Introduction

Crocus sativus L. is a sterile triploid plant belonging to Iridaceae family (Fernandez et al. 2011). The dehydrated red stigmas of the flower of Crocus comprise the world’s costliest spice, saffron (Rakesh et al. 2009). The distinctive feature on account of which Crocus stands out from other plants is the massive and unparalleled accumulation of apocarotenoids like crocin, picrocrocin, and saffranal (Frusciante et al. 2014). These apocarotenoids impart typical color, outstanding taste, and inimitable aroma to saffron because of which it has been cherished since millennia in different cuisines (Kyriakoudi et al. 2015). Saffron apocarotenoids also exhibit multiple biological actions including anticancer, anti-inflammatory, antimicrobial, and antioxidant activities (Kaefer and Milner 2008). The most noteworthy health benefits relate to its activity against various types of cancers (Gutheil et al. 2012; Zhang et al. 2013), gastric disorders (Kianbakht and Mozaffari 2009), cardiovascular diseases (Zheng et al. 2006), alzheimer disease (Moshiri et al. 2015), and mild-to-moderate depression (Lopresti and Drummond 2014).

The apocarotenoids build up throughout the growth of Crocus stigmas, reaching a maximum at fully developed scarlet stage and decreasing thereafter (Rubio et al. 2010; Mir et al. 2012). The market value of saffron is exceptionally high and the demand is soaring for various purposes. However, the production is dismally low, rather on the decline (Jain et al. 2016). Therefore, a great interest has surfaced with regards to understanding and eventually manipulating the biosynthesis and production of saffron apocarotenoids. However, so far, most of the studies have focused only on the spice-related properties and very little attention has been paid to the regulation of biosynthesis and accumulation of apocarotenoids in C. sativus (Ahrazem et al. 2015). Although some of the genes involved in the biosynthetic pathway have been identified and characterized (Baba et al. 2015a; Frusciante et al. 2014), very little is known about the genes that regulate tissue and developmental stage-specific production of these compounds. There is only a single report where an ultrapetala transcription factor has been shown as a potential candidate regulating apocarotenoid biosynthesis in Crocus (Ashraf et al. 2015).

To gain a comprehensive understanding about apocarotenoid biosynthetic pathway and its regulation, we generated a transcriptome map of Crocus (Baba et al. 2015b). A search for genes encoding transcription factors resulted in identification of many transcription factor families among which zinc-finger gene family was an abundant one. Furthermore, differential expression profiling between stigma and rest of the Crocus flower suggested that many zinc-finger genes were induced in stigma which is the actual site of apocarotenoid biosynthesis. This prompted us to design a detailed study on this class of transcription factors. In this context, we did transcriptome wide identification, phylogenetic analysis, and expression profiling of zinc-finger family of transcription factors from Crocus so as to elucidate their possible role in regulation of stigma development and apocarotenoid biosynthesis.

Zinc-finger domains are ubiquitously dispersed throughout the eukaryotic genomes (Gourcilleau et al. 2011) and mediate important regulatory roles (Iuchi 2001). Zinc-finger family of transcription factors is one of the largest and versatile transcription factor gene families in plants (Hall 2005). Based on the number and order of Cys and His residues, zinc-finger proteins have been categorized into at least 14 families (Klug and Schwabe 1995; Arnaud et al. 2007). On the basis of plant transcription factor database (http://planttfdb.cbi.edu.cn/) (Zhang et al. 2011), it has been shown that out of the total transcription factors, the percentage of zinc fingers (ZNFs) in different plant species such as Arabidopsis, potato, rice, maize, and wheat is almost similar (16–18%). It is speculated that ZNFs are probably conserved in plant evolution, and their key biological functions are most likely comparable in different plant species (Li et al. 2013). ZNFs mediate important regulatory functions by interacting with DNA or chromatin, RNA and other proteins (Iuchi 2001; Yang et al. 2006). ZNFs also affect various other functions, such as plant growth and development (Luo et al. 1999; Yun et al. 2002), phytohormone response (Molnar et al. 2002), and biotic and abiotic stresses responses (Tian et al. 2010).

In this study, we identified 81 ZNFs from Crocus transcriptome data-set developed in our laboratory (Baba et al. 2015b). We performed comparative phylogenetic analysis of CsZNFs with Arabidopsis zinc-finger homologs, and investigated the transcript levels of selected genes in different tissues, flower developmental stages, and in response to different phytohormones and various abiotic stresses using qRT-PCR. CsSAP09 was found to be highly expressed in stigma, the site of apocarotenoid biogenesis, at anthesis stage. Since Crocus apocarotenoids are synthesised only in stigma in a developmental stage-specific fashion reaching their pleatue at anthesis stage, therefore, CsSAP09 appears a credible candidate from the point of regulation of apocarotenoid biosynthesis. Therefore, CsSAP09 was cloned and characterized. The gene was found to be localized in nucleus and showed activated reporter gene transcription in yeast. These results provide insights about the phylogenetic relation and functions of Crocus zinc fingers in general and CsSAP09 in particular. Furthermore, the study may act as a foundation towards elucidation of functionality of Crocus zinc-finger genes.

Materials and methods

Identification of zinc-finger transcription factors from C. sativus transcriptome

Based on BLAST homology results, 173 putative zinc-finger transcripts were identified from a transcriptome data-set of C. sativus published from our laboratory (Baba et al. 2015b). The EMBOSS package of Transeq program was used to translate zinc-finger transcription factor DNA sequences. The amino-acid sequences of the longest open reading frame (ORF) in six ORFs were selected. Based on the domain search in NCBI CDD and Interproscan domain search, 81 sequences containing the ZnF domain were finally sorted out for analysis. The sequences of Arabidopsis zinc-finger proteins were obtained from (http://www.Arabidopsis.org).

Phylogeny and classification of Crocus zinc-finger proteins

To analyze the phylogenetic relationship amongst CsZNF gene family members, zinc-finger domains from all CsZNFs were aligned by CLUSTAL W (Thompson et al. 1994). Multiple sequence alignment and phylogenetic analyses were carried out by MEGA software version 6 (Tamura et al. 2013). The evolutionary history was deduced by the neighbour-joining (NJ) method. The phylogenetic tree with a bootstrap of 1000 replications was finally used to represent the evolutionary history of the genes.

To infer the phylogenetic relationship between major ZNF gene subfamilies of Crocus sativus and corresponding Arabidopsis thaliana zinc-finger gene families, separate phylogenetic trees based only on zinc-finger domains of the two plants were constructed for C2H2, C3H and A20/AN1 (SAPs) using the MEGA 6 software (Tamura et al. 2013).

Plant material and treatments

Crocus sativus L., plants grown in an experimental plot at the Indian Institute of Integrative Medicine (IIIM), Srinagar, India (longitude: 34°5′24′′N; latitude: 74°47′24′′ and altitude 1585 m amsl), were used as the source of plant material. Flowers were harvested at different developmental stages, viz., 3 days prior to anthesis, on the day of anthesis and 2 day post-anthesis. Further different tissues including tepals, stigmas, anthers, leaves, and corms were also collected, wrapped in silver foil, snap frozen in liquid nitrogen, and stored in −80 until required.

Phytohormone and stress treatments were given as described earlier (Baba et al. 2015a). Briefly, C. sativus corms were grown in pots containing soil under greenhouse conditions at 26 ± 2 °C till the flowering stage. For phytohormone treatments, plants were sprayed with 0.1 mM salicylic acid (SA), 50 µM 2, 4-D, 0.1 mM methyl jasmonate (JA), and 0.1 mM gibberellic acid (GA3). For oxidative, salt and dehydration stresses, 5 µM methylviologen (MV), 200 mM NaCl, and 200 mM mannitol were added to soil intermittently till the soil got saturated. Cold and UV stresses were given by transferring the plants to 4 °C and UV-B light of 1500 J/m2, respectively. Control plants were grown under optimal conditions. Tissue samples were collected after 24 h of the treatments and stored till further use.

Isolation of nucleic acids and cDNA synthesis

RNA isolation was done by RNeasy RNA isolation kit (Qiagen) following the manufacturer’s instructions. RNA quality was assessed by separation on 1% agarose gel. Furthermore, RNA was quantified using NanoDrop® ND-1000 spectrophotometer (NanoDrop Technologies). The samples were then treated with DNase I (Fermentas) to remove any genomic DNA contamination. cDNA was synthesised by reverse transcription kit (Fermentas) as per manufacturer’s guidelines. Genomic DNA was isolated from corms using DNeasy Plant mini kit (QIAGEN) according to the manufacturer’s protocol.

qRT-PCR analysis

Quantitative RT-PCR was used to analyze the transcript levels of Crocus zinc-finger proteins. qRT-PCR was performed in triplicates in ABI Step One Real time (Applied Biosystems). The reaction was performed in a total volume of 20 μl which included 10 μL of 2X SYBR Green Master Mix, 0.2 μM gene-specific primers, and 100 ng of template cDNA. The cycling conditions were 95 °C for 20 s, followed by 40 cycles of 95 °C for 15 s and 60 °C for 1 min. Transcript abundance of zinc-finger genes was detected in different tissues including corms, stigmas, tepals, anthers, and leaves; flower development stages; and also in plants exposed to different stresses and phytohormones. The sequence of primers used is given in supplementary table S1. The analysis was done by relative quantification method ddCT (ΔΔΔΔCT). The amplification of Crocus actin gene was used as an endogenous control.

Cloning and sequence analysis of CsSAP09 gene

Gene-specific primers (CsSAP09 F and CsSAP09 R) based on the gene sequence from the transcriptome data (Baba et al. 2015b) were used to clone CsSAP09 gene. The cycling parameters were 3 min at 94 °C, 30 cycles (30 s at 94 °C, 30 s at 55 °C and 45 s at 72 °C) and a final extension of 7 min at 72 °C. The amplified clone was purified with gel extraction kit (Qiagen), cloned in pGEM-T Easy vector (Promega), and sequenced. A search for homologous nucleic acid and protein sequences was performed using the BLAST algorithm (http://www.ncbi.nlm.nih.gov/). A multiple sequence alignment was generated by ClustalW (http://www.ebi.ac.uk/clustalw/). The deduced amino-acid sequence was analyzed by ProtParam Programs at ExPASy proteomics server (http://ca.expasy.org/) to calculate pI/Mw. The phylogenetic tree was constructed using the MEGA 6.06 software (Tamura et al. 2013).

Subcellular localization

The subcellular localization of CsSAP09 was determined by transient expression assay in onion epidermal cells. CsSAP09 with restriction sites for XhoI and EcoRI was amplified using CsSAP09YFP-F and CsSAP09YFP-R primer pair and fused in frame with YFP reporter gene in PAM-PAT-35S. The cycling parameters were the same as described elsewhere. The fusion construct of CsSAP09-YFP was bombarded on to the onion peels using biolistic gene delivery device PDS-1000/He (Bio-Rad). The onion peels were incubated for 24 h in dark before visualization under confocal microscope.

Transactivation assay

CsSAP09 was cloned in yeast (Saccharomyces cerevisiae) expression vector pGBKT7 (Clontech) at NdeI-EcoRI site to express the CsSAP09 protein fused to GAL4 DNA-binding domain (GAL4-BD). The primers used were CsSAP09PGBKT-F and CsSAP09PGBKT-R and cycling conditions were 3 min at 94 °C, 30 cycles (30 s at 94 °C, 30 s at 55 °C, and 45 s at 72 °C), and final extension for 7 min at 72 °C. The resulting construct was transformed into yeast strain Y187. The positive transformants were selected onto synthetic media lacking tryptophan, adenine, and histidine.

Amplification of the core promoter of CsSAP09

CsSAP09 promoter sequence was isolated by the genome-walking method using the GenomeWalker Universal Kit (Clontech) following the manufacturer’s instructions. For library preparation, genomic DNA was digested with four different blunt-end generating restriction endonucleases (DraI, PvuII, EcoRV, and StuI). Each set of digested genomic DNA samples was purified and ligated to the GenomeWalker AP adaptor (provided with the kit) independently to generate four adapter-ligated libraries. To amplify the promoter region of CsSAP09, two-round PCR was performed using the adapter primer (provided with the kit) and gene-specific primers CsSAP- SP1 and CsSAP- SP2. In silico analysis was performed through the PlantCare database (http://bioinformatics.psb.ugent.be/webtools/plantcare/html/).

Results

Identification of zinc-finger proteins from Crocus transcriptome and their phylogenetic analysis

In this study, we provide information on the zinc-finger (ZNF) proteins of Crocus sativus. Initially a total of 173 unigenes were predicted from Crocus transcriptome (Baba et al. 2015b), possibly containing ZNF domains. These sequences were further analyzed by InterProScan and NCBI CDD search which confirmed the presence of ZNF domains in only 81 sequences and these were taken for further analysis.

To study the phylogenetic relationship amongst the C. sativus zinc-finger proteins (CsZNFs), a neighbour-joining unrooted tree with bootstrap analysis of 1000 replicates was constructed by MEGA6 from the alignment of zinc-finger domains of the protein sequences. The phylogenetic analysis divided the proteins into eight subfamilies with C2H2, C3H, and A20/AN1 as the major subfamilies containing 29, 20, and 14 sequences, respectively (Fig. 1). The other subfamilies included GRF, NAM, B box, Swim, and Zn-HD containing proteins and all these had less than five gene sequences. Since C2H2, C3H, and A20/AN1 emerged as the major subfamilies, we focussed our study on these three classes. Generic names (C2H201-C2H229, C3H01-C3H20, and SAP01-SAP14) were assigned to distinguish each member of the three major subfamilies for the purpose of further analysis.

Fig. 1
figure 1

Phylogenetic analysis of CsZNF genes. The phylogenetic tree was generated by the neighbour-joining (NJ) algorithm using Molecular Evolutionary Genetics Analysis (MEGA) software version 6.06 (Tamura et al. 2013). The numbers on major branches indicate bootstrap percentages for 1000 replicate analyses. CsZNF genes have been divided into eight subfamilies

Full size image

Phylogenetic and evolutionary relationship of Crocus zinc-finger proteins with Arabidopsis homologs

To understand the phylogenetic relationships of CsZNFs with Arabidopsis homologs and to gain insights about their putative functions, unrooted phylogenetic trees were built based on the alignment of zinc-finger domains of proteins from Crocus and those from Arabidopsis. To simplify the comparison, phylogenetic trees were generated separately for the C2H2, C3H, and A20/AN1 (also called SAPs) subfamilies (Figs. 2, 3, 4). The results showed that although in some branches, Crocus and Arabidopsis homologs are very closely related, but this was not the case in most of the other branches. For example, in a comparative phylogenetic analysis between Crocus and Arabidopsis C2H2 zinc fingers, it was observed that ten CsC2H2 genes including CsC2H203, CsC2H204, CsC2H205, CsC2H206, CsC2H208, CsC2H210, CsC2H211, CsC2H213, and CsC2H214 formed a separate clad, while others grouped with their Arabidopsis homologs (Fig. 2). Likewise, in case of C3H subfamily, many Crocus C3H genes grouped with their Arabidopsis counterparts. For example, CsC3H05 and CsC3H06 shared a common clad with AT3G12680.1 and AT1G10320.1, while CsC3H01, CsC3H01, CsC3H02, CsC3H03, and CsC3H04 grouped with AT2G32930.1 and AT3G48440.1. However, many others like CsC3H16, CsC3H19, and CsC3H20 formed a separate clad (Fig. 3). Furthermore, phylogenetic analysis of Crocus and Arabidopsis SAP proteins showed that while most of the genes from Crocus grouped with Arabidopsis genes, CsSAP012 formed a separate clad (Fig. 4).

Fig. 2
figure 2

Phylogenetic tree of C2H2 proteins of Crocus sativus and Arabidopsis thaliana. A phylogenetic tree was generated from zinc-finger domains of C2H2 proteins of C. sativus and A. thaliana using MEGA software version 6.06. The numbers on major branches indicate bootstrap percentages for 1000 replicate analyses

Full size image
Fig. 3
figure 3

Phylogenetic tree of C3H proteins of Crocus sativus and Arabidopsis thaliana. A phylogenetic tree was constructed from zinc-finger domains of C3H proteins of C. sativus and A. thaliana using MEGA software version 6.06. The numbers on major branches indicate bootstrap percentages for 1000 replicate analyses

Full size image
Fig. 4
figure 4

Phylogenetic tree of AN1/A20 proteins of Crocus sativus and Arabidopsis thaliana. A phylogenetic tree was constructed from AN1 zinc-finger domains of SAPs of C. sativus and A. thaliana using MEGA software version 6.06. The numbers on major branches indicate bootstrap percentages for 1000 replicate analyses

Full size image

Tissue and developmental stage-specific expression of CsZNF genes

To get first round information about the biological function of CsZNFs, the expression patterns of one representative gene from three major subfamilies (CsC2H216, CsC3H13, and CsSAP09) were analyzed by quantitative RT-PCR. The selected CsZNFs showed varying expression levels in different tissues and flower development stages as depicted in (Fig. 5a–f). CsC2H216 exhibited highest expression in tepal at post-anthesis stage of flower development (Fig. 5a, d). For CsC3H13, the highest expression was observed in anther at post-anthesis stage of flower development (Fig. 5b, e). Furthermore, CsSAP09 displayed peak expression in stigma and anthesis stage of flower development alluding to its probable role in apocarotenoid production (Fig. 5c, f).

Fig. 5
figure 5

Spatial and temporal expression profiling of CsZNFs. Relative expression ratio of three CsZNF genes analyzed by qRT-PCR (a–c): in five different tissues and (d–f): at three different flower development stages. The relative expression level of each gene was calculated relative to its expression in control sample. qRT-PCR data were normalized using Crocus actin gene. Experiment was done in triplicates. Error bars indicate standard deviation

Full size image

Stress inducible expression of CsZNF genes

To evaluate the role of CsZNF genes in abiotic stresses, the expression of three selected genes was analyzed by qRT-PCR under various stress treatments. Overall, transcript levels of all the three genes tested were detected to respond to more than one treatment. CsC2H216 was significantly induced by cold and dehydration stress (Fig. 6a); CsC3H13 was highly induced by cold and salt stress (Fig. 6b), whereas CsSAP09 displayed enhanced expression in response to methylviologen, dehydration, and salt stress (Fig. 6c).

Fig. 6
figure 6

Stress induced and hormone responsive expression profiling of CsZNFs. Relative expression levels of three CsZNF genes analyzed by qRT-PCR (a–c): under different stresses and (d–f): in response to phytohormone treatments. The relative expression ratio of each gene was calculated relative to its expression in control sample. qRT-PCR data were normalized using Crocus actin gene. (Error bars indicate standard deviation)

Full size image

Hormone responsive expression of CsZNF genes

The expression kinetics of CsZNF genes was also investigated in response to various phytohormones. Transcript abundance revealed that the genes exhibited differential expression in response to various phytohormones. All the three genes responded to abscisic acid and methyl jasmonate (Fig. 6d–f). CsC2H216 showed maximum induction in response to methyl jasmonate followed by ABA (Fig. 6d). CsC3H13 highly responded to auxin followed by gibberellic acid and ABA (Fig. 6e), while CsSAP09 exhibited highest induction in response to ABA followed by salicylic acid and methyl jasmonate (Fig. 6f).

Cloning and analysis of CsSAP09

As mentioned in the above sections, CsSAP09, was highly expressed in stigma of Crocus flower and its expression surged upto the day of anthesis and later showed a decline. This expression pattern matched with the accumulation pattern of apocarotenoids, therefore, indicating probable role of this gene in regulating apocarotenoid metabolism. To move further towards characterizing the gene, we cloned full-length CsSAP09. The complete cDNA of CsSAP09 was 0.489 kb long. The open reading frame (ORF) encoded for a protein of 162 amino acids with an A20 and AN1 zinc-finger domain at N and C termini. The predicted molecular weight and pI of the protein was 17.4 kDa and 7.9, respectively. The ProtParam analysis suggested that the protein is hydrophilic with GRAVY value of −0.396.

The database search using the amino-acid sequence as query showed homology to several AN1/ A20-ZNFs from plants including Musa acuminata subsp. malaccensis, Phoenix dactylifera, Sesamum indicum, etc. All these proteins showed homology to CsSAP09 in the AN1-type and A20-like zinc-finger region (Fig. 7b). Phylogenetic analysis of CsSAP09 with other plant A20/AN1-ZNFs revealed that CsSAP09 showed close similarity to Musa acuminata subsp. malaccensis stress-associated protein 5-like (MASAP5), Phoenix dactylifera stress-associated protein 5-like (PdSAP5), and Sesamum indicum stress-associated protein 4 (SiSAP4) (Fig. 7c). Furthermore, the comparison of full-length cDNA of CsSAP09 with a genomic DNA fragment generated by PCR amplification and sequencing revealed the coding region of the genomic clone of CsSAP09 to be continuous, without any intron.

Fig. 7
figure 7

Sequence alignment and phylogeny of CsSAP09. a Diagrammatic representation of A20 and AN1 zinc-finger domains, b multiple sequence alignment and c phylogenetic tree of CsSAP09 and other A20/AN1-zinc-finger protein homologs. The neighbor-joining tree was constructed using MEGA 6.06 with 1000 bootstrap replications. The GenBank accession numbers of different sequences used are as follows: Phoenix dactylifera (XP_008812241.1), Musa acuminata subsp. malaccensis (XP_009393382.1), Elaeis guineensis (XP_010911244.1), Brachypodium distachyon (XP_003575065.1), Eucalyptus grandis (XP_010031502.1), Oryza sativa Japonica (XP_015639390.1), Fragaria vesca subsp. vesca (XP_004303001.1), Setaria italic (XP_004969762.1), Nelumbo nucifera (XP_010271706.1), Sesamum indicum (XP_011094369.1), Arachis duranensis (XP_015951458.1), Amborella trichopoda (XP_006851496.1), and Vitis vinifera (XP_010654069.1). Values at the node denote percentage bootstrap value. The genetic distances are indicated by the horizontal bar

Full size image

CsSAP09 localizes in the nucleus

To understand the mechanism underlying the regulatory role of CsSAP09, its subcellular localization was examined. Programs like NucPred and NLStradamus (Brameier et al. 2007; Nguyen et al. 2009) predicted a nuclear localization signal in CsSAP09. To ascertain the nuclear localization, CsSAP09 was fused in frame to the N-terminus of the YFP reporter gene of PAM-PAT-35S vector. CsSAP09-YFP fusion driven by the 35 S promoter of cauliflower mosaic virus (CaMV35S) promoter was bombarded onto onion (Allium cepa) epidermal cells by gold particle bombardment and visualized under confocal microscope. While the control YFP fluorescence was observed throughout the cell, CsSAP09-YFP fluorescence was localized only to the nucleus (Fig. 8a) confirming nuclear localization of CsSAP09.

Fig. 8
figure 8

Subcellular localization and transactivation assay of CsSAP09. a Upper panel shows YFP accumulation throughout the cell, while the lower one shows CsSAP09-YFP is localized to the nucleus. b Fusion protein of the GAL4 DNA-binding domain and CsSAP09 were expressed in yeast strain Y187. The empty pBD (pGBKT7) vector (negative control) and the pGAL4 vector (positive control) were used. The culture solution of the transformed yeast was dropped onto SD/-Trp-ade-his plates. The plates were incubated for 3 days and then subjected to β-galactosidase assay

Full size image

CsSAP09 functions as a transcriptional activator in yeast

The transcriptional activity assay of CsSAP09 was examined in yeast strain Y187. For this, ORF of CsSAP09 was fused in frame to the GAL4 DNA-binding domain (BD) to produce pGBKT7-CsSAP09-BD construct. The GAL4 DNA-binding domain-CsSAP09 fusion protein was expressed in yeast strain Y187. Yeast cells transfected with the pGBKT7-BD vector, lacking the activation domain, as the negative control did not grow on the SD medium deficient in His and Trp (SD/-Trp-Ade-His). However, the cells carrying pGBKT7-BD- CsSAP09 or pGBKT7-53 + pGADT7-T (the positive control) grew well on the SD/-Trp-Ade-His medium (Fig. 8b). CsSAP09 provided the activation domain and the cells exhibited activity of β-galactosidase reporter gene upon addition of X-gal on Whatman filter paper (Fig. 8b). Thus, these results confirm that CsSAP09 is transcriptionaly active in yeast.

Isolation of core promoter of CsSAP09

The expression of SAP genes is known to be induced by various abiotic factors including drought, salt, cold, light/UV light, mechanical wounding, and plant hormones (Giri et al. 2011, 2013). These factors modulate the gene expression by interacting with the cis-regulatory elements in the upstream regions of SAPs genes. To get an idea about the putative cis elements in the upstream region of CsSAP09, the region flanking the 5′ end was isolated. Genome-walking approach isolated a 565 bp fragment upstream of the start codon, corresponding to the putative promoter region of CsSAP09 (Fig. 9). Computational analysis using PlantCare database revealed several important cis-acting regulatory elements within the promoter region of CsSAP09, as shown in Table 1. Consistent with the established role of SAPs in abiotic stress, a cis-acting regulatory element essential for the anaerobic induction (ARE element) was found at the position 93. An MBS cis element which is an MYB-binding site involved in drought inducibility (Urao et al. 1993) is also present at positions 65,262 and 460. Besides, many light responsive elements were also found.

Fig. 9
figure 9

Isolation and analysis of CsSAP09 promotor. DNA sequence of C. sativus CsSAP09 promoter (565 bp) amplified from the StuI genome-walking library. The putative TSS, TATA box, CAAT box, and other important cis-regulatory elements are marked out

Full size image
Table 1 Putative cis-acting regulatory elements identified in the promoter of CsSAP09 by PLANTCARE databases (http://bioinformatics.psb.ugent.be/webtools/plantcare/html/)
Full size table

Discussion

Transcription factors are central elements that regulate spatio-temporal expression of genes, thus ensuring their accurate functioning in an organism. Proper identification and classification are necessary for functional characterization of any new gene. In this study, 81 ZNFs belonging to 8 subfamilies were identified from Crocus transcriptome. C2H2 was the largest family with 29 genes followed by C3H with 20 genes and A20/AN1 or SAP with 14 members (Fig. 1). Since these families formed the most abundant classes, our further study was focused on them. In Arabidopsis, 176 C2H2 and 68 C3H genes, while in rice, 189 C2H2 and 67 C3H genes have been identified (Agarwal et al. 2007; Ciftci-Yilmaz and Mittler 2008; Wang et al. 2008). Furthermore, A20/AN1-type zinc-finger genes are represented by 14 members in Arabidopsis and 18 in rice. In our study, the number of genes identified was lesser than other plants. This may be because an initial search of transcriptome may underestimate the number of CsZNF genes for these being transcription factors, are often expressed at low levels (Arora et al. 2007). Another reason might be the fact that, in this study, only flower tissue was used for generation of transcriptome and hence does not include the genes expressed in other plant parts. Therefore, more genes containing the ZNF domain may come to the fore as more transcriptome and genome sequence information is made available.

To know about the evolutionary relationship of CsZNFs with their Arabidopsis homologs, comparative phylogenetic analysis was done. In case of all the three families, while many Crocus ZNFs clustered with their Arabidopsis homologs, others either formed a separate clad or some of the genes were seen as outliers (Figs. 2, 3, 4). This might be because of the fact that Arabidopsis and Crocus belong to distant families. Furthermore, less number of sequences from Crocus might also be a reason for this pattern.

Gene expression patterns are an essential aspect of study of gene functions (Gu 2004). The expression profile of a gene is usually indicative of its function. We studied expression pattern of three representative genes (CsC2H216, CsC3H36, and CsSAP09) one each from C2H2, C3H, and SAP classes. The three genes were chosen, because we had the longest sequence available for them among all the members present in each of these classes. These selected genes displayed a complex expression pattern in different tissues and at different flower developmental stages (Fig. 5). Among the five tissue types tested, CsC2H216 showed highest expression in tepal; CsC3H13 exhibited peak expression in anther while CsSAP09 was induced more in stigma. This expression pattern may indicate their tissue specific roles. Furthermore, CsSAP09 displayed peak expression at anthesis stage of flower development. The expression pattern of CsSAP09 corroborates with accumulation pattern of apocarotenoids which are produced in stigma and their levels increase upto the anthesis stage and then show a decline (Ashraf et al. 2015). This indicates possible role of this gene in regulating apocarotenoid biosynthesis in Crocus.

Growth, development, and metabolism of plants are usually challenged by different stresses from the environment, such as salinity, drought, osmotic, and oxidative stresses during their life cycle. A myriad of stress-related genes is induced whose expression products are required to adapt to these environmental insults (Kasuga et al. 1999; Coram and Pang 2006). Zinc-finger proteins are also known to mediate various biotic and abiotic stress-related adaptations in different plants (Tian et al. 2010). In this context, we investigated the expression of the selected CsZNFs in response to various stresses. We observed increase in expression of CsC2H216 in response to cold and dehydration (Fig. 6a) and of CsC3H13 in case of cold and salt stress (Fig. 6b). Earlier reports have also demonstrated involvement of C2H2 and C3H family proteins in various abiotic stresses (Li et al. 2013). This suggests that role of these gene families in stress tolerance remains conserved across plant species. Furthermore, expression of CsSAP09 was also induced in response to salt, methylvialogen, and dehydration stresses (Fig. 6c). Almost all SAP family genes have been implicated in various stress responses (Giri et al. 2011, 2013). Overexpression of SAPs has been shown to enhance tolerance to multiple abiotic stresses in a number of transgenic plants (Mukhopadhyay et al. 2004).

The stress responses in plants are mediated by phytohormones. Therefore, we investigated expression of CsZNFs in response to various phytohormone treatments. Our results showed that all the three genes were induced by different phytohormones, albeit to varied levels. CsC2H216 showed maximum induction in response to methyl jasmonate followed by ABA (Fig. 6d). CsC3H13 was highly elicited in response to auxin followed by gibberellic acid and ABA (Fig. 6e), while CsSAP09 exhibited highest expression in response to ABA followed by salicylic acid and methyl jasmonate (Fig. 6f). All these hormones play role in controlling and fine tuning various stress responses. Moreover, induction of CsSA09 expression in response to ABA further establishes its role in abiotic stress, since ABA is known to regulate tolerance to abiotic stress (Wani and Kumar 2015).

AN20/AN1, an emerging class of zinc-finger proteins, also called as stress-associated proteins (SAPs), shows considerable levels of structural conservation and functional homology from diverse plant species. These proteins were, for the first time, identified in rice as products of multiple stress responsive genes (Mukhopadhyay et al. 2004; Vij and Tyagi 2006), but their functional repertoire has been expanding. Since expression profile of CsSAP09 suggested its role in regulating apocarotenoid metabolism, its full-length gene was cloned and its sequence analyzed. It has an AN1-type zinc-finger region at the C terminus of the protein spanning amino acids 87 to 157. It has a consensus sequence of Cx2-4Cx9-12Cx2Cx4Cx2Hx5HxC, where x represents any amino acid. The conserved cysteine and histidine residues form and stabilize the zinc finger. There are four conserved cysteine residues at the amino-acid positions 14, 18, 30, and 33 toward the N-terminus of the protein (Fig. 7). This region is akin to the A20-like zinc-finger protein (Dixit et al. 1990). Comparison of full-length cDNA of CsSAP09 with a genomic fragment generated by PCR amplification and sequencing revealed that the genomic clone is continuous, without any intron. The O. sativa stress-associated protein 1(OsSap1), identified as the first plant A20/AN1 zinc-finger as well as tea, Camellia sinensis, and CsZfp were also without any intron (Mukhopadhyay et al. 2004; Paul and Kumar 2015).

Recently, various research groups have reported that A20/AN1-like proteins localize to various subcellular compartments (Hishiya et al. 2006; Kanneganti and Gupta 2008). The NLStradamus and NucPred Programs (Brameier et al. 2007; Nguyen et al. 2009) were used to predict the nuclear localization signal in CsSAP09. This analysis revealed the presence of a putative nuclear localization sequence. Congruently, our study also revealed that CsSAP09 is localized in the nucleus (Fig. 8a). AtSAP3 and AtSAP5 are also nuclear localized (Kim et al. 2015). Furthermore, it was found that CsSAP09 was transcriptionally active in yeast strain Y187 (Fig. 8b).

The promoter isolation led to the amplification of 565 bp long fragment which contained many cis elements including MBS, a Myb-binding site; ARE; and some light responsive elements (Fig. 9). Myb genes are known to regulate gene expression in response to many developmental cues as well as biotic and abiotic stress responses (Liu et al. 2015; Li et al. 2016). Furthermore, Myb genes are also involved in regulating plant secondary metabolism (Chezem and Clay 2016). Since expression profile of CsSAP09 hinted at its involvement in apocarotenoid metabolism and stress response, the presence of MBS and ARE elements in its promoter region further supports its role. Furthermore, a number of reports have shown that light regulates carotenoid metabolism in a number of plants (Ahrazem et al. 2016). The presence of light responsive elements in CsSAP09 promoter suggests that the gene might be involved in light responsive regulation of apocarotenoid metabolism.

In conclusion, we did transcriptome wide identification and phylogenetic analysis of zinc-finger genes in Crocus in which C2H2, C3H, and A20/AN1 or SAPs emerged as major classes. The transcript abundance of one representative member from each of these classes was investigated which showed that the genes are induced in response to different abiotic stresses and phytohormones. The transcript profile of CsSAP09 corroborated with known metabolite profile of apocarotenoids, therefore, indicating its possible role in regulating apocarotenoid metabolism. Further studies will elucidate the exact role of CsZNFs in general and CsSAP09 in particular in Crocus apocarotenoid metabolism.