Evolutionary assessment of SQUAMOSA PROMOTER BINDING PROTEIN-LIKE genes in citrus relatives with a specific focus on flowering

Abstract

Phase transition and floral induction in citrus requires several years of juvenility after germination. Such a long period of juvenility has been a major hindrance to its genetic improvement program. Studies have shown that miR156 along with its downstream genes SQUAMOSA PROMOTER BINDING PROTEINS (SBP) and SBP-LIKE (SPL) mediate the phase transition and floral induction process in plants. Our current study has systematically analyzed SPLs in 15 different citrus-related species, systematically annotated them based on their close homology to their respective Arabidopsis orthologs, and confirmed the functional attributes of the selected members in floral precocity. The majority of the species harbored 15 SPLs. Their cis-element assessment suggested the involvement of the SPLs in diverse developmental and physiological processes in response to different biotic and abiotic cues. Among all, SPL5, SPL9, and SPL11 stood out as consistently differentially expressed SPLs in the adult and young tissues of different citrus-related species. Independent overexpression of their F. hindsii orthologs (FhSPL5, FhSPL9, and FhSPL11) brought an enhanced expression of endogenous FLOWERING LOCUS T leading to the significantly precocious flowering in transgenic Arabidopsis lines. Future study of the genes in the citrus plant itself is expected to conclude the assessments made in the current study.

Graphical Abstract

Core

By systematically identifying the members of the SPL family and comparing their expression levels in in the adult and young tissues of different citrus-related species, it was found that SPL5, SPL9, and SPL11 play an important role in the regulation of childhood in citrus. Independent overexpression of their F. hindsii orthologs (FhSPL5, FhSPL9, and FhSPL11) brought an enhanced expression of endogenous FT leading to the significantly precocious flowering in transgenic Arabidopsis lines.

Gene & accession numbers

Most of their genomes are included in the local Citrus Pan-genome to Breeding database (CPBD; http://citrus.hzau.edu.cn/). Some genomic data has not been published, please contact the author if necessary. Sequence data from this article can be found in the database of the National Center for Biotechnology Information (NCBI) under the accession numbers: FhSPL5(OR234301), FhSPL9 (OR234302), FhSPL11 (OR234303), CsActin (Cs1g_pb000860), AtFT (NM_001334207.1), AtAcin (NM_112764.4).

Introduction

Juvenile citrus plants are characterized by thorniness and vigorous growth unlike their frequently flowering adult counterparts (Furr et al. 1947; Hendry 1982). As compared to the annual herbaceous plants, the very first floral induction (after phase transition) in citrus-like perennial plants take a relatively much longer time after germination. While some citrus species start flowering at age as early as 4-6 months after germination in a tropical climate, others spend 4-15 years of juvenility before flowering (Table 1). The long juvenility of the majority of citrus species has been a serious hindrance to rapid yield gain and breeding cycle. Hence, several attempts have been made and practiced to circumvent the issue since long (Furr et al. 1947; Moss 1969; Goldschmidt et al. 1985; Endo et al. 2005; Soares et al. 2020). However, there has not been a significant breakthrough in reducing juvenility in the species yet. It still necessitates better know-how on the molecular and physiological processes behind phase transition and floral induction.

Table 1 Number of SPLs discovered in citrus members and relatives

Flowering in higher plants like citrus and Arabidopsis is tightly regulated by the interplay of diverse molecular players. Often multiple pathways (vernalization, autonomous, photoperiod, hormone, and age pathways) converge during the floral induction (Teotia and Tang 2015; Kim 2020). Compared to others, age-dependent floral induction is fail-safe in plants as it occurs even in non-inductive conditions. However, it essentially depends on the de-repression of miR156-regulated SQUAMOSA PROMOTER BINDING-LIKE (SPL) genes (Wu et al. 2009). Such change is apparently threshold dependent as the progressive decrease in miR156 transcript abundance in the subsequent leaves facilitates leaf morphogenesis (He et al. 2018). miR156 and one of its target SPL member, SPL9 oppositely regulate the age-dependent activation of SUPPRESSOR OF OVEREXPRESSION OF CONSTANS 1 (SOC1), a crucial MADS-box floral activator in Arabidopsis (Wang et al. 2009). An orthologous transgenic study with C. sinensis derived SOC1 showed that its overexpression in Arabidopsis leads to precocious flowering and delayed senescence of the flowers in the plants (Tan and Swain 2007), indicating the functional conservation of CsSOC1. Whether CsSPL9 too is involved in its regulation in citrus was yet to be unraveled.

Apart from their role in age-dependent flowering, citrus studies have shown the involvement of miR156 in the positive regulation of somatic embryogenesis (Wu et al. 2011; Wu et al. 2015; Long et al. 2018), cellular starch accumulation (Liu et al. 2017), potential involvement in male sterility (Fang et al. 2016), etc. However, the detailed mechanism behind how miR156-SPL module shapes such phenotypes is yet to be elucidated. Moreover, even though there had been some studies on individual citrus species, they lack consensus on gene nomenclature.

In the current study, we have carried out a systematic assessment of SPLs from 15 representative citrus-related species (Aurantioideae family members) that included the earliest flowering Fortunella hindsii to late flowering citrus wild relative Aegle marmelos. Our study showed that SPLs are highly conserved among citrus members in their number and basic sequence features. Among 15 SPLs discovered in the majority of species, SPL7/8 and SPL3/4/5 were found to be the putatively most distant and the most recently evolved members respectively. We further assessed and confirmed the potential of F. hindsii derived SPLs (FhSPL5, FhSPL9, and FhSPL11) in floral precocity via orthologous overexpression in Arabidopsis.

Results

Citrus SPLs exhibit highly conserved sequence features

SPLs were retrieved from respective genomes of 15 citrus-related species, most of which are included in the local Citrus Pan-genome to Breeding Database (CPBD; http://citrus.hzau.edu.cn/) (Liu et al. 2022). Unlike Arabidopsis, which harbors 17 SPLs (including two identical copies of AtSPL13s), most of the citrus-related members (13 out of 15 species) appear to harbor 15 SPLs (Tables 1, S1). Interestingly, a manual homology search revealed that 14 SPL harboring M. paniculata too harbors the genomic region for its missing SPL (putative MpSPL4). However, it constitutes multiple deleterious point mutations within its putative exons. We analyzed the target sites of miR156 in the SPL gene family taking an earlier study by Liu et al. (2017) as a reference (Figure S1, Table S2). The results showed that except for SPL1, SPL7, SPL8, SPL12, and SPL14, all other SPLs harbor putative miR156 target sites (Table 1).

Based on the studies in Arabidopsis (Guo et al. 2008; Xing et al. 2010), we divided the miR156-targeted citrus SPLs into two distinct subgroups- members with smaller peptides- represented by SPL3 (SPL3, SPL4, and SPL5), and those with larger peptides- represented by SPL9 (SPL2, SPL6A, SPL6B, AtSPL9, SPL10, SPL11, SPL13A, SPL13B, and SPL15). All SPL proteins harbored the 79 amino acid long signature domain comprising two zinc finger binding motifs (C3H or C4 and C2HC) and a bipartite nuclear localization signal (KR-X11-RRR/K) overlapped with the latter zinc finger (Figs. 1 and S2). Overall, sequences of each SPL-types were highly identical among most of the assessed species except for A. marmelos, the most distant relative taken in the study. Notably, SPL8 and SPL10 were located tandemly in a tail-to-tail fashion with just ~7 kb intermediary region in all species.

Fig. 1

SBP domain-specific motifs are conserved in citrus-derived SPLs. a Conserved SBP domains in all SPL types derived from a representative member Aegle marmelos. The domain-specific conserved motifs are denoted on the top. b A higher degree of overall sequence conservation within a particular SPL type (SPL5 as a representative member) derived from the citrus relatives

SPLs exhibit distinct evolutionary patterns among citrus relatives

The phylogenetic tree was developed from the SBP domains of the citrus-related species and Arabidopsis. Based on their clustering, we divided the SPLs into three groups- Cluster-I (constitutes SPL7s), Cluster-II (constitutes citrus SPL3s), and Cluster-III (constitutes all remaining SPLs) (Fig. 2a). Interestingly, some SPL orthologs have specifically been duplicated and diversified among citrus members as compared to their Arabidopsis counterparts which include SPL4/SPL5 (close orthologs of AtSPL5), SPL6A/SPL6B (close orthologs of AtSPL6), and SPL13A/SPL13B (close orthologs of fully identical AtSPL13A/AtSPL13B) (Fig. 2a). Some SPL members showed reduced duplication among citrus-related species, which include SPL3 (close ortholog of AtSPL3/AtSPL4), SPL14 (close ortholog of AtSPL14/AtSPL16), SPL11 (close ortholog of AtSPL2/AtSPL10/AtSPL11), and SPL9 (close ortholog of AtSPL9/AtSPL15) (Fig. 2a). Among all, one SPL member, SPL10, appeared to be unique to the citrus-related species which was not clustered with any of the AtSPLs indicating its unique evolution at least among the citrus-related members. SPL1 and SPL12 were clustered together instead of clustering with their respective AtSPL orthologs suggesting for the independent duplication/diversification among citrus-related species and Arabidopsis. Interestingly, two SPL members, SPL7 and SPL8 were accompanied by their respective AtSPLs indicating for their evolutionary and functional conservation in Arabidopsis and citrus relatives.

Fig. 2

Evolutionary relationship, gene structure, and protein motif assessment of SPLs. a Neighbor-Joining phylogeny of the SBP domains of citrus-related species and Arabidopsis (1000 bootstrap, Poisson correction method). The area of each compressed sub-cluster is relative to the diversity of the compressed sequences. n represents the number of species harboring the respective SPL genes. b Gene structures (upper) and conserved protein motifs (lower) of each SPL type from a representative species F. hindsii. Blue squares, yellow squares, and black lines of the gene structure represent UTRs, exons, and introns respectively; each colored box in the SPL proteins represents a motif, and its size corresponds to the motif length. c Sequence information of the most conserved 10 protein motifs among the SPLs in b

Gene structure and motif conservation assessment of the SPLs

To ascertain whether the phylogenetic relationship was also by the gene structure and protein motifs similarities, we assessed such features in all SPLs. As expected, the closely clustered SPLs, exhibited similar gene (exon/intron) structures and conserved motifs (Figs. 2b, S3, and S4). Of 15 SPLs, the SPL1/SPL7/SPL12/SPL14 are the longest SPLs in both gene and peptide length. Among the most conserved 10 motifs predicted, motif4+motif1+motif2 encompassed the SBP domain, which was conserved in all SPLs as expected (Fig. 2b and c). Additionally, a C-terminally located motif-6 constitutes part of the protein sequence putatively involved in the transmembrane binding of the SPL1s, SPL7s, SPL12s, and SPL14s. It is notable that SPL6A too harbors motif-6, even though no transmembrane helix was predicted in it (Figs. 2b, c, and S5). While it is very plausible that the other six conserved motifs (motif3, motif5, motif7, motif8, motif9、motif10) play role in the proper functioning of the SPLs harboring them (all SPLs in our study), their functional relevance at this point is yet unclear.

Cis-regulatory element (CRE) assessment

SPL members are involved in various development processes. To have a general overview of their nature of expression, we fed the 2.0 kb regions upstream of the translation start site of each SPL to the PlantCARE database and retrieved potential transcription factor binding cis-element within. They were later visualized using TBtools (Figure S6) and boxplots (Fig. 3). Citrus SPL promoters harbor cis-elements responsive to light, methyl jasmonate, abscisic acid, auxin, salicylic acid, wound, drought, defense, gibberellin, low temperature, circadian rhythm, and other response factors. Among the predicted cis-elements those responsive to light were the most prevalent in all SPL promoters with SPL10 and SPL13 promoters harboring them at the highest and lowest number respectively (Fig. 3a). Interestingly, among all, SPL11 and SPL13 promoters harbor a relatively higher number (2) of circadian-associated cis-elements as compared to others (0-1) (Fig. 3b). Additionally, most of the SPL promoters harbor just a single temperature-responsive element except for that of SPL8. Promoters of SPL3/5/13A/13B promoters, on the other hand, constitute none (Fig. 3c).

Fig. 3

The number of cis-acting elements predicted on the promoter of SPL genes in 15 citrus-related species

Interestingly, the SPL promoters harbored very few cis-elements potentially responsive to either defense/stress, salicylic acid (SA), wound, methyl jasmonate (MeJA), and/or drought (Fig. 3d-h). While the majority of the SPL promoters harbored single auxin and gibberellin responsive elements, they constitute a relatively higher number of abscisic acid (ABA) responsive elements with the SPL10 promoters harboring them in the highest number (Fig. 3i-k).Most of the SPL promoters harbor just a single cis-element each associated with the meristem, endosperm, and seed-specific regulation. However, SPL6B promoters lacked any meristem regulation related; SPL1/4/8/9/10/14 lacked any endosperm regulation related, and SPL1/4/6A/9/12/13A/14 lacked any seed-specific regulation related cis-elements (Fig. 3l-n). The functional relevance of the cis-regulatory elements in their expression is yet to be explored in citrus.

Expression profiling of the SPL orthologs in citrus members

To assess which of the SPLs exhibit uniquely different expression patterns in young and adult tissues (buds and leaves) as well as in flowers of four representative members of the cultivated citrus species, namely C. reticulata, C. maxima ‘Majiayou’, F. hindsii, and C. sinensis, and we detected the expression level of SPLs gene through Real-time quantitative PCR. Some of the representative SPLs relatively highly expressed in adult leaves as compared to young leaves include C. sinensis (CsSPL3, CsSPL4, CsSPL11, and CsSPL12), Fortunella hindsii (FhSPL5, FhSPL6A, and FhSPL12), C. maxima ‘Majiayou’ (CmjSPL1, CmjSPL3, CmjSPL4, and CmjSPL14) and C. reticulata (CrSPL3, CrSPL4, and CrSPL10). Notably, FhSPL6B and CmjSPL14 exhibited 74- and 31 times higher expression respectively in adult leaves as compared to their young counterparts. Similarly, the SPLs exhibiting relatively higher expression at the adult buds as compared to the young ones include C. sinensis (CsSPL3 and CsSPL11), F. hindsii (FhSPL3, FhSPL4, FhSPL5, FhSPL6B, FhSPL10 and FhSPL11), C. maxima ‘Majiayou’ (CmjSPL3 and CmjSPL11) and C. reticulata (CrSPL3, CrSPL5, CrSPL6A, and CrSPL13A). CrSPL3, in particular, was up-regulated 19 times higher in the adult buds as compared to the young ones. Furthermore, SPL5 was relatively highly expressed in the flower tissues of four species (Fig. 4).

Fig. 4

SPL orthologs exhibit species-specific expression variation. a Different organizations. the bars represent 1cm. b Gene expression detection. YL, juvenile leaf; AL, adult leaf; YB, juvenile bud; AB, adult bud; FL, fully open flower; The darker the blue, the higher the expression level. Values represent relative expression levels. Three biological replicates were conducted for each group of experiments, and the expression level of Actin was used as an internal control

Since the change in meristem identity is crucial to flowering in plants, we opted to screen the SPLs exhibiting relatively higher expression in adult buds as compared to their young counterparts.

While several SPL genes were up-regulated in buds of different citrus members, SPL3 and SPL11 were up-regulated in all four species, and SPL4, SPL5, and SPL9 were up-regulated in all but Majiayou. Adult leaves, on the other hand, exhibited a higher level of SPL4 expression in all four species. Overall, SPLs of F. hindsii exhibited the widest, those of C. sinensis exhibited the narrowest, and those of the remaining two species exhibited intermediary expression variations (closer to C. sinensis) (Fig. 4). Among all, we selected SPL5 (exhibiting higher expression in adult leaves of the majority of the species), SPL9 (exhibiting higher expression in adult buds of the majority of the species), and SPL11 (exhibiting higher expression in adult buds of all species) from Kumquat to assess their florigenic potential.

Independent overexpression of FhSPL5, FhSPL9 and FhSPL11 induced precocious flowering in Arabidopsis

The expression of plant SPL is generally regulated by upstream miR156. Therefore, we predicted its target site within the sequence of SPL genes of F. hindsii. The results showed that FhmiR156 targeted the 5’-UTR region of FhSPL5(Fh4g16520) and the CDS region of FhSPL9(Fh6g17020) and FhSPL11(Fh7g25680) (Table S3). We constructed independent overexpression vectors of FhSPL5, FhSPL9, and FhSPL11 using respective CDS (Table S4) after synonymous mutations of miR156 target sites of FhSPL9 and FhSPL11. Three independent positive transgenic lines of each gene were selected and grown to their T3 generation through subsequent progeny selection in the selection media (Fig. 5a). The transgenic lines were confirmed by RT-qPCR of ectopically expressed FhSPLs (Fig. 5b).

Fig. 5

Citrus SPLs exhibit cross-species functional conservation. a Representative images of WT and respective transgenic lines of FhSPL5, FHSPL9, and FhSPL11 at 3 weeks after germination. b Confirmation of the ectopic FhSPL-overexpressing transgenic Arabidopsis lines; c Quantitative FT expression analysis showed that all FhSPL transgenic lines exhibited its higher expression with the FhSPL5 transformants showing the highest and that of FhSPL9 showing the lowest. d All SPL-transgenic lines flowered at a significantly short time as compared to WT and the transgenic lines constituted a significantly lesser number of rosette leaves. The groups (or lines) with the same letter above bars were not statistically significantly different (Student’s T-test, p ≤ 0.05). e Correlation of endogenous FT expression (ΔΔC_t-value) to leaf number and flowering time in the transgenic Arabidopsis thaliana (single-tailed T-test; p ≤ 0.05)

The results showed that the average flowering time of the wild type was about 28.4 days after sowing, and the average number of rosette leaves was 9.8 (Fig. 5c). The transgenic lines overexpressing FhSPL5 had a significantly reduced flowering time of 21.2-23.8 days and a rosette leaf number of 6.0-6.2 (Fig. 5c). Similar observations were made for the overexpression lines of FhSPL9 (22.0-23.2 days of flowering time; 5.8-6.2 number of rosette leaves) and FhSPL11 (23.8-24.6 days of flowering time and 6.8-8.0 number of rosette leaves) (Fig. 5c). We further confirmed that the overexpression of FhSPLs led to the upregulation of endogenous AtFT(), a crucial gene involved in floral induction, and assessed its correlation with the precocious flowering in the transgenic Arabidopsis lines. Even though all transgenic lines exhibited a negative correlation of FT expression to days-to-flowering and total leaf number, it was statistically significant only for the FhSPL11 transgenic lines (Fig. 5d). The data showed a relatively stronger flowering promotion effect of SPL5 and a very similar effect of SPL9 as compared to SPL11 in Arabidopsis. An earlier study by Shalom et al. (2015) with the C. clementina derived SPL5 (annotated as CiSPL5 in the study) comes in agreement with our observation. Gene sequence analysis revealed that SPL5 is a protein composed of only 131 amino acids. In combination with the expression data (Fig. 4), it is very likely that SPL5 plays a positive role in shortening juvenility and/or promoting flowering in citrus.

Discussion

Characterization of SPL family in citrus-related species

A total of 221 SPL genes have been identified among 15 citrus-related species in this study. Gene duplication followed by their diversification and/or neofunctionalization has often been regarded as the driver behind SPL multiplication in plants (Guo et al. 2008; Ren et al. 2022). Our assessment showed that certain SPLs (SPL4/5, SPL13A/13B, and SPL6A/6B) have been specifically multiplied and diversified among citrus-related species as compared to those in Arabidopsis (AtSPL5, AtSPL13, and AtSPL6 respectively) suggesting for their functional diversification among citrus related species. However, citrus members harbor a relatively lesser number of SPLs (15) as compared to Arabidopsis (17). The infrequent reproductive cycle caused by gametic sterility, apomixis, and vegetative propagation in combination with their long juvenility in citrus has been suggested to be part of the reason (Xu et al. 2013; Zeng et al. 2019). Some citrus SPLs shared relatively higher sequence similarities and were tightly clustered with each other (SPL1/SPL12, SPL6A/SPL6B, and SPL13A/SPL13B), to which earlier study proposed for their segmental duplication during the evolutionary process (Zeng et al. 2019). The study further argued for the occurrence of a more recent whole genome duplication event, which was not found in citrus, as a likely contributor behind the relatively higher levels of expansion of SPL numbers in plants like Arabidopsis (17), poplar (30), apple (30), etc. Our study has shown that the SPL number has been fixed among the citrus-related species, although one wild relative, M. paniculata, had a loss of functional SPL14 ortholog.

The phylogenetic tree developed from the SBP domains of the respective SPLs showed that the members with relatively similar gene structure and conserved motif patterns had been clustered together, which was partly corroborated by the studies of Song et al. (2021), Zeng et al. (2019) Shalom et al. (2015), and Liu et al. (2017) on independent citrus species. We additionally found that four of the miR156 non-targeted SPLs (SPL1, SPL7, SPL12, and SPL14) harbor transmembrane domains near their C-termini. While the studies in orthologous species suggest their crucial role in SPL localization and/or function (Stone et al. 2005; Chao et al. 2017; Yang et al. 2022), their functional relevance in citrus species itself has not been reported yet.

Evolutionary diversification and functional relevance of citrus SPL genes

In Arabidopsis, 12 out of 17 SPLs are targeted by miR156 (Cardon et al. 1999; Rhoades et al. 2002; Xu et al. 2016). Our assessment showed that such a feature is conserved among the assessed member species as well. Like in Arabidopsis, except for SPL1, SPL7, SPL8, SPL12, and SPL14, all other SPLs harbor miR156-recognition sites in citrus-related species (Table 1). This comes in agreement with earlier reports on C. clementina (Shalom et al. 2015; Zeng et al. 2019), C. sinensis (Liu et al. 2017). F. hindsii (Long et al. 2018), etc. Furthermore, except for the SPL8, all other miR156 non-targeted SPLs are the ones with the highest exon number (10) and the longest gene/peptide size.

SPLs lacking miR156 target site

Each citrus-related member harbored a uniquely single SPL ortholog, which clustered with AtSPL7 (Fig. 2a). Studies have reported the involvement of this miR156 target site lacking SPL in copper deficiency response, which depends on the cleavage of its C-terminally located transmembrane domain followed by its nuclear transport (Garcia-Molina et al. 2014; Schulten et al. 2019) suggesting for similar fate and role of SPL7 in citrus members.miR156 target site lacking SPL8 citrus orthologs were clustered uniquely with the AtSPL8. Studies in Arabidopsis vaguely indicate that it may play a positive role in the pollen-specific brassinosteroid signaling during anther sac formation and pollen development in coordination with miR156-regulated SPLs (Unte et al. 2003; Xing et al. 2013). Whether citrus orthologs are involved in a similar role is yet unclear. Interestingly, a study in C. clementina showed that its ortholog is upregulated in the buds upon de-fruiting (Shalom et al. 2015). Heavy fruit load is known to inhibit flowering, the removal of which promotes the process in the subsequent season (Muñoz-Fambuena et al. 2012). A More recent study by Zeng et al. (2019) showed that the SPL8 ortholog in C. clementina (annotated as CclSBP7) uniquely transcribed three splice variants (α, β, and γ), overexpression of which brought precocious flowering in Arabidopsis by elevating the expression of FT, FRUITFULL (FUL), APETALA1 (AP1), and LEAFY (LFY). Additionally, the transgenic lines produced relatively shorter siliques (Cao et al. 2019; Zeng et al. 2019). An earlier study in Arabidopsis reported that AtSPL8 plays a localized tissue-specific role in response to gibberellin (GA) signaling (Zhang et al. 2007). Hence, it requires further assessment of the florigenic potential of the miR156 non-targeted SPL8 in citrus.

Additional miR156 non-targeted members, SPL1 and SPL12 citrus orthologs were clustered together- close to AtSPL1/AtSPL12. Like their Arabidopsis orthologs, all but CrSPL12 harbored putative transmembrane binding domain near respective C-termini. An Arabidopsis study reported their involvement in thermotolerance at the reproductive stage (Chao et al. 2017). Furthermore, a study in rice showed that its SPL12 ortholog, OsSPL6 is crucial for the suppression of ER stress conditions thereby avoiding cell death in developing panicles (Wang et al. 2018). An earlier study in C. clementina by Zeng et al. (2019) also suggested their (annotated as CclSBP5 and CclSBP14 in the study) involvement during drought stress response and floral induction in citrus based on their expression profile. A study in Tamarix chinensis also suggests the involvement of the SPL1/12 as well as SPL14 orthologs in coping with salt stress, as the study showed their elevated expression of salt stress (Wang et al. 2019). SPL14 orthologs were clustered close to AtSPL14/16 in our study. Its positive involvement in juvenile phase extension and sensitivity to the fungal toxin fumonisin B1 had been reported in Arabidopsis (Stone et al. 2005). Based on their expression profile change upon draught treatment, Zeng et al. (2019) suggested the involvement of SPL14 (annotated as CclSBP12 in the study) in the revival of vegetative growth in C. clementina. It requires further assessment to conclude the potential involvement in stress response and juvenility in citrus.

SPLs harboring miR156 target sites

miR156 harboring SPLs are attributed to their involvement in plant development in an age-dependent manner and are of particular interest to flowering-related studies. Among the 10 miR156-targeted citrus SPLs, SPL10 was the only sub-clade that was not clustered with any Arabidopsis orthologs. An orthologous transgenic study of the C. clementina derived SPL10 (annotated as CclSBP6 in the study) showed that its ectopic overexpression significantly delays flowering time in Arabidopsis plants due to the reduced expression of FT and SPL2/3/4/5/9. The overexpression lines additionally exhibited dwarf growth, slender leaves, smaller flowers, shorter siliques, and longer root phenotypes under long-day conditions. Due to the similarity of the transgenic plants to the miR156 overexpression lines, the study suggested the potential of SPL10 being not targeted by miR156 (Zeng et al. 2019). It requires additional study in citrus to conclude the regulation of this uniquely citrus-specific SPL and its potential role in citrus juvenility.

SPL13, which is present in two identical copies in Arabidopsis and is reportedly involved in the negative regulation of the transition from cotyledonary to vegetative leaf stage (Martin et al. 2010), is present in pair (SPL13A and SPL13B) in citrus. A study in Arabidopsis showed that its overexpression leads to the stunted growth and formation of distorted branches, while its silencing delays flowering (Gao et al. 2018). Tomato plants transformed with its SPL13 ortholog exhibit precocious flowering (Cui et al. 2020). A study in kumquat reported that overexpression of csi-miR156a or independent knock-down of SPL13B and SPL5 (annotated as CsSPL14 and CsSPL3 respectively in the study) significantly enhanced the somatic embryogenesis competence of its callus (Long et al. 2018). Their more recent study suggested the reduced starch accumulation led by the overexpression of starch biosynthesis gene repressors, TOE1.1 and TOE1.2 in the miR156 repressed citrus calli as one of the key factors behind their lower SE efficiency (Feng et al. 2022). Their finding came close to an earlier study by Liu et al. (2017), which reported increased starch content in the miR156 overexpressed citrus calli with a higher degree of downregulation of SPL10 and SPL13B (annotated as CsSPL2 and CsSPL14 in the study). These studies link the decline in embryogenetic potential with maturity often observed in plants (Isah 2016). Our expression assessment showed that both versions of SPL13 exhibit their higher expression at the buds in all of the species except for C. sinensis. However, the transcripts of SPL13A were much higher in the adult buds while that of APL13B were almost at the similar level in both young and adult buds suggesting for their functional differences at the apical meristems. It requires further study to conclude their potential role in SAM fate determination.

Citrus-related species additionally harbor SPL6 in pairs (SPL6A and SPL6B). Studies in other species suggest their involvement in defense-related gene activation in planta (Padmanabhan et al. 2013). A study in C. sinensis showed that the expression of all SPLs decreases with time after infection with Diaporthe citri. However, SPL6B (annotated as CsSBP6 in the study) exhibits a relatively lesser degree of decrease (Song et al. 2021). It is also notable that unlike herbaceous plants like Arabidopsis and rice, tree species like citrus, apple, poplar, hybrid cherry tree (Prunus × yedoensis), etc. have the duplication and diversification of SPL6 orthologs (Li et al. 2013; Zeng et al. 2019; Gao et al. 2022). While it is plausible, it requires further assessment to elaborate on the potential involvement of these SPLs in pathogen defense/resistance and the mechanism behind it.

A study in apple reported that MdSPL2 (close ortholog of CsSPL13B) and MdSPL33 (close ortholog of CsSPL9) promote anthocyanin accumulation in fruit peel by hindering the suppressive effect of miR156 mediated by lncRNAs (MLNC3.2 and MLNC4.6) (Yang et al. 2019). Contrastingly, a transgenic study in poplar showed that the miR156 overexpressing lines exhibits reduced abundances of SPL8/11/12/17/28/29 and higher accumulation of anthocyanin in its shoot (Wang et al. 2020). Earlier independent studies in Arabidopsis (Gou et al. 2011) also reported the negative regulation of anthocyanin accumulation by a miR156 targeted AtSPL9 in its stem. A study in blueberry showed similarly reduced expression of its six miR156 targeted SPLs in ripe whole fruit. The abundance of the SPLs would have otherwise repressed the expression of DFR, a gene responsible for anthocyanin biosynthesis (Li et al. 2021). Whether SPL-dependent anthocyanin accumulation is species-dependent or tissue-specific is not fully clear yet. Moreover, the anthocyanin content of many citrus fruit changes with their maturity both inside and out. It is likely but needs substantial evidence to conclude its association with the SPL expression in the plants.

Studies in Arabidopsis suggest the involvement of SPL3, SPL4, and SPL5 in integrating age pathway to the photoperiod and GA signals to promote flowering (Jung et al. 2012; Jung et al. 2016). They are involved in floral meristem promotion (Xu et al. 2016). They redundantly interact with FD and facilitates/enhances its binding to the promoters of AP1, LFY, and FUL promoters, which gradually increase with plant age (Jung et al. 2016). A study in C. clementina showed that fruit load represses the expression of some SPLs and the removal of fruits leads to their increased expression at buds. Those SPLs include SPL5 (annotated as CiSPL5 in the study), SPL10, and SPL8 (Shalom et al. 2015). SPL9, on the other hand, plays role in integrating GA-signaling into the age pathway to promote shoot maturation (Schwarz et al. 2008), juvenile-to-adult phase transition (Yu et al. 2012; Hyun et al. 2016), axillary bud formation (Zhang et al. 2020), and flowering (Wang et al. 2009). The SPL11 orthologs, AtSPL2/10/11 reportedly regulate shoot maturation and proper development of the lateral organs in the reproductive phase in Arabidopsis. They additionally suppress root regeneration with age by binding to AP2/ERF promoters (Ye et al. 2020) and are partly involved in phase transition by regulating FUL (Shikata et al. 2009). A recent study in C. clementina showed that SPL11 (annotated as CiSPL11 in the study) positively regulates the expression of CiKN6, a citrus homolog of KNOTTED1-LIKE HOMEOBOX (KNOX) family genes, by binding to its promoter. Furthermore, the study showed that CiKN6 complexes with CiKN1, which in turn suppresses the expression of miR164a by binding to its promoter and modulates leaf development in citrus (Zeng et al. 2022). In the current study, based on their relatively unique tissue expression pattern, we chose SPL5, SPL9, and SPL11 orthologs from Kumquat for their functional assessment.

Confirmation of the florigenic potential of selected SPLs

An earlier study by Zhu et al. (2019) showed that FhSPL5, FhSPL8, FhSPL10, and FhSPL3 (annotated as FhSPL1, FhSPL7, FhSPL8, and FhSPL9 in their study) exhibited expression correlation with the flowering related genes and suggested for their functional redundancies in floral induction. It should be noted that Zhu et al. (2019) reported higher number of FhSPL members earlier, most likely due to the less refined genome data used in the study. In the current study, we assessed the expression profiles of the SPLs in young and adult leaves and buds as well as flowers of four different species. Based on our recent assessment, we selected FhSPL5 (expressed in flower buds at a relatively higher degree), FhSPL9, and FhSPL11 (both expressed relatively highly in adult buds) for their florigenic potential assessment. Their independent overexpression led to significantly precocious flowering in Arabidopsis confirming their promotional role in floral induction (Fig. 5). An earlier study by Shalom et al. (2015) with the C. clementina derived SPL5 (annotated as CiSPL5 in the study) comes in agreement with our observation. Their study additionally reported that the overexpression of CiSPL5 with an intact miR156 target site brought only a slight change to the flowering time. Our study showed that, among the three, FhSPL5 overexpression led to the highest endogenous FT expression in plants. The floral precocity effect of FhSPL5 is likely more direct than those of FhSPL9 and FhSPL11 since their Arabidopsis orthologs are reportedly involved in the regulation of flowering time, phase change/axillary bud formation, and shoot maturity respectively (Shikata et al. 2009; Hyun et al. 2016; Jung et al. 2016). However, it requires further experimental evidence in citrus to conclude as such for the citrus orthologs.

Conclusion and perspective

Flowering in a majority of citrus species requires a long period of juvenility after germination, which has remained to be a serious hindrance to researchers and farmers alike. The phase transition process in a plant is often linked to the reciprocal change in the abundance of miR156 and miR172. Several studies have shown that such a process is coordinated by the miR156 targeted SPLs. Our current study has methodically assessed and freshly annotated such SPLs from 15 different citrus-related species. We confirmed that most of them harbor a full set of 15 SPL members. Cis-regulatory element assessment of the SPL promoters suggested the involvement of the gene in the diverse developmental and physiological process via their responsive expression to hormones, defense/stress, wound, light/temperature, etc. We found that the expression profile of SPLs among different species varies significantly. Interestingly FhSPL5 showed higher expression at the flower buds and FhSPL9/FhSPL11 exhibited higher expression at adult buds (compared to young) of all species assessed. Their independent overexpression in Arabidopsis brought precocious flowering by upregulating the endogenous FT expression. Our report is the first to document such an occurrence with the ectopic expression of citrus-derived SPL9/SPL11. Being a woody species, the observations made on the Arabidopsis could still be suggestive for the citrus species. Future study on the species is expected to conclude the observations and predictions made in the current study.

Materials and methods

Plant materials

Seeds were collected from the mature fruits of C. reticulata, C. maxima ‘Majiayou’, F. hindsii, and C. sinensis. They were incubated at 28℃ for a week to accelerate the germination and moved to the greenhouse for growth until next spring. Samples were collected when the seedlings were about 5 months old. The leaves were sampled from the fully unfolded leaves near the top (juvenile leaf, YL), and the buds were sampled from the top buds and three additional buds below (juvenile bud, YB). The adult bud (AB) was sampled at the stage before physiological differentiation, adult leaf (AL) was sampled from the fully unfolded leaf near the bud, and the flower (FL) was sampled of the fully open flower. Three biological replicates are selected for each group of samples. All citrus varieties came from the germplasm resource nursery at Huazhong Agricultural University (Wuhan, China). For the SPL expression assessment, the leaves, buds, and flowers of adult trees, as well as the leaves and buds of young trees, were collected. They were frozen in liquid nitrogen after sampling and stored at - 80 ℃ until use.

Identification of SPL gene family members

A total of 15 citrus-related species (2n) were selected as the representative Aurantioideae members for the study- Aegle marmelos, Murraya paniculate, Atalantia buxifolia, Clausena lansium, Citropsis gilletiana, Poncirus trifoliata, Fortunella hindsii, Citrus mangshannensis, Citrus ichangensis, Citrus sinensis, Citrus reticulata ‘Pokan’, Citrus hongheensis, Citrus maxima ‘Majiayou’, Citrus maxima ‘Zipi’, and Citrus medica). Most of their genomes are included in the local Citrus Pan-genome to Breeding database (CPBD; http://citrus.hzau.edu.cn/). The hidden Markov model (HMM) and Basic Local Alignment Search Tool (BLAST) were used to screen the SPL gene in their respective genomes. The protein family database (Pfam) derived SBP domain profile PF03110, was used during the process (Finn et al. 2014). Arabidopsis-derived amino acid and nucleotide sequences of its 16 SPL genes were retrieved from the TAIR database (https://www.arabidopsis.org/). Homologous citrus SPLs were assessed with blastp. The SPL number was determined based on the results of HMM and blastp followed by their nomenclature based on the closest Arabidopsis SPLs.

Phylogenetic Analysis

To study the evolutionary relationship between SPLs of different citrus varieties and members of the SPL gene family in Arabidopsis, we took their respective SBP domain to construct the phylogenetic trees. Multiple alignment was carried out using ClusterW in MEGA7.0, and an evolution tree was prepared using the Neighbor-Joining (NJ) method with default parameters (bootstrap = 1000). The iTOL was used (https://itol.embl.de/) for the curation of the developed phylogenetic tree.

Analysis of gene structure, motif, and cis-acting elements

The gene structure (exon and intron composition) of SPL genes among 15 citrus-related species was assessed using GSDS 2.0 (http://gsds.gao-lab.org/) (Hu et al. 2014). Respective protein motifs were predicted using the MEME Suite (https://meme-suite.org/meme/index.html) with default parameters (for a maximum of 10 motifs) (Bailey et al. 2015). The gene structure and motif data were visualized using TBtools (Chen et al. 2020). To compare and analyze the cis-acting elements of the SPL promoters, 2.0 kb long promoter sequences of respective SPLs were retrieved, and the cis-elements were predicted using PlantCARE (http://bioinformatics.psb.ugent.be/webtools/plantcare/html/) (Lescot et al. 2002). GraphPad Prism 9 was used to prepare the boxplots of major cis-elements.

RNA extraction and real-time quantitative PCR

The earlier protocol described by He et al. (2022) was used for total RNA extraction from the plant samples, their reverse transcription, and real-time quantitative PCR. Actin was used as an internal control and each sample was subjected to three repeats. The primers used for the quantitative PCR have been provided in Supplementary Table S5.

Transgenic plant regeneration and analysis

According to the prediction made for the miR156 target site, FhSPL5 harbor it at its 5’-UTR, while FhSPL9 and FhSPL11 harbor such site within their CDS region (Table S4, Figure S7). After synonymous mutations of FhSPL9 and FhSPL11 miR156-target sites, respective CDS were incorporated into pK7WG2D overexpression vector. The plasmid was transformed into Agrobacterium strain GV3101, which was then used for the transformation of Arabidopsis (Col-0) using the floral dip method as described by reference (He et al. 2022). T3 transgenic plants were generated via antibiotic resistance screening. Arabidopsis flowering time and rosette leaf number were used for the statistical analysis as described by reference (Zeng et al. 2019).