stage significantly promotes panicle branching and increased grain yield [18]. OsSPL16 is also a regulator of grain size, shape, and quality [19]. OsSPL3 regulates crown root development [20]. For maize, SBP proteins encoding genes, unbranched2 and unbranched3, affect plant architecture and yield traits by regulating the lateral primordia initiation [21].

Numerous studies have revealed that many development processes mediated by SBP proteins are closely related to miR156. It is reported that miR156 in Arabidopsis can complementarily bind to the 3′ UTR of SPL3 mRNA, and reduce its expression level through translation repression or transcript cleavage [10, 11]. In rice, overexpression of OsmiR156 decreased the expression of SPL genes, indicating the conserved interaction relationship between SPL and miR156 [22]. Similarly, miR156 targeted OsSPL16 and OsSPL13 control grain shape, size and quality in rice [19, 23]. In switchgrass, miR156/SPL4 module controls aerial axillary bud formation and biomass yield [24].

Sugarcane (Saccharum spontaneum), one of the most economically valuable plant, is a perennial tropical or subtropical crop, contributing up to about 80% of sugar production and 40% biofuel feedstock in the world [25]. Since 2000 years ago, sugarcane has been cultivated as sugar crop in China and India [26]. This domesticated sugarcane cultivar is a cross between species S. officinarum and S. spontaneum and accounts for the major genome information to modern sugarcane cultivars [27]. Although the genome information of S. spontaneum L. is available [28], little progress has been made in sugarcane germplasm improvement through sexual propagation due to the degeneration of sugarcane reproductive organs [29]. Therefore, unveiling the fundamental mechanism of the sugarcane reproductive developmental process is necessary to develop improved varieties [30].

Concerning recent findings of SBPs roles in Arabidopsis, rice, and other plants, analysis of SBP gene function in sugarcane will undoubtedly accelerate sugarcane germplasm improvement. In this present study, we systematically analyzed the SBP gene family of sugarcane for their gene structure, phylogeny, motif and domain composition, miR156 target site, and expression pattern in various tissues and organs. Besides, the interaction between the SBP genes and miR156 was critically examined to study their functional relationship during the reproductive stage in sugarcane.

Results

Identification and characterization of SBP genes in S. spontaneum

To identify of SBP genes in sugarcane, the HMM profile of the SBP domain was used as a query to search the sugarcane genome database and BLASTP program. Initially, 66 putative SBP proteins were identified from the sugarcane genome database. All the resulting sequences were further checked by SMART and pfam tools to confirm SBP domain. Sixteen proteins without SBP (Cys-Cys-His-Cys, Zn2) motif or with incomplete SBP domain were removed. Finally, 50 SBP proteins were identified and used for further analysis. Among them, 13 SBP genes had 2, 3 or 4 alleles, including 7 SsSBPs with 2 allelic genes, 5 SsSBPs with 3 allelic genes and 1 SsSBP with 4 allelic genes. We named these SsSBPs as SsSBP1 to SsSBP30 based on their chromosomal locations and added − 1 to − 4 for their alleles (Table 1). To futher investigate the conserved status of the SBP domain, 30 SBP protein sequences from sugarcane were aligned to predict conserved domains. The alignment results showed that all SsSBP proteins contained the complete SBP domain and possessed the typical characteristics of SBP domain with two Zinc motifs (Zn1 and Zn2) and one nuclear localization signal (NLS) (Fig. 1).

Table 1 The characteristics of identified SBP genes in sugarcane
Full size table
Fig. 1
figure 1

Multiple alignment of the highly conserved SBP domains

Full size image

The detailed information about the SsSBPs was deduced by ExPASy server, including protein length, molecular weight (MW), theoretical isoelectric point (pI) and the grand average of hydropathicity (GRAVY). The length of the SsSBPs ORF region varied from 570 bp (SsSBP5) to 3195 bp (SsSBP3–3) and the protein lengths ranged from 189 to 1064 amino acids. The MW of the proteins ranged from 19.3214 to 116.09958 kDa. The pI ranged from 5.37 to 10.32, and the values of GRAVY were all negative, suggesting that all SsSBPs are hydrophilic. Moreover, the subcellular localization of 50 SsSBP proteins was predicted by ProtComp software and found that all SsSBP proteins localized in the nucleus except SsSBP4 and SsSBP12 proteins, which have no NLS signal, and localize in the cell membrane (Fig. S1; Table 1).

Phylogenetic analysis of the SBP gene families

We selected a total of 293 SBP homologs from 17 representative species from 7 green plant families, including chlorophytes, bryophytes, lycophytes, gymnosperms, basal magnoliophytes, eudicots and monocots, for phylogenetic analysis of SBP. Among them, both Ostreococcus sp. RCC809 and Ostreococcus Lucimarinus in green algae had only one SBP gene; 46 SBP genes exist in soybean. In comparison to the number of genes in these species, SBP genes in S. spontaneum showed an obvious expansion in the number of genes (Fig. 2). To gain further insight into the phylogenetic relationship of SsSBP genes, a phylogenetic tree was constructed using SBP proteins from Arabidopsis, Vitis vinifera, Ananas comosus, Sorghum bicolor and Oryza sativa (Fig. 3). SBP genes from these different species could be classified into 8 groups (I to VIII), and SBP proteins also tend to cluster the similar group. As expected, SsSBPs exhibited a closer relationship with the SBP proteins from S. bicolor and O. sativa. Group V and VII contained maximum SsSBP genes, where SBP genes from S. bicolor and O. sativa were also grouped. While the group I contained only 2 members of SsSBP genes formed the smallest group. This result was in agreement with the conservation analysis of the SBP proteins in other plants like Arabidopsis, grape, rice and sorghum. For example, a relatively high homologous genes, AtSPL6 / SbSBP5 / OsSPL1 / SsSBP8 / SsSBP9 clustered in one evolutionary branch (Fig. 3A). In addition, a ML phylogenetic tree was also constructed based on gene sequence similarity of 50 SsSBP proteins. The result indicated that the alleles of each SsSBP gene cluster in the same group, indicating that their sequences have high homology (Fig. 3B).

Fig. 2
figure 2

Distribution of SBP genes in 17 species

Full size image
Fig. 3
figure 3

Maximum likelihood phylogenetic tree of SBP gene family

Full size image

Structure characterization of SBP genes in S. spontaneum

To better understand the genetic diversity of the SsSBP genes, the coding sequence of each gene was compared with their corresponding genomic sequence. The result revealed that the exon of SsSBP genes ranged from 2 to 13 in number (Fig. 4). SsSBP genes in group I contain 4–7 introns, group V contains 2–4 introns, group VI contains 1–2 introns, most of SsSBP members in group VIII contain 2 introns. Interestingly, alleles of SsSBP genes (SsSBP1–1/2, SsSBP3–1/2/4, SsSBP4–1/2, SsSBP15–2/3, SsSBP17–1/2, SsSBP18–1/2, SsSBP20–1/2, SsSBP21–1/2 and SsSBP24–1/2/3) had the same number of exon/introns, although the length of introns varied. Some alleles possessed different exon/intron numbers (SsSBP2–1/2/3, SsSBP7–1/2/3, SsSBP10–1/2, SsSBP11–1/2). The SsSBP3 allele gene SsSBP3–3 possessed the largest number (13) of exons, but the SsSBP3 alleles SsSBP3–1/2/4 only had 3 exons (Fig. 4). These results indicate that the number of exon and intron are diverse in different groups yet nearly consistent within the same group.

Fig. 4
figure 4

The exon/intron structure of sugarcane SBP genes

Full size image

The SsSBP genes clustered into the same group exhibited similar structure and possessed a similar motif sequence. A total of 20 motifs were identified in SsSBP proteins, designated as motif 1–20 (Fig. 5, Fig. S2). The result expectedly showed that all SBP members contain Motif 1, Motif 2, Motif 3, Motif 5 and Motif 6, which was annotated as the SBP domain. Most of SBP members within the same group exhibited similar motif composition and exon-intron structure. In contrast, some motifs were found to be specific to one or two groups of SsSBP proteins. Motif 8 only appeared in group VIII. Motif 14, Motif 17 and Motif 19 were only present in group II, suggesting that proteins in these groups may have a specific biological function under a given condition. At the same time, the divergence among the different groups indicated their diverse functions.

Fig. 5
figure 5

Motif composition of sugarcane SBP proteins. The motif numbers from 1 to 20 are displayed in different colored boxes

Full size image

Chromosome distribution and gene duplication of SsSBP genes

The chromosome distribution information of SsSBP genes revealed that 49 of the 50 SsSBP genes are located to the eight chromosomes of S. spontaneum, with the SsSBP30 mapped to the unanchored scaffolds (Fig. 6). On chromosomes 1 and 7, only two SsSBP genes were found. Chromosomes 3 and 5 contain five SsSBP genes. Chromosome 2 had the maximum number of SsSBP genes with 12 members, followed by chromosome 6 with 11 SsSBP genes. In addition, 27 synteny gene pairs were identified in sugarcane using MCScanX software, with 24 pairs of alleles and 3 pairs of nonalleles. It should be defined as a tandem duplication event if a chromosomal region within 200 kb containing two or more genes [31]. According to this criterion, only two tandem duplications (SsSBP4–3/SsSBP6 and SsSBP17–2/SsSBP19) were noticed (Fig. 6A, Table S3). These results indicate that segmental duplication events might significantly contribute to the SsSBP gene expansions than tandem duplication.

Fig. 6
figure 6

Gene location and synteny in sugarcane. A Synteny analysis of SBP genes within sugarcane. B Synteny analysis of SBP genes between sugarcane and sorghum

Full size image

To further analyze the evolutionary process of SsSBP genes, a comparative analysis of genome synteny blocks between S. spontaneum and Sorghum bicolor was conducted. Sorghum is the closest related diploid to sugarcane, and the comparison of gene structures between these two species provided clues to the evolutionary gene events caused by polyploidization. A total of 37 syntenic gene pairs between S. spontaneum and S. bicolor were found (Fig. 6B, Table S3). To further understand the evolutionary forces on SsSBP genes, the ratio of the synonymous (Ks) and nonsynonymous (Ka) substitutions rate (Ka/Ks) was calculated for estimating the selection pressure of homologous genes, where Ka/Ks < 1 indicates purifying selection, Ka/Ks = 1 means neutral selection and Ka/Ks > 1 indicates positive selection [32]. In this study, with the exception of three gene pairs SsSBP4–1/SsSBP5 (1.0379), SsSBP3–2/SsSBP3–4 (1.98497), SsSBP24–2/SsSBP24–3 (1.00893), Ka/Ks ratios of SsSBP homologous genes were less than 1, indicating that these genes probably underwent a purifying selection (Table S2). Similarly, most Ka/Ks values of sorghum genes were also less than 1, suggested that SBP genes of these two close species underwent a strong purifying selection to reduce adverse mutations after duplication during the evolutionary process (Table S3).

miR156 target prediction, distribution and expression pattern analysis

Previous studies showed that miR156 complementarily binds to SBP genes either at the coding or 3’UTR region and reducs gene expression level through translation repression or transcript cleavage [11, 24]. In this present study, 22 SsSBP genes were found as the targets of miR156, and these genes were mainly distributed in groups V, VI, VII and VIII (Fig. 7A). Among these SsSBP genes, miR156 complementary sequences were at their coding regions except SsSBP5 where miR156 binding in the 3′-UTR (Fig. 7A; Fig. S3). Interestingly, the allelic genes SsSBP3–1, SsSBP3–2, SsSBP3–3 and SsSBP3–4 were all the targets of miR156 (Fig. 7A). Although there are 20 miRNA members of Saccharum sp., only one putative miRNA156 in S. spontaneum is represented in the miRbase database (https://www.mirbase.org/ v 22.1). Therefore, we performed a genome-wide study and found 29 members of miR156 genes in S. spontaneum genome. These Ssp-miR156 genes were mainly distributed on chromosomes 2, 3, 4, 5, 6, 8, except for chromosomes 1 and 7 (Fig. S4; Table S4). Ssp-miR156a, Ssp-miR156f, Ssp-miR156j and Ssp-miR156k as alleles were localized on chromosome 2A. Chromosome 3A possesses 4 Ssp-miR156a alleles. Chromosome 3B contains 3 miR156 members (2 Ssp-miR156a and 1 Ssp-miR156l). Chromosome 4A has 4 miR156 members (3 Ssp-miR156a and 1 Ssp-miR156i), followed by 2 and 3 members on chromosome 4B (Ssp-miR156a and Ssp-miR156d) and 5A (2 Ssp-miR156a and 1 Ssp-miR156e), respectively. Two Ssp-miR156 were on chromosome 8A (Ssp-miR156e and Ssp-miR156k) and 8D (2 Ssp-miR156b). Only 1 Ssp-miR156 was found on chromosome 2C (Ssp-miR156b), 2D (Ssp-miR156b), 3C (Ssp-miR156a), 6A (Ssp-miR156a) and 6B (Ssp-miR156a). However, no Ssp-miR156 members was found on chromosomes 1 and 7 (Fig. S4; Table S4).

Fig. 7
figure 7

Sequence alignment and expression patterns of miR156 compare with SsSBP genes. A Sequence alignment of the miR156 complementary sequences with the target sites in SsSBP genes. B The expression patterns of miR156 and their targets in different tissue samples. C-J The confirmation of miR156s and their targets by degradome ananlysis

Full size image

To further gain insight into the role of miR156 during female gametophyte development, we studied the miR156-SBP module during female gametophyte development. The results showed that the expression level of miR156 was mostly enriched in the mature stage of female reproductive development, and relatively low expression levels were found during the stages of AC (Archesporial Cell) to MMC (Megaspore Mother Cell). Generally, the expression level of miR156 increased from the initial stage to the mature stage of the female gametophyte by sRNA-seq analysis (Fig. 7B; Table S5). In addition, the expression profiles of Ssp-miR156 precursors were also quantitatively verified using RT-PCR and qRT-PCR analysis. The results for the relative expression of Ssp-miR156 were consistent with the sRNA-seq data (Fig. S5). On the contrary, the expression level of target SsSBP genes was mostly decreased during the female gametophyte development stages, such as the target SsSBP11–1, SsSBP21–2, SsSBP22 and SsSBP30 (Fig. 7B).

To verify the authenticity of miR156-SBP module in sugarcane, we performed the degradome analysis and found that miR156 family members target the SsSBPs. The miR156k could bind the site 1400 bp of its target SsSBP2–1 (Fig. 7C). Similarily, miR156a binds on the SsSBP3–2 (site 862) (Fig. 7D), SsSBP7–1 (site 850) (Fig. 7E), SsSBP7–3 (site 1110) (Fig. 7F), SsSBP10–1 (site 568) (Fig. 7G), SsSBP10–2 (site 1096) (Fig. 7H), SsSBP11–1 (site 1140) (Fig. 7I) and SsSBP21–1 (site 1065) (Fig. 7J). Taken together, these results suggest that miR156-SBP module is highly conserved, and the regulation pattern has diverged in different species.

Expression profiles analysis of SsSBP genes

To study spatiotemporal expression patterns of SsSBP genes, RNA-seq data of different organs and tissues were analyzed. The expression level of SsSBP genes of leaf development and female reproductive organs is shown by heatmap representation (Fig. 8). As illustrated in Fig. 8A, SsSBP4, SsSBP6, SsSBP13, SsSBP14, SsSBP18, SsSBP21 and SsSBP26 sustained low expression level in sugarcane leaf gradient segments, while SsSBP1 and SsSBP20 showed high expression in the leaf gradient segments. The transcript levels of SsSBP7, SsSBP10, SsSBP19, SsSBP22, SsSBP28 and SsBP29 decreased gradually from base to mature zone of leaf in sugarcane, showed that gene expression decreased following the maturing leaf (Fig. 8A, Table S6).

Fig. 8
figure 8

Expression profiling of SsSBP genes in different sugarcane tissues. The color bar shows the gene expression with log2 (FPKM+ 1). Green colors indicate low expression and red colors indicate high expression

Full size image

To investigate the SsSBP genes involvement in sugarcane female reproductive organ development, the transcription level of all SsSBP genes was extracted from RNA-seq data of sugarcane female reproductive organs. The heat map represents expression levels in the lines at five developmental stages shown in Fig. 8B. Many SsSBP genes showed different expression patterns among these five development stages. SsSBP1 and SsSBP10 were highly expressed in different stages of female gametophyte development. The transcripts of 7 SsSBP genes (SsSBP13, SsSBP14, SsSBP15, SsSBP16, SsSBP17, SsSBP18, SsSBP19) were zero in all these samples. The expression level of SsSBP7 and SsSBP30 showed differential expression during the female gametophyte development, revealing that these two genes may play an important role in AC and MMC stages (Fig. 8B, Table S7).

We also performed qRT-PCR experiments to confirm the expression level of some SsSBP genes in those different female developmental stages. As shown in Fig. S6, the results of qRT-PCR data were highly consistent with the RNA-seq data for the relative expression of SsSBP genes during the female gametophyte development. Further studies may focus on the role of these genes on female reproductive development.

Discussion

Sugarcane (S. spontaneum) has been widely domesticated and cultivated for thousands of years for its excellent economic values. It has become essential industrial material for sugar sources [25, 26]. The high quantity of genome data and abundance of increasing high-throughput transcriptome data make it possible to explore gene functions in non-model plants like Saccharum spp. Although the genome information of S. spontaneum L. is available, little progress has been made in sugarcane germplasm innovation and improvement due to the degeneration of sugarcane reproductive organs [28, 29]. Previous studies revealed that SBP genes play crucial roles in plant development, especially in flower development, signaling transduction, and vegetative to reproductive phase transition [13,14,15]. However, the functions of S. spontaneum SBP genes remain unknown, although 17 SPLs were identified in sugarcane without taking alleles into account [33]. As for sugarcane genomic autopolyploidization, we conducted the genome-wide identification of SBP genes and their alleles in S. spontaneum, which resulted in the identification of 30 SBP genes (Fig. 1, Table 1). The number of SBP genes in S. spontaneum was similar to that in P. trichocarpa (28), O. zativa (19), and S. bicolor (19), but smaller than that in G. max (46), indicating that SBP genes in different species underwent different gene duplication events. Based on phylogenetic and gene structure analysis, SsSBP genes could be divided into eight groups (group I-VIII), which is consistent with the results of previous studies on SBP genes [34].

In general, the members of SBP genes clustered into a subgroup shared similar gene structure and functions, suggesting these genes underwent common evolutionary origins. In other words, gene duplication events (segmental and tandem duplication) are the major driving forces for evolution and gene expansion by which many paralogous gene pairs are produced and could help organisms cope with different developmental processes [35]. In our study, a total of 27 duplication events were found in SsSBP genes, consisted of segmental duplications and tandem duplication (Fig. 6). The Ka/Ks ratio is reported as the criterion for estimation the gene duplication. The Ka/Ks ratio of a given > 1 means that the gene has experienced positive selection, = 1 suggests neutral selection and < 1 indicates purifying selection. Based on the values of Ka/Ks ratio, all the SsSBP gene pairs were duplicated under purifying selection except gene pairs SsSBP4–1/SsSBP5 (1.0379), SsSBP3–2/SsSBP3–4 (1.98497), SsSBP24–2/SsSBP24–3 (1.00893) (Table S3). The diversity of SsSBP genes is likely to be motivated by gene duplication and genomic structure variation during the evolutionary process.

Up to now, there is little functional information on the SBP genes of sugarcane. Generally, the gene functions, to a large extent, are correlated to their expression patterns. In this present study, the expression levels of 30 SsSBP genes were examined across the four different leaf gradient segments and five female gametophyte development stages (Fig. 7). Most SsSBPs were predominantly expressed in the initial developmental stages of either leaf development or female gametophyte development. These results were similar with other species in the apical meristem, including apical buds inflorescences and flower buds [9, 10, 22]. Among the SBP genes in Arabidopsis, AtSPL1 and AtSPL12 expressed highly in inflorescences and overexpression of these two genes enhanced the inflorescence thermotolerance [36]. AtSPL2, AtSPL9, AtSPL10, AtSPL11, AtSPL13 and AtSPL15 were reported to control the determination of leaf shape and the transformation of vegetative to reproductive stages [37]. Interestingly, the evolutionary analysis showed that AtSPL1 and AtSPL12 are highly orthologous to SsSBP genes in group II, including SsSBP1, SsSBP20 and SsSBP25. AtSPL2, AtSPL9 and AtSPL10 are orthologous to SsSBP genes in the group V with SsSBP10, SsSBP11, SsSBP12, SsSBP16, and SsSBP29. Based on the belief that homologous genes perform similar functions. SsSBP1 and SsSBP10, which were expressed highly in female gametophyte stages, would be involved in the development of female reproductive organs in sugarcane. Three genes SsSBP1, SsSBP20, SsSBP25 grouped with SsSBP16, SsSBP29 and SsSBP30, which are orthologous to AtSPL2, AtSPL9 and AtSPL10, expressed highly in the sugarcane leaves, confirming their roles in the regulation of leaf development. Certainly, additional studies need to be performed to confirm the potential roles in female gametophyte development (for SsSBP1 and SsSBP10) and leaf development (for SsSBP1, SsSBP20, SsSBP25, SsSBP16, SsSBP29 and SsSBP30).

In addition, miR156/SBP module has been reported to govern many aspects of plant growth and development [10, 17, 24, 38]. Overexpression of miR156 in Arabidopsis significantly repressed the SPL transcription and resulted in the loss of apical dominance, leading to dwarfism, an increase in total leaf number, and plant biomass [39]. Meanwhile, the expression levels of the target SBP genes of miR156 were suppressed in the miR156 overexpressing plants [10, 37]. In the present study, the transcript level of miR156 was abundant in the mature stage of female reproductive development (Fig. 7B). In contrast, most putative target SsSBP genes predicted miR156 target sites showed lower expression levels in these tissues (Fig. 7B). These results suggested that the transcript of miR156 is negatively correlated with the expression of most SsSBP genes (Fig. 8). All together, our results revealed that miR156/SBP module could be used as an important tool to genetically improve crop architecture and productivity.

Conclusion

A total of 30 SBP genes were identified in sugarcane (S. spontaneum) by genome-wide analysis. These SsSBP genes were comprehensively characterized and classified into eight groups. The phylogenetic analysis showed that these genes shared orthologous relationships of SBP members from Arabidopsis and rice. The spatiotemporal expression patterns of these SsSBP genes in different tissues indicate that SsSBP genes may regulate the leaf and female gametophyte development. Our results also showed that miR156 targeted many SsSBP genes. The expression level of miR156 was enriched in the female reproductive mature stages. The different expression levels between the miR156 and SsSBP genes in diverse tissues suggested that miR156/SBP module plays a crucial role in the leaf and female gametophyte development processes (Fig. 9). Taken together, our study provides the foundation for future in-depth elaboration of the potential functions of the SBP genes in the growth and development of sugarcane.

Fig. 9
figure 9

A summary work model of this study for SBP genes in sugarcane

Full size image

Methods

Identification and annotation of SBP genes in sugarcane

Sugarcane genome data, CDS, protein sequence and annotation data were downloaded from the sugarcane Genome database (http://sugarcane.zhangjisenlab.cn/sgd/html/index.html) [28]. Arabidopsis and other species sequences were searched and downloaded from Phytozome v13 (https://phytozome.jgi.doe.gov/pz/portal.html) [40]. To identify the SBP genes in sugarcane, the HMM profile of the SBP domain (PF03110) was downloaded from the Pfam database (http://pfam.xfam.org/) [41] and used as the query to search the sugarcane genome database. SBP homologs were obtained by running a local BLASTP search using the Arabidopsis and rice SBP sequence as a query against the given protein database with an E-value cutoff of 10− 5. The candidate genes were further confirmed by SMART server (http://smart.embl-heidelberg.de/). Sequences without the complete SBP domain were deleted. Finally, all the candidates were confirmed by multiple sequence alignments using DNAMAN software to ensure they contained the SBP domain. ExPASy (https://www.expasy.org/) [42] server was used to calculate the detailed information about the SsSBPs in sugarcane, such as molecular weights (MW), isoionic point (pI), and the grand average of hydropathicity (GRAVY). The subcellular localization of the SBP proteins identified was obtained using the ProtComp (v.9.0) software (http://www.softberry.com/).

Gene structure, sequence alignments and phylogenetic analysis of SsSBP genes

The exon/intron structure of SBP genes was analyzed using Gene Structure Display Server (http://gsds.cbi.pku.edu.cn/index.php) [43] by comparing their coding and genomic sequences. Using BLASTP program to search homologous gene pairs among sugarcane and sorghum with the parameter of e-value = 1e-10. The estimation of selection and substitution rates, the non-synonymous (Ka), synonymous (Ks) and Ka/Ks substitution ratios of the homologous gene pairs of sugarcane and sorghum were calculated by the easy Ka/Ks calculation program. MCScanX software [44] was used to detect the gene synteny and collinearity in sugarcane, and the SBPs locations were shown using Circos software [45]. Multiple sequence alignment of SBP protein sequence from Arabidopsis, rice, and sorghum was conducted using the MUSCLE in MEGA (v.6.0) [46]. A phylogenetic tree was constructed using RAxML software (http://www.phylo.org/index.php/) using the maximum likelihood (ML) method with bootstrap 1000 replications. The phylogenetic tree was displayed and manipulated using the Interactive tree of life (iTOL, https://itol.embl.de/) [47,48,49].

Conserved motif identification, miR156 target site prediction and distribution

The conserved motifs of SsSBP proteins were identified using the online program MEME (http://meme-suite.org/tools/meme) [50] with the default setting parameters: maximum number of motifs to find was 20; minimum width of motif was 6 and maximum width of motif was 50. The sequence logos of the SsSBP domain were showed by TBtools [51]. To predict the putative target sites of miR156, the cDNA sequences of SsSBP genes were analyzed using psRNATarget tool (http://plantgrn.noble.org/psRNATarget/). The chromosome location information of the Ssp-miR156s and SBPs were searched in sugarcane genome databases, and MapInspect software was used to generate chromosomal distribution information.

Plant material and sample preparation

The sugarcane (S. spontaneum L.) cultivar Yuetang 91–976 was grown and collected by State Key Laboratory for Conservation and Utilization of Subtropical Agro-Bioresources (Guangxi, China), and all samples from this cultivar was adopted for all experiment. When the plants reached the age of florescence stage, five different stages of the sugarcane female gametophyte development (i.e., AC, MMC, Meiosis, Mitosis and Mature) were collected. All samples were harvested as three biological replicates, which were quick-freeze with liquid nitrogen and stored at ultra-low temperature to facilitate the extraction of RNA.

RNA extraction, expression profiles and qRT-PCR analysis

Total RNA was isolated by the Omega Total RNA kit II (R6934–02, USA). The evaluation of RNA quality was performed by the gel electrophoresis and 2000 spectrophotometer assessment at 260 nm (NanoDrop, Thermo Fisher Scientific), and Illumina sequencing was done using the method of Zhao et al. (2018).

For qRT-PCR analysis, the cDNA was synthesized using the ThermoScript RT-PCR kit (Life Technologies) in a 20 μL volume reaction under the program: 42 °C for 15 min and 85 °C for 15 s. According to the SYBR Premix RT reagent kit system (TaKaRa, Dalian, China), the reaction contains 1 μg RNA prior to qRT-PCR.

To understand the expression profiles of SBP genes, the RNA-seq data of leaf development were downloaded from the Saccharum Genome database (http://sugarcane.zhangjisenlab.cn/sgd/html/index.html). The RNA-seq data of female reproductive development have been deposited in the European Nucleotide Archive (ENA, accession number PRJEB44944). Different leaf developmental stages, including basal zone, a transitional zone, a maturing zone, a mature zone [52, 53], and the female reproductive development stages, AC, MMC, Meiosis, Mitosis and Mature, were used for the study. The RNA-seq raw reads were filtered by Trimmomatic software with default parameters to obtain clean reads. The clean reads were mapped to the reference genome using Hisat2 [54]. Gene expression was calculated by Cufflinks software [55]. The log2-transformed RPKM value of the expression patterns of SsSBP genes was used to generate the heatmap using the pheatmap package in R software. The expression pattern of miR156 was calculated by count values according to the miRNA-seq data.

To further confirm the expression profiles of the SsSBP genes, qRT-PCR assays were performed in different female reproductive development stages. qRT-PCR was conducted in CFX96 Real-Time System (Bio-Rad) using SYBR Green (TaKaRa) according to the instructions. Each reaction contains 12.5 μL SYBR mixture, 1.0 μL specific primer and 1 μg sample template. Three replicate reactions were performed for each sample under the following program: 95 °C for 30s; 40 cycles of 95 °C for 5 s; 60 °C for 30 s. The primers used in this study are listed in Table S1.