Identification, characterization and expression analysis of calmodulin and calmodulin-like proteins in Solanum pennellii

In plants, the calmodulin (CaM) proteins is an important calcium-binding protein, which play a crucial role in both regulating plant growth and development, as well as in the resistance mechanisms to various biotic and abiotic stresses. However, there is limited knowledge available on the CaM family functions in Solanum pennellii, a wild tomato species utilized as a genetic resource for cultivated tomatoes. In this study, 6 CaM (SpCaM) and 45 CaM-like (SpCML) genes from Solanum pennellii were selected for bioinformatics analysis to obtain insights into their phylogenetic relationships, gene structures, conserved motifs, chromosomal locations, and promoters. The results showed that the 6 SpCaM proteins contained 4 EF-hand domains each, and the 45 SpCML proteins had 2-4 EF-hand domains. The 51 CaM and CaM-like genes contained different intron/exon patterns and they were unevenly distributed across the 12 chromosomes of S. pennellii. The results of the analysis of the conserved motifs and promoter cis-regulatory elements also indicated that these proteins were involved in the responses to biotic and abiotic stresses. qRT-PCR analysis indicated that the SpCaM and SpCML genes had broad expression patterns in abiotic stress conditions and with hormone treatments, in different tissues. The findings of this study will be important for further investigations of the calcium signal transduction mechanisms under stress conditions and lay a theoretical foundation for further exploration of the molecular mechanisms of plant resistance.


Results
Biochemical characteristics of the SpCaM and SpCML proteins. In previous studies, the amino acid sequences of the CaM and CML proteins in Arabidopsis and rice have been reported. In this study, 51 non-redundant sequences were identified in the Solanum pennellii genome, including 6 SpCaM and 45 SpCML 33 . All of the SpCaM and SpCML proteins were named according to their amino acid identity percentage with true canonical CaM7 (AtCaM7) 12,13 . Then, the biochemical characteristics of these proteins were predicted using the ExPASy proteomics server 38 and Wolf PSORT program 39 (Table 1). These SpCaM proteins shared more than 90% sequence similarity with AtCaM7. The number of amino acids (aa), molecular weight, isoelectric point (pi), and percentage of methionine in all the SpCaM proteins, except for SpCaM3, was 149, 16.8 kDa, 4.1, and 6.0%, respectively. The number of amino acids in the SpCML proteins varied from 129 to 282, except for SpCML43, which contained 340 amino acids. The molecular weights of the SpCML proteins ranged from 14.7 to 36.1 kDa, and their pi and percentage methionine ranged from 3.9 to 9.5 and 0.9 to 8.6%, respectively. Except for the absence of cysteine in SpCML5, SpCML9, SpCML12, and SpCML19, the rest of the SpCaM and SpCML proteins contained both cysteine and lysine. All SpCML and SpCaM proteins lacked the N-myristoylation sites, except for SpCML5, SpCML8, and SpCML36. SpCaM1-SpCaM6 possessed a standard structure characterized by 4 EF-hand type calcium-binding regions. The number of SpCML EF-hand domains varied from 1 to 4. The predicted results for the protein subcellular localizations of SpCaM and SpCML are listed in Table 1.

Phylogenetic analysis of SpCaM and SpCML families. The phylogenetic relationships between
the CaM and CML family members of the Solanum pennellii, Arabidopsis, and rice were analyzed using the neighbor-joining method of MEGA6.0 40 . The CaM and CML of the three species were divided into five groups (Fig. 1). The 6, 6, and 5 CaM proteins of the Solanum pennellii, Arabidopsis, and rice, respectively, were individually classified into group V, which was closest to group IV, which was made up of 5 SpCML, 9 AtCML, and 9 OsCML. Only one CML (OsCML-1) existed in group V. In the phylogenetic tree, groups I and III were the largest and the smallest with 59 and 12 CML proteins, respectively. Group I consisted 23 SpCML, 25 AtCML, and 11 OsCML, while group III consisted 5 SpCML, 4 AtCML, and 3 OsCML. Group II consisted 12 SpCML, 12 AtCML, and 8 OsCML.
The dendrogram showed that the proteins of the Solanum pennellii were generally closer to the proteins of Arabidopsis than those of rice, suggesting the phylogenetic relationship between Solanum pennellii and Arabidopsis is relatively closer.

Genomic distribution of SpCaM and SpCML genes.
To determine the distributions of the 6 SpCaM and 45 SpCML on the chromosomes, their physical locations were searched using the NCBI database and were mapped to 12 chromosomes using online MapGene2Chrom program 41 . As can be seen in Fig. 2, the 6 SpCaM and 45 SpCML were unevenly distributed across the 12 chromosomes. Chromosomes 5, 7, and 8 contained only one gene (SpCML), while chromosome 3 contained the most genes (7 SpCML and 1 SpCaM). Chromosomes 1, 4, and 11 all contained seven genes. The respective number of genes located on chromosomes 2, 6, 9, 10, and 12 www.nature.com/scientificreports www.nature.com/scientificreports/ were 6, 4, 2, 5, and 2, respectively. The 6 SpCaM genes were distributed on five chromosomes (chromosomes 1, 3, 10, 11, and 12, which contained 1, 1, 2, 1, and 1 gene, respectively. There was only one pair of SpCaM paralogous genes and two SpCaM genes (SpCaM3 and SpCaM4) on chromosomes 11 and 12, respectively. There were two pairs of paralogous genes (SpCML10/SpCML17 and SpCML11/SpCML31) on chromosome 2, while the other two Genetic structure analysis of the SpCaM and SpCML genes. The exon-intron structures of the genes can provide significant evidence to support the phylogenetic relationships within a gene family 42 , and so genetic structure analysis of the SpCaM and SpCML were carried out using tools available with online website GSDS 43,44 (Fig. 3). The analysis of the exons and introns of the CaM and CML genes enabled the genetic structure of these genes to be further understood. Five groups in the SpCaM and SpCML families were observed, which were consistent with the respective corresponding phylogenetic relationships depicted in Fig. 1. Fifteen genes in group I, all members of group II (except SpCML18), and all members of group III contained only one exon each. The SpCML18 gene (group II) contained one intron and two exons, as did SpCML4, SpCML26, and SpCML40 of group I. In group I, SpCML36, SpCML38, SpCML43, SpCML44, and SpCML45 formed a small cluster containing 4-5  www.nature.com/scientificreports www.nature.com/scientificreports/ exons and 3-4 introns. Group IV was different from the other four groups and could be divided into two subgroups: one subgroup (SpCML1-3) contained four exons and three introns; the other had only one exon and no intron. All SpCaM genes belonged to group V, which involved 2-4 exons and 1-3 introns. Group I and IV genes possessed complex structures, suggesting that gene divergence occurred during evolution.

Conserved motif analysis of the SpCaM and SpCML proteins.
To ascertain the feature sequences of the SpCaM and SpCML protein families, the program MEME 45 was used to analyze the conserved motifs of the 51 genes based on their phylogenetic classifications, and 15 conservative motifs were identified in these proteins (Fig. 4, Table 2). The motifs 1, 2, 3, 4, 5 and 9 were annotated as EF-hand domains by the InterProScan, and the EF-hand domains in motifs 1 and 3 were more complete than those in motifs 5 and 9 ( Table 2). As shown in Fig. 4, in addition to SpCML2, SpCML8, SpCML5, SpCML25, SpCML31, SpCML37, SpCML41, and SpCML42 lacked a EF-hand domain, and SpCML11, SpCML13, SpCML38, SpCML45, and SpCML43 lacked two EF-hand domain, the remaining SpCML and SpCaM all contained four EF-hand domain. The degenerate EF-hands in SpCML did not correspond to motif 6, 7, 8, 10, 11, 12, 13, 14, and 15. The motif structures of the SpCML proteins in groups I and II showed diversity and complexity. All SpCML proteins from group I, except for SpCML40, SpCML43, SpCML38, and SpCML45, contained motifs 1, 2, and 4; motifs 5, 6, 8, 9, 10, and 14 occurred only in group I. Motifs 11 and 15 appeared only in groups I and II, and only in two genes, SpCML43 (group I) and SpCML15 (group II) contained motif 15. Five SpCML, including SpCML43 (group I), SpCML45 (group I), SpCML11 (group II), SpCML31 (group II), and SpCML19 (groupIII), Figure 4. The conserved motifs of SpCaM and SpCML proteins. The phylogenetic tree was constructed using the full-length protein sequences of 6 SpCaM and 45 SpCML. The conserved motifs of SpCaM and SpCML proteins were grouped according to the phylogenetic classification. All motifs were identified by MEME. The motifs 1, 2, 3, 4, 5, and 9 were annotated as EF-hand domains. EF-hands were marked in red border rectangle. (2020) 10

Cis-Element analysis of SpCaM and SpCML genes.
To investigate the mechanisms of the stress-induced gene expression, the online database PlantCARE 46 was used to analyze the cis-elements of the 2000 bp upstream sequences of the promoter regions for the SpCaM and SpCML gene coding sequences ( Table 3). The results revealed that cis-acting elements associated with responses to phytohormones, such as abscisic acid (ABRE), salicylic acid (TCA-element and W-box), gibberellin (GARE-motif), methyl jasmonate (CGTCA-motif), ethylene (ERE) and auxin (TGA-element); adversity, such as anoxia stress (ARE), low temperature (LTR), light (Sp1 and I-box), drought (MBS), dehydration (DRE); and defense and stress-related elements (TC-rich repeats) occurred widely in the promoter regions of the SpCaM, and SpCML. 78.4% of the 51 genes contained ARE, while 72.5% contained ABRE, and 70.6% contained CGTCA-motif and ERE. Other cis-elements (W-box, MBS, I-box, TC-rich repeats, TCA-element, TGA-element, LTR) accounted for a relatively small proportion of these genes, about 30 to 58%. In addition, there were three cis-elements that accounted for less than 18%, including the GARE-motif (17.6%), DRE (7.8%), and Sp1(2.0%). The fact that SpCaM and SpCML genes had the same or different cis-acting elements suggested that these genes may be simultaneously regulated in response to stress sometimes, or specifically regulated at other times when plants resist adverse external environments. These genes are involved in responses to different stresses. Fig. 5, transcripts of 51 genes were tested in all tissue samples, which revealed various expression levels of genes. Heat map displayed the expression level of the SpCaM and SpCML genes in stems (S) and roots (R) relative to leaves (Log2 = 0). One SpCaM and two SpCML genes-SpCaM3, SpCML17, and SpCML38-were upregulated in both stems and roots. SpCML14, SpCML17, and SpCML23 showed high expression levels in stems. And SpCaM1, SpCaM4, SpCML1, SpCML2, SpCML3, SpCML5, SpCML7,SpCML9, SpCML10, SpCML18, SpCML30, SpCML31, and SpCML32 were also highly expressed in roots. The tissue-based expression results indicated that SpCaMs and SpCMLs showed the specificity of gene function during plant growth and development.
The expression levels of SpCaM and SpCML genes under ABA, GA, and SA treatments are depicted in Fig. 7. ABA treatment upregulated (Log2 > 0) 24, 8, and 16 genes in leaves, stems, and roots, respectively. GA treatment induced (Log2 > 0) 40, 19, and 18 genes in leaves, stems, and roots, respectively. SA treatment upregulated (Log2 > 0) 44, 37, and 46 genes in leaves, stems, and roots, respectively. These results indicated that a higher number of SpCaM and SpCML genes were upregulated in response to SA than in response to ABA and GA. Notably, in leaves, 16, 31, and 39 genes showed strong expression (Log2 > 1) upon ABA, GA, and SA treatments. In stems, 5, 13, and 28 genes were strongly induced (Log2 > 1) by ABA, GA, and SA treatments. In roots, 11, 11, and 38 genes were also overexpressed (Log2 > 1) by ABA, GA, and SA treatments. Thus, the total number of strongly upregulated genes was higher in leaves than in stems and roots. The analysis also showed that 27 genes were induced simultaneously (Log2 > 0) in leaves, stems, and roots by SA treatment, while 2 (SpCML4 and SpCML44) and 4 genes (SpCML13, SpCML25, SpCML34, and SpCML44) were upregulated simultaneously in leaves, stems, and roots, respectively, by ABA and GA treatments. These results suggested that SpCaM and SpCML genes may be associated with plant resistance to abiotic stress and regulatory hormones and that different members may play different roles in response to different stimuli.

Discussion
Ca 2+ , as a multifunctional signaling molecule, is at the core of complex antistress signaling pathways in response to adverse environmental conditions 47 . Calmodulin (CaM) is the main calcium sensor in all eukaryotes and can sense changes in the concentration of Ca 2+ . Change of intracytoplasmic free Ca 2+ level is the earliest response of cells to various abiotic and biological stresses 5 . Recent studies have found that CaM and CMLs are key components of stress signal transduction. For instance, heat shock proteins (HSPs) induced by high temperatures are regulated by heat shock transcription factors (HSFs). HSFs are activated by phosphorylation of protein kinases. Compared with wild type, the activity of HSFs in the AtCaM3 mutant was decreased, which inhibited www.nature.com/scientificreports www.nature.com/scientificreports/ the biosynthesis of HSPs and reduced the heat resistance of plants 48 . The overexpression of the soybean GmCaM4 gene activates the pathogenesis-related (PR) gene and accumulates jasmonic acid (JA), which increases soybean resistance to the oomycete Phytophthora sojae, Alternaria tenuissima and Phomopsis longicolla. However, the silencing of the GmCaM4 gene significantly inhibited the expression of the PR gene 49 . MYB2 (CaM binding transcription factor) contains a Ca 2+ -dependent CaM binding domain and regulates the expression of salt and dehydration response genes in Arabidopsis 50 . It has been confirmed that the interaction of GmCaM4 and MYB2 regulates the expression of salt-responsive genes and improves tolerance to high-salt environments 49 . AtCML42 mutant increases the expression of JA responsive gene, thus enhancing the plant defense against herbivorous www.nature.com/scientificreports www.nature.com/scientificreports/ insects. In addition, JA-induced Ca 2+ elevation and root growth inhibition are more pronounced in AtCML42 mutants. The above results indicate that AtCML42 is an important component connecting Ca 2+ and JA signals, and plays a negative regulatory role. AtCML42 is also involved in abiotic stress responses. AtCML42 mutant decreases resistance to ultraviolet radiation B (UV-B) and accumulates abscisic acid content under drought stress 51 . The difference is that the accumulation of JA in ATCML37 mutants is significantly reduced, which indicated ATCML37 plays an active regulatory role in Ca 2+ signaling pathway 52 . These data suggest that CaM and CML mediate multiple defense signaling pathways. Solanum pennellii possesses excellent resistance to stress, which is an important germplasm resource to cultivate high-quality tomato 53 . However, the structural characteristics of CaM and CML genes and their responses to various stresses have not been systematically studied in Solanum pennellii. In this study, we identified and systematically analyzed the two families.
We identified 6 CaM and 45 CML from the Solanum pennellii genome sequence. This is roughly consistent with the number of genes identified in other plant species previously reported, such as Arabidopsis 9 , Oryza sativa 13 , Gossypium raimondii 11 , and Nicotiana Benthamiana 8 ( Table 4). The results showed that there were differences in the number of genes in CaM and CML families ( Table 4). The differences in sizes of genes in these families may be due to their ploidy levels and their involvement in different vital cellular processes. In principle, adding or evolving more genes or genomes is the inevitable result of and the correct direction for plant evolution. This phenomenon may occur because ecological strategies of different plants to cope with different environments are related to the adaptation and expansion of gene families [54][55][56] . Over the course of evolution, variations in gene family size are mainly caused by natural variation in different species and their adaptation to complex growth environments 56,57 .
As a result of the abundant selective splicing of genes and the post-translational modification of proteins, the functional and chemical complexity of proteins is enhanced. The post-translational modification event myristoylation has extremely diverse biological functions associated with signal transduction, protein transport, protein localization, extracellular communication, and protein regulation and metabolism. The analysis showed that SpCaM was without myristoylation sites (Table 1). Palmitoylation and myristoylation are sometimes interrelated and interdependent, and the absence of myristoylation may lead to the disappearance of palmitylation 58 .
Phylogenetic trees were constructed to understand the evolution of SpCaM and SpCML (Fig. 1). The SpCaM and SpCML were classified into five groups (I, II, III, IV, and V). The results showed that CML groups (I, II, III, IV) dominated the phylogenetic tree. SpCaM and SpCML evolved together from their common ancestors, and these SpCML evolved before SpCaM. This is why there are more SpCML genes than SpCaM genes in the genome, and SpCML genes were diversified more. Location analysis of these 51 genes on the chromosome revealed that these genes were not evenly distributed on the chromosome (Fig. 2). Chromosomes 3 had the most genes (8 genes), followed by chromosomes 1, 4, and 11 (7 genes each) and chromosome 2 (6 genes). Chromosomes 1, 3, 10, 11, and 12 contained both CaM and CML genes ( Table 5).
The conserved motif and gene structure analyses of SpCaM and SpCML genes showed that each group shared similar exon-intron structures and motifs, which provided further evidence for their classification (Figs. 3 and 4). Gene structure analysis showed that most SpCML genes lacked introns, while SpCaM contained only one long intron (Fig. 3); these findings were in accordance with findings on the exon-intron structure of CaM and CML genes in Arabidopsis 9 , Nicotiana benthamiana 8 , Brassica rapa L. 16 , and Solanum tuberosum 8 . However, some SpCML genes contained 1, 3, or 4 introns. At present, studies on the evolution of introns have found that intron loss is more likely to occur than intron gain during evolution 59 . Based on these insights, it can be hypothesized that the majority of SpCML without introns are older than SpCaM. The few SpCML genes with introns possibly evolved from their closest SpCaM. This explains why group IV SpCML and group V SpCaM are the closest in the evolutionary tree (Fig. 1). The conserved motif is also a key index to evaluate protein function 60 . The exon-intron distribution analysis reflected the conservatism and functional differences among different proteins. Conserved motif analysis suggested all SpCaM proteins contain 4 EF-hand type calcium-binding domains, and all SpCML contain at least 1 EF-hand type calcium-binding domain (Fig. 4).
qRT-PCR analysis of SpCaM and SpCML indicated that the expression levels of SpCaM and SpCML genes were affected in Solanum pennellii under abiotic stress and hormone treatments. The expression profiles of SpCaM and SpCML genes in different tissues showed different expression levels of SpCaM and SpCML genes (Fig. 5). The  SpCaM4, SpCML1, SpCML2, SpCML3, SpCML9, SpCML18, and SpCML31 showed significantly higher expression level (Log2 > 1) in roots than in other tissues, while expression level of SpCML17 in stems was significantly higher (Log2 > 1) than in other tissues, suggesting that different SpCaM and SpCML gene members have distinct expression levels in various tissues. The diversified expression of these SpCaM and SpCML genes revealed that they might play a significant role in different plant tissues 61 .  Table 4. The number of CaM and CML in different species.
The expression levels of SpCaM and SpCML genes under abiotic stress and hormone treatments in different tissues indicated that the expression of SpCaM and SpCML genes were affected (Figs. 6 and 7). Under cold stress, the expression of SpCaM4 in leaves was significantly increased (Log2 > 1), while down-regulation of SpCaM4 expression was found under ABA treatment, revealing that SpCaM4 may be involved in Ca 2+ transport under cold stress. The results are not entirely consistent with previous studies. Delk et al. found Arabidopsis CML24 was expressed in all major organs and upregulated under cold stress and ABA treatment 62 . It is has been reported that AtCML9 was induced under salt stress and ABA treatment, and involved in salt stress tolerance by affecting ABA-mediated pathways 63 . In Solanum pennellii, the expression levels of 11 genes were obviously upregulated (Log2 > 1) under salt stress, ABA and GA treatments, including 3 (SpCML28, SpCML35 and SpCML37), 4 (SpCaM6, SpCML10, SpCML24 and SpCML44) and 4 (SpCML3, SpCML23, SpCML40 and SpCML44) genes in leaves, stems and roots. These ten genes might participate in salt stress via ABA and GA-mediated pathway.
Under drought, salt, and ABA treatments, 4 (SpCaM6, SpCML10, SpCML24 and SpCML44) and 3 (SpCML3, SpCML23 and SpCML40) genes in stems and roots showed strong expression (Log2 > 1) (Figs. 6 and 7). The results are consistent with the study by Xu et al. who reported OsMSR2 (Oryza sativa l. multi-stress response gene 2), a novel CML gene, was strongly upregulated under drought and salt stress in different tissues at different stages of development, and enhanced tolerance to salt and drought via ABA-mediated pathway in rice 61 .
Conversely, Arabidopsis AtCML37, AtCML38, and AtCML39 showed greater sensitivity to drought and salt than to ABA and SA, suggesting that these proteins may act as Ca 2+ transducers in signaling pathways independent of ABA and SA 64 . OsCML4 confers drought tolerance through ROS-scavenging in an ABA independent manner in rice 65 . This phenomenon also exists in this study. The expression of the SpCML20 gene in stems was significantly induced (Log2 > 1) under drought and salt than under ABA and SA (Figs. 6 and 7). These results suggested that SpCaM and SpCML genes have diverse functions in different tissues in response to different stimuli, and may play a role as stress response genes to improve stress tolerance.
In this study, a total of 6 CaM and 45 CML genes were identified in the Solanum pennellii genome. These 51 genes were unevenly located on 12 chromosomes. SpCaM and SpCML were classified into five groups via phylogenetic analysis. Further analysis of their conserved motifs and gene structure revealed their evolutionary relationship, wherein it was suggested that SpCML evolved earlier than SpCaM. Analysis of cis-acting elements of these genes implied that they play crucial roles in response to multiple signaling pathways related to stress resistance. This study provides important insights into the evolution and function of Solanum pennellii genes, which lays a good foundation for the genetic improvement of stress-resistant tomato cultivars.

Materials and Methods
Identification of SpCaM and SpCML. All CaM and CML protein sequences of Arabidopsis and rice were obtained from the TAIR database (http://www.arabidopsis.org/) and rice Database (http://rice.plantbiology.msu. edu/), respectively. The whole protein and nucleotide sequences of Solanum pennellii 33 were obtained from NCBI (https://www.ncbi.nlm.nih.gov/genome/).

Phylogenetic analysis and chromosomal localization. CaM and CML protein sequences of
Arabidopsis (6 and 50, respectively) 12 and rice (5 and 32, respectively) 13 and SpCaM and SpCML protein sequences of Solanum pennellii were aligned by the MUSCLE program of MEGA6.0 40 , with default settings. Then, phylogenetic trees were constructed using the neighbor-joining method of MEGA6.0, in which bootstrap value was set to 1000. The chromosomal location information of 51 genes of Solanum pennellii was obtained from the NCBI database. The online MapGene2Chrom program was used to map their chromosomal locations (http:// mg2c.iask.in/mg2c_v2.0/) 41 . Sequence analysis. Physicochemical parameters of SpCaM and SpCML proteins, including theoretical isoelectric point (pi), molecular weight, amino acid sequence length (AA), and the N-terminal myristoylation were predicted using the ExPASy proteomics server (http://web.expasy.org/myristoylator/), with default settings 38 . The ScanProsite tool of ExPASy was used to retrieve the EF-hand domain, and calcium-binding region. The subcellular localization of proteins was predicted using the Wolf PSORT (http://www.genscript.com/psort/ wolf_psort.html) program 39 . The structure of these genes was analyzed using tools available with online website GSDS (http://gsds.cbi.pku.edu.cn/) 43,44 . Genomic DNA sequences of SpCaM and SpCML were downloaded from the NCBI database.

Cis
Abiotic stress and hormone treatments. Thirty-day-old seedlings were used to explore the responses of the plant to abiotic stress and hormone treatments. For cold stress, salt stress and drought stress, the seedlings were respectively placed in 1/2 full nutrient solution at 4 °C, with 100 mmol/L NaCl, and with 10% polyethylene glycol (PEG) 6000. Drought stress was simulated by decreasing osmotic potential. For the hormone treatments, the seedlings were respectively grown in 1/2 full nutrient solution with 150 µmol /L gibberellic acid (GA), with 100 mmol /L abscisic acid (ABA), and with 100 µmol/L salicylic acid (SA). The seedlings were collected at 1 h after treatments. All the treatments collected three biological samples, which were immediately frozen in liquid nitrogen and stored at −80 °C for further analysis.
RNA extraction and qRT-PCR assays. TRIzol reagent (Tianmo biotech, Beijing, China) was used to extract total RNA from the roots, stems, and leaves according to the manufacturer's instructions. Then, DNase I treatment was used to removing genomic DNA contamination from total RNA. Two micrograms of total RNA were used for the first-strand cDNA synthesis using the 5X All-In-One RT MasterMix (with AccuRT Genomic DNA Removal Kit) (Applied Biological Materials, Zhenjiang, China). For qRT-PCR analysis, the reactions were performed using the Bestar ® Sybr Green qPCR Master Mix (DBI, Shanghai, China) in an ABI7500 qRT-PCR system according to the manufacturer's instructions The primers used for qRT-PCR analysis are listed in Table 6. For all analyses, actin was used as an internal control. Three technological replicates of each sample were assayed. The relative quantification of specific mRNA levels was calculated from the cycle threshold (Ct) using the 2 −ΔΔCt method 67 .  Table 6. Primers used in qRT-PCR analysis.