Identification of sodium homeostasis genes in Camelus bactrianus by whole transcriptome sequencing.

Salt dietary intake is tightly coupled to human health, and excessive sodium can cause strokes and cardiovascular diseases. Research into the renal medulla of camels exhibiting high salt resistance may aid identification of the mechanisms governing resistance to high salinity. In this study, we used RNA sequencing (RNA-seq) to show that in the renal medulla of camels under salt stress, 22 mRNAs, 2 long noncoding RNAs (lncRNAs), and 31 microRNAs (miRNAs) exhibited differential expression compared with the free salt-intake diet group. Using fluorescence in situ hybridization and dual-luciferase reporter assays, we demonstrated that the lncRNA LNC003834 can bind miRNA-34a and thereby relieve suppression of the salt-absorption-inhibiting SLC14A1 mRNA from miRNA-34a, suggesting that the above lncRNA-miRNA-mRNA act as competing endogenous RNAs (ceRNAs). We subsequently performed short hairpin RNA and small RNA interference and reactive oxygen species (ROS) detection assays to show that SLC6A1, PCBP2, and PEX5L can improve the antioxidation capacity of renal medulla cells of camel by decreasing ROS levels. Our data suggest that camels achieve sodium homeostasis through regulating the expression of salt-reabsorption-related genes in the renal medulla, and this involves ceRNAs (SLC14A1 mRNA, LNC003834, and miRNA-34a) and antioxidant genes (SLC6A1, PCBP2, and PEX5L). These data may assist in the development of treatments for diseases induced by high salt diets.

competing endogenous RNAs

cardiovascular diseases

dihydroethidium

Dulbecco’s modified Eagle’s medium/nutrient mixture F‐12

fluorescence in situ hybridization

γ‐aminobutyric acid

long noncoding RNAs

microRNAs

mutant type

negative control

RNA sequencing

reactive oxygen species

small RNA

wild‐type

Salt is a mineral that is mainly composed of sodium chloride and normally exists in two natural forms including halite and sea salt. Sodium serves as an essential molecule in mammalian physiology for its roles as an electrolyte and osmotic solute [1, 2, 3]. The need to satisfy the sense of salty taste is innate and important [4], and the intake of an adult per day is less than 2000 mg of sodium (or 5 grams of salt) recommended by the World Health Organization [5]. But in many Western countries, the typical intake of salt is about 10 grams per day, which is higher than that in many countries in Asia and Eastern Europe [6]. Compared with Westerners, Asians are sensitive to salt and generally have higher blood pressure readings in the morning and at night [7]. According to guidelines by the United States, people with hypertension, middle‐aged, and older adults should control their intake to no more than 1500 mg of sodium per day [8]. High sodium in diets has been proved to be one of the most outstanding dietary risks [9]. Excessive dietary salt consumption will raise the risk of kidney damage, stroke, autoimmune diseases, including rheumatoid arthritis disease [10, 11], and cardiovascular diseases (CVDs) like hypertension [2, 6, 12, 13]. In addition, hypertonic sodium ion accumulation in peripheral tissues is associated with aging, hypertension, diabetes, chronic kidney disease, and heart failure [14]. In the United States and other westernized nations, CVDs usually trigger the death of adults over the age of 65 years old [15], while cutting salts from diet can lower blood pressure and decreases the risk of death from CVDs in people with hypertension [16, 17] and in adults and children with no acute illness [10, 17, 18, 19]. For instance, if sodium intake reduces to 1000 mg per day, the incidence of CVDs will decrease by about 30 percent [1].

Animals that live in the desert, such as the Gobi Desert, frequently confront the lack of food and tolerate high salinity water for a long period of time [20, 21]. Due to evolution and natural selection, camels (Camelus bactrianus) developed resistance to heat, drought, and salt for long‐term exposure in the desert [22]. The amount of daily salt needed to meet the requirements is 2 to 5 g in pigs as well as 3.9 g in goats [23, 24], for example. Regarding high salt consumption, camels should have 1.5 or 2 oz. (roughly 42.5 or 56.7 g) of salt daily or 4.5 to 6 oz. (approximately 127.6 to 170.1 g) of salt every three days [25]. As a salt‐metabolism‐related essential organ, the kidney primarily participates in the reabsorption of nutrients, waste filtration, and electrolyte homeostasis in mammals [26]. In the kidney, the renal medulla generally maintains the salt and water balance of the blood and consists of the vasa rectae (both spuria and vera), the venulae rectae, the medullary capillary plexus, the loop of Henle, and the collecting tubule [27, 28].

RNAs transcribed from DNAs are a class of vital molecules that have diverse biological roles. mRNAs take charge of directing protein synthesis and noncoding RNAs, such as miRNAs and lncRNAs, are involved in post‐transcriptional regulation [29, 30, 31, 32]. The studies including mammalian hibernation, frog and insect freeze tolerance, and turtle and marine snail anoxia tolerance have reported that the post‐transcriptional regulation mediated by noncoding RNAs is linked to biotic resistance under stress, for instance, that miRNA can rapidly halt the translation of unnecessary proteins [33]. Implicated in preventing sodium from entering the cell, the SLC6A19 mRNA aimed by miR‐193b and miR‐542‐5p as well as the putative competing endogenous LNC001438, LNC003417, LNC001770, miR‐199c, and TENM1 mRNA were considered as salt‐resisting modulators [34, 35].

Hence, taking transcriptome sequencing as a basis, our research targeting differential gene expression in the renal medulla of camels under salt stress will help scientists comprehending how the mammals achieve sodium homeostasis through natural instinct. The result of this study may offer an alternative investigation direction for solving the health problems due to excessive salt intake.

Materials and methods

Camels for salt stress treatment

The six female camels from 7.8 ± 0.4 years old (mean ± SD age) living in the Alxa Gobi Desert of northern China were penned and randomly divided into two groups (three camels per group): The experimental group was under salt stress condition, and the control group was under free salt‐intake diet, both groups were assessed for the consecutive periods of 24 days. The mean weight of the six camels was 352.6 ± 7.1 kg (Table S1). In the salt stress group, the extra salt intake base was 200 g·day−1 and increased by 100 g every 3 days, and the daily intake of salt was 900 g at the end of the experiment, following free water and feeding. Camels ate and drank freely in control group, with daily 50 g salt consumption. Procedures involving animals, their care, and the humane kill were conducted in conformity with the Guidelines on the Humane Treatment of Laboratory Animals (HTLA Pub. Chapter 2‐6, revised 2006 in China) and were approved by the Animal Care and Use Committee of the Inner Mongolia Agricultural University (IMAU‐IACUC‐2019‐31560313). The written informed consent of the experimental research was achieved by S. Wang (as an animal owner).

Tissue sample collection

After injecting 0.5 mg·kg−1 xylazine (Table S1), the six Alxa bactrian camels of the two groups were sacrificed by bleeding of the carotid artery. Renal medulla tissues for the study were sampled and stored in 1.5 mL frozen tubes and then immediately deposited in liquid nitrogen (−196 °C) for preservation and continuous experiments.

RNA library construction and bioinformatics analysis

Total RNAs were extracted from renal medulla tissues of camels using the Qiagen RNeasy Mini Kit (QIAGEN, Germany). Potential RNA degradation and contamination were checked on 1% agarose gels. The purity, concentration, and integrity of total RNAs were determined by the NanoPhotometer spectrophotometer (IMPLEN, USA), the Qubit RNA Assay Kit in Qubit 2.0 Fluorometer (Life Technologies, USA), and the RNA Nano 6000 Assay Kit of the Agilent Bioanalyzer 2100 system (Agilent Technologies, USA), respectively. Three micrograms of total RNA per sample were utilized for small RNA and lncRNA library construction. The sequencing libraries of small RNA and lncRNA were individually created using the NEBNext Multiplex Small RNA Library Prep Set for Illumina (NEB, USA) and the NEBNext Ultra Directional RNA Library Prep Kit for Illumina (NEB, USA). These steps were preceded by the removal and cleaning up of ribosomal RNA using the Epicentre Ribo‐zero rRNA Removal Kit (Epicentre, USA) and ethanol precipitation. Pooled RNA‐seq was applied to the constructed libraries of small RNA and lncRNA using an Illumina HiSeq 2500 platform and 4000 platform in turn [36]. The miRNA data were analyzed by quality control, reads mapping to Camelus bactrianus genome (NCBI reference genome: Ca_bactrianus_MBC_1.0; RefSeq: GCF_000767855.1) using Bowtie (parameters: ‐v 0 ‐k 1) [37], and novel miRNA prediction with miREvo (parameters: ‐i ‐r ‐M ‐m ‐k ‐p 10 ‐g 50000) [38] and mirdeep2 (parameters: quantifier.pl ‐p ‐m ‐r ‐y ‐g 0 ‐T 10) [39]. The differential expression of miRNA was analyzed via TPM [40] and DEGseq with qvalue < 0.01 and |log2(foldchange)| > 1 [41]. Small RNA annotation was investigated using the Bos taurus database. The prediction of target genes was performed via miRanda (parameters: ‐sc 140 ‐en −10 ‐scale 4 ‐strict ‐out) and RegRNA2.0 (parameters: score ≥ 170; free energy ≤ −15). The lncRNA data were analyzed by quality control, reads mapping to Camelus bactrianus genome (NCBI reference genome: Ca_bactrianus_MBC_1.0; RefSeq: GCF_000767855.1) using TopHat v2.0.9 (parameters: ‐‐library‐type fr‐firststrand) [42] and Cufflinks (parameters: min‐frags‐per‐transfrag = 0) transcript assembly [43]. lncRNA and mRNA were sorted by CNCI v2 with default parameters [44], CPC 0.9‐r2 (parameters: ‐evalue 1e‐10) [45], Pfam Scan v1.3 (parameters: ‐E 0.001 ‐‐domE 0.001 ‐pfamB) [46] and PhyloCSF v20121028 (parameters: ‐‐orf = ATGStop; ‐frames = 3; ‐removeRefGaps) [47]. Annotations of mRNA and lncRNA were accomplished using Camelus bactrianus and Bos taurus databases, and the differential expression analysis of lncRNA and mRNA was exerted by Cuffdiff (http://cole‐trapnell‐lab.github.io/cufflinks/cuffdiff/index.html). The prediction of target gene was carried out by lncRNA gene upstream/downstream 100 kb and Pearson correlation coefficient with |Pearson correlation| > 0.95 [48]. Gene Ontology and Kyoto Encyclopedia of Genes and Genomes pathway enrichment analyses of target genes and differentially expressed mRNAs were conducted by GOseq [49] and KOBAS 2.0 (parameters: blastx 1e‐10 and padjust: BH) [50].

RNA fluorescence in situ hybridization (FISH) of camel renal medulla tissues

FISH was performed on renal medulla tissues from salt stress and control camels in accordance with the protocol of Fluorescence in situ Hybridization Kit (GEFAN, China) through fluorescent groups (FAM and Cy3) labeled RNA probes and DAPI stain. The sequences of fluorescent group labeled RNA probes are listed in Table S2.

Cell isolation and cultivation

The sampled renal medullas of deceased camel were removed and cleansed by Dulbecco’s phosphate‐buffered saline (Solarbio, China) with 1% penicillin and 1% streptomycin (NCPC, China). Following chopping, the small pieces of renal medulla tissue were seeded on the T25 culture bottle (Corning, Germany) and cultivated with Dulbecco’s modified Eagle’s medium/nutrient mixture F‐12 (DMEM/F‐12) (Hyclone, USA) containing 10% fetal bovine serum (Sigma, USA), 0.1% penicillin, and 0.1% streptomycin (NCPC, China) in a humidified chamber with 5% CO2 at 37 °C. The cells migrated out around tissues were initiated into digestion via using trypsin‐EDTA solution (Solarbio, China) at a cell growth confluence of 80%. The renal medulla cells of camels those in good growth condition were selected for the subsequent experiments after passage. Besides renal medulla cells, the 293T cells were also cultured in the above‐modified DMEM/F‐12 with 5% CO2 at 37 °C.

Dual‐luciferase reporter assay

The formula of high transfection efficiency was explored in 293T cells, using NC‐FAM oligo and pcDNA3.1‐GFP plasmid (GenePharma, China). Different co‐transfected combinations, miRNA‐34a oligo (GenePharma), NC oligo (GenePharma), LNC003834 plasmid vector including wild‐type (WT) and mutant type (MUT) (LNC003834S1 WT, LNC003834S2 WT, LNC003834S1 MUT, and LNC003834S2 MUT, GenePharma), SLC14A1 luciferase reporter plasmid WT and MUT (SLC14A1 WT and SLC14A1 MUT, GenePharma) were transfected into the 293T cells by lipofectamine 2000 with a 70% transfection efficiency in line with the proportion of fluorescent cells. The cells were lysed after 48 h and the fluorescent intensity of luciferase was detected using Dual‐Luciferase Reporter Assay System (Promega, USA). The sequences of the corresponding RNAs are listed in Table S3 and Table S4. The functional sequences of LNC003834S1 and LNC003834S2 were as follows:

a. LNC003834S1 WT: 5’‐AGGAACATAGGGACACTGCCC‐3’,

b. LNC003834S2 WT: 5’‐TTGTCATTCTTTTGCACTGCCC‐3’,

c. LNC003834S1 MUT: 5’‐TCCTTGTATCCCTGTGACGGG‐3’,

d. LNC003834S2 MUT: 5’‐AACAGTAAGAAAACGTGACGGG‐3’.

Short hairpin RNA and small RNA interference

For shRNA‐ and sRNA‐mediated SLC6A1, PEX5L, and PCBP2 suppression, the packaging shRNA lentivectors and small RNA oligos with attached GP‐siRNA‐Mate Plus reagent were purchased from GenePharma (China). 48 h after the transfection into camel renal medulla cells, the fluorescent intensity was detected according to the manufacturer’s protocol. The synthesized shRNA and sRNA sequences are listed in Table S5 and Table S6.

RNA isolation and qRT‐PCR

RNAs were extracted from renal medulla tissues and cells of camel using mirVana miRNA Isolation Kit (Ambion, USA). The primers were designed by Primer Express 3.0.1 (Table S7) and qRT‐PCRs were exerted on the 7900 HT Sequence Detection System (ABI, USA).

Western blot

The proteins were extracted by M‐PER Mammalian Protein Extraction Reagent (Thermo, USA). SLC6A1, PCBP2, and PEX5L protein expression were severally examined using antibodies against‐SLC6A1 (ab426 and ab72448, Abcam, UK) and PCBP2 (ab96169, Abcam, UK) and PEX5L (nbp1‐91484, Novus Biologicals, USA). The signals were monitored via Pierce ECL Western Blotting Substrate (Sangon Biotech, China).

Reactive oxygen species detection

ROS staining was arranged after RNA interference in renal medulla cells by ROS fluorescence probe‐DHE (KeyGEN, China). The red fluorescence levels were analyzed by BD FACSDiva 8.0.1.1 on BD FACS flow cytometry (BD Biosciences, USA).

Statistical analysis

Aside from RNA‐seq data statistics based on negative binomial distribution, Student’s t‐test was used to compare the mean values between groups and a P‐value of less than 0.05 was considered as statistically significant.

Results

Differentially expressed genes in renal medulla under salt load

RNA‐seq analysis identified 22 mRNAs (15 upregulated and 7 downregulated), 2 lncRNAs (1 upregulated and 1 downregulated), and 31 miRNAs (18 upregulated and 13 downregulated) that were differentially (P < 0.05) expressed in camel renal medulla by comparing high‐salt group and control group (Fig. 1A and B; Table S8 and Table S9). We noticed that SLC14A1, a molecule involved in salt absorption [51, 52], displayed significantly higher expression intensity grade in salt stress (Fig. 1C). Other upregulated molecules, such as SLC6A1, PEX5L, and PCBP2 (Fig. 1C), are speculated as prospective mediators against reactive oxygen species (ROS) [53, 54, 55, 56, 57]. Moreover, SLC14A1 mRNA and a novel lncRNA LNC003834 (XLOC_182366, NW_011544901.1) were identified as underlying targets of downregulated miRNA‐34a, suggesting that SLC14A1 mRNA, LNC003834, and miRNA‐34a could be competing endogenous RNAs (ceRNAs). The ceRNAs function has been clarified in regulating RNA transcripts via competing for shared miRNAs at the post‐transcriptional level [58, 59]. Therefore, the above 4 protein‐coding genes (SLC14A1, SLC6A1, PEX5L, and PCBP2) and the 2 noncoding genes (LNC003834 and miRNA‐34a) were screened out as candidate genes of salt resistance in camel renal medulla. When considering qRT‐PCR results, their different levels were consistent with RNA‐seq gene expression analyses (Fig. 1D).

Fig. 1

Differentially expressed genes in renal medulla of camels subjected to salt stress or free diet. (A) Differential log10 (FPKM+1) values of mRNA and lncRNA and log10 (TPM+1) values of miRNA. (B) Number and (C) log2 (fold change) value of significantly differential mRNA, lncRNA, and miRNA genes. (D) Average 2‐ΔCt value of genes assessed by qRT‐PCR. Data are presented as mean ± SD (n = 3 camels) by Student’s t‐test. Symbols: *P < 0.05. β‐actin and U6 were used as loading control for mRNA and lncRNA, and miRNA, respectively.

Colocated LNC003834 and miRNA‐34a in extranuclear region

Using FISH techniques, the cellular localizations of LNC003834 and miRNA‐34a were tracked in camel renal medulla tissues. FAM fluorescent group‐labeled lncRNA and Cy3 fluorescent group‐conjugated miRNA probes were designed and synthesized according to the sequences of LNC003834 (Table S10) and miRNA‐34a (miRBase, Accession: MIMAT0004340, ID: bta‐miR‐34a). By fluorescence microscopy, LNC003834 (green) and miRNA‐34a (red) presented nonoverlapping fluorescence against DAPI‐stained nuclei (blue) (Fig. 2A). The results signify that the localization of potential interaction between LNC003834 and miRNA‐34a is extranuclear.

Fig. 2

Competing endogenous SLC14A1 mRNA, miRNA‐34a, and LNC003834. (A) Immunofluorescence microscopy image of camel renal medulla under salt stress and free diet (20× and 40×). LNC003834, miRNA‐34a, and nucleus were independently labeled by FAM (green), Cy3 (red), and DAPI (blue). Scale bar, 500 μm. (B) The inserted sequence in SLC14A1 luciferase reporter plasmid and LNC003834 (LNC003834S1 and LNC003834S2) plasmids including two potential binding regions. (C) The experimental conditions with 70% transfection efficiency of NC‐FAM oligo and pcDNA3.1‐GFP plasmid (green) in 293T cells in vitro (200×). Scale bar, 100 μm. (D–F) Mean Firefly/Renilla ratio of miRNA‐34a mimic, SLC14A1 WT and SLC14A1 MUT, LNC003834S1 WT and LNC003834S1 MUT, and LNC003834S2 WT, and LNC003834S2 MUT. WT and MUT separately refer to wild‐type and mutant type. Data are presented as mean ± SD (n = 3) by Student’s t‐test. Symbols: *P < 0.05; **P < 0.01.

Competing endogenous LNC003834, miRNA‐34a, and SLC14A1

In the dual‐luciferase reporter assay, SLC14A1 wild‐type and mutant type luciferase reporter plasmids (SLC14A1 WT and SLC14A1 MUT), which contain original or base‐altered for miRNA‐34a potential binding sequence (SLC14A1 WT: 5’‐TGCCTGCGCTCTGTTCACTGCCT‐3’, SLC14A1 MUT: 5’‐ACGGACGCGAGACAAGTGACGGA‐3’) (Fig. 2B), were transfected into 293T cells by liposome. The synthetic miRNA‐34a (miRBase, Accession: MIMAT0004340, ID: bta‐miR‐34a) oligo, LNC003834 WT and MUT plasmids (LNC003834S1 WT, LNC003834S2 WT, LNC003834S1 MUT, and LNC003834S2 MUT), including the material or base‐changed the miRNA‐34a potential binding sequences (Fig. 2B), were also transfected into 293T cells. A 70% transfection efficiency of NC‐FAM oligo and pcDNA3.1‐GFP plasmid was obtained with 40 pmol oligo with 1 L lipofectamine 2000 and 0.4 g plasmid with 2 L lipofectamine 2000 (Fig. 2C). The test results indicated that the average value of the Firefly/Renilla luciferase ratio of the ‘SLC14A1 WT + miRNA‐34a mimic’ group was significantly (P < 0.01) lower than the ‘SLC14A1 WT + negative control (NC) mimic’ group (Fig. 2D). However, the Firefly/Renilla ratio of the ‘SLC14A1 MUT + miRNA‐34a mimic’ group and the ‘SLC14A1 MUT + NC mimic’ group were not statistically significant (Fig. 2D). The results suggest that miRNA‐34a inhibits SLC14A1 expression by its binding to target sequence on SLC14A1 mRNA. The ‘SLC14A1 WT + miRNA‐34a mimic + LNC003834S1 WT’ group showed no significant difference with the Firefly/Renilla ratio when compared with the ‘SLC14A1 WT + NC mimic + LNC003834S1 WT’ group and markedly (P < 0.05) higher Firefly/Renilla ratio value than the ‘SLC14A1 WT + miRNA‐34a mimic + LNC003834S1 MUT’ group (Fig. 2E). In terms of the prospective LNC003834S2, the ‘SLC14A1 WT + miRNA‐34a mimic+ LNC003834S2 WT’ group exhibited significantly (P < 0.01) lower Firefly/Renilla ratio than the ‘SLC14A1 WT + NC mimic + LNC003834S2 WT’ group, while there was no apparent difference of Firefly/Renilla ratio between the ‘SLC14A1 WT + miRNA‐34a mimic + LNC003834S2 WT’ group and the ‘SLC14A1 WT + miRNA‐34a mimic + LNC003834S2 MUT’ group (Fig. 2F). The results demonstrate that LNC003834S1 reduces the inhibition of SLC14A1 expression by emulous binding miRNA‐34a, but not LNC003834S2. Generally, these results supply evidence that miRNA‐34a can target and bind the seed region of SLC14A1 mRNA and depress SLC14A1 expression, while LNC003834S1 weakens the inhibitory effect by competitively binding miRNA‐34a.

Establishment of stable SLC6A1, PEX5L, and PCBP2 short hairpin RNAs (shRNAs) expressing camel renal medulla cell lines

Based on the sequences of SLC6A1, PEX5L, and PCBP2, the corresponding shRNAs were designed and synthesized. In line with 70% transfection efficiency (lentivirus fluid/cell media = 1/5, total volume = 1 mL) (Fig. 3A), shRNAs against SLC6A1, PEX5L, and PCBP2 were transfected into cultured renal medulla cells of camel by lentivirus packaging. The qRT‐PCR results showed that the expression levels of SLC6A1, PEX5L, and PCBP2 were decreased following transfections (Fig. 3B), representing that the expression of SLC6A1, PEX5L, and PCBP2 can be inhibited effectively by lentivirus‐mediated shRNAs expression. Regretfully, western blot results only detected PCBP2 protein expression (Fig. 3C), SLC6A1 and PEX5L had no clear bands. The possible main reason is that the SLC6A1‐ and PEX5L‐antibodies are not suitable for detecting the two proteins in camel cells with the absence of specific antibodies.

Fig. 3

RNA interference and ROS detection. (A) The experimental formulas of achieved 70% transfection efficiency of shRNA lentiviral vectors targeting SLC6A1, PCBP2, and PEX5L in renal medulla cells of camel in vitro with lentivirus‐NC/media = 1/5 (200×). Scale bar, 100 μm. (B) Expression levels of SLC6A1, PCBP2, and PEX5L in the light of the ΔΔCt method in different groups. GAPDH was used as a loading control (intrinsic control), and expression levels were normalized with SLC6A1, PCBP2, and PEX5L mRNA levels of NC and RNA interference groups. (C) Western blot targeting PCBP2. β‐actin was used as a loading control. (D) ROS detection after shRNA interference in renal medulla cells of camel by red fluorescent imaging DHE probe (100×). Scale bar, 100 μm. (E,F) Fluorescence intensity of ROS after shRNA interference against SLC6A1, PCBP2, and PEX5L, and after small RNA interference against mixed SLC6A1, PCBP2, and PEX5L (sRNA‐MIX). Data are presented as mean ± SD (n = 3) by Student’s t‐test. Symbols: ***P < 0.001. shSLC6A1, shPCBP2, and shPEX5L represent short hairpin RNAs of SLC6A1, PEX5L, and PCBP2 in turn. IC refers to intrinsic control with GAPDH, NC indicates negative control with nonsense shRNA, and BC means blank control.

Surging ROS levels in SLC6A1, PEX5L, and PCBP2 shRNAs and small RNA (sRNA)‐MIX renal medulla cells

The ROS levels were detected by dihydroethidium (DHE) fluorescent probe in renal medulla cells. Fluorescence imaging and intensity reflected that ROS levels were significantly higher (P < 0.001) in SLC6A1, PEX5L, and PCBP2 shRNAs and sRNA‐MIX cells (three oligos of SLC6A1, PEX5L and PCBP2 small RNA co‐interference) when compared with the negative and blank controls (Fig. 3D–F; Fig. S1). The results manifest that defective expression of SLC6A1, PEX5L, and PCBP2 cause increased ROS levels in camel renal medulla cells.

Discussion

In the previous work of our laboratory, the camel’s complex features coupled with desert adaptations such as water and fat metabolism had been unveiled through comparative genomic analysis, as well as the unique osmoregulation, osmoprotection, and compensatory mechanisms for water reservation responding to aridity stress based on transcriptomic evidences [60]. This study detected 22 mRNAs, 2 lncRNAs, and 31 miRNAs that were differentially expressed in renal medulla tissues of camels subjected to salt stress or free diet by RNA‐seq. In the identified protein‐coding genes, SLC14A1, SLC6A1, PEX5L, and PCBP2 were predicted to positively respond to salt stress, and SLC14A1 mRNA may compete for endogenous RNA with a novel lncRNA LNC003834 and miRNA‐34a. According to a former research [61], SLC14A1 encodes for the urea transporter B and is expressed in endothelial cells of the descending vasa recta of mammalian kidneys. In the renal medulla, the urea transporter B can improve urinary concentration ability and water reabsorption via its urea transmembrane transporter activity, and the single channel water permeability is similar to that of aquaporin 1 [62, 63]. In comparison to Dahl salt‐resistant rats or Dahl salt‐sensitive rats with normal salt diet, the SLC14A1 mRNA levels were appreciably reduced in the choroid plexus of Dahl salt‐sensitive rats on high salt diet, indicating that reduced SLC14A1 expression was correlated to increased sodium ion levels in the cerebrospinal fluid and elevated average arterial pressure [52]. Additionally, based on prior studies, SLC6A1 is involved in sodium‐ and chloride‐dependent γ‐aminobutyric acid (GABA) transportation [53, 64]. GABA can influence salt tolerance of Arabidopsis and actively responds to heat stress in four‐year‐rice and salt stress in maize plumule by strengthening antioxidant ability [54, 65], whereas rare are animal studies that explored the relationship between GABA and environmental stress resistance. PCBP2‐deficient mice and primary mouse embryonic fibroblasts emerge ROS production increment and cellular senescence acceleration [55]. PEX5L is engaged in the peroxisome matrix targeting signal‐1 binding and the peroxisome targeting sequence binding in which the peroxisome dysfunction is associated with fatal oxidative damage [56, 66, 67]. Furthermore, we verified the expression levels of candidate salt‐resistance‐related molecules including SLC14A1, SLC6A1, PEX5L, PCBP2, LNC003834, and miRNA‐34a to be consistent with RNA‐seq outcomes by qRT‐PCR.

For ceRNAs studies which contained SLC14A1 mRNA, LNC003834, and miRNA‐34a, several experiments were performed. The fluorescence distribution of the FISH images appeared that the functional areas of LNC003834 and miRNA‐34a were localized in the extranuclear space of the camel renal medulla. Relying on dual‐luciferase reporter assay system, we subsequently found that LNC003834 can act as a molecular sponge for miRNA‐34a to regulate SLC14A1 expression and presented that: (a) SLC14A1 expression was evidently restrained by miRNA‐34a due to the densely bound 23 seed matches of 3’ tail sequence of SLC14A1 mRNA; (b) The newly recognized 21 seed matches of LNC003834 can reverse the miRNA‐34a‐mediated SLC14A1 suppressed expression via competitively binding miRNA‐34a. These experiment results demonstrated relationships between LNC003834, miRNA‐34a, and SLC14A1 mRNA and that were identified as competing endogenous RNAs with maintaining sodium homeostasis.

At present, the causes of salt sensitivity can be mainly explained by two theories [68]. In the historical natriuretic handicap theory, salt sensitivity originates from a renal functional impairment, which represents the phenomena of slow excretion and sodium retention. In the contemporary vasodysfunction theory, salt‐induced hypertension results from abnormal vascular resistance responses under high salt intake. Notably in this research of biological salt resistance, we discovered a critical protein‐coding gene which contributes to enhancing vascular resistance in renal medulla with high salt consumption. The upregulated expression of SLC14A1 may inhibit sodium ion reabsorption from the renal tube to the descending vasa recta through competing endogenous LNC003834, miRNA‐34a, and SLC14A1 mRNA in camel renal medulla (Fig. 4). The examination of plasma Na+ concentration provided further evidence that no statistical significance was attained by comparing salt stress group and control group (Table S11).

Fig. 4

Salt‐resistance metabolism combined with competing endogenous RNAs (LNC003834, miRNA‐34a, and SLC14A1) and transcribed antioxidant genes (SLC6A1, PCBP2, and PEX5L) in renal medulla of camel.

Following shRNA knockdown of SLC6A1, PEX5L and PCBP2, and sRNA‐MIX interference, we learned that ROS levels were palpably increased in the renal medulla cells of camel. The occurrences meant that upregulated expression of SLC6A1, PEX5L, and PCBP2 may ameliorate antioxidative effects in positive response to salt resistance (Fig. 4). It also hints that camel as one of mammal species can invoke SLC6A1 gene to meet effortlessly to environmental challenges like plants do.

In summary, the present work exhibits that the inhibition of sodium reabsorption is activated by stronger SLC14A1 expression which induced by ceRNAs comprising LNC003834 and miRNA‐34a; while, SLC6A1, PEX5L, and PCBP2 reduce ROS levels via increased expression to achieve salt resistance in renal medulla of camel. It indicates that although with high dietary intake of salt, the reabsorption of sodium ion is inhibited. Herein, we reveal a mechanism for maintaining sodium homeostasis in camel and envision that our work can empower people to seek out a new way for solving human diseases triggered by high salt intake.

Conflict of interest

The authors declare that there is no conflict of interest regarding the publication of this article.

Author contributions

DZ conceived and designed the study. DZ, JP, CL, FM, and YC carried out the experiments. YC analyzed the data and wrote the manuscript. YZ, JC, and HZ provided valuable insights in the discussion and revision of the manuscript. All authors read and approved the final manuscript.

Fig. S1. ROS detection after sRNA‐MIX interference in renal medulla cells of camel by red fluorescent imaging DHE probe (100×). sRNA‐MIX indicates three classes of small RNAs (sSLC6A1, sPEX5L and sPCBP2) co‐interference against SLC6A1, PEX5L and PCBP2. Scale bar, 100 μm.

Table S1. The details of six camels.

Table S2. The sequences of fluorescent group labeled RNA probes.

Table S3. The inserted SLC14A1 sequences in SLC14A1 luciferase reporter plasmids.

Table S4. The inserted LNC003834 sequences in LNC003834S1 and LNC003834S2 plasmids.

Table S5. The sequences of different short hairpin RNAs (shRNAs).

Table S6. The sequences of different small RNAs (sRNAs).

Table S7. The primer sequence of candidate genes of salt‐resistance.

Table S8. The significantly differential mRNAs and lncRNAs in the renal medulla of camel under salt stress.

Table S9. The significantly differential miRNAs in the renal medulla of camel under salt stress.

Table S10. The sequence of novel lncRNA LNC003834 gene.

Table S11. Plasma Na+ concentration.

Click here for additional data file.