Introduction

Currently, ecological deterioration, soil erosion and desertification increase have seriously restricted the development of local economy in karst area of southwestern China. The use of genetic engineering to repair desertification is an important method in vegetation recovery1. As one of typical species in southwestern China, the study on the adaptation mechanisms of Lonicera confusa to karst environment, especially to karst calcium-rich environment, can provide important theoretical basis for repairing desertification environment by plant genetic engineering.

Calcium ions (Ca2+), one of the most abundant metal elements, is available in most soils2,3. It is involved in the growth, development and the adaptation to the environment in plants. It is also one of the most important second messenger upon environmental stimulation4. Through the cytoplasmic concentration of Ca2+ cyclical change, plants response to outside stimulation and produce calcium signals. The calcium target protein, such as calmodulin (CaM), calcium dependence protein kinase (CDPKs) and calcineurin B (CBL), pass down the signal to regulate plant growth, development, photosynthesis and stress resistance5,6,7. On the other side, high concentration of Ca2+ in the cell can cause toxicity8,9. Excessive Ca2+ restricted the growth of many plants in calcareous soils10. For instance, high Ca2+ interfered with various crucial cell processes, including Ca2+ dependent signaling, phosphate-based energy metabolism and microskeletal dynamics6,11,12. Previous studies indicated that plant cells have evolved to possess fine mechanism to adjust free Ca2+ concentration in the cytoplasm in response to environmental changes, which mainly through the body Ca2+ transport system, including Ca2+-ATPase and Ca2+ channels8,13,14,15,16,17,18. For example, some plant cells can actively transport Ca2+ from the cytosol into the vacuole, endoplasmic reticulum, mitochondria, plastids and cell walls8,13,15,19,20,21, and other plants could accumulate crystalline Ca oxalate and deposit it in the storage parenchyma, bundle sheath cell, epidermal trichomes or chlorenchyma12,22,23,24.

L. confusa belongs to the family Caprifoliaceae. As one medicinal plant with ecological value25,26,27,28,29,30,31,32,33, L. confusa could adapted to karst calcium-rich area of southwest China34,35. Our previous results showed that L. confusa could excrete excess Ca salts via stomata and store the excess Ca2+ in glands and trichomes34. However, the underlying molecular mechanisms of L. confusa in respond to the excess Ca2+ remain to be solved. The research of the adaptability of L. confusa in response to the karst calcium-rich environment will benefit to the exploration of theoretical knowledge and ecological restoration.

In this study, the ultrastructure of mature leaves treated with high or low Ca2+ was observed by transmission electron microscope (TEM). Moreover, differentially expressed genes (DEGs) are identified under different Ca2+concentrations by employing RNA-seq technology and GeneFishing PCR. The results showed that multiple DEGs were related to Ca2+ transport and accumulation in L. confusa. Furthermore, protein-protein interaction network among DEGs identified some important regulatory genes and transcription factors. In summary, the present results provided the groundwork for revealing the molecular mechanism for the adaptation to calcium-rich environment in L. confusa.

Results

Calcium in soil could be transported to trichomes, glands, and stoma in the leaves of L. confusa with higher Ca2+ treatment

The mature leaves of L. confusa planted in Nongla Karst Experimental Site (108°19′E, 23°29′N) was performed for microwave digestion and flame atomic absorption spectrometry. The results revealed that the average calcium content was about 5.93 mg · g−1, which was much higher than that planted in sandstone soil36. In order to identify the distribution law of calcium in the leaves of L. confusa, the 0, 25, 50, 75, 100 and 125 mg/L gradient concentration of calcium chloride were used for pouring the plant materials, respectively. The leaves with different concentration of external Ca2+ treatment were labeled with Fluo-3/AM ester and examined under LSCM.

The previous studies indicated that Fluo-3 fluorescence intensity is proportional to the changes of Ca2+ concentration37. The LSCM results in this study revealed that the fluorescent intensity improved with increased external Ca2+ treatment (Fig. 1A–J), which indicated that the calcium intake in L. confusa and calcium concentration in the soil solution are directly related. Therefore, the leaves in high calcium environment may absorb much more calcium than in normal environment. L. confusa, as a calcium resistance plants in karst areas, should possess some mechanisms to avoid excessive Ca2+ in the cytoplasm to affect the normal signal transmission and the formation of the cytoskeleton dynamics process. Further observation showed that the fluorescent intensity was mainly distributed in trichomes (Fig. 1, red arrows), glands (Fig. 1, white arrows) and stoma (Fig. 1, pink arrows). This result indicated that the excess of Ca2+ absorbed from soil could transport to trichomes, glands, and stoma, which is consistent with our previous results in mature leaves that planted in Nongla Karst Experimental Site34. Moreover, the fluorescent intensity further increased with the Ca2+ treatment time prolonged (Fig. 1G,K,M,O,Q).

Figure 1
figure 1

Confocal image and the relative density of the fluorescence of L. confusa leaves loaded with Fluo-3/AM. (A,C,E,G and I) represent the confocal images of L. confusa abaxial surface of leaves after 25, 50, 75, 100 and 125 mg/L Ca2+ treatment for 36 h. (B,D,F,H and J) represent the perspective images of (A,C,E,G and J), respectively; (K and L) respectively represent the confocal and perspective images of glands after 100 mg/L Ca2+ treatment for 36 h; (M and N) represent the confocal and perspective images of abaxial surface of leaves after 100 mg/L Ca2+ treatment for 72 h; (O and Q) represent the confocal images of glands in leaves after 100 mg/L Ca2+ treatment for 72 h; (P and R) represent the perspective images of glands in leaves after 100 mg/L Ca2+ treatment for 72 h; (S and T) represent the confocal and perspective images of stomata in the leaves after 100 mg/L Ca2+ treatment for 72 h. Bar = 100 μm.

Full size image

The excess of Ca2+ could transport to trichomes, glands, and stoma, which was also revealed by using the energy-dispersive X-ray spectrometer for chemical component analysis of the leaves of L. confusa. The results showed that the Wt% of calcium in epidermal cell, trichomes, glands, stoma and especially in their surrounding cells of L. confusa cultivated in calcareous soil was much higher than that cultivated in sandstone soil (Fig. 2A–D,G). Compared with L. confusa in sandstone soil, there were 2.81, 2.68, 2.30 and 6.44-fold increasement of Wt% in epidermal cell, trichomes, glands, stoma and their surrounding cells of L. confusa in calcareous soil. The content of calcium in the different leaf structure was also measured by using linear scanning method and we found that the ROT count of calcium in trichomes part was much higher than that of other parts (Fig. 2F,H). These results again indicated the excess of Ca2+ absorbed from calcium rich environment could transport to trichomes, glands, and stoma.

Figure 2
figure 2

The calcium content analysis by using EDX. (A and B) represent the SEM (Scanning Electron Microscopy) map of leaves cultivated in calcareous and sandstone soil (white and red square frames represent the map scanning area of epidermal cells, stomata and its surround cells); (C and D) represent the EDX map of calcium; (E and F) represent the linear scanning area and the different element change in different scanning point (the different color curve represent different element); (G) The Wt % of calcium in different structures cultivated in calcareous and sandstone soil. (H) The amplified calcium change curve in different scanning point from (D).

Full size image

Many calcium channels and pools are found in the leaves of L. confusa in high calcium environment

As mentioned above, the calcium content in leaves of L. confusa planted in calcium rich soil was higher than that in calcium poor soil. In order to make clear how the calcium distributed in cells of L. confusa, potassium antimonate was used to localize the calcium in cells by using TEM technique. The smaller and fewer calcium antimonate precipitate were observed in the cells of in L. confusa treated with lower concentration of Ca2+ (25 mg/L, Fig. 3A). In contrast, the relatively larger and orderly arranged stored calcium were observed along the cell wall in the materials treated with higher level of Ca2+ (125 mg/L, Fig. 3B). This result indicated that formation of the common crystalline formation (such as calcium oxalate) arranged across the cell wall was another important mechanism for L. confusa to adapt the higher Ca2+ environment except for three other mechanisms we reported before34.

Figure 3
figure 3

TEM analysis of L. confusa leaves with high and low Ca2+ treatment. (A and B) represent TEM analysis of antimonite-labeled cells located in white square frames in Fig. 2b (low Ca2+ treatment, 25 mg/L) and Fig. 2a (high Ca2+ treatment, 125 mg/L), respectively (Arrow indicates the calcium antimonate precipitate), (C and D) represent the TEM analysis of antimonite-labeled cells near to glands, stoma in low and high Ca2+ treatment leaves, respectively. The antimonite-labeled cells were treated with EGTA and the calcium antimonate precipitates were chelated (Arrows indicate the calcium pools and arrow heads indicate the calcium channels).

Full size image

Calcium antimonate can be chelated by EGTA. Therefore, the antimonite-labeled cells near to glands, stoma in low and high Ca2+ treatment leaves were treated with EGTA. TEM results showed that there are obviously more Ca2+ pools in the leaves of L. confusa treated with higher Ca2+ for a certain time (Fig. 3D), and fewer Ca2+ pools in lower Ca2+ treated materials (Fig. 3C). Very interestingly, we also observed that more calcium ion channels in the L. confusa treated with higher Ca2+ than in lower Ca2+ treated materials (Fig. 3C,D, arrow heads), which suggested that L. confusa could store the Ca2+ in the calcium pools. Taken together, our results indicated more calcium channels and pools were distributed in the leaves of L. confusa in high-Ca2+ environment than in low-Ca2+ environment.

Identification of differentially expressed genes by RNA sequencing in higher and lower calcium-treated L. confusa

The L. confusa treated with 125 mg/L calcium chloride solution and pure water (as control) for 24 hours and 30 days were used for DEG analysis. Two biological replicates of RNA-seq of L. confusa leaf tissues treated with different level of Ca2+ were performed. The two sets of the corresponding sequencing samples were combined, which generated 57393938, 69001767 and 60782634 clean reads, respectively. The obtained clean reads were then mapped to the assembled transcriptome of L. confusa38. The results showed that 35584241, 44851148 and 38596972 clean reads can be mapped to the assembled unigenes38, respectively. Among the mapped reads, 24908968, 32292826 and 28171538 clean reads were uniquely mapped to the unigenes, respectively. 10675273, 12558322 and 10425434 clean reads were mapped to multiple locations of unigenes, respectively. After mapping, 94607, 90574 and 86396 expressed unigenes for 0 h, 24-hour and 30-day Ca2+-treatment were obtained, respectively. To identify the DEGs between control and calcium-treated L. Confusa, we set the expression level of unigenes of control as a control and investigated the up- or down-regulated unigenes of calcium-treated L. confusa. The analysis showed that 322,69 unigenes were differentially expressed between control and 24-hour calcium-treatment L. confusa, and 43,148 unigenes were differentially expressed between control and 30-day calcium-treatment L. confusa. Also, we observed that a total of 34,155 unigenes were differentially expressed between 24-hour and 30-day calcium-treatment L. confusa samples, which indicated that short-term (24-hour) and long-term (30-day) Ca2+ treatment to L. confusa may induce different physiological status of L. confusa. As shown in Supplementary Fig. 1, these DEGs between control and calcium-treated L. Confusa were further annotated with GO terms. Furthermore, the GO-enriched DEGs between control and calcium-treated L. Confusa were displayed by heatmap analysis (Fig. 4A). We next performed GeneFishing PCR to validate the DEGs identified in RNA sequencing analysis. The L. confusa treated with 125 mg/L calcium chloride solution and pure water (as control) for 30 days were used for GeneFishing analysis. In all, 24 DEGs were identified between control and high Ca2+-treated L. confusa (Supplementary Table 1). The DEGs were further confirmed by semi-quantitative RT-PCR experiments (Supplementary Fig. 2), which was very consistent with the RNA sequencing results.

Figure 4
figure 4

(A) Heatmap analysis of differentially expressed genes among control (no Ca2+ treatment), 24-hour Ca2+ treatment and 30-day Ca2+ treatment L. Confusa. (B) Protein-protein interaction network of differentially expressed genes in L. Confusa constructed by Cytoscape software. Red circles indicate the differentially expressed genes identified in Genefishing experiments. Green circles indicate the identified transcription factors.

Full size image

Furthermore, Cytoscape software39 was used to construct a complex protein-protein interaction network among DEGs in L. confusa. As shown in Fig. 4B, there were 253 nodes and 762 edges obtained in the network. Some important transcription factors, such as MYB59, WRKY19, WRKY51 and WRKY70 were involved in the network. Also, Tua3 (alpha-3 tubulin), one of the important cytoskeleton components, interacts with 9 targets including EB1C, EB1A, EB1B, ZWI, EMB2804, KIS and so on, in which all target proteins are associated with the formation and biological functions of microtubule. Therefore, the results indicate that Tua3 and target proteins may form a complex network to regulate cell morphology and intercellular transportation L. confusa.

Functional Enrichment of differentially expressed genes

For functional annotation of DEGs of short-term and long-term Ca2+ induction L. confusa, the DEGs were further analyzed using Cytoscape EnrichmentMap (http://www.cytoscape.org/). As shown in Fig. 5, EnrichmentMap analysis of short-term and long-term Ca2+ induction DEGs generated 300 and 404 nodes, respectively. These nodes were classified into different categories. The common terms between short-term (24 hours) (Fig. 5A) and long-term (30 days) (Fig. 5B) Ca2+ induction DEGs were “Response to stimulus”, “Developmental process”, “Biological regulation”, “Metabolic process”, “Cellular process” and “Transport”. Very importantly, some DEGs were enriched in different categories. For example, EnrichmentMap analysis showed that short-term Ca2+ induction DEGs were also clustered in “Signaling”. This result indicated that short-term Ca2+ induction in L. confusa might activate some important signal transduction pathways and/or defense responses. Nevertheless, long-term Ca2+ induction DEGs were mainly clustered in “Cellular cation homeostasis”, “Component organization” and “Localization”. Interestingly, the enrichment of “Cellular cation homeostasis” suggested that L. confusa gradually adapted to high- Ca2+ environment and kept a new cation balance after long-term Ca2+ treatment. Meanwhile, “Transport” process was enriched both in short-term and long-term Ca2+ induction L. Confusa, which suggested that Ca2+ treatment induced the activation of relative Ca2+ transport pathways in L. Confusa.

Figure 5
figure 5

Biological Process analysis of differentially expressed genes (DEGs) between short-term (24 hours) Ca2+ -treated and long-term Ca2+ treated L. confusa. GO modules enriched with short-term (24 hours) Ca2+ -treated DEGs (A) and long-term (30 days) Ca2+ treated DEGs (B) were visualized by the EnrichmentMap in Cytoscape. The red and yellow circles indicate the common and different biological processes between short-term (24 hours) Ca2+ -treated and long-term Ca2+ -treated L. confusa, respectively.

Full size image

KEGG pathway analysis

To reveal biological functions of the identified genes, the DEGs between control (no Ca2+ treatment) and 30-day Ca2+ treatment L. Confusa were further assigned for KEGG pathway analysis. As shown in Supplementary Dataset 1, a total of 99 KEGG pathways were ranked by p-value. The top 5 pathways included “Ribosome” (91 unigenes, P = 3.97 × e−7), “Phenylpropanoid biosynthesis” (61 unigenes, P = 1.74 × e−6), “Methane metabolism” (P = 2.39 × e−6), “Cyanoamino acid metabolism” (29 unigenes, P = 3.46 × e−5) and “Phenylalanine metabolism” (50 unigenes, P = 7.06 × e−5). Furthermore, we extracted the Ca2+ metabolism-related pathways for analysis. As shown in Supplementary Fig. 3, the Ca2+ metabolism-related pathways were classified into five categories including Cellular Processes, Environmental Information Processing, Genetic Information Processing, Metabolism and Organismal Systems. Among them, Cellular Processes category enriched 142 genes in which 32 genes are involved in the pathway of “Transport and catabolism”. More importantly, we observed that a lot of (265) genes are involved in “Signal transduction” pathway in Environmental Information Processing category. These results indicated long-term Ca2+ treatment on L. Confusa may induce Ca2+ transport and related signal transduction processes.

Discussion

L.confusa is a woody perennial, evergreen and twining vine. It is widely cultivated in eastern Asia as an important medicinal plant. L.confusa is also processed into food and a healthy beverage, which facilitates the rapid increase of its commercial value in herbal medicine markets. However, the economy development has been greatly impeded by ecological deterioration, soil erosion and desertification in karst area of southwestern China. L.confusa is one of typical species in southwestern China. Therefore, this study focused on the adaptation mechanism of L. confusa to karst calcium-rich environment, which will reveal important theoretical basis for repairing desertification environment.

Various studies have revealed that Ca2+ will accumulate to a very high level over time in cells when water evaporates from the surface of plants. Moreover, the excess Ca2+ can be precipitated as inactive Ca oxalate22. Within plants, most long-distance Ca2+ transports through plant tissues have been demonstrated to follow apoplastic pathways40,41. After the Ca2+ transfer to the plant cell, Ca2+ can be transported from the cytosol into the vacuole, endoplasmic reticulum, mitochondria, plastids and cell walls where Ca2+ can be captured by H+/Ca2+ antiporters (CAX) and Ca2+-ATPases (ACA)8,21. CAX1 has a primary role in Ca2+ accumulation in the leaf mesophyll and it is important in controlling apoplastic Ca2+, stomatal aperture, and growth41,42. In this study, we also observed high-Ca2+ treatment induced the transport of Ca2+ to trichomes, glands, and stoma in the leaves of L. Confusa. Moreover, more calcium channels and calcium pool are accumulated in the leaves of L. confusa in high calcium environment. Although the signaling role of Ca2+ in organisms has developed rapidly3,21, fundamental information regarding the mechanisms that regulate Ca2+ transport and storage in plants still remains elusive43.

RNA-seq technology is an effective transcriptome analysis tool, which can detect novel and rare RNA transcripts and accurately quantify gene expression level. In this study, we identified multiple DEGs (differentially expressed genes) between control and calcium-treated L. Confusa. Bioinformatics analysis showed that 322,69 unigenes were differentially expressed between control and 24-hour calcium-treatment L. confusa, and 43,148 unigenes were differentially expressed between control and 30-day calcium-treatment L. confusa. Importantly, the different gene expression level between 0-hour and 24-hour or 30-day Ca2+-treated L. confusa might be induced not only by the Ca2+ treatment but also by own natural growth of L. confusa. Therefore, we could not fully exclude the effect of 24-hour or 30-day natural growth of L. confusa on the gene expression level. Nevertheless, the samples used for RNA-seq were mature leaves of L. confuse. We supposed that the gene expression level in the mature leaves is very stable in 24 hours or 30 days, and the different gene expression level between 0-hour and 24-hour or 30-day Ca2+-treated L. confusa were mainly induced by Ca2+ treatment. Furthermore, the identified DEGs were validated by GeneFishing PCR. For example, Tua3, CaM, CDPKs, CBL and some other important genes were identified. Tua3 is one of the important cytoskeleton components. Our network analysis showed that Tua3 interacts with 9 target genes, in which all target genes are associated with the formation of microtubule. Taken together, these results indicate that Tua3 and target genes may regulate cell morphology and intercellular transportation by a complex gene interaction network in L. confusa. In the meanwhile, it has been reported that CaM, CDPKs and CBL are involved in cytoskeleton formation and movement, osmotic stress resistance44,45. Calcium, as the second messenger upon environmental stimulation in plant, plays critical roles in response to outside signals and activates the expression of related genes. For Ca2+ transport into the vacuole in plant cells, both Ca2+-ATPases (ACA) and Ca2+/H+ antiporters (CAX) play important roles46. Therefore, transcript abundance of transporters could be up-regulated under conditions of adequate Ca supply. Also, proteins that modify activity of transporters, such as CAM or CXIP, could also be regulated in a similar fashion47,48. For instance, compared with epidermal cells, mesophyll cells of Arabidopsis have the higher capacity to store Ca2+ in their vacuoles by virtue of the higher expression of the Ca2+/H+ antiporter on the tonoplast membrane41. Our DEGs analysis in this study showed that CaM and CAX were upregulated in long-term Ca2+ treated L. Confusa, which might reflect the adaptive changes of L. Confusa to high-Ca2+ environment. In addition, EnrichmentMap analysis showed that multiple DEGs were involved in “Cellular cation homeostais” and “Transport” processes. The results suggested that Ca2+ treatment induced the activation of relative Ca2+ transport pathways and L. confusa gradually adapted to high-Ca2+ environment and kept a new cation balance after long-term Ca2+ treatment, which is very consistent with our LSCM and TEM results. Collectively, identification of these important DEGs will help us to explore the possible regulatory mechanism of the adaptation to high-Ca2+ environment in L. confusa.

Materials and Methods

Plant materials

L. confusa cultivars were taken from typical Nongla Karst Experimental Site (108°19′E, 23°29′N) as previously described34. The L. confusa cultivars were planted in the soils transported from Nongla Karst Experimental Site. In order to remove the calcium ions in the soils, the soils were washed with pure water for several times. Furthermore, 24 basin of L. confusa materials were divided into six groups and treated with different concentrations of Ca2+, respectively.

Detection of Ca2+ on the leaves of L. confusa by LSCM

The Ca2+ images in the leaves of L. confusa were detected as previously described34,49. The L. confusa leaves growing in soils with different level Ca2+ were soaked in Fluo-3/AM ester solution at 4 °C for 2 h. These L. confusa leaves were then washed with PBS solutions and observed under Olympus FV1000 LSCM (OLYMPUS, JAPAN).

TEM analysis

TEM analysis of the leaves of L. confusa was performed as previously described34,50 except for small modification. First, 5 independent leaf samples of L. confusa with different level Ca2+ treatment were isolated, washed with phosphate buffer solution and cut into 0.2 cm × 0.2 cm slices. Furthermore, these samples were fixed and dehydrated through graded ethanol concentrations once for 10 min and soaked in 100% acetone twice for 30 min. Subsequently, the samples were treated with 2% potassium pyroantimonate and dehydrated. Finally, the samples were embedded with Spurr epoxy resin and sectioned with superfine section machine. The obtained sections were dyed and examined using TEM (Dutch FEI Tecnai F20-Twin, Netherlands).

RNA isolation and GeneFishing PCR

The total RNA isolation of the L. confusa leaves treated with different level of Ca2+ was performed as previously described51. GeneFishing kit (Seegene, Inc.) was used for differential display PCR analysis. GeneFishingTM PCR were done with dT-ACP2 and 20 pairs of random primers, the PCR products were run on 2% agarose gel and the cDNA bands ranged from 100 and 1.5 kb were used for cloning and sequencing (Supplementary Fig. 2). QIAquick Gel extraction kit (Qiagen) was used to isolate the differentially expressed bands. The isolated DNA fragments were cloned into pGEM®-T Easy vector (Promega) and sequenced. Further Semi-quantitative RT- PCR experiments were utilized for the validation of the GeneFishing results. The Primers used for GeneFishing PCR and Semi-quantitative RT- PCR were shown in Supplementary Tables 2 and 3, respectively.

RNA sequencing and DEG analysis

The RNA-seq cDNA libraries of L. confusa leaves treated with different level of Ca2+ were constructed using TruSeqTM RNA Sample Preparation Kit (Illumina, Inc.). Shortly, oligo (dT) magnetic beads (NEB) were used to capture poly-A mRNAs from the isolated total RNA. The captured mRNAs were fragmented into 200–500 bp for cDNA synthesis and adaptor ligations. The synthesized cDNAs were PCR-amplified, quantified and sequenced on Illumina HiSeq 2000 using 2 × 100 bp pair-end sequencing protocol. The generated clean reads were deposited in NCBI Sequence Read Archive (SRA) Sequence Database (Accession number: SRR6024635). Bioinformatics analysis of differentially expressed genes of RNA-seq samples was performed as described37. Briefly, gene expression level of each RNA-seq sample was measured by RPKM (Reads Per kb per Million reads). The genes with RPKM ≥1 were regarded to be expressed in the RNA-seq analysis. The edgeR software was then used to identify the differentially expressed genes. The differentially expressed genes among L. confusa leaves treated with different level of Ca2+ were selected using a log FC (log-fold expression change) >2 or <−2, a false discovery rate (FDR < 0.001) and p-value < 0.005 as the threshold value.

Pathway analysis and interaction analysis

The functions of the DEGs were analyzed by using Kyoto Encyclopedia of Genes and Genomes (KEGG) database (http://www.genome.jp/tools/blast). Cytoscape_2_6_3 (http://www.cytoscape.org/plugins/index.php) was used to analyze the protein-protein interaction among DEGs.

Accession number

RNA sequencing data have been deposited in NCBI Sequence Read Archive (SRA) Sequence Database with accession number SRR6024635.