Genome-Wide Analysis and Expression of Cyclic Nucleotide–Gated Ion Channel (CNGC) Family Genes under Cold Stress in Mango (Mangifera indica)

The ‘king of fruits’ mango (Mangifera indica) is widely cultivated in tropical areas and has been threatened by frequent extreme cold weather. Cyclic nucleotide–gated ion channel (CNGC) genes have an important function in the calcium-mediated development and cold response of plants. However, few CNGC-related studies are reported in mango, regardless of the mango cold stress response. In this study, we identified 43 CNGC genes in mango showing tissue-specific expression patterns. Five MiCNGCs display more than 3-fold gene expression induction in the fruit peel and leaf under cold stress. Among these, MiCNGC9 and MiCNGC13 are significantly upregulated below 6 °C, suggesting their candidate functions under cold stress. Furthermore, cell membrane integrity was damaged at 2 °C in the mango leaf, as shown by the content of malondialdehyde (MDA), and eight MiCNGCs are positively correlated with MDA contents. The high correlation between MiCNGCs and MDA implies MiCNGCs might regulate cell membrane integrity by regulating MDA content. Together, these findings provide a valuable guideline for the functional characterization of CNGC genes and will benefit future studies related to cold stress and calcium transport in mango.

Mango, known as the 'king of fruits', is one of the most popular fruits [23]. Its annual fruit yield ranks fifth around the world [24]. The main cultivated Mangifera species in the tropical areas around the world is Mangifera indica [25]. As a typical tropical plant, mango is sensitive to cold temperature [26], especially frequent extreme cold weather, which has significantly threatened to mango production in recent years [27]. However, few studies have revealed the molecular basis of cold stress response in mango trees, even if the cold storage of detached fruits has been well studied [28]. Considering that overexpression of CNGC genes promotes rice cold tolerance, we conducted genome-wide analysis of the CNGC family in mango and evaluated its expression in mango tissues, as well as that under cold stress [29]. The results showed that expression of several MiCNGCs was upregulated under cold temperature and highly correlated to leaf damage index malondialdehyde (MDA) contents, implying their beneficial roles in regulating mango cold tolerance. Therefore, our study will offer guidance for functional characterization of CNGC genes and benefit future studies in mango.

Characterisation of CNGC Family in Mango
The Arabidopsis and rice CNGC genes were selected to search homologous genes in the genomes of Amborella trichopoda, sweet orange (Citrus sinensis) and mango [24,30,31]. Because sweet orange belongs to the Sapindales order together with mango, it was selected as a nearby reference [24]. Amborella was chosen as the outgroup reference to the Sapindales order since it is the basal angiosperm [30]. As a result, 7, 33 and 43 CNGC genes were identified in above three species (Tables 1 and S1). In mango species, these genes ranged from 399 to 2343 bp with predicted protein lengths of 132-780 aa. Their molecular weights and theoretical pI ranged from 14907.27 to 89883.31 Da and 4.88 to 9.67, respectively. Most of them (36 of 43) were predicted to be located at the plasma membrane, contain 3-7 transmembrane helices ( Figure S1). Five and two were predicted to be nuclear and extracellular, respectively. There was no more than one transmembrane helix in these seven genes. A total of 39 mango CNGC genes were distributed at 13 chromosomes. Two major gene clusters, including 10 and 12 CNGC genes at chromosomes 9 and 15, respectively, were labeled ( Figure 1).

Phylogenetic Relationships of Mango CNGC Genes
The CNGC proteins of Arabidopsis, rice, Amborella, sweet orange and mango were selected to construct the maximum-likelihood phylogenetic tree ( Figure 2). Finally, four groups (I-IV) were generated, where almost the same quantities of CNGC proteins from the four species existed in each group. Interestingly, only sweet orange and mango CNGC proteins were clustered into group IV-C. Mango CNGC proteins were further aligned to identify their conserved domains. The results showed that most of them contained the CNBD domain Figure S2). However, nine mango CNGC proteins lost the N-termini due to evolution issues, namely MiCNGC10, 11, 15, 21, 25, 27, 28, 35 and 36. We further calculated the values of synonymous substitutions (Ks), nonsynonymous site (Ka) and their ratio (Ka/Ks). The results indicated that all the Ka/Ks values were below 1 in homologous gene pairs (Table S2).

In Silico Expression of CNGC Genes in Mango Tissues
We further examined the expression levels of mango CNGC genes in leaf, fruit peel and fruit flesh based on published transcriptome data. The expression data were normalized into fragments per kilobase of exon model per million mapped fragments (FPKM) (Table S3) (Table S3).

Phylogenetic Relationships of Mango CNGC Genes
The CNGC proteins of Arabidopsis, rice, Amborella, sweet orange and mango were selected to construct the maximum-likelihood phylogenetic tree ( Figure 2). Finally, four groups (I-IV) were generated, where almost the same quantities of CNGC proteins from the four species existed in each group. Interestingly, only sweet orange and mango CNGC proteins were clustered into group IV-C. Mango CNGC proteins were further aligned to identify their conserved domains. The results showed that most of them contained the CNBD domain Figure S2). However, nine mango CNGC proteins lost the N-termini due to evolution issues, namely MiC-NGC10, 11, 15, 21, 25, 27, 28, 35 and 36. We further calculated the values of synonymous substitutions (Ks), nonsynonymous site (Ka) and their ratio (Ka/Ks). The results indicated that all the Ka/Ks values were below 1 in homologous gene pairs (Table S2).

In Silico Expression of CNGC Genes in Mango Tissues
We further examined the expression levels of mango CNGC genes in leaf, fruit peel and fruit flesh based on published transcriptome data. The expression data were normalized into fragments per kilobase of exon model per million mapped fragments (FPKM) (Table S3). MiCNGC17, 19 and 22 were relatively expressed at high levels in all three tis-

In Silico Expression of CNGC Genes under Cold Stress in Mango Fruit Peel
In order to evaluate their molecular functions under cold stress in fruit peel, moderate (12 °C) and extreme (5 °C) temperatures were chosen to compare CNGC expression patterns. The two treatments allow us to better understand the expression tendency of CNGC genes under cold stress. The results indicated that most CNGC genes (33 of 43) showed no or relatively low expression levels (FPKM < 10) under cold treatments of both 5 and 12 °C, implying that they might play minimal roles for cold tolerance (Table S3). However, 10 CNGC genes showed higher expression triggered by cold stress (Figure 4). Among these, three of them showed moderate expression patterns under 12 °C, but more than three-fold upregulated expression level under 5 °C (MiCNGC4, 9 and 34). Five genes showed similar expression patterns whether at 5 or 12 °C (MiCNGC13, 17, 19, 31 and 36), while MiCNGC13 and 36 were up and downregulated over 3-fold under 5 °C after prolonged 14-day cold treatment, respectively. Moreover, MiCNGC22 and MiCNGC33 showed absolutely opposite expression patterns under the two degrees.

In Silico Expression of CNGC Genes under Cold Stress in Mango Fruit Peel
In order to evaluate their molecular functions under cold stress in fruit peel, moderate (12 • C) and extreme (5 • C) temperatures were chosen to compare CNGC expression patterns. The two treatments allow us to better understand the expression tendency of CNGC genes under cold stress. The results indicated that most CNGC genes (33 of 43) showed no or relatively low expression levels (FPKM < 10) under cold treatments of both 5 and 12 • C, implying that they might play minimal roles for cold tolerance (Table S3). However, 10 CNGC genes showed higher expression triggered by cold stress (Figure 4). Among these, three of them showed moderate expression patterns under 12 • C, but more than three-fold upregulated expression level under 5 • C (MiCNGC4, 9 and 34). Five genes showed similar expression patterns whether at 5 or 12 • C (MiCNGC13, 17, 19, 31 and 36), while MiCNGC13 and 36 were up and downregulated over 3-fold under 5 • C after prolonged 14-day cold treatment, respectively. Moreover, MiCNGC22 and MiCNGC33 showed absolutely opposite expression patterns under the two degrees.

Expression Profiles of CNGC Genes under Cold Stress in Mango Leaf
To better understand how the CNGC genes affect cold stress response in mango plant, 10 mango CNGC genes were selected for qRT-PCR validation in mango leaf. These genes include five with expression changes over threefold under cold stress in fruit peel and five with FPKM values over 10 in leaf. The results indicated that most genes were upregulated after cold treatment in leaf except MiCNGC4 ( Figure 5). Among these, six were up-regulated more than threefold, namely MiCNGC9, 13, 17, 22, 31 and 38. Additionally, the expression of MiCNGC13 was not significantly affected when temperature was higher than 4 • C, indicating that MiCNGC13 might have leaf-and fruit-peel-specific temperature sensitivity.
Plants 2023, 12, x FOR PEER REVIEW 8 of Figure 4. In silico expression of CNGC genes under cold stress of (blue) and 12 °C (red) in man fruit peel. The x-axis represents the samples collected at 0, 2, 7 and 14 days after treatment. The axis represents FPKM values. The error bar represents the standard error. * represents that the pression level was up or downregulated over 3-fold.

Expression Profiles of CNGC Genes under cold Stress in Mango Leaf
To better understand how the CNGC genes affect cold stress response in man plant, 10 mango CNGC genes were selected for qRT-PCR validation in mango leaf. Th genes include five with expression changes over threefold under cold stress in fruit p and five with FPKM values over 10 in leaf. The results indicated that most genes w upregulated after cold treatment in leaf except MiCNGC4 ( Figure 5). Among these, were up-regulated more than threefold, namely MiCNGC9, 13, 17, 22, 31 and 38. Ad tionally, the expression of MiCNGC13 was not significantly affected when temperatu was higher than 4 °C, indicating that MiCNGC13 might have leaf-and fruit-peel-spec temperature sensitivity. . In silico expression of CNGC genes under cold stress of (blue) and 12 • C (red) in mango fruit peel. The x-axis represents the samples collected at 0, 2, 7 and 14 days after treatment. The y-axis represents FPKM values. The error bar represents the standard error. * represents that the expression level was up or downregulated over 3-fold.

Positively Correlated Mango CNGC Genes with MDA Contents
The content of malondialdehyde (MDA) is usually considered as a lipid peroxidation index that indicates the damage of stress. We therefore further measured MDA in mango leaf to evaluate the physiological effects of cold stress. As expected, MDA contents were significantly increased under cold treatment, specially under 2 • C ( Figure 6A). This result suggested that temperatures under 2 • C might cause irreversible damage to the mango plant. To further determine whether CNGC genes regulate MDA level or not, we performed correlation analysis to explore CNGC genes that were highly correlated with MDA contents. The results showed that three and five CNGC genes were positively correlated with MDA contents with significant correlation coefficients at 0.05 (R > 0.532) and 0.01 (R > 0.661) cut-offs, respectively.

Positively Correlated Mango CNGC genes with MDA Contents
The content of malondialdehyde (MDA) is usually considered as a lipid peroxidation index that indicates the damage of stress. We therefore further measured MDA in mango leaf to evaluate the physiological effects of cold stress. As expected, MDA contents were significantly increased under cold treatment, specially under 2 °C ( Figure 6A). This result suggested that temperatures under 2 °C might cause irreversible damage to the mango plant. To further determine whether CNGC genes regulate MDA level or not, we performed correlation analysis to explore CNGC genes that were highly correlated with MDA contents. The results showed that three and five CNGC genes were positively correlated with MDA contents with significant correlation coefficients at 0.05 (R > 0.532) and 0.01 (R > 0.661) cut-offs, respectively.

Species-Specific Expansion of CNGC Family in Mango
In the present study, we have successfully identified 7, 33 and 43 CNGC genes in

Species-Specific Expansion of CNGC Family in Mango
In the present study, we have successfully identified 7, 33 and 43 CNGC genes in Amborella, sweet orange and mango (Tables 1 and S1). As a basal angiosperm, Amborella has a relatively small number of CNGC genes compared with other four species (Figure 2). Beyond this, these genes are almost evenly distributed into three groups (I, II and III) and two subgroups (IV-A and IV-B), indicating the conserved evolution patterns and pressure among these groups [4]. Interestingly, there is a subgroup (IV-C) containing CNGC genes from either mango (23) or sweet orange (21). Most of them are distributed at chromosomes 9 and 15 in mango (18) and chromosome 9 in sweet orange (21) (Figure 1), verifying their Sapindales classification. However, the species-specific expansion of CNGC genes among mango and sweet orange also emphasizes the species divergence during evolution [32]. In addition, mango CNGC genes are classified into two major gene clusters, which might be divided from the same cluster during the hypothetical auto-diploidization [24]. Seven mango CNGC genes contain short coding regions (<1000 bp), such as MiCNGC10, MiC-NGC28 and so on (Table 1), which might be caused by frequent chromosomal recombination (Table S3) [33]. They might lose the function of Ca 2+ -permeable channels without complete transmembrane structure ( Figure S1), which also affected their subcellular localizations predicted by CELLO.

Candidate CNGC Genes in Cold Stress Response in Mango
The in silico expression of CNGC genes in mango tissues illustrates that mango CNGC genes have tissue-specific expression patterns (Figure 3). The high and constitutive expression of MiCNGC17 indicate that it might affect the whole mango development period. We further evaluated the expression of CNGC genes under cold stress to identify candidate regulators that contribute to cold tolerance. As shown, five MiCNGCs in fruit peel and five in leaf displayed more than three-fold upregulation of gene expression; specifically, MiCNGC9 and MiCNGC13 are induced in both tested tissues suggesting their potential functions under cold stress (Figures 4 and 5). Since MDA is the main indicator of cell membrane integrity [34], we observed that MDA content is significantly increased at 2 • C compared with 4, 6 and 8 • C ( Figure 6A), which means that lower temperature leads higher damage. Then we revealed that eight MiCNGCs are positively correlated with MDA by correlation analysis (Figure 6B), emphasizing that MiCNGCs might regulate MDA content to adjust cell membrane integrity. Therefore, these eight genes could be considered as early cold-responsive markers in the mango leaf, among which MiCNGC13 ranks first as the early cold-responsive maker gene in the mango leaf and fruit peel.

Bioinformatic Analysis of CNGC Genes
CNGC protein sequences of Arabidopsis and rice were selected as queries to search homologous proteins using the Blastp method [4,35]. The genomes of Amborella, sweet orange and mango were set as targets for sequence retrieval [24,30,31]. The details of CNGC genes in Ambrella and sweet orange are listed in Table S1. The accession numbers and chromosome positions of mango CNGC genes are in Table 1 and were further used to illustrate their distribution in chromosomes using TBtools software [36]. The sequence features of length, molecular weight and pI were predicted using the ProtParam tool with subcellular localization predicted using CELLO software [37,38]. The TMHMM2.0 software was selected to predict the transmembrane helices in CNGC proteins [39]. A maximumlikelihood phylogenetic tree was constructed for phylogenetic analysis with bootstrap values of 1000 using MEGA 7.0 software [40]. The mango CNGC proteins were further aligned by DNAMAN7 software to examine the conserved domains [41]. The Ks, Ka and Ka/Ks values were calculated using TBtools software [36].

In Silico Gene Expression Analysis
The raw data of transcriptome sequencing were downloaded from the Sequence Read Archive (SRA) database [42]. The raw data of leaf (SRR3288569), fruit peel (SRR2960401) and fruit flesh (SRR11060165) were selected to evaluate the expression patterns of CNGC genes in different mango tissues. The raw data of cold-treated fruit peel (SRP066658) were selected to evaluate the expression patterns of CNGC genes in mango fruit peels under cold stress, which were generated at 0, 2, 7 and 14 days after storage at 5 and 12 • C [43]. All raw data were trimmed to generate clean reads, which were further mapped to mango CNGC genes to calculate FPKM values for each gene with the RSEM software [44,45]. The heat map was generated using Morpheus software [46]. The hierarchical clustering method was selected to cluster the FPKM values at the levels of tissues and genes [47].

Plant Materials and Treatment
The mango variety Hongyu was selected for cold treatment. Experiments were carried out in a mango garden (109.11 • E, 19.22 • N) at Jishi Village, Changjiang, China. The branches of 15-year-old trees were put into RR-CTC806C incubators (Rainroot Scientific, Beijing, China) for cold treatments at four temperatures, namely 2, 4, 6 and 8 • C. After that, leaves were collected at 0, 1, 2 and 4 h. Leaves from two branches were put together as one sample and each sample was repeated three times as a biological replicate. All the samples were stored at −80 • C before further analysis.

Quantitative PCR and MDA Assay
Total RNA was isolated from each sample using the Tiangen RNA prep Pure Plant Kit (Tiangen Biomart, Beijing, China) and then transcribed into cDNA using the GoScript Reverse Transcription System (Promega, Madison, WI, USA) [48]. Quantitative PCR was conducted with the QuantStudio 6 Flex Real-Time PCR System (Thermo Fisher Scientific, Waltham, MA, USA). The reaction program contained three stages of initiation (94 • C for 30 s), 40 cycles (94 • C for 5 s and 60 • C for 30 s) and dissociation. The TransStart Tip Green qPCR SuperMix (Transgen Biotech, Beijing, China) was selected as the reaction solution, containing 10 µL supermix, 0.4 µL Passive Reference Dye, 1 µL cDNA, 0.5 µL of two primers and 7.6 µL nuclease-free water. Each sample was repeated three times as technical replicates. The Primer 3 software was used for primer design, and MiActin was selected as a reference gene (Table 2) [49]. The relative expression levels were calculated with the ∆∆Ct method as previously described [50,51]. The MDA contents were examined using the Malondialdehyde (MDA) Content Assay Kit (Solarbio, Beijing, China) according to the manufacturer's instruction [52]. About 0.1 g of each sample was broken into powder in liquid nitrogen and isolated with 1 mL Extraction reagent, which was fully homogenized in ice and centrifuged for supernatant (8000× g for 10 min at 4 • C). A total of 100 µL supernatant was mixed with 300 µL MDA working reagent and 100 µL Reagent III. Distilled water was selected as a blank reference. The mixtures of samples (T) and distilled water (B) were incubated for 60 min under 100 • C, which were cooled and centrifuged (10,000× g for 10 min) for supernatant at room temperature. The supernatants of mixtures (200 µL) were placed in a 96-well flat-bottom plate for the detection of absorbance (A) at 450, 532 and 600 nm with the Biotek Synergy H1 system (Agilent Technologies, Lexington, MA, USA). The MDA content was calculated with the following formula, MDA (nmol/g) = (12.9 × (∆A532 − ∆A600) − 1.12 × ∆A450) × Vrv ÷ (W × Vs ÷ Vsv) = 5 × (12.9 × (∆A532 − ∆A600) − 1.12 × ∆A450) ÷ W. Vrv, Vs, Vsv and W represent total reaction volume (0.5 mL), sample volume (0.1 mL), the volume of Extraction reagent (1 mL) and sample weight, respectively. ∆A450 = A450(T) − A450(B), ∆A532 = A532(T) − A532(B), ∆A600 = A600(T) − A600(B).

Conclusions
Extreme cold weather is a significant threat to the tropical fruit tree mango. Thus, we carried out a genome-wide analysis and assessed the expression of mango CNGC genes under cold stress, due to their importance in calcium-mediated development and cold response. The result revealed 43 CNGC genes with species-specific expansion in the mango genome. In silico expression analysis indicated their tissue-specific expression patterns and five differentially expressed CNGC genes under cold stress in mango fruit peel. The results of the qRT-PCR validation and MDA assay revealed five differentially expressed CNGC genes under cold stress in the mango leaf, which are also positively correlated with MDA contents. These results indicate a candidate early cold-responsive marker gene, MiCNGC13, in the mango leaf and fruit peel, which will be helpful to future studies related to cold stress in mango.

Supplementary Materials:
The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/plants12030592/s1, Figure S1: Transmembrane topology analysis for mango CNGC proteins; Figure S2: Alignment of CNGC proteins in mango; Table S1: Details of CNGC genes in amborella and sweet orange; Table S2: The values of synonymous substitutions (Ks), nonsynonymous site (Ka) and their ratio (Ka/Ks) in homologous gene pairs.; Table S3: FPKM values of mango CNGC genes in different tissues and under cold stress.

Conflicts of Interest:
The authors declare no conflict of interest.