Plasma exosome-derived microRNAs expression profiling and bioinformatics analysis under cross-talk between increased low-density lipoprotein cholesterol level and ATP-sensitive potassium channels variant rs1799858

Exosome-derived microRNAs (exo-miRs) as messengers play important roles, in the cross-talk between genetic [ATP-sensitive potassium channels (KATP) genetic variant rs1799858] and environmental [elevated serum low-density lipoprotein cholesterol (LDL-C) level] factors, but the plasma exo-miRs expression profile and its role in biological processes from genotype to phenotype remain unclear. A total of 14 subjects with increased LDL-C serum levels (≥ 1.8 mmol/L) were enrolled in the study. The KATP rs1799858 was genotyped by the Sequenom MassARRAY system. The plasma exo-miRs expression profile was identified by next-generation sequencing. 64 exo-miRs were significantly differentially expressed (DE), among which 44 exo-miRs were up-regulated and 20 exo-miRs were down-regulated in those subjects carrying T-allele (TT + CT) of rs1799858 compared to those carrying CC genotype. The top 20 up-regulated DE-exo-miRs were miR-378 family, miR-320 family, miR-208 family, miR-483-5p, miR-22-3p, miR-490-3p, miR-6515-5p, miR-31-5p, miR-210-3p, miR-17-3p, miR-6807-5p, miR-497-5p, miR-33a-5p, miR-3611 and miR-126-5p. The top 20 down-regulated DE-exo-miRs were let-7 family, miR-221/222 family, miR-619-5p, miR-6780a-5p, miR-641, miR-200a-5p, miR-581, miR-605-3p, miR-548ar-3p, miR-135a-3p, miR-451b, miR-509-3-5p, miR-4664-3p and miR-224-5p. Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) analyses were subsequently implemented to identify the top 10 DE-exo-miRs related specific target genes and signaling pathways. Only 5 DE-exo-miRs were validated by qRT-PCR as follows: miR-31-5p, miR-378d, miR-619-5p, miR-320a-3p and let-7a-5p (all P < 0.05). These results firstly indicated the plasma exo-miRs expression profile bridging the link between genotype (KATP rs1799858) and phenotype (higher LDL-C serum level), these 5 DE-exo-miRs may be potential target intermediates for molecular intervention points.

which become a serious public health problem [1]. The increased LDL-C plasma concentration (≥ 1.8 mmol/L) and its related ASCVD is the result of a combination of genetic and environmental factor influences. These environmental factors include lifestyle risk factors (e.g., unhealthy diet, smoking, physical inactivity and obesity, etc.), chemical and physical hazards (e.g., hyperlipidemia, hypertension and hyperglycemia, etc.), and their interactions. Comprehensive management and control of those multiple modifiable environmental risk factors is significant related to lower LDL-C level and lower cardiovascular events risk [2], but those benefits may be dampened by high genetic risk [3]. The genetic susceptibility factor, as an inherent and lifetime risk factor, is powerful independent predictor of high LDL-C level and its related ASCVD. Indeed, the ATP-sensitive potassium channels (KATP) variant rs1799858 was a genetic risk factor for higher LDL-C plasma concentration (≥ 1.8 mmol/L) and its related macro-/micro-vascular arteriosclerotic event risk [4]. However, the mechanism of elevated LDL-C plasma level and its induced ASCVD under specific genetic background of KATP variant rs1799858 remains unclear.
Co-evolution of the genetic and environment factors leads to the development of higher LDL-C serum levels and its related ASCVD. Non-coding RNA, especially the plasma exosome-derived microRNAs (exo-miRs), as the bridge between environmental factors and genetic factors, plays a critical role in this crosstalk process. Exosomes are lipid bilayer extracellular vesicles with a diameter of 30-150 nm secreted by almost all nucleated cells, which mediate cell-cell communication through their components, including microRNAs (miRs), mRNA, DNA, proteins and lipids. The miRs are small and endogenous RNAs (containing about 22-25 nucleotides), which take part in regulating multiple target genes at the post-transcriptional level. Exo-miRs are involved in kinds of physiological or pathological processes in occurrence and development of elevated LDL-C level and its related ASCVD. However, the plasma exo-miRs expression profiles under cross-talk effect between genetic and environment factors remain largely unknown. In this study, using a next-generation sequencing method, we sought to characterize the circulating exo-miRs expression profile in subjects with increased LDL-C level (≥ 1.8 mmol/L) under specific genetic background of KATP polymorphism rs1799858. We then performed Gene Ontology (GO) annotation and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis based on predicted target genes. The top 10 DE-exo-miRs were then confirmed by individual quantitative real-time polymerase chain reaction (qRT-PCR) in 50 increased LDL-C levels (≥ 1.8 mmol/L) subjects with T-allele of rs1799858 and 50 same subjects with counterpart CC genotype. This helped us to facilitate our understanding of the molecular processes from genotype (KATP rs1799858) to phenotype (higher LDL-C serum level).

Study subjects
A total of 14 subjects with only increased LDL-C serum concentration (≥ 1.8 mmol/L) were recruited into the study from South China. Subjects with other types of dyslipidemia were excluded from the study, including increased levels of triglyceride (TRIG ≥ 1.7 mmol/L), total cholesterol (TC ≥ 5.2 mmol/L) or (and) apolipoprotein B (Apo B ≥ 80 mg/dL), and (or) decreased levels of high-density lipoprotein cholesterol (HDL-C < 1.0 mmol/L) and apolipoprotein AI (Apo AI < 120 mg/dL). All participants with different types of dyslipidemia were newly diagnosed according to guidelines [5]. All subjects combined with smoking, drinking, hypertension (HTN), coronary atherosclerotic heart disease (CAD), type 2 diabetes mellitus (T2D), stroke, abnormal liver function [alanine aminotransferase (ALT) or (and) aspartate aminotransferase (AST) more than 3 times upper limit of normal], abnormal kidney function [estimated glomerular filtration rate (eGFR) less than 90 ml/min·1.73 m 2 ], or (and) any other medical conditions or drugs that may influence blood lipid levels were also excluded from the study. All blood biochemistry analysis was conducted on enrollment to the study by using standard analytical techniques.

Exosome isolation, exo-miRs sequencing and sequencing data analysis
The test process was carried out according to the following procedure: (1) Sample collection: The whole-blood samples were collected into anticoagulation vacuum tube with EDTA (3 mg/mL) on enrollment, but after a 12-h fasting and a light, low-fat meal the night. Hemolyzed samples were excluded from the experimental workflow. The freshly whole-blood samples were centrifuged within an hour from collection (3000g × 15 min, 4 ℃) to separate plasma. The upper plasma was transferred to a new Eppendorf tube, and then centrifuged (2000g × 20 min, 4 ℃) to remove additional cellular fragments. The cleared supernatant was cautiously transferred to another new Eppendorf tube and stored at − 80 ℃. (2) Isolation exosomes from plasma: Exosomes from the supernatant were isolated with exoEasy Maxi kit (Qiagen, Dusseldorf, Germany) according to the manufacturer's protocol with modifications described in Stranska et al. [6]. The eluates were collected to low protein binding tubes and stored at − 80 ℃. (3) Extraction RNA from exosomes: Exo-miRs were extracted by HiPure Liquid miRNA Kit/HiPure Serum/Plasma miRNA Kit (Megan, China) according to the manufacturer's instructions. The quantity and integrity of exo-miRs yield was assessed by using the Qubit ® 2.0 (Life Technologies, Carlsbad, USA) and Agilent 2200 TapeStation (Agilent Technologies, Carlsbad, USA) separately. (4) Exo-miRs sequencing: exo-miRs sequencing was performed using Illumina platforms (Illumina, Carlsbad, USA) at Ribobio Co. (Guangzhou, China). Briefly, RNAs were successively ligated with 3′-and 5′-RNA adapter. The adapter-ligated RNAs were then submitted to reverse transcription reaction and amplified with a low-cycle. The PCR products were PAGE-size-selected according to manufacturer's protocol of NEBNext ® Multiplex Small RNA Library Prep Set for Illumina (New England BioLabs Inc., Massachusetts, USA). The purified exo-miRs library products were assessed using the Agilent 2200 TapeStation, and then sequenced using an Illumina HiSeq2500 with single-end 50 bp. (5) Sequencing data analysis: The clean reads were acquired after quality control and preprocessing of FASTQ. The miRDeep2 was performed to determine known mature exo-miRs based on miRBase21 (https ://www.miRBa se.org) and predict novel exo-miRs. The expression of exo-miRs was calculated by reads per million (RPM) values. The differential expression of exo-miRs in subjects with different genotypes of rs1799858 was calculated by edgeR algorithm according to the criteria of |log2 (Fold Change)|≥ 1 and P value < 0.05. The online softwares (miRDB, miRTarBase, miRWalk and TargetScan) were performed to predict exo-miRs related targets gene. KOBAS 2.0 software was used to further analysis of GO and KEGG pathway. (6) Validation of top 10 DE-exo-miRs: qRT-PCR array was performed for top 10 DE-exo-miRs in a new verification cohort.

Statistical analysis
All analysis for baseline characteristics was performed with SPSS version 24 (SPSS, Chicago, USA). Categorical variables were presented as frequencies. Continuous variables were presented as mean ± SD. The differences on continuous variables between the two genotypes (CC vs. TT + CT) of rs1799858 in subjects with increased LDL-C serum concentration (≥ 1.8 mmol/L) were assessed by independent-sample t-test while categorical variables by Chi-square test. The DE-exo-miRs between the two different genotypes was also analyzed with edgeR software. Both GO category and KEGG pathway analyses were evaluated by Chi-square test or Fisher's exact test. The false discovery rate (FDR) was calculated to adjust the P values. If an adjusted P value is less than 0.05, the result is considered as significant.

Clinical baseline characteristics of study subjects
The clinical features among subjects with different genotypes of KATP rs1799858 in this study are shown in Table 1.

GO analysis of Enriched biological processes, cellular component and molecular functions regulated by CTGs of top 10 DE-exo-miRs.
GO analysis was used to identify the biological processes, cellular component and molecular functions for top 10 DE-exo-miRs related CTGs. As shown in Fig. 2, the top 10 DE-exo-miRs related CTGs in subjects carrying T-allele (TT + CT) of rs1799858 were obviously linked to regulation of signaling, apoptotic process, vesicle-mediated transport, homeostatic process, protein complex subunit organization, lipid metabolic process, autophagy, angiogenesis, oxidation-reduction process, response to hypoxia, inflammatory response, and microtubule-based process, relating to cellular components such as vesicle, membrane-bounded organelle (e.g., endoplasmic reticulum, mitochondrion, lysosome), membrane protein complex, microtubule organizing center, transcription factor complex, and transmembrane transporter complex. The molecular functions of the top 10 DE-exo-miRs related target genes were correlated with protein binding, ion binding, DNA binding, enzyme binding, sequence-specific DNA binding, ATP binding, transcription factor activity, kinase activity, lipid binding, oxidoreductase activity and gated channel activity.

KEGG analysis of enrichment pathway regulated by CTGs of top 10 DE-exo-miRs
KEGG analysis was also used to identify the for top 10 DE-exo-miRs related CTGs were evidently enriched in 75 pathways. As shown in Fig. 3 and Additional file 1: Figure S2, the top 30 pathways were involved in environmental information processing (e.g., signaling pathways of PI3K-Akt, MAPK and Ras, etc.), genetic information processing (e.g., protein processing in endoplasmic reticulum), human diseases (e.g., insulin resistance and nonalcoholic fatty liver disease), metabolism (e.g., metabolic pathways), and organismal systems (e.g., signaling pathways of insulin, chemokine, platelet activation and T cell receptor).

Discussion
It's well known that genetic or germline variants (e.g., genotype) have a huge impact on the phenotypic landscape of a population. Complex disease (e.g., cardiovascular disease, cancer, etc.) is defined as a phenotype that is caused by many individual gene events, with a *The top signaling pathways were related to metabolism-related diseases or inflammation including those related to human diseases such as insulin resistance, non-alcoholic fatty liver disease, metabolic/insulin/chemokine/T cell receptor signaling pathways and platelet activation; those related to environmental information processing functions including PI3K-Akt/MAPK/Ras/FoxO/Rap1/Hippo/Wnt signaling pathways; and those related to genetic information/cellular processes such as endocytosis and protein processing in endoplasmic reticulum. Our analysis indicated that these pathways may be involved in the molecular processes from genotype (KATP rs1799858) to phenotype (higher LDL-C serum level) significant contribution from environmental factors. Germline variants influence clinic outcomes of complex disease. Recent studies found that functionally mutations of natural killer cells were positively associated with cancer risk [7,8]. Similarly, in our previous study found that the KATP SNP rs1799858 was associated with increased risk of elevated LDL-C serum concentration (≥ 1.8 mmol/L) and its related macro-/micro-vascular arteriosclerotic events. However, the underlying mechanism of genetic effects on phenotype mediated by genotype-environment interactions remains elusive. exo-miRs, as the leading factor in inducing genetic susceptibility changes, are messengers in the cross-talk between environmental factors and genetic factors. The exo-miRs expression profile and its role of genetic variants on cellular signaling pathways will facilitate our understanding of the relationships between genotype (e.g., KATP SNP rs1799858) and phenotype (e.g., LDL-C serum concentration ≥ 1.8 mmol/L and its related ASCVD). This is the only study to reveal a distinct exo-miRs expression in subjects with LDL-C serum concentration ≥ 1.8 mmol/L under specific genetic background of KATP polymorphism (rs1799858). (1) We identified the DE-exo-miRs, and respectively screened up-/down-regulating of the top 20 DE-exo-miRs, whose CTGs were identified. (2) GO and KEGG pathway analyses were implemented on those exo-miRs related CTGs.
To further investigate the function of exo-miRs under cross-talk status between higher LDL-C level and different genotypes of rs1799858, GO and KEGG analyses were performed for the 1045 CTGs of top 10 DE-exo-miRs in increased LDL-C ≥ 1.8 mmol/L subjects with T-allele (TT + CT) of rs1799858. GO analyses suggested that enrichment of CTGs played crucial roles in biological processes, cellular component and molecular functions (Fig. 2), consistent with a regulatory role on dyslipidemia and related ASCVD [46] for these exo-miRs in the transcription/translation processes [47]. Many differentially regulated KEGG pathways were identified. The results showed that PI3K-Akt signaling pathway, metabolic pathways, and MAPK signaling pathway were the top 3 differentially regulated pathways (Fig. 3). Importantly, PI3K-Akt pathway, which plays an essential role in cellular physiology by regulating growth factor signals during organismal growth and critical cellular processes (e.g., lipid metabolism, glucose homeostasis, protein synthesis, cell proliferation and survival) in normal physiology and morbid conditions (e.g., obesity and T2D) [48]. MAPK pathway, which is known as an important signal transmitter that transmit signals from receptors on the surface to DNA in the nucleus of the cell, is essential in regulating of lipid homeostasis as well as many other cellular processes (e.g., inflammation, cell differentiation, cell division, cell proliferation, motility, apoptosis and stress response) [49]. Both signaling pathways are not only involved in the regulation of lipid homeostasis but also related to atherosclerosis and ASCVD, manifesting the characteristics of synchronous activation [50]. These top 10 exo-miRs were interacted with target genes, forming a network that was influenced by significantly differently regulated exo-miRs in higher LDL-C (≥ 1.8 mmol/L)level patients with T-allele (TT + CT) of rs1799858 (Fig. 4), but only 5 exo-miRs were further comfirmed based on a new verification cohort (Fig. 5). These findings indicated that these DE-exo-miRs could play an important role in increased LDL-C plasma concentration (≥ 1.8 mmol/L) and its related ASCVD by regulating these two pathways.

Study strengths and limitations
The strength of the study was that this is the first time to characterize the circulating exo-miRs expression profile in biological processes from genotype (KATP variant rs1799858) to phenotype (increased LDL-C serum levels), indicating that the potential role of exo-miRs related epigenetic modification on co-evolution of the genetic and environment factors leads to the development of higher LDL-C serum levels and its related ASCVD. There are some limitations in this study. Firstly, the sample size is small so that large-scale, prospective population-based cohort studies will be conducted to confirm the reported results, such as the relationships between these DE-exo-miRs plasma levels and LDL-C exception level related ASCVD events under specific genetic background of KATP variant rs1799858. Secondly, this was only a preliminary bioinformatics analysis so that the possible miss-distance effect and non-specific effect could not be excluded due to lack of validation at the cellular and molecular levels. Finally, due to the validation of DE-exo-miRs by qRT-PCR at the clinical level is incomplete, it is necessary to carry out further functional verification on those 5 validated DE-exo-miRs and its related pathways at the cellular and molecular levels as follows: to study the function of these exo-miRs at the translational or transcriptional levels based on luciferase system or fluorescent microscopy, to identify the specific nucleotide sequences and exo-miRs binding, and to observe its impact on overall functions of signal pathways after mutation. Therefore, results must be interpreted carefully.

Conclusion
This study firstly indicated that the plasma exo-miRs expression profile bridging the link from genotype (KATP rs1799858) to phenotype (higher LDL-C serum level), and these DE-exo-miRs (especially top 10 DE-exo-miRs) may be potential target intermediates for development of novel diagnosis, prevention, and treatment of LDL-C exception level and its related atherosclerotic vascular disease, warrant further research.
Additional file 1: Figure S1. Volcano map of DE-exo-miRs between different genotypes of KATP rs1799858 in subjects with elevated LDL-C (≥1.8 mmol/L) serum level. Figure S2. Bubble map for KEGG analysis of enrichment pathway regulated by CTGs of top 10 DE-exo-miRs.