Mol Genet Genomics DOI 10.1007/s00438-014-0858-9

Original Paper

Comprehensive analysis of CCCH‑type zinc finger gene family in citrus (Clementine mandarin) by genome‑wide characterization Shengrui Liu · Muhammad Rehman Gul Khan · Yongping Li · Jinzhi Zhang · Chungen Hu 

Received: 8 January 2014 / Accepted: 19 April 2014 © Springer-Verlag Berlin Heidelberg 2014

Abstract  The CCCH-type zinc finger proteins comprise a large gene family of regulatory proteins and are widely distributed in eukaryotic organisms. The CCCH proteins have been implicated in multiple biological processes and environmental responses in plants. Little information is available, however, about CCCH genes in plants, especially in woody plants such as citrus. The release of the whole-genome sequence of citrus allowed us to perform a genome-wide analysis of CCCH genes and to compare the identified proteins with their orthologs in model plants. In this study, 62 CCCH genes and a total of 132 CCCH motifs were identified, and a comprehensive analysis including the chromosomal locations, phylogenetic relationships, functional annotations, gene structures and conserved motifs was performed. Distribution mapping revealed that 54 of the 62 CCCH genes are unevenly dispersed on the nine citrus chromosomes. Based on phylogenetic analysis and gene structural features, we constructed 5 subfamilies of 62 CCCH members and integrative subfamilies from citrus, Arabidopsis, and rice, respectively. Importantly, large numbers of SNPs and InDels in 26 CCCH genes were identified from Poncirus trifoliata and Fortunella japonica using whole-genome deep re-sequencing. Furthermore, citrus CCCH genes showed distinct temporal and spatial

Communicated by S. Hohmann. Electronic supplementary material  The online version of this article (doi:10.1007/s00438-014-0858-9) contains supplementary material, which is available to authorized users. S. Liu · M. R. G. Khan · Y. Li · J. Zhang · C. Hu (*)  Key Laboratory of Horticultural Plant Biology (Ministry of Education), College of Horticulture and Forestry Science, Huazhong Agricultural University, Wuhan 430070, China e-mail: [email protected]

expression patterns in different developmental processes and in response to various stress conditions. Our comprehensive analysis of CleC3Hs is a valuable resource that further elucidates the roles of CCCH family members in plant growth and development. In addition, variants and comparative genomics analyses deepen our understanding of the evolution of the CCCH gene family and will contribute to further genetics and genomics studies of citrus and other plant species. Keywords  CCCH zinc finger · Phylogenetic analysis · Gene structure and conserved motifs · SNPs and InDels · Expression patterns · Clementine mandarin

Introduction Citrus trees are the most widely grown and economically important fruit crop in the world. Most citrus species are diploid (2n = 2x = 18) with highly heterozygous, relatively small genomes and over 30,000 predicted genes (Gmitter et al. 2012). All citrus species belong to the Rutaceae family, which includes six closely related genera: Citrus, Fortunella, Poncirus, Eremocitrus, Microcitrus and Clymenia (Swingle and Reece 1967). Despite the economic importance of citrus, genome-wide research resources and studies have been limited. However, the recent availability of the complete genome sequences and assemblies of clementine mandarin (Gmitter et al. 2012) and sweet orange (Xu et al. 2013) now allow for comprehensive bioinformatics analyses and the characterization of numerous known or novel gene families. So far, only the MADS-Box family, the members of which are involved in plant development and signal transduction, has been comprehensively analyzed (Hou et al. 2014). The CCCH zinc finger proteins associate

13



with RNA and are important regulators involved in plant development and stress responses, but there have been no previous studies to functionally characterize CCCH genes in citrus. Therefore, we took advantage of recent genomic advancements to analyze the CCCH genes in the whole citrus genome. As one of the largest transcription factor families in plants, the zinc finger transcription factors are important regulators involved in plant development and biological processes. They are characterized by the presence of common zinc finger motif, which play critical roles in interactions with other molecules (Takatsuji 1998; Moore and Ullman 2003). Most zinc finger transcription factors previously identified in plants, such as the RING-finger, Dof, WRKY, ERF, and LIM families, regulate gene expression by binding DNA or proteins (Kosarev et al. 2002; Lijavetzky et al. 2003; Zhang and Wang 2005; Nakano et al. 2006; Arnaud et al. 2007). In contrast, CCCH zinc fingers are specific and distinct from other zinc finger transcription factors, which regulate gene expression by binding to mRNA (Wang et al. 2008a). The CCCH domain proteins contain 1–6 typical C3Htype motifs and were originally defined as C–X6–14–C– X4–5–C–X3–H (Berg and Shi 1996), but redefined as C–X4–17–C–X4–6–C–X3–H recently (Peng et al. 2012). The well-studied mammalian protein tristetraprolin, a member of the TIS11 family, contains two CCCH zinc fingers and can directly bind to AU-rich elements within the 3′-untranslated region of the target transcripts to facilitate mRNA degradation (Lai et al. 1999, 2000, 2003; Baou et al. 2009). The identified Zfp36l2 protein in mouse, which is an mRNA-binding and destabilizing protein, plays a vital role in the physiological control of female fertility at the level of early embryonic development (Ramos et al. 2004; Stumpo et al. 2009). In Caenorhabditis elegans, PIE-1 and POS-1 are characterized as two CCCH proteins that can control germ cell fate by inhibiting transcription or the activation of protein expression from maternal RNAs (Tenenhaus et al. 2001; Ogura et al. 2003). Another CCCH protein, zinc finger antiviral protein, has been isolated from Rat2 fibroblasts, can directly bind to specific viral RNA sequences through its CCCH motifs and inhibit retroviral RNA production (Gao et al. 2002). Although many CCCH proteins have been characterized in animals, only a small number have been functionally characterized in plants. CCCH proteins have a large degree of functional diversity and are involved in a wide range of biological processes, such as embryo development (Li and Thomas 1998), floral morphogenesis (Li et al. 2001), FRIGIDA-mediated winter-annual habit (Schmitz et al. 2005), salt stress responses (Sun et al. 2007), seed germination (Kim et al. 2008), and secondary cell wall biosynthesis (Ko et al. 2009) in Arabidopsis. In rice, many CCCH genes

13

Mol Genet Genomics

are involved in processes such as leaf senescence (Kong et al. 2006) and plant architecture determination (Wang et al. 2008b). In particular, OsC3H12 (Os01g68860) has shown that enhanced resistance to bacterial blight disease accompanied by the accumulation of jasmonic acid (JA) and induced expression of JA signaling genes (Deng et al. 2012), as well as OsTZF1 confers delayed senescence and stress tolerance by regulating stress-related genes (Jan et al. 2013). Besides, CsSEF1 was characterized, which encoding putative CCCH-type zinc finger protein expressed during cucumber somatic embryogenesis (Grabowska et al. 2009). Another CCCH protein from cotton, GhZFP1, interacts with GZIRD21A and GZIPR5 to enhance tolerance to drought, salt, salicylic acid, and fungal disease stresses in transgenic plants (Guo et al. 2009). The current study was endeavor toward genomic identification of all CCCH domain-containing genes in citrus to enhance our understanding of their functions. We identified 62 CCCH genes by studying sequence phylogeny, genome organization, chromosomal location, gene structure, and conserved motifs. We also identified SNPs and InDels in CCCH genes between different citrus species. In addition, 24 genes were selected for investigation of their expression patterns in different tissues. The expression patterns of several genes were further surveyed under drought and ABA stress conditions. Our results provide a subset of candidate genes that may be used for future investigation into the functions of CCCH genes in citrus development and stress response.

Methods Identification of CCCH genes in citrus Citrus clementine genome database (http://www.phytozome. net/clementine.php) was employed to identify CCCH motif-containing proteins using the Basic Local Alignment Search Tool algorithms (BLASTP) and TBLASTN with the published Arabidopsis, rice, Populus, human and Trypanosoma CCCH proteins as query sequence and with e-value cutoff set as 1e−005 (Hudson et al. 2004; Wang et al. 2008a; Kramer et al. 2010). The Hidden Markov Model of Simple Modular Architecture Research Tool (SMART) (Letunic et al. 2004) and Pfam (Finn et al. 2006) were used to examine all obtained protein sequences for the presence of CCCH motif. The sequences identified by SMART (Sm00356) and Pfam (PF00642) were considered to be citrus CCCH proteins. Online web server FGENESH (Salamov and Solovyev 2000) was used to correct the predicted genes by manual re-annotation. The sequences were further examined for the CCCH domain using the InterProScan program (Quevillon et al. 2005) with stringent parameters.

Mol Genet Genomics

We identified 62 unique CCCH motif-containing genes and named them CleC3H01 to CleC3H62, following the nomenclature proposed by Hu et al. (2010). The number of CCCH genes we identified is far more than that predicted in PlantTFDB (Zhang et al. 2011a). To further characterize the citrus CCCH domain-containing proteins, an online ExPasy program (http://web.expasy.org/protparam/) was used to calculate the length, molecular weight, and isoelectric point of each protein. Phylogenetic analysis of citrus CCCH genes The ClustalX (version2.1) program was used to generate multiple alignments of the 62 amino acid sequences. Errors were manually corrected. The phylogenetic trees were generated with MEGA4.0 using the Neighbor-Joining (NJ) algorithm (Tamura et al. 2007). Bootstrap analysis with 1,000 replicates was used to evaluate the significance of the nodes. Pairwise gap deletion mode was used to ensure that the divergent domains could contribute to the topology of the NJ tree. Functional assignments and sequence properties of citrus CCCH genes To assign putative functions to the CCCH genes, the Blast2go program was run locally to BLAST against a reference database that stores UniProt entries, Gene Ontology, Enzyme Commission, and Kyoto Encyclopedia of Genes and Genomes annotation (Schmid and Blaxter 2008). The amino acid sequences of CCCH proteins were analyzed for physicochemical parameters (ProtParam). The exon/intron organization of CCCH genes were identified by comparing the coding sequences with their corresponding genomic sequences using Gene Structure Display Server program (Guo et al. 2007). SMART (Letunic et al. 2004) and Multiple EM for Motif Elicitation were used to identify conserved motif structures of CCCH protein sequences.

Plant material and treatments Clementine mandarin was grown in the greenhouse (16 h light/8 h dark, 25 ± 2 °C) in a 3:1 soil: sand mixture. Seedlings that germinated after 6 weeks were used for sample collection from different tissues and for different stress treatments. For drought stress, the seedlings were treated with 300 mM mannitol and samples were collected at 0, 1, 6, 12, 24 and 48 h after treatment. For ABA treatment, the seedling leaves were sprayed with 100 μM ABA and samples were collected at 0, 0.5, 1, 2, 4 and 6 h after treatment. Treatment with deionized water was performed as a control at 0 h point. The control and stress-treated plants were collected, frozen in liquid nitrogen, and stored at −80 °C to await further analysis. At least three biological replicates were performed for each sample at all developmental stages. Real‑time PCR verification Total RNA was isolated using Oligotex mRNA mini kit (Qiagen, USA) according to the manufacturer’s instructions. Total RNA was treated with DNase I and first strand synthesis of cDNA was performed using RT Primer Mix and Primescript RT Enzyme Mix I. The gene-specific primers were designed using Primer 5.0 with melting temperatures of 58–60 °C, primer lengths of 19–20 bp and amplicon lengths with 91–242 bp. Real-time RT-PCR was conducted on LightCyclerW480 Detection System (Roche, Germany), as described previously (Zhang et al. 2011b). At least three replicates were performed for each gene. The expression level of the citrus β-actin was used as the internal reference gene. Relative gene expression with respect to β-actin was determined as described previously (Livak and Schmittgen 2001). One-way ANOVA was performed by SPSS to obtain the P values.

Results

Chromosomal location of citrus CCCH genes

Characterization and analysis of citrus CCCH genes

To determine the physical location of CCCH genes, the starting position of all CCCH genes on each chromosome was confirmed by BlastN searches against the local database of the complete sequence of the sweet orange genome (Xu et al. 2013). MapInspect software (http://www.plantb reeding.wur.nl/uk/softwaremapinspect.html) was used to confirm chromosomal location of citrus CCCH genes and results were revised manually. The duplication events of CCCH genes in citrus were searched based on previous parameters: e-value 90 % (Song et al. 2013).

All the proteins containing the motif C–X4–15–C–X4–6–C– X3–H were selected in citrus genome database. This motif covers both the conventional (C–X7–C–X5–C–X3–H and C–X8–C–X5–C–X3–H) and the recently defined non-conventional CCCH motifs. The HMM profiles of the CCCH genes from Arabidopsis, rice, Populus, human, and Trypanosoma were used as query to identify the CCCH motifcontaining genes using BLASTP and TBLASTN programs. A total of 62 non-redundant CCCH genes (named CleC3H01 to CleC3H62) were obtained. The number of identified citrus CCCH genes (62) differs from other

13



representative species, such as Populus, Arabidopsis, rice, maize, mouse, human and Trypanosoma brucei, which contain 91, 68, 67, 68, 58, 55 and 48 previously predicted CCCH genes, respectively (Hudson et al. 2004; Liang et al. 2008; Wang et al. 2008a; Kramer et al. 2010; Chai et al. 2012; Peng et al. 2012). The 62 identified citrus CCCH genes are presented in Table 1, which includes the length of the Open Reading Frame (ORF), the number of exons, molecular weights, isoelectric points (PIs), and the CCCH motif number of each protein. These CCCH genes are protein coding and the length of amino acid (aa) sequences ranges from 121 (CleC3H25) to 2,165 (CleC3H07) with an average of 556 aa. The isoelectric points varied from 4.75 (CleC3H20) to 9.67 (CleC3H61). Further details on the sequences and motifs of the 62 CCCH genes are provided in Online Source 1. Previous studies indicate that members of the CCCH gene families in both animals and plants have between one and six CCCH motifs (Hudson et al. 2004; Wang et al. 2008a; Kramer et al. 2010; Chai et al. 2012). In this study, we identified the motif characteristics of CCCH genes in citrus, Arabidopsis, rice and Populus (Fig. 1). Similar to the other three species (Arabidopsis, rice and Populus), all of the citrus CCCH genes have between one and six CCCH motifs and 58.1 % of them have at least two CCCH motifs. A total of 132 CCCH motifs were identified (Table 2), which were fewer than those found in Arabidopsis (152), rice (150), Populus (211) and maize (180). Two conventional CCCH motifs (C–X7–C–X5–C–X3–H and C–X8–C– X5–C–X3–H) accounted for 82.3 % of all the citrus CCCH motifs and constituted the largest two groups, similar to those found in Arabidopsis (82.2 %), rice (78.7 %), Populus (82.0 %) and maize (79.4 %), followed by C–X5–C– X4–C–X3–H and C–X7–C–X4–C–X3–H. It is noteworthy that the C–X7–C–X6–C–X3H motif only has been identified in Arabidopsis, rice and maize as compared with in citrus and Populus. Surprisingly, we did not find unique motifs in citrus as compared with Arabidopsis, while the C–X10– C–X5–C–X3H motif is unique to citrus as compared with Populus. Chromosomal location and gene duplication of citrus CCCH genes Previous studies have shown that the clementine mandarin is a hybrid of the ‘Mediterranean’ mandarin (C. reticulata)  × sweet orange (tangor) (Nicolosi et al. 2000; Ollitrault et al. 2012; Xu et al. 2013). Therefore, the CCCH genes were located in the sweet orange genome. According to the starting position of each gene, 54 of the 62 CCCH genes were found to be unevenly distributed across the 9 citrus chromosomes, whereas the chromosomal location of the 8 remaining genes (CleC3H04, 12, 28, 35, 36, 37, 56

13

Mol Genet Genomics

and 58) is unknown (Fig. 2). Among these 54 genes, chromosome 2 contains the largest number of CCCH genes (12). By contrast, chromosomes 1, 3, 6, and 9 each have fewer CCCH genes; 2, 4, 5 and 3, respectively. Chromosomes 4 and 7 each contain 8 CCCH genes while chromosomes 5 and 8 both have 6 (Fig. 2). Inspection of the phylogenetic tree topology uncovered several pairs of CCCH proteins with a high degree of homology, suggesting that they are putative paralogous pairs (homologous genes within a species that diverged by gene duplication) (Fig. 3). The identification of paralogs was based on the following criteria: (1) the length of aligned sequence covers >80 % of the longer gene; and (2) the similarity of the aligned regions is >70 % (Gu et al. 2002; Yang et al. 2008). In total, only six pairs (CleC3H28/CleC3H29, CleC3H35/CleC3H36, CleC3H08/CleC3H56, CleC3H46/ CleC3H58, CleC3H21/CleC3H22 and CleC3H57/CleC3H60) of putative paralogous CCCH proteins were identified, accounting for 90 %) based on stringent criteria. The distance between the two genes is