The GIGADB Dateset¹ provided the basic data, including protein numbers, cDNA sequence numbers and open reading frame lengths, as well as the genomic sequences and their scaffold positions. The published rice CCCH proteins were originally used as probes to search for CCCH members in the moso bamboo genome with BLASTP and TBLASTN having an E-value cutoff set as 1e-005, and the sequences were detected again with SMART² (Letunic et al., 2004) and Pfam³ (Finn et al., 2008) to ensure the existence of the CCCH domain. ExPASY provided information about molecular weights and isoelectric points (Gasteiger et al., 2003).

The ClustalX (version 1.83) program (Chai et al., 2012) was used for the multiple sequence alignment of CCCH sequences of moso bamboo. The neighbor-joining method of MEGA 6.0 software⁴ was used to construct the phylogenetic trees. Bootstrapping of each branch was performed using 1,000 replications. MEGA 6.0 and ClustalX programs were used to construct the integrated phylogenetic tree of moso bamboo, maize and rice. The Gene Structure Display Server program (Han et al., 2014) was used to study the intron/exon structure of PeC3H genes.

The MEME⁵ program was used to determine the conserved motifs of the 119 amino acid sequences of the PeC3H genes with the following parameters: the maximum number of motifs was 15 and the optimum width was 18–200 residues (Ma et al., 2014). In addition, sequence logos of the CCCH zinc finger motifs were produced using online WebLogo software (Crooks et al., 2004).

The genomic location of PeC3H genes were identified according to the information provided by moso Bamboo SMRT database, and the collinearity of CCCH gene family of bamboo were analyzed by using the default parameters of MCScanX (Wang et al., 2012) and shown by Circos (Krzywinski et al., 2009). To study relationships between orthologous PeC3H genes and other selected species, the homolinear analysis maps were constructed using Dual Synteny Plotter software⁶. DnaSP 5 software was used to calculate the PeC3H’s K_a and K_s values. The duplication time (T) was calculated as T = K_s/2λ × 10^–6 million years ago (MYA) (λ = 6.5 × 10^–9 for grasses) (Gaut et al., 1996).

To study the cis-elements of PeC3H genes’ promoter sequences, the cis-elements upstream of the transcription initiation site (2,000 bp of genomic DNA sequence) were analyzed using the PlantCARE program (Liu et al., 2009).

Bamboo seedlings gathered from Guilin City, Guangxi Province, China and were cultured in an artificial climate incubator for 7 weeks (26 ± 2°C, 16 h light) in preparation for subsequent experiments. Sow the moso bamboo seeds in a round pot (diameter 20.5 cm, height 19 cm). Afterward, the seedlings were sprayed independently with H₂O, 100-μM ABA, 100-μM Me-JA, and 100-μM GA solutions (10 ml) at 0 h. Untreated seedlings served as the controls. Leaf samples were taken at different times (1, 3, 6, 12, and 24 h). Subsequently, total RNAs of these samples were extracted using the TRIzol method. RNA was reversed transcribed into cDNA using a PrimeScript^® RT Reagent Kit (TaKaRa, Dalin, China).

To study the expression levels of PeC3Hs under different hormone-treatment conditions, 12 genes were selected for qRT-PCR analysis. The specific primers for the 12 selected PeC3H genes were designed using Primer 5.0 software, and the intrinsic membrane protein 41 (TIP41) was used as the reference gene (Fan et al., 2013). The program settings for the qRT-PCR were as follows: 95°C for 30 s; 40 cycles of 95°C for 5 s, 55–60°C for 10 s, and 72°C for 10 s. The relative expression levels of relevant genes were calculated using the following formula: 2^–ΔΔCT [ΔCT = C_T, _Target − C_T,_TIP41. ΔΔC_T = ΔC_T − ΔC_T, _H2O]. (Schmittgen and Livak, 2008). GraphPad5 was used to analyze the data.

The full-length coding sequence of PeC3H74 was obtained using specific primers (F: ATGAC TCCCTTGACTGGTTTCTTGACTGTT, R: GATAAGCAGA TTGGTAGCCTAGCA ACATAG) and cloned into pCAMBIA1305, containing CAMV35S and green fluorescent protein (GFP), using XbaI and Sam I restriction sites. PeC3H74-GFP was introduced into Agrobacterium tumefaciens EHA105, then injected into Nicotiana tabacum and observed using a confocal laser scanning microscope (Carl Zeiss LSM710, Germany) (Zhou et al., 2013).

The pGBKT7 vector (Clontech, Palo Alto, CA, United States) was used to study the transcriptional activity of the PeC3H74 protein in yeast. The full-length PeC3H74 was PCR amplified using gene-specific primers (F: GGAATTCATGACTCCCTTGACTGGTTT; R: ACGCGTCGACGATAAGCAGATTGGTAGCCT) and cloned into pGBKT7 vector containing a GAL4 DNA-binding domain. The empty plasmids (pGBKT7) and pGBKT7-53 + pGADT7-T antigen as negative and positive experimental controls. pGBKT7-PeC3H74, pGBKT7 and pGBKT7-53 + pGADT7-T were transformed in yeast strain by the lithium acetate method, respectively. The transformants were dropped on various SD selective media, namely SD/–Trp and SD/–Trp/–His/–Ade/X-α-GAL and incubated at 30°C for 3–5 days.

To determine drought tolerance, 3-week-old plants (OE-6, OE-7, OE-9, WT) were withholding water for a week. The electrolyte leakage was measured according to previous studies (McKersie et al., 1996). MDA and H₂O₂ content was measured using Biochemical Assay Kit (BC0025 and BC3595, Liandong U Valley A85, Tongzhou District, Beijing, China). According to the method provided by (Kumar et al., 2014), DAB staining of the leaves after drought treatment. PeC3H74-OE transgenic Arabidopsis seedlings were grown on 1/2 MS medium for 3 days, and then transferred to 1/2 MS medium containing 10 μM ABA for vertical cultivation for 7 days, and the change in root length was observed.

The mature Arabidopsis leaves were soaked in (0 and 1 μM) ABA for 6 h, and then transferred to a solution containing 25% propanetriol and 2 g/mL chloral hydrate for 4 days. After removing chlorophyll, observe and measure under a fluorescence microscope. The stomata ratios of width to length were >0.5 (open), 0.5–0.2 (partially closed) and <0.2 (closed), respectively.

Statistical significance was determined using a paired Student’ s t-test⁹. The mean ± standard error from the mean (SE) of at least three replicates are presented, and significant differences relative to controls are indicated at ^∗p < 0.05 and ^∗∗p < 0.01.

Through a BLASTP preliminary screening, 123 putative CCCH-encoding genes of moso bamboo were obtained. The present of the CCCH domain was determined by Pfam and SMART, and 119 potential CCCH sequences were ultimately identified as CCCH genes (PeC3H1-119) (Liu et al., 2011). Detailed information of PeC3H genes are given in Table 1. The identified PeC3H genes encoded proteins ranging from 73 to 2,039 amino acids, with a mean length of 533 amino acids. The greatest and lowest molecular weights were 224.86 and 8.12 kDa, respectively.

We compared the sequences of 150 C-X₈-C-X₅-C-X₃-H and 62 C-X₇-C-X₅-C-X₃-H motifs separately, and then generated the sequence logos (Figure 1). To study the characteristics of the CCCH motifs of the PeC3H genes, the sequences were compared with CCCH motifs of rice and A. thaliana. In addition, the compound sequence logos of the two motifs were generated in the same way, and the motif sequences were highly conserved. Four of the amino acids in the three patterns were completely conserved, similar to the sequence diagrams provided by Pfam and SMART. There were also some differences in the sequence logos among the three plants. For example, lysine occurred more often at C1 + 6 than arginine in both moso bamboo and rice B motif logos, while the reverse case was observed in the A. thaliana B motif logo.

To evaluate the phylogenetic relationships among the 119 predicted moso bamboo CCCH zinc finger proteins, we used the 119 putative protein sequences to construct an unrooted phylogenetic tree with 1,000 bootstrap replicates (Wang D. et al., 2008) (Supplementary Figure 1). We divided them into 13 subfamilies (CCCH A to M) based on bootstrap values above 100. However, the Pe3H24 gene was not included in any of the 13 subfamilies because its bootstrap values with other genes were less than 100, which is a phenomenon that also occurs among the CCCH genes of other plants (Chai et al., 2012). Among the 13 subfamilies, subfamily L contained the greatest number of CCCH (29), followed by subfamilies C (25), F (15), A (12), and I (9). Both G and K subfamilies had six members, H and J had four members, and the B, D, E, and M had only two members each.

The MEME server identified 15 conserved motifs in the CCCH protein of moso bamboo (Supplementary Figure 2), and details of conserved amino acids are provided in Supplementary Table 1. Motif 1, 4, and 15 are typical CCCH motifs, and each PeC3H gene contains at least one CCCH motif (motif 1, 4, and 15). PeC3H11, −15, −28, −33, −51, and −81 in I subfamily all contain SWIB motif (motif 13), and the SWI/SNF protein is required for proper protein complex formation in yeast. However, there are also some genetic differences within the same subfamily. In subfamily A, only PeC3H110 contained motif 6, meanwhile, motif 5 only appears in PeC3H4 and -8.

By analyzing the Gene Structure Display Server diagram (Supplementary Figure 3), we found that there were differences in the numbers of introns in the different genes, ranging from 0 to 13, with an average of 4. In total, 15 PeC3H genes had no introns, while 37 genes contained only one or two introns. Some subfamilies had similar numbers of introns in each member, but some subfamilies had significant differences. For example, PeC3H9, −28, and −111 contain 1, 7 and 11 introns, respectively, and all they are all members of the I subfamily.

The strongly conserved microsynteny of the CCCH gene was observed by comparing CCCH genes with other genes in moso bamboo (Figure 2), the synteny pairs details were shown in Supplementary Table 2. Pe_Scaffold1 did not contain any genes and is, therefore, not shown in Figure 2. In total, 63 collinear gene pairs were identified. We found that of the 119 CCCH genes, a pair of genes (PeC3H88/-89) in scaffold 6 were at a distance of less than 100 kb, which may be caused by a tandem duplication (Indicated by the red line) (Figure 2) (Lin et al., 2014). Both PeC3H88 and PeC3H89 belonged to the C subfamily, and their bootstrap values reach 1,000, indicating a highly conserved. The analysis of the collinear gene pairs of CCCH genes showed that 62 gene pairs remained in the conserved positions in the segmental duplicated blocks, indicating that gene duplication played an important role in the CCCH gene expansion of moso bamboo.

The gene duplication mechanisms of CCCH gene family in moso bamboo was studied by constructing a comparative syntenic maps with four representative species [one dicots (Arabidopsis) (Supplementary Figure 4A) and three monocots (Brachypodium distachyon, rice and maize) (Supplementary Figure 4B)]. The syntenic relationship between CCCH genes of bamboo and four species, the most is maize (118), followed by rice (107), B. distachyon (93) and Arabidopsis (2), indicating that in comparison with monocotyledonous plants, CCCH genes of moso bamboo show a high evolution divergence with dicotyledonous plants. PeC3H4 and PeC3H8 have syntenic pairs in all four plants, and these two genes may play a key role in the evolution of CCCH genes family. Through to K_a/K_s ratios calculations of the CCCH genes synteny pairs (Supplementary Tables 2–6), K_a/K_s < 1 of most of the synteny pairs, show that the moso bamboo CCCH genes family during evolution may experience a strong purifying selection pressure, and the PeC3H genes experienced a large-scale duplication event, probably 5.13–271.79 million years ago (MYA) by the formula T = K_s/2λ (λ = 6.5 × 10^–9) (Gaut et al., 1996).

The promoter region of a gene usually contains multiple cis-elements that play key roles in responses to different stresses (Dung Tien et al., 2012). Cis-elements directly influence gene regulation involved in stress-responsive gene expression (Bilas et al., 2016). Various interactions between cis-acting elements and transcription factors function as molecular switches for transcription to determine transcription initiation events (Bilas et al., 2016). Therefore, identifying cis-acting elements in promoter region is very important to understand the role of transcription factors in stress response. The cis-acting elements in promoter regions of 119 PeC3H genes were detected to prepare study their regulatory mechanism. We focused on three types of cis-elements (ABA, Me-JA, and GA) (Figure 3 and Supplementary Table 7), and there were a large number of cis-elements related to these three hormones among the PeC3Hs. ABA-responsive elements (ABREs), the cis-acting elements of ABA, existed in many PeC3H genes. Therefore, we speculated that most PeC3H genes were regulated by ABA stress responses. In total, 85.7% (102/119) of PeC3H gene promoter regions contained an ABA-responsive element (Figure 3 and Supplementary Table 7). Meanwhile, we found 440 ABRE elements in PeC3Hs, and the largest number of three elements. The CGTCA/TGACG-elements are cis-acting elements of the Me-JA response and have regulatory effects on plant leaf senescence (He et al., 2002). The CGTCA/TGACG-elements were found in promoter regions of 101 (84.8%) PeC3Hs (Supplementary Table 7), and 300 CGTCA/TGACG-elements in PeC3Hs were discovered. The GARE/P-box/TATC-box-element associated with GA was present in 62 of 119 (52.1%) promoter regions of PeC3Hs, and 104 GARE/P-box/TATC-box-element in PeC3Hs were discovered. In addition, none of the three cis-acting elements was found in PeC3H1. The analysis of cis-acting elements in the PeC3Hs will aid in further studies of the tolerance of moso bamboo.

There are some CCCH genes were positively or negatively regulated by ABA, GA and Me-JA in rice (Huang et al., 2012). Twelve genes (PeC3H2, −7,−11, −20, −21, −26, −56, −34, −74, −99, −100, and −110) were selected to study the expression level using qRT-PCR. The specific primers used in qRT-PCR analysis of these genes are shown in Supplementary Table 8. At the same time, the expression levels of these genes under water treatment were analyzed (Figure 4). Under water treatment, although the gene expression level changes, it is basically low expression.

The ABA treatment of moso bamboo, 11 genes reached their maximum expression levels at different times, except PeC3H20 was inhibited compared with the control (Figure 4). In particular, gene PeC3H11 had an expression level that was 32-fold greater than that of the control at 24 h (Figure 4).

After the GA treatment, the expression profiles of five genes (PeC3H7, −21, −26, −56, and −99) were suppressed, while those of other PeC3H genes were up-regulated, but the expression levels were lower than those of the control at some time. For example, PeC3H74 reached its peak at 6 h and was greater than the control; however, its expression levels at other times were lower than the control (Figure 4). After the GA treatment, the expression levels of PeC3H genes were not significantly different from those of the control. For example, the expression levels of PeC3H2 and PeC3H56 were initially up-regulated, reaching their maximum levels, which were no more than fivefold those of the control. Furthermore, the expression levels of five genes (PeC3H7, −11, −20, −100, and −110) were greatest at 1 h.

Next, the expression levels of CCCH genes after Me-JA treatment was analyzed, and 4 genes (PeC3H20, −21, −99, and −110) were suppressed, while those of other PeC3H genes were up-regulated (Figure 4). At 24 h, the expression levels of four genes (PeC3H7, −11, −26, and −56) were the greatest. In addition, PeC3H2 and PeC3H100 reached the highest expression levels at 1 h. PeC3H34 and PeC3H74 reached their highest expression levels at 6 h.

Under ABA treatment, only the expression of PeC3H20 was suppressed, but under the other two hormone treatments, at least 4 genes were suppressed. Only PeC3H74 gene under the treatment of three hormones, the highest expression level exceeds fivefold, suggesting that this gene plays a role in resisting stress during plant growth and development. Thus, most PeC3H genes were up-regulated under stress treatments, indicating that they play key roles in abiotic and biotic stress responses.

To study the subcellular localization of PeC3Hs, a PeC3H-GFP vector was constructed and transiently expressed in N. tabacum leaves. 35S:GFP served as a control (Figure 5A). PeC3H74 gene expression had higher induction under ABA, GA, and Me-JA treatment, so it was further analyzed. Based on the GFP signal, PeC3H74 was localized to the cytomembrane.

The Y2H yeast strain was used to study the transcriptional activities of PeC3Hs. The pGBKT7-PeC3H74, positive control plasmids pGBKT7-53 and pGADT7-T, and pGBKT7 (the negative control plasmid) transformed into the Y2H yeast strain, independently. The transformants were cultured on SD/–Trp medium, and they all produced white colonies (Figure 5B). On the SD–Ade/–His/–Trp/X-α-GAL medium, strains containing PeC3H74 and positive control turned blue, while the negative control did not grow (Figure 5B). These results suggest that PeC3H74 can function as a transcriptional activator.

Because the transgenic technology of moso bamboo is still immature, we transferred the PeC3H74 gene into A. thaliana and studied whether PeC3H74-OE was related to drought stress through transgenic Arabidopsis strains (OE-6, OE-7, and OE-9). Leaves of transgenic Arabidopsis strains grown for 2 weeks drive GUS activity (Supplementary Figure 5). After 6 days of cultivation on 1/2 MS medium, Arabidopsis seedlings were placed on 1/2 MS medium containing different concentrations of ABA (0 and 10 μM). The root lengths of wild type and PeC3H74-OE on 0 μM ABA were not significantly different, while the root length of PeC3H74-OE on 10 μM ABA was significantly different from that of wild type (Figures 6A,B).

Subsequently, we examined the drought tolerance of PeC3H74-OE strain. 3 weeks of WT and PeC3H74-OE plants were not irrigated for 7 days, WT withered more than PeC3H74-OE plants (Figure 7A). After re-watering for 3 days, all PeC3H74-OE plants were survived, the survival rate of the WT was only 16.7 percent. In addition, before drought treatment, the electrolyte leakage (EL) and malondialdehyde (MDA) of WT and PeC3H74-OE plants are not much different (Figures 7B,C). After 7 days of drought, the content of EL and MDA of WT Increased, significantly different from PeC3H74-OE plants. The results showed that after drought treatment, PeC3H74-OE plants suffered less membrane damage than WT. Drought stress can lead to accumulation of reactive oxygen species (ROS). Before drought treatment, DAB staining showed that H202 accumulated less in WT and PeC3H74-OE plants (Figure 7D). After 10 days of drought treatment, the accumulation of H₂O₂ in PeC3H74-OE plants was significantly less than that of WT. The detection of H₂O₂ content before and after treatment showed that after treatment, the difference in H₂O₂ between WT and PeC3H74-OE plants was significant, which was consistent with DAB staining results (Figure 7E). These results indicate that PeC3H74 enhances the drought tolerance of Arabidopsis.

The opening and closing of stomatal are affected by ABA, and the loss of water on the leaf surface is closely related to the regulation of stomatal (Osakabe et al., 2014). In order to determine the regulating effect of PeC3H74-OE on plant stomatal size. Observe the changes in stomatal size of WT and overexpressing strains without treatment or 1 μM ABA treatment, respectively. Observation by fluorescence microscopy revealed that the stomatal size of WT and overexpressing strains were not significantly different under untreated conditions. After 6 h of 1 μM ABA treatment, the rate of stomatal closure and partial closure of PeC3H74-OE plants was significantly higher than that of WT (Figures 8A–C). The results showed that PeC3H74 may induce stomatal closure through ABA to achieve the purpose of drought resistance.