Analysis of Fecal Microbiota in Patients with Hypertension Complicated with Ischemic Stroke

Ischemic stroke is a disease with a very high incidence in the clinic, and hypertension is the most important variable risk factor of ischemic stroke. Studies have shown that intestinal microbes are involved in the occurrence and development of various diseases. This study aims to explore whether intestinal microbes play an important role in the pathogenesis of ischemic stroke in a hypertensive population. In this study, the inpatients in the Department of Neurology and Cardiology of the Second Affiliated Hospital of Shandong First Medical University in April 2021 were selected, including seven patients with hypertension complicated with ischemic stroke and only seven patients with hypertension. After collecting the stool samples of patients, the gene sequence of the samples was detected by 16S rRNA sequencing technology, and the double-ended 2 × 150 bp sequencing was carried out. After sequencing, the results were analyzed by diversity analysis, species difference analysis, species function difference analysis, and other bioinformatics tests. According to the test results, serum proteomics and biochemical blood tests were carried out to verify. There was no significant difference in α diversity and β diversity between hypertension complicated with the cerebral infarction and hypertension groups. LEfSe analysis showed that at the genus level, compared with the hypertension group, Bacteroides, UCG_009, and Eisenbergiella had significantly increased relative abundance. The genera with relatively significantly reduced abundance are Ruminococcus_gnavus_group, Sutterellaceae, Burkholderia, and Prevotella and the LDA score of Prevotella is <  − 4, which indicates that there are significant differences. Compared with the blood biochemical indexes, the results showed that the level of APOA1 in hypertensive patients with ischemic stroke was significantly higher than that in hypertensive patients (p < 0.05), but there was no significant difference in total cholesterol (CHOL), triglyceride (TG), high-density lipoprotein (HDL), low-density lipoprotein (LDL), apolipoprotein B (APOB), and free fatty acid (NEFA). Proteomic analysis showed that there were 89 up-regulated genes and 51 down-regulated genes in the serum of the two groups, and the expression of APOC2 and APOC3 in the cerebral infarction group with hypertension was significantly higher than that in the hypertension group (p < 0.05). The intestinal diversity of patients with hypertension complicated with stroke is similar to that of patients with hypertension, but there are differences in microbiota, among which Prevotella is the most significant. Prevotella could affect lipid metabolism so that APOC2 and APOC3 in the blood are significantly increased, leading to cerebral artery atherosclerosis and, finally, ischemic stroke. This provides a new idea for preventing and treating ischemic stroke in patients with hypertension, but the mechanism of Prevotella acting on apolipoprotein needs further verification by basic medical research.


Introduction
Ischemic stroke refers to the general name of brain tissue necrosis caused by narrow or occluded blood supply arteries (carotid artery and vertebral artery) and insufficient blood supply in the brain.As the second leading cause of death and the third leading cause of disability in the world, ischemic stroke has always been a common concern of human beings.In recent years, with the continuous extension of human life and the aging of the population, the incidence of ischemic stroke has also increased significantly (Roy-O' Reilly et al. 2018), gradually becoming an important public health burden, so ischemic stroke starting with its various risk factors is the most direct means to Yitong Jiang and Chunhua Liu contributed equally to this work and shared the first authorship.
Extended author information available on the last page of the article prevent stroke.Hypertension is the most important variable risk factor (Sarikaya et al. 2015) of stroke.Some studies have shown that treating hypertension can effectively reduce the risk of stroke (Paul et al. 2021).Therefore, it is very important to recognize the relevant mechanism between blood pressure and ischemic stroke.
The number of microorganisms in the human body is now understood to approximate the number of cells in a reference human body.There are approximately 3 × 10 13 cells in the human body and the human microbiome is estimated to contain 3.9 × 10 13 microbes (Sender et al. 2016).With the development of amplicon sequencing of 16S rRNA gene and whole genome sequencing, we can further understand the relationship between intestinal microbiota community structure and the human body.Intestinal microbiota is involved in a series of physiological processes that are vital to host health, including energy homeostasis, metabolism, intestinal epithelial health, immune activity, and neurobehavioral development.The metabolic capacity endowed by the genome exceeds that of a single host organism, which makes the intestinal microorganisms an active participant in host physiology (Barko et al. 2018).Intestinal microbiota can also lead to a variety of diseases, such as inflammatory bowel disease (IBD), inflammatory skin diseases (such as psoriasis and atopic dermatitis), autoimmune arthritis, type 2 diabetes, obesity, and atherosclerosis (Singh et al. 2017).In recent years, more and more studies have shown that intestinal microbiota and its metabolites play an important role in the risk factors of ischemic stroke, such as hypertension, diabetes, and hyperlipidemia.The prevalence of hypertension is increasing year by year.To improve the prevention and control level of hypertension in China, the diagnostic threshold of hypertension has been lowered from systolic blood pressure ≥ 140 mmHg and/or diastolic blood pressure ≥ 90 mmHg to systolic blood pressure ≥ 130 mmHg and/or diastolic blood pressure ≥ 80 mmHg in China Clinical Practice Guide, which shows that people pay more and more attention to hypertension.Clinically, we found that hypertension is very common in the stroke population, and hypertension can promote ischemic stroke by increasing shear stress, endothelial dysfunction, and arteriosclerosis (Konukoglu et al. 2017;Cipolla et al. 2018).Hypertension also promotes cerebral small vessel diseases through several mechanisms, including insufficient perfusion, decreased self-regulation ability, and local increase of blood-brain barrier permeability (Diener et al. 2020;Price and Kasner 2014;Santisteban et al. 2020).
More and more studies have shown that hypertension can promote the occurrence of ischemic stroke (Cipolla et al. 2018).However, we know little about the relationship between the occurrence of ischemic stroke and intestinal microbiota in the hypertensive population.Therefore, this study applied 16S rRNA gene sequencing, proteomics, and other technologies to compare the microbiota differences between hypertensive and hypertensive patients with ischemic stroke and explore the role of different microbiota in disease development, to further understand the internal relationship between the occurrence and development of ischemic stroke and intestinal microbes, and to provide reasonable data reference and research support for finding and treating ischemic stroke from intestinal microbes.

Research Object
In this experiment, we enrolled patients hospitalized in the Department of Neurology and Cardiology of the Second Affiliated Hospital of Shandong First Medical University in April 2021, including seven patients with hypertension complicated with ischemic stroke and only 7 patients with hypertension.
The admission criteria of hypertension complicated with ischemic stroke group are as follows: (1) age over 18 years old; (2) clinical diagnosis of acute cerebral infarction, and the onset time is less than 10 days; (3) volunteer to participate in this research.The exclusion criteria are as follows: (1) connective tissue disease; (2) diabetes; (3) suffering from acute and chronic digestive diseases; (4) patients who are delirious and unable to collect stool samples correctly.
The admission criteria of the hypertension control group are as follows: (1) age over 18 years old; (2) volunteering to participate in this research.The exclusion criteria are as follows: (1) previous history of cerebral infarction; (2) connective tissue disease; (3) diabetes; (4) suffering from acute and chronic digestive diseases; (5) patients who are delirious and unable to collect stool samples correctly.
The clinical information of the experimental group and the control group was collected, including gender, age, current medical history, and past history.During the hospitalization, the feces were collected, and the blood lipid, blood glucose, blood routine, and other test indexes were recorded.This study was approved by the ethics committee of our hospital, and all participating researchers were informed.

Collection of Feces
The sterile cotton swab is used to dip 50 mg of the middle stool of each of the patients with hypertension complicated with stroke and those with hypertension only, immerse it in the sampling tube of cereal microbiota storage buffer, rotate and stir it for 30 s, then tighten the cover of the sampling tube, shake it left and right for 10 s to complete the mixing, record the name and collection time of the sample, store it at room temperature, and send it for inspection within 1 week.

DNA Extraction
The DNA extraction kit GHFDE100 (Zhejiang Hangzhou Equipment Preparation 20,190,952) is used to extract the metagenomic DNA from the sample according to the manufacturer's instructions.

PCR Amplification and 16S rRNA Detection Method
The V4 region of bacterial 16S rRNA gene was amplified by PCR using forward primer 515F (5′-GTG CCA GCMGCC GCG GTAA-3′) and reverse primer 806R (5′-GGA CTA CHVGGG TWT CTAAT-3′).Sample-specific paired-end 6-bp barcodes are integrated into the TruSeq adapter for multiple sequencing.The polymerase chain reaction consists of 25 µl Phusion High-Fidelity PCR Master Mix, 3 µl (10 µm) forward and reverse primers, 10 µl template DNA, 3 µl dimethyl sulfoxide, and 6 µl ddH 2 O.The heat cycle includes initial denaturation at 98 ℃ for 30 s and then 25 cycles, including denaturation at 98 ℃ for 15 s, annealing at 58 ℃ for 15 s, renaturation at 72 ℃ for 15 s, and extension at 72 ℃ for 1 min.PCR amplification products were purified with Agencourt AMPure XP magnetic beads and quantified with PicoGreen dsDNA Assay Kit.After the individual quantitative step, the amplified fragments were mixed in equal amounts, and the double-ended 2 × 150 bp sequencing was carried out using the Illumina NovaSeq6000 platform of Guhe Information Technology Co., Ltd.

Statistics and Analysis of OTU Data
Raw sequencing reads with exact matches to the barcodes were assigned to respective samples and identified as valid sequences.The low-quality sequences were filtered through following criteria (Gill et al. 2006;Chen and Jiang 2014): sequences that had a length of < 20%, sequences that contained ambiguous bases, and sequences that contained mononucleotide repeats of > 8 bp.Paired-end reads were assembled using Vsearch V2.22.1 (-fastq_mergepairs -fastq_minovlen 5).The Amplicon Sequence Variant (ASV) picking using Vsearch v2.22.1 included dereplication (-derep_fulllength), performing quality control and denoising sequences with UNOISE2 algorithm (-cluster_unoise) (Edgar 2016), chimera removal (--uchime3_denovo) (Rognes et al. 2016), and mapping to ASVs with 100% similarity threshold (-usearch_global).The Quantitative Insights Into Microbial Ecology (QIIME2 2020.6)pipeline was employed to process the OTUs and sequencing data, as previously described (Bolyen et al. 2019).OTU taxonomic classification was conducted by searching the representative sequences set against the SILVA138 database (Quast et al. 2012).
Furthermore, the OTU table is generated to record the abundance of each OTU in each sample and the classification of these OTUs.All samples were discarded, which contained less than 0.001% OTU of the total sequence.In order to minimize the difference in sequencing depth among samples, an average, rounded, thin OTU table is generated by averaging 100 uniformly resampled OTU subsets at the minimum sequencing depth of 90% for further analysis.

Bioinformatics and Statistical Analysis
Sequence analysis is mainly carried out by using QIIME2 and R software packages (version 3.2.0).The OTU table in QIIME2 is used to calculate OTU-level α diversity index, such as Chao1 richness estimate, ACE measure (coverage estimate based on abundance), PD_whole_tree, Shannon diversity index, and Simpson index.An ordered abundance curve at OTU level is generated to compare the quantity and uniformity of OTUs among samples.UniFrac distance measurement was used for β diversity analysis to investigate the structural changes of microbial communities in the models.It was visualized by principal coordinate analysis (PCoA) and non-metric multidimensional scale (NMDS).Student's T-test and Monte Carlo permutation test with 1000 permutations were used to determine the difference of UniFrac distance between groups and visualized by box diagram.Principal component analysis (PCA) is also based on the composition profile of the genus level.The R package "vegan" was used to evaluate the significance of microbial community structure differences among groups by PER-MANOVA (displacement multivariate analysis of variance).Venn diagram is generated by using the R package Venn diagram according to the occurrence of OTU in samples/ groups (regardless of its relative abundance) to show the shared and unique OTU between samples or groups.
The abundance of taxa at the phylum, class, order, family, genus, and species in a sample or group was statistically compared by Kruskal-Wallis test in the R stats software package.LEfSe (Linear Discriminant Analysis Effect Size) is executed with default parameters to detect the difference between different groups and enrich the classification groups.Random forest analysis is used to distinguish samples from different groups using the R package "Random Forest," which has 1000 trees and all default settings.Ten times cross-validation is used to estimate the generalization error.The expected "baseline" error is also included, which is obtained by simply predicting the classifiers of the most common category labels.Co-occurrence analysis was carried out by calculating Spearman rank correlation among dominant taxonomic groups.
Based on high-quality sequences, the microbial function was predicted by PICRUSt 2 (phylogenetic investigation of the community by reconstructing unobserved state, https:// github.com/ picru st/ picru st2/).The bacterial function database and gut-brain module (GBMs) were also analyzed by using the human gut metabolism module.For each sample, the function is analyzed using OMixer-RPM version 1.0 (https:// github.com/ raesl ab/ omixer-rpm) with KO redundancy predicted from PICRUSt 2.

Proteinomics Analysis
The blood serum of patients hospitalized in the Second Affiliated Hospital of Shandong First Medical University in September 2022 was collected, including 2 patients with hypertension complicated with ischemic stroke and only 3 patients with hypertension.Two-microliter serum sample was added to a 96-well plate, and further diluted with 0.1 M NH 4 HCO 3 (Sigma, MO, USA), 6 M urea (Sigma, MO, USA), and 2 M thiourea (Sigma, MO, USA) with lysis buffer containing the following components.The disulfide bond was reduced at 33 ℃ for 45 min at 600 rpm, and 10 mM tris(2-carboxyethyl) phosphine hydrochloride (TCEP) was added and alkylated with iodoacetamide.All crude protein extracts were precipitated with acetone at 20 ℃ for 2 h, and then re-dissolved with 200 µl TEAB buffer.Then, it was digested with trypsin at 37 ℃ overnight, and then, TFA was added to extinguish it.The Sep-Pak column of Waters Co was used to desalt these peptides.
Thermo Proteome Discoverer (PD, 2.4.1.15)was used to carry out feature detection, database search, and protein/polypeptide quantification on all original documents.The identity of MS polypeptide sequence and protein was determined by matching fragment patterns in the UniProt mouse database (downloaded on March 23, 2021, containing 17,040 sequences).Aminomethylation of cysteine residues is set as fixed modification.There are only two missing cleavages per peptide.The fragment quality tolerance is 0.02 Da.Label-free quantification (LFQ) was calculated in each parameter group.Select unique polypeptide and razor polypeptide for protein quantification.Mass spectrometry proteomics data has been stored in the ProteomeXchange consortium through Pride partner repository with data set identifier PXD030026.Bioinformatics analysis was carried out using David, Cytoscape, and OmicStudio tools.Only protein with p value < 0.05 was included in GO concentration analysis and network formation.
Analysis of Ischemic Stroke Score NIHSS and mRS scores of 7 stroke patients were collected and compared with the relative abundance of intestinal microorganisms.Spearman correlation test was used to judge the correlation between intestinal microbiota and ischemic stroke.

Research Object
This experiment enrolled the inpatients in the Department of Applied Neurology and Cardiology of the Second Affiliated Hospital of Shandong First Medical University from 2021 to 2022, including 72 patients with hypertension complicated with ischemic stroke and 46 patients with hypertension only.The NIHSS of ischemic stroke patients before and after treatment was recorded.
The admission criteria of the hypertensive stroke group are as follows: (1) age over 18 years old; (2) the clinical diagnosis is acute cerebral infarction, and the onset time is less than 10 days.The exclusion criteria are as follows: (1) connective tissue disease; (2) diabetes; (3) suffering from acute and chronic digestive diseases; (4) patients who are delirious and unable to collect stool samples correctly.The admission criteria of the hypertension control group include (1) age over 18 years old.The exclusion criteria are as follows: (1) previous history of cerebral infarction; (2) connective tissue disease; (3) diabetes.The blood lipid, blood sugar, blood routine, and other test indexes were recorded.The clinical information of the experimental group and the control group was collected, including gender, age, current medical history, and past history.This study was approved by the ethics committee of our hospital, and all participating researchers were informed.

Detection Methods
After fasting for 12 h, 5 ml venous blood was collected at 7-9 am the next morning; serum was centrifuged, and seven blood lipids were analyzed by immunoturbidimetry.BECK-MAN COULTER AU5800 automatic biochemical analyzer is adopted, and the operation procedures are strictly followed.The daily maintenance and quality control of the instrument are well done.

Statistical Analysis
SPSS25.0 software is used, and the data is expressed by the mean standard deviation.If the data is normal, two independent samples T-test is adopted; a nonparametric ranksum test is adopted.

ELISA Detection
This experiment enrolled the inpatients in the Department of Applied Neurology and Cardiology of the Second Affiliated Hospital of Shandong First Medical University in 2022, including 6 hypertensive patients with ischemic stroke and 4 hypertensive patients.
The admission criteria of the hypertensive stroke group are as follows: (1) age over 18 years old; (2) the clinical diagnosis is acute cerebral infarction, and the onset time is less than 10 days.The exclusion criteria are as follows: (1) connective tissue disease; (2) diabetes; (3) suffering from acute and chronic digestive diseases; (4) patients who are delirious and unable to collect stool samples correctly.The admission criteria of the hypertension control group include (1) age over 18 years old.The exclusion criteria are as follows: (1) previous history of cerebral infarction; (2) connective tissue disease; (3) diabetes.After fasting all night (at least 10 h), blood samples were collected from the subjects' anterior elbow vein from 7:30 am to 9:00 am.The collected blood was coagulated at 4 ℃ for 30 min and then centrifuged at 3000 rpm at 4 ℃ for 20 min.The analytical samples were immediately stored at − 20 ℃ to avoid repeated freeze-thaw cycles.ELISA kit (Jiangsu Enzyme-free Industry Co., Ltd., Jiangsu, China, catalog number E1766H1) was used to determine the concentration of APOC2 and APOC3, and the operation was carried out according to the manufacturer's plan.

Analysis of α Diversity and β Diversity of Intestinal Microorganisms in the Hypertension Group and Hypertension Complicated with Stroke Group
A total of 1372 OTUs were obtained from 14 samples, including 1013 OTUs in both groups, 100 unique OTUs in the hypercholesterolemic group, and 299 unique OTUs in the hypertensive group with stroke, which indicates that compared with the hypertensive group, the hypertensive group with stroke has more specific OTU (Fig. 1A).The dilution curve shows that the curve tends to be flat with the increase of sequencing quantity, which indicates that the sequencing quantity of the sample is sufficient (Fig. 1B).In this study, we used different multi-α diversity indexes to evaluate the difference in species richness and evenness between the two groups.The results showed that (Fig. 1C) there was no significant difference in Chao 1, Shannon, and Simpson indexes between the hypertension group and hypertension complicated with stroke group (p > 0.05).Compared with hypertensive patients, the richness and uniformity of intestinal microbial species in hypertensive patients with stroke did not change significantly.
According to Anosim analysis (Fig. 1D), there was no significant difference between and within the hypertension group and hypertension complicated with the stroke group (R < 0, p > 0.05).To compare the similarities or differences between the two groups of samples, we conducted PCoA analysis and inter-group NMDS analysis.According to the results (Fig. 1E-H), there was no apparent aggregation between the same group of samples, so there was no noticeable difference in species composition between the two groups.

Analysis of Intestinal Microbiota Difference Between Hypertension Group and Hypertension Complicated with Stroke Group
At the phylum level, the first three groups of dominant bacteria in the two groups are Bacteroides, Firmicutes, and Proteobacteria (Fig. 2A).At the level of two genera, the top three genera with the highest abundance in the hypertension group are Prevotellaceae, Bacteroidaceae, and Lachnospiraceae.Bacteroidaceae, Lachnospiraceae, and Ruminococcaceae are the top three with the highest abundance in the hypertensive stroke group (Fig. 2B).The cluster heat map (Fig. 2C, D) of species composition between different species and between groups shows the species with high abundance in the sample more intuitively.It represents the trend of relative abundance of species in the hypertension group and hypertension complicated with the stroke group at the genus level.Post-cluster thermogram showed that Bacteroides had the highest abundance, followed by Faecalibacterium.LEfSe analysis can find the species with significant differences in abundance among groups.The experimental results show that (Fig. 2E, F), at the order level, compared with the hypertension group, Bacteroides, UCG_009, and Eisenbergiella have significantly increased relative abundance in the hypertension complicated with the stroke group respectively; the genera with relatively significantly reduced abundance are Ruminococcus_gnavus_group, Sutterellaceae, Burkholderia, and Prevotella, and the LDA score of Prevotella is 4.34 (p < 0.05), which indicates that there are significant differences.

Correlation Between the Relative Abundance of Prevotella and Disease Severity in Patients Before and After Stroke Treatment
Firstly, the sex and age of the two groups were compared, and there was no significant difference (Table 1).NIHSS and mRS scores of 7 stroke patients were collected and compared with the relative abundance of intestinal microorganisms by Spearman correlation test (Table 2).Results show there was no significant correlation between Prevotella and mRS score.The relative abundance of Prevotella in the intestinal microbiota of patients with ischemic stroke was negatively correlated with its post-treatment NIHSS value (r s = − 0.757, p = 0.049), further illustrating that the absence of Prevotella makes the condition more complex (Table 3).

Analysis of the Difference in Intestinal Microbiota Function Between the Hypertension Group and Hypertension Complicated with Ischemic Stroke Group
Based on the functional difference analysis of the GMM metabolic module between the two groups (Fig. 3A), the results showed that the difference in tryptophan degradation (p = 0.009, p < 0.01) and lysine I degradation (p = 0.044, p < 0.05) in hypertension combined with stroke group was significantly higher than those in hypertension group.Tryptophan can be hydrolyzed by tryptophanase into indole, ammonia, and pyruvic acid, the latter being the critical precursor to form acetyl-CoA, which is used to generate energy through the citric acid cycle (Bellmaine et al. 2020).The first step of the lysine catabolism pathway is to form a sugar crown base, followed by 2-aminoadipic acid.Through decarboxylation, 2-aminoadipic acid is converted into glutaryl-CoA and crotonyl-CoA, and finally oxidized into acetyl-CoA, which enters the tricarboxylic acid cycle (Matthews 2020).To sum up, the intestinal microbiota of hypertensive patients with stroke is significantly higher than that of hypertensive patients in providing energy for the body.
The prokaryotic function of culturable bacteria between the two groups was predicted by FAPROTAX (Fig. 3B).The results showed that the reduction effect of nitric acid in the hypertension group was significantly higher than that in hypertension complicated with the ischemic stroke group (p = 0.035).Studies have shown that nitric acid reduction can dilate blood vessels (Sindler et al. 2014).
Based on the functional difference analysis of the Megacycle metabolic pathway database (Fig. 3C), the results showed that in the hypertension complicated with stroke group, there were super pathways of lipopolysaccharide biosynthesis (p = 0.0379), butyric acid II produced by acetyl-CoA fermentation (p = 0.0379), keto gluconic acid metabolism (p = 0.002), and butyl acetate produced by L-lysine fermentation (p = 0.002).Bacterial lipopolysaccharide, also known as endotoxin, is an important component of the outer membrane of the cell wall of Gram-negative bacteria, which can produce a variety of inflammation-related factors and participate in inflammatory reactions.Studies have shown that keto gluconic acid metabolism can enhance the toxicity of some bacteria, such as Streptococcus pneumonia (Hu et al. 2019).To sum up, through functional prediction, hypertension with stroke group may be stronger than The dilution curve shows that the curve tends to be flat with the increase of sequencing quantity, which indicates that the sequencing quantity of the sample is sufficient.c Comparison of Chao1, Shannon, and Simpson indexes of intestinal microorganisms between groups.Wilcoxon test showed that p(Chao1) = 0.7104 and p(Shannon, Simpson) = 0.1649, both of which were greater than 0.05, indicating that there was no obvious difference in species richness and uniformity between the two groups, that is, there was no obvious difference in α diversity.d Anosim test results of β diversity between the two groups.Anosim test is a nonparametric test to compare the statistical differences between the two groups.The upper picture shows the unweighted distance Anosim analysis, p = 0.93, and the lower picture shows the weighted distance Anosim analysis, p = 0.524, suggesting that there is no significant difference between and within the hypertension group and hypertension complicated with the cerebral infarction group.e Inter-group NMDS (non-metric is mostly calibration) analysis.It shows that there is no obvious difference in β diversity between the two groups, and the stress function value is 0.0233.f-h Beta diversity of intestinal microbes among groups was based on PCoA results of Bray-Curtis distance.Figure 3

Changes of Serum Protein in Hypertensive Patients with Ischemic Stroke
We collected the serum of 2 patients with hypertension complicated with stroke and only 3 patients with hypertension.On the platform of proteomics, we used Orbitrap Eclipse mass spectrometer and Easy-NLC 1200 system to collect the data.The Thermo Proteome Discoverer (PD, 2.4.1.15)was used to detect the features of all the original documents, search the database, and quantify protein/polypeptide.The hierarchical cluster analysis (Fig. 4A, C) of 131 LFQ protein intensities showed 89 up-regulated genes and 51 downregulated genes.APOC2 and APOC3 were expressed in the hypertension complicated with the stroke group.The sample size of proteomics experiment is small, and the conclusion may not be representative, so we improved the ELISA test to further verify the results.
To further verify the results of the proteomic analysis, we selected the serum of 6 patients with hypertension complicated with cerebral infarction and four patients with hypertension only, and compared APOC2 and APOC3 of the two groups by ELISA.The results show that the level of APOC2 (p = 0.043) and APOC3 (p = 0.002) in hypertension complicated with stroke is significantly higher than that in the hypertension group (Fig. 4D, E).

Changes of Hematological Indexes in Patients with Hypertension Complicated by Stroke
To further explore the difference of related genes in the cholesterol metabolism pathway between the two groups, we selected 72 hypertensive patients with cerebral infarction and 46 hypertensive patients only, collected 5 ml venous blood at 7-9 am, centrifuged the serum, and made seven blood lipid analyses by immunoturbidimetry (Table 4).The results in the table show the changes.The results show that the level of APOA1 in hypertension complicated with stroke is significantly higher than that in the hypertension group (p = 0.040).Still, there is no significant difference in blood glucose (GLU), total cholesterol (CHOL), triglyceride (TG), high-density lipoprotein (HDL), low-density Fig. 2 Abundance analysis of intestinal microbial species in the hypertension group and hypertension complicated with cerebral infarction group.a The relative abundance between the two groups at the phylum level.The first three groups of dominant microbiota in the two groups at the phylum level are Bacteroides, Firmicutes, and Proteobacteria.b It shows the relative abundance between the two groups at the genus level, and the top three with the highest abundance in the hypertension group are Prevotellaceae, Bacteroidaceae, and Lachnospiraceae.The highest abundance of Bacteroidaceae, Lachnospiraceae, and Ruminococcaceae in the hypertensive cerebral infarction group.c A cluster heat map showing the horizontal species composition of each species.Each small square represents the species corresponding to each sample, and its color represents the content of the species (red represents high abundance and blue represents low abundance).Each row represents the expression of each species in different samples, and each column represents the expression of different species in each sample.It represents the trend of relative abundance of species in the hypertension group and hypertension complicated with the cerebral infarction group.d Post-cluster thermogram showed that Bacteroides had the highest abundance, followed by Enterobacter.e Evolutionary tree of species markers in each group of LEfSe, in which different colors represent significant difference species between different samples or groups.f Species markers of LEfSe groups.Compared with the Kruskal-Wallis rank-sum test, linear discriminant analysis (LDA) determined the most characteristic species between the two groups at the genus level (LDA score > 2, p < 0.05), and the abundance of Bacteroides, UCG_009, and Eisenbergiella was higher in the hypertension group with cerebral infarction.In the hypertension group, the abundance of Ruminococcus_gnavus_group, Sutterellaceae, Burkholderia, and Prevotella was higher, and the LDA score of Prevotella was less than − 4, which suggested that there was a significant difference.Some literature showed that Prevotella was related to cholesterol metabolism

Analysis of Therapeutic Effect on Ischemic Stroke
To evaluate the therapeutic effect, 72 patients with ischemic stroke were scored, and it was found that there was a significant difference in NIHSS score before and after treatment (p < 0.01).And we found that the NIHSS scores of speech and exercise were also significantly different before and after treatment (p < 0.01) (Table 5).

Discussion
The number of microorganisms in the human intestine is > 10 13 , including bacteria, viruses, fungi, and protozoa.There are more than 1000 kinds of bacteria in the intestine, and the intestinal microbiota plays a particularly important role in maintaining the health of the host (Gomaa 2020).For example, Treg cells in the intestine maintain the antiinflammatory environment of the intestine by inhibiting the differentiation of T helper (Th) 17 cells and the proliferation of γδ T cells (Benakis et al. 2016).At the same time, intestinal microbes can also lead to a variety of diseases, such as inflammatory bowel disease (IBD), inflammatory skin diseases (such as psoriasis and atopic dermatitis), autoimmune arthritis, type 2 diabetes, obesity, and atherosclerosis (Singh et al. 2017).Studies have also shown that intestinal microbes are related to ischemic stroke.After a stroke, the ecological imbalance of intestinal microbes leads to the increase of intestinal permeability and the activation of the intestinal immune system, which in turn leads to heterotopia of intestinal bacteria and pro-inflammatory cells, which enter the brain tissue through the damaged blood-brain barrier (Hu et al. 2022).The intestinal microbiome also promotes the development of cerebral atherosclerosis by producing intermediate metabolites, including trimethylamine oxide (TMAO), lipopolysaccharide (LPS), phenylacetylglutamine (PAGln), and reduced butyrate (Chen et al. 2020a).According to the known risk factors and chronic diseases, the subjects with no previous stroke history and older than 60 years old were divided into low-risk, medium-risk, and high-risk groups by researchers such as Zeng.The results showed that the enrichment of conditionally pathogenic bacteria (such as Enterobacteriaceae and Vironidae) and lactic acid-producing bacteria (such as Bifidobacterium and Lactobacillus) in the high-risk group was significantly higher than that in the low-risk group.In contrast, butyric acid-producing bacteria (such as Lactobacillus and Lactococcus) decreased relatively (Zeng et al. 2019).Li et al. found that the α diversity and structure of intestinal microbes in ischemic stroke patients were similar to those in healthy people.Still, many bacteria could produce short-chain fatty acids in the intestinal microflora of stroke patients, including Corynebacterium, Ackerman, Ruminococcus UCG005, and Brucella (Li et al. 2019).
Zhang and other researchers found that the abundance of Pseudomonas aeruginosa, Coccisphinctenophilus, and Klebsiella Adler in patients with cerebral infarction was significantly higher than that in healthy people, and the levels of IL-4, tumor necrosis factor-β (TN-β), IL-1β, and C-reactive protein in patients with cerebral infarction were significantly higher than those in healthy people.This experiment showed that intestinal microbiota played a vital role in inducing cerebral infarction, but the specific mechanism was unknown (Zhang et al. 2020).
There are more and more articles about the relationship between ischemic stroke and intestinal microbes, and some of them have studied its pathogenesis from the perspective of essential medicine.Studies have shown that intestinal symbiotic microorganisms can favorably regulate regulatory T Table 3 Correlation between the relative abundance of Prevotella and disease severity in patients before and after stroke treatment r s , Spearman's rank order correlations test.The Spearman correlation test showed that the relative abundance of Prevotella in the intestinal microbiota of patients with ischemic stroke had a negative correlation with the NIHSS value after treatment (r s = − 0.757, p = 0.049).
There was no significant correlation between the relative abundance of Prevotella and mRS (r s = − 0.299, p = 0.515).There was no significant correlation between the relative abundance of Prevotella and NIHSS value before treatment (r s = − 0.356, p = 0.434) *p < 0.05 and **p < 0.01 compared with the sham group Variable r s p NIHSS before treatment − 0.356 p = 0.434 NIHSS after treatment − 0.757 p = 0.049* mRS − 0.299 p = 0.515 Fig. 3 Hypertension group and hypertension complicated with cerebral infarction group have different functions due to different microbiota species composition.a Analysis of species and functional differences between two groups based on GMM metabolic module.Turkey test showed that there were significant differences in tryptophan degradation (p = 0.009) and lysine I degradation (p = 0.044) between the two groups.b The prokaryotic function of culturable bacteria between the two groups was predicted by FAPROTAX.Using a nonparametric rank-sum test, the results show that there are obvi-ous differences in nitric acid reduction between groups (p = 0.035).Some literature shows that nitric acid reduction can cause vasodilation.c Analysis of species and functional differences between groups based on the Megacycle metabolic pathway database.Turkey test showed that there were obvious differences between the two groups in the super pathway of lipopolysaccharide biosynthesis (p = 0.0379), butyric acid II produced by acetyl-CoA fermentation (p = 0.0379), keto gluconic acid metabolism (p = 0.002), and butyl acetate produced by L-lysine fermentation (p = 0.026) cells (Treg) and γδ T cells, both related to cerebral ischemia injury.γδ T cells, a significant lymphocyte group with innate immune function, located on the epithelial surface, can aggravate ischemic brain/injury by secreting IL-17 and generating chemotactic signals for peripheral myeloid cells.
Effector T cells contribute to ischemic injury, but Treg cells can promote neuroprotection by down-regulating inflammation after ischemia.After the acute stage of ischemic stroke, Treg cells appear in ischemic tissues and play a neuroprotective role by secreting the anti-inflammatory cytokine IL-10.Treg cells in the intestine are essential to maintain the anti-inflammatory environment of the intestine by inhibiting the differentiation of T helper (Th) 17 cells and the proliferation of γδ T cells (Benakis et al. 2016).Singh et al. made a mouse model of massive cerebral infarction (fMCAo) and intervened and detected it.After FMCAo, the microflora was out of balance, and Th1 and Th17 responses, which were mainly pro-inflammatory, were induced, which was related to the increase of infarct volume.However, the expression of interferon-β and IL-17 cytokines, markers of Th1 and Th17 T cell polarization (Singh et al. 2016), significantly increased in mice treated with fMCAo.Hypertension is the most important variable factor leading to ischemic stroke.The incidence of cerebral infarction in hypertensive patients is increasing year by year.So what kind of intestinal microbiota of hypertensive patients is likely to lead to acute cerebral infarction?Or what is the specific mechanism by which hypertension increases the incidence of acute cerebral infarction through intestinal microorganisms?
In this study, by comparing the composition and function of intestinal microflora between hypertensive patients and hypertensive patients with ischemic stroke, we found that there was no noticeable difference in α diversity and β diversity of intestinal microorganisms between the two groups.However, there were still some different genera at the order level.Compared with hypertensive patients with cerebral infarction, the relative abundance of Bacteroides, UCG_009, and Eisenbergiella was significantly increased.The genera with relatively significantly reduced abundance are Rumi-nococcus_gnavus_group, Sutterellaceae, Burkholderia, and Prevotella, and the LDA score of Prevotella is < − 4, which indicates that there are significant differences.
After a stroke, the intestinal microbiota of patients with cerebral infarction changed significantly, with a significant decrease in resident bacteria, such as Prevotella and Bacteroides (You et al. 2018).Some studies have found that patients with stroke and transient ischemic attack have fewer commensal or beneficial genera, including Bacteroides, Prevotella, and Faecalibacterium (Yin et al.2015).Many studies have confirmed that the occurrence of ischemic stroke was negatively correlated with the abundance of Prevotella in the intestine.Our research further verified this point and found that it still applies to hypertensive people.
Prevotella is a Gram-negative obligate anaerobic bacterium belonging to Bacteroidetes, one of the largest bacterial phylum known to be the main branch of Bacteroides (Accetto et al. 2021).It can appear in many parts of the human body, such as the skin, oral cavity, vagina, and gastrointestinal tract.Prevotella is associated with many diseases, especially autoimmune diseases, oral infections, or other infections (Tett et al. 2021).We have found a negative correlation Fig. 4 Quantitative proteomic analysis shows different characteristics of protein in different populations.a The hierarchical cluster analysis of 131 LFQ protein intensities (Log2) showed that 89 genes were up-regulated and 51 genes were down-regulated, which represented the gene expression trend in the hypertension group and hypertension complicated with the cerebral infarction group.The abundance between genomes above the dotted line is p < 0.05, the horizontal axis is disease grouping, and the vertical axis is a differential gene, and its color indicates the content of the species (red is a relatively high expression gene, blue is relatively low expression gene).b KEGG analysis of gene pathways between groups showed that there were genes involved in cholesterol metabolism.c Volcano map of relative abundance of differential gene expression between the two groups.The gene with p < 0.05 (the part above the dotted line) is considered to be significantly differentially expressed.Scattered points describe the difference scores of p values of these items after − log10 conversion.Red indicates the genes that are up-regulated compared with hypertension, and blue indicates down-regulated genes.It can be seen from the figure that the expression of APOC2 and APOC3 in cerebral infarction with hypertension is significantly higher than that in hypertension.d, e Quantitation of APOC2 and APOC3 levels, as measured using ELISA.The normal distribution test showed that APOC3 was normal, and two independent samples T-test was adopted, p < 0.005, indicating that there was a significant statistical difference in APOC3.However, APOC2 does not conform to the normal distribution.The nonparametric rank-sum test shows that p (APOC2) is < 0.05, indicating that there is a significant statistical difference in APOC2 between the two groups, *p < 0.05 and **p < 0.01 ◂ Table 4 Comparison of blood lipid and blood glucose between hypertensive patients and hypertensive patients with ischemic stroke The normal distribution test showed that low-density lipoprotein, apolipoprotein B, and free fatty acid were normal, and two independent samples T-test was adopted, p > 0.05, indicating that there was no significant statistical difference in low-density lipoprotein, apolipoprotein B, and free fatty acid between the two groups.However, alkaline phosphatase (ALP), blood glucose (glucose), TC, TG, HDL, APOA1, LP, and HCY do not conform to the normal distribution.The nonparametric rank-sum test shows that p (apolipoprotein A1) is < 0.05, indicating that there is a significant statistical difference in apolipoprotein A1 between the two groups, and the remaining p values are all greater than 0.05, with no significant statistical difference between the abundance of Prevotella and the condition of patients with ischemic stroke after treatment; that is, the absence of Prevotella may lead to a worse prognosis, but the specific mechanism is still unclear.Studies have shown that Prevotella can regulate blood lipid, and reduce weight and cholesterol (Liang et al. 2021;Guevara-Cruz et al. 2019;Medina-Vera et al. 2019;Chen et al. 2021;Christensen et al. 2019;Eriksen et al. 2020;Xia et al. 2017).By feeding Sprague-Dawley rats with a whole-grain diet (WGQ), some researchers found that the levels of serum total cholesterol, low-density lipoprotein cholesterol, and non-high-density lipoprotein cholesterol in rats decreased, and the abundance of Prevotella was significantly increased (Xia et al. 2017).
Studies have shown that Prevotella can regulate blood lipids, and reduce weight and cholesterol.Christensen and other researchers found a negative correlation between the abundance of Prevotella and weight change through a 6-week, parallel, and random experiment on overweight people (Christensen et al. 2019).Chen and other researchers found that several risk factors (waist circumference, body mass index, diastolic blood pressure, systolic blood pressure, fasting blood glucose, glycated hemoglobin, total cholesterol, triglyceride, and low-density lipoprotein cholesterol) and inflammatory markers (white blood cell count and absolute value of neutrophils) in carotid atherosclerosis (CAS) group were significantly higher than those in the control group.However, the abundance of Prevotella in the control group was significantly higher than that in the CAS group, which indicated that the abundance of Prevotella was negatively correlated with hyperlipidemia (Chen et al. 2021).There is more and more evidence that Prevotella has the function of lowering blood lipid, but the specific mechanism is still unclear.
We can consider whether it is because of the lack of Prevotella in the intestines of some patients with hypertension, which leads to uncontrolled blood lipids, thus leading to cerebral atherosclerosis and finally leading to ischemic stroke.Therefore, we compared serum proteomics between hypertensive patients and hypertensive patients with ischemic stroke and found that there are 89 up-regulated genes and 51 down-regulated genes, and 8 genes involved in cholesterol metabolism, namely APOA1, APOA2, APOA4, and APOC1.
To further verify the hypothesis, we compared the blood biochemical indexes of patients in the hospital and found that there was a significant difference in APOA1 between the two groups.Then, we selected APOC2 and APOC3 with obvious differences in proteomics results for ELISA.In the end, we concluded that the lack of Prevotella in the intestines of some patients with hypertension caused the blood lipid level not to be adjusted in time, which promoted the development of atherosclerosis, and finally led to the increased risk of ischemic stroke.However, there are limitations to this study.The small experimental cohort might not accurately reflect the enormous patient population.
Second, even though we discovered a potential connection between Prevotella and blood lipids, we did not investigate the precise mechanism.This experiment provided insights for preventing the occurrence of ischemic stroke, enhanced understandings of intestinal microbes, explored its mechanism, and finally suggested a role of intestinal microbes in disease development to the clinic.We found that almost all patients with ischemic stroke hospitalized in the Second Affiliated Hospital of Shandong First Medical University took citicoline orally, and the NIHSS changed significantly before and after treatment.Citicoline is the generic name of the pharmaceutical substance that chemically is cytidine-5′-diphosphocholine (CDP-choline), which is identical to the natural intracellular precursor of phospholipid phosphatidylcholine.Following injection or ingestion, citicoline is believed to undergo quick hydrolysis and dephosphorylation to yield cytidine and choline, which then enter the brain separately and are used to resynthesize CDP-choline inside brain cells.The neuroprotective activity of citicoline has been repeatedly shown in preclinical models of brain ischemia and trauma (Grieb 2014).Citicoline makes the conversion of choline in citicoline to trimethylamine (TMA) and its putative atherogenic N-oxide (TMAO) less likely than the choline moiety in other food sources such as phosphatidylcholine (Synoradzki and Grieb 2019).TMAO is closely associated with the risk of cardiovascular adverse events (Tang et al. 2013).Studies have shown that the absence of Prevotella in the intestinal microbiota of patients with ischemic stroke leads to the increase of TMAO, leading to the occurrence of cerebrovascular accidents (Yin et al. 2015).We speculated that the application of citicoline is made up for the negative effect caused by the absence of beneficial microflora such as Prevotella.However, whether the two are related still needs further investigation.
In addition, we found that Prevotella produces shortchain fatty acids (SCFA) by decomposing carbohydrates.SCFA can not only provide energy for intestinal epithelial cells, but also affect the production of mucin, and physiologically affect the blood flow of colonic mucosa, the interaction between liquid and electrolyte, the autonomic nervous system, and the secretion of intestinal hormones (Ohira et al. 2017).Studies have shown that the increase of circulating SCFA is related to the direct decrease of adipocyte and adipogenesis (Morrison and Preston 2016), and the regulation of blood lipids is helpful to slow down the occurrence and development of atherosclerosis.Studies have also shown that SCFAs directly or indirectly regulate microbiome-gut-brain interactions, including immune, endocrine, neural, and humoral pathways (Dalile et al. 2019).Studies have found that ischemic stroke reduces the level of SCFAs in the intestine, and transplantation of fecal microflora rich in these metabolites is an effective way to treat the disease.Butyric acid supplementation can effectively treat the ischemic stroke model by remodeling the intestinal microbiome, enriching beneficial lactobacillus, and repairing leaky intestines (Chen et al. 2020b).In our study, the prevalence of Prevotella in the intestines of hypertensive patients is significantly higher than that of ischemic stroke, which indicates that the lack of SCFA caused by different microbiota may be an important factor in the occurrence of ischemic stroke in hypertensive patients.
When FAPROTAX technology was used to predict the prokaryotic function of bacteria in patients, we found that the reduction effect of nitric acid in hypertensive patients was significantly higher than that in hypertension complicated with stroke.Some studies have shown that the reduction effect of nitric acid can dilate blood vessels (Sindler et al. 2014), so we suspect that this may be due to different intestinal microbiota, which leads to different degrees of vasodilation, and ultimately leads to different risks of ischemic stroke in hypertensive people.This conjecture still needs further experimental verification.

Fig. 1
Fig.1Diversity of intestinal microbiota in the hypertension group and hypertension complicated with cerebral infarction group.a Wayne diagram of OTU comparison between groups.There are 1013 identical OTUs between the two groups.b The dilution curve shows that the curve tends to be flat with the increase of sequencing quantity, which indicates that the sequencing quantity of the sample is sufficient.c Comparison of Chao1, Shannon, and Simpson indexes of intestinal microorganisms between groups.Wilcoxon test showed that p(Chao1) = 0.7104 and p(Shannon, Simpson) = 0.1649, both of which were greater than 0.05, indicating that there was no obvious difference in species richness and uniformity between the two groups, that is, there was no obvious difference in α diversity.d Anosim test results of β diversity between the two groups.Anosim test is a nonparametric test to compare the statistical differences between the two groups.The upper picture shows the unweighted distance Anosim analysis, p = 0.93, and the lower picture shows the weighted distance Anosim analysis, p = 0.524, suggesting that there is no significant difference between and within the hypertension group and hypertension complicated with the cerebral infarction group.e Inter-group NMDS (non-metric is mostly calibration) analysis.It shows that there is no obvious difference in β diversity between the two groups, and the stress function value is 0.0233.f-h Beta diversity of intestinal microbes among groups was based on PCoA results of Bray-Curtis distance.Figure3represents three different grouping factors.Figure F shows PC1-PC2, Figure G shows PC1-PC3, and Figure H shows PC2-PC3.Each dot in the figure represents a sample, and the dots of the same color come from the same grouping.The similarity of distance samples.PERMANOVA statistical method was used to test that the p value (group significance) of PC1-PC2 = 0.556, R value (explanatory degree) = 0.0677, p value (group significance) of PC1-PC3 = 0.587, R value (explanatory degree) = 0.0677, and p value of PC2-PC3 = 0.577 Fig.1Diversity of intestinal microbiota in the hypertension group and hypertension complicated with cerebral infarction group.a Wayne diagram of OTU comparison between groups.There are 1013 identical OTUs between the two groups.b The dilution curve shows that the curve tends to be flat with the increase of sequencing quantity, which indicates that the sequencing quantity of the sample is sufficient.c Comparison of Chao1, Shannon, and Simpson indexes of intestinal microorganisms between groups.Wilcoxon test showed that p(Chao1) = 0.7104 and p(Shannon, Simpson) = 0.1649, both of which were greater than 0.05, indicating that there was no obvious difference in species richness and uniformity between the two groups, that is, there was no obvious difference in α diversity.d Anosim test results of β diversity between the two groups.Anosim test is a nonparametric test to compare the statistical differences between the two groups.The upper picture shows the unweighted distance Anosim analysis, p = 0.93, and the lower picture shows the weighted distance Anosim analysis, p = 0.524, suggesting that there is no significant difference between and within the hypertension group and hypertension complicated with the cerebral infarction group.e Inter-group NMDS (non-metric is mostly calibration) analysis.It shows that there is no obvious difference in β diversity between the two groups, and the stress function value is 0.0233.f-h Beta diversity of intestinal microbes among groups was based on PCoA results of Bray-Curtis distance.Figure3represents three different grouping factors.Figure F shows PC1-PC2, Figure G shows PC1-PC3, and Figure H shows PC2-PC3.Each dot in the figure represents a sample, and the dots of the same color come from the same grouping.The similarity of distance samples.PERMANOVA statistical method was used to test that the p value (group significance) of PC1-PC2 = 0.556, R value (explanatory degree) = 0.0677, p value (group significance) of PC1-PC3 = 0.587, R value (explanatory degree) = 0.0677, and p value of PC2-PC3 = 0.577 ◂

◂ Table 1
Baseline characteristics of hypertensive patients and hypertensive patients complicated with stroke.There is no significant difference in age and gender between the two groups

Table 2
Severity of ischemic stroke and relative abundance of Prevotella in intestinal microbiota NIHSS National Institutes of Health Stroke Scale.The NIHSS was used to assess the severity of cerebral infarction in patients, and this score was positively correlated with the severity of ischemic stroke.mRS modulate RANK score.mRS is a scale used to evaluate the neurological functional recovery of patients with stroke

Table 5
Analysis of therapeutic effect on ischemic strokeThe statistics by the NIHSS of patients with ischemic stroke before and after treatment.The nonparametric rank-sum test shows that p is < 0.01, indicating that there are significant differences in NIHSS before and after treatment *p < 0.05 and **p < 0.01 compared with the sham group