For this study, a total of 72 patients with a history of hypertension were selected and peripheral blood samples were collected from February 2021 to August 2021 at the Department of Cardiology, the First Affiliated Hospital of Guangxi Medical University. The informed consent forms used in this study were approved by the Ethics Committee of the first affiliated Hospital of Guangxi Medical University. All experiments were performed by the principles and regulations established by the ethics committee.

All the participants were registered according to the following inclusion criteria: (a) diagnosis of essential hypertension, and the diagnostic criteria of definite SSH as previously reported by Citterio et al. (19). (b) Age between 30 and 70 years, (c) while the exclusion criteria were: individuals with severe diabetes, heart failure, stroke, peripheral artery disease, congenital heart disease, acute myocardial infarction, liver and kidney disease, or cancer patients. All 72 patients were categorized into two groups including SSH group with 29 patients and SRH group have 43. A modified Sullivan acute NS-loading test was used to distinguish SRH and SSH patients, and mean arterial pressure (MAP) was also measured with the systolic blood pressure (SBP) and diastolic blood pressure (DBP) for each time point. Blood pressure was measured at two-time points: baseline BP (before and after sodium-loading). After sodium-loading, two times for each MAP value (MAP1 and MAP2) were calculated from the corresponding 2 BP values. MAP was considered to be the DBP plus one-third of pulse pressure (PP). Patients with MAP2 − MAP1 ≥ 5 mm Hg were diagnosed with SSH and others with SRH. This diagnostic method has widely been used to differentiate the SSH from SRH and was also validated to have the same accuracy as the long-period sodium-loading test. Furthermore, five patients from each group were randomly selected for subsequent RNA-seq analysis.

Under aseptic conditions, the collected whole blood was mixed with an equal volume of PBS, and Ficoll-Hypaque density gradient centrifugation was used to isolate the peripheral blood mononuclear cells (PBMC). Then, TRIzol reagent (Invitrogen) was added to the PBMC at a ratio of about 1 × 10⁷ cells/1.5 ml to degrade the proteins and the RNA was extracted from the cells. In addition, the Agilent 2100 Bioanalyzer (Agilent DNA 1000 Reagents) was also used to estimate the total RNA concentration, RNA integrity number (RIN), 28S/18S, and fragment size to evaluate the yield, purity, and integrity of RNA.

MGISEQ-2000 sequencing technology was being used for RNA-Seq sequencing. The qualified library was denatured into the single-stranded library by adding NaOH and then the solution was diluted to a certain concentration according to the expected amount of data on the machine. Afterward, the denatured and diluted library was loaded to Flow Cell for hybridization with the adapter, and then the cluster generation platform cBot was used to complete the bridge PCR amplification. Finally, the prepared Flow Cell was sequenced using Illumina sequencing system SBS reagents. The sequencing data was aligned to the reference gene sequence (reference species: Homo sapiens; data source: NCBI; reference genome version: GCF_000001405.39_GRCh38.p13) for clean read using Bowtie2 (version: v2.2.5) (20) and RSEM (Version: v1.2.8) (21) was used to calculate the gene expression levels of each sample.

DESeq2 package (version 1.36.0) in R software (version 3.6.0) for estimate variance-mean dependence in count data from high-throughput sequencing assays and test for differential expression based on a model using the negative binomial distribution (22). The FDR < 0.05 and |log2 (fold change)| > 1 values were used as a criterion to find the differentially expressed genes (DEGs). In addition, pheatmap package in R software was performed to plot the heatmap of DEG (23).

Gene ontology (GO) annotation contained three parts: BP, CC, and MF. For the GO and KEGG analyses, all candidate genes were mapped to each item in the Gene Ontology database¹, and the number of genes in each item was calculated (24). Furthermore, hypergeometric test was applied to find the items that were significantly enriched in candidate genes compared with all background genes in this species. The P value is calculated using the base function phyper (version 4.3.0) of R software. Subsequently, multiple tests are carried out on P value to be positive, and the calibration software package is Q value (version 2.28.0).

The gene set enrichment analysis (GSEA) software (version 4.0.3) was used on all detected genes with FDR < 0.05 as the criterion, and the top three of biological process (BP) and pathway enriched gene sets were selected for analysis. The gene sets in this analysis were collected from the Molecular Signatures Database (MSigDB, version 6.2)² (25).

Through the STRING database (version 11.5)³, we selected protein-protein interaction (PPI) analysis with a comprehensive score greater than 0.9 to construct a PPI network. Cytoscape (version 3.9.0) was used to visualize the PPI network and then analyzed the modules with higher scores in the PPI network through Molecular Complex Detection (MCODE). In addition, we used 12 algorithms in Cytohubba including MCC, DMNC, MNC, Degree, EPC, BottleNeck, Eccentricity, Closeness, Radality, Betweenness, Stress, and Clustering coefficient to calculate the total weight of each gene, and ultimately selected top 10 most frequently occurring 10 genes among the 12 algorithms as hub genes.

The connectivity map (CMAP) database⁴ is an open resource that links diseases, genes, and drugs through similar or opposite gene expression profiles (26). CMAP database was also used to predict the potential small molecule compounds that can reverse the changes in DEG expression. Mean ≤ 0.4 and P < 0.01 were set as the screening criteria.

Whole transcriptome RNA-seq analysis was carried out on PBMCs isolated from five SSH and five SRH patients randomly selected from 72 patients.

The baseline characteristics of all patients with SRH and SSH are shown in Table 1. These include gender, age, blood pressure, heart rate, and so on. The results from two independent sample t-test showed that age, BMI did not significantly differ between the SSH and SRH groups by SPSS (version: SV26.0) (P > 0.05). In addition, we found significant differences in LDL-C, MCH, α-HBD (P > 0.05).

Table 2 shows the saline-loading test of the study participants. The results show that SSH accounts for 36% and SRH for 64%.

A total of 19,813 genes were identified in PBMC samples and 431 genes were found to be differentially expressed between SSH and SRH samples of which 294 were up-regulated and 137 down-regulated genes in the SSH group. The criteria for DEGs were selected to q-value < 0.05 and |log2Foldchange| > 1. The heatmap and volcano plots for DEGs are shown in Figures 1A,B. The Venn diagram shows the number of genes unique and shared between the two groups. The shared number of genes shared between SSH and SRHare16578 (Figure 1C).

Gene ontology functional enrichment analysis for these DEGs were performed. The results indicated that these DEGs were significantly enriched in eight enriched cellular component (CC), such as plasma membrane, nuclear nucleosome, and hemoglobin complex (Figure 2A). For molecular functions (MF), these DEGs were mainly enriched in hemoglobin alpha binding, chemokine activity, and inward rectifier potassium channel activity (Figure 2B). In addition, these DEGs were most significantly enriched in the BP of immune response, cell adhesion, and inflammatory chemotaxis (Figure 2C). The enriched KEGG pathways were performed in the top five signal pathways, including Rheumatoid arthritis, Systemic lupus erythematosus, Salmonella infection, Alcoholism, and Graft-versus-host disease (Figure 2D).

Gene set enrichment was performed to identify gene sets with statistically significant differences between the SSH group and the SRH group. The top three of GO_BP sets are shown. The most significant enriched up-regulated GO_BP sets included regulation of nuclear-transcribed mRNA catabolic process deadenylation dependent decay, Regulation of megakaryocyte differentiation, and Receptor signaling pathway (Figure 3A). The most significantly enriched down-regulated GO_BP sets included ATP synthesis coupled electron transport, Oxidative phosphorylation, Antigen processing, and presentation of exogenous peptide antigen via MHC class I (Figure 3B). As illustrated in Figure 4, to determine signaling pathways associated with SSH and SRH, GSEA pathway analysis was performed on DEGs. The top three most significantly enriched gene sets positively correlated with the SSH group were ABC transporters, ERBB signaling pathway, and TGF beta signaling pathway (Figure 4A). The top three most significantly enriched gene sets negatively correlated with the SSH group were PARKINSONS disease, oxidative phosphorylation, and ribosome (Figure 4B).

We used the STRING database to construct the PPI network, which consisted of 396 nodes and 89 edges (Figure 5). Nodes represent DEGs, and edges represent interactions between DEGs. High-scoring modules in the PPI network were analyzed by MCODE (Figure 6). The ten identified hub genes (IL6, CCL2, IL1A, CXCL1, BUB1, CCL3L3, CDC20, CENPA, CXCL2, and TTK) were showed in Table 3 and Figure 7. The expression level of the hub gene in samples were shown in Figure 8.

Analysis of the CMAP database showed that iopromide and iloprost compounds potentially exerted a negative regulatory effect on DEGs (Table 4). We deduced that these two drugs are potential candidates to alleviate SRH progression to SSH.