CN114533024A

CN114533024A - Biphasic affective disorder biomarkers and uses thereof

Info

Publication number: CN114533024A
Application number: CN202111649172.8A
Authority: CN
Inventors: 胡少华; 来建波; 张佩芬; 泮艳梦; 蒋佳俊; 汤安英; 张旦华; 吴冕; 路静; 牟婷婷; 任非凡
Original assignee: Zhejiang University ZJU
Current assignee: Zhejiang University ZJU
Priority date: 2021-12-30
Filing date: 2021-12-30
Publication date: 2022-05-27
Also published as: WO2023123613A1

Abstract

The invention discloses a bipolar affective disorder biomarker and application thereof. Wherein the intestinal microbiome is analyzed by metagenomic sequencing, the serum metabolome is analyzed by non-targeted mass spectrometry, the brain function is analyzed by resting-state functional magnetic resonance imaging (rs-fMRI), and as a result, four intestinal microorganisms, Akkermansia muciniphila, Citrobacter spp, Yersinaceae spp and Enterobacter spp, are found to have remarkably reduced abundance in patients with bipolar affective disorder, and intestinal signals are transmitted to the brain through neuroactive metabolites. Neuroactive metabolites of microbial origin in serum including pantothenic acid, riboflavin, folic acid, pyridinol, kynurenic acid, GABA and SCFA are markedly dysregulated in bipolar disorder; the functional connectivity between the marginal zones of bipolar affective disorder is generally reduced and the microorganisms associated with serum neuroactive metabolites are also associated with specific brain networks, notably associated with the linguistic zone, thalamus and striatum, sensorimotor region, hippocampal formation and amygdala.

Description

Biphasic affective disorder biomarkers and uses thereof

Technical Field

The invention relates to the study of intestinal microbiology, serum metabonomics and brain functional omics of bipolar affective disorder, in particular to a bipolar affective disorder biomarker and application thereof.

Background

Bipolar Disorder (BD) is a major affective Disorder manifested primarily as recurrent depressive or (hypo) manic episodes as a clinical feature. According to the mental health survey statistics of the world health organization, the lifetime prevalence rate of BD is 2.4%, wherein the prevalence rate of BD-I is 0.6%, the prevalence rate of BD-II is 0.4%, and the prevalence rate of BD is 1.4%. The latest epidemiological survey report of mental diseases in China shows that the lifetime prevalence rate of BD is 0.6%. The suicide rate of BD patients is 20-30 times of that of healthy people, suicide risk is the first of all mental diseases, and great harm is brought to families and society of patients. Currently, clinical diagnosis of BD is mainly based on diagnostic criteria of symptomatology, and the rate of missed diagnosis and misdiagnosis are high. Therefore, the intensive elucidation of the pathogenesis of BD and the search for effective control strategies and new therapeutic targets are urgent problems to be solved.

Current research has found that BD is a systemic disease with changes in both intestinal flora and serum metabolism. More and more studies have shown that the microbial-gut-brain axis is involved in the pathogenesis of bipolar disorder. The gut flora may communicate bi-directionally with the central nervous system via the gut-brain axis. In one aspect, signals from the brain can affect physiological effects of the gut, including motility, secretion, and immune function. On the other hand, the information sent by the intestinal tract can generate the regulation and control effect on the function of the brain.

Blood and its metabolic profile act as a bridge for two-way communication between brain and intestinal microorganisms. Current research has found that the serometabolome of BD patients differs from that of healthy people and is characterized by alterations in specific amino acids, lipids, citric acid cycle, pathways involved in polyunsaturated fatty acid metabolism.

Bipolar disorder is currently generally considered to be the result of the combined action of genetic and environmental factors, but its diagnosis still depends on symptomatic assessment, lacking objective biomarkers. In view of the core role of the gut flora in host metabolism, comprehensive analysis of the gut microbiome and the serum metabolome can reveal not only the relationship between them, but also reflect the pathophysiological basis of BD.

Disclosure of Invention

In order to overcome the deficiencies of the prior art, it is an object of the present invention to provide bipolar affective disorder biomarkers and uses thereof. The present invention integrates multidimensional datasets of gut microbiome, serum metabolome and brain function connectivity based on clinical and questionnaire data from 100 depressed bipolar disorder patients and 49 healthy control groups matched by age, BMI (Body Mass Index) and gender. The gut microbiota was analyzed by shotgun metagenomic sequencing, the serum metabolome was analyzed by non-targeted mass spectrometry, and brain function was analyzed by whole brain resting state functional magnetic resonance imaging (rs-fMRI). The invention aims to research the intestinal-cerebral axis disorder of a patient suffering from bipolar disorder without medication from the aspect of omics, provides screening and application of comprehensive landscape and biomarkers of acute-phase bipolar depression based on the intestinal-cerebral axis, and provides guidance for pathological mechanism research, clinical recognition, development trend prediction and accurate treatment of the bipolar disorder.

The bipolar affective disorder biomarker, four gut microbiota, Akkermansia muciniphila, Citrobacter spp, Yersiniaceae spp and Enterobacter spp, is significantly reduced in abundance in patients with bipolar affective disorder, with gut signaling to the brain via neuroactive metabolites.

The bipolar disorder biomarker, the four gut microbiologically derived neuroactive metabolites in serum, including pantothenic acid, riboflavin, folic acid, pyridinol, kynurenic acid, GABA, and SCFA, are dysregulated in bipolar disorder patients;

disorders significantly associated with the language zones "network-3", "network-4", "network-10", thalamus and striatum "network-20", sensorimotor zone "network-16", hippocampal formation and functional connectivity of the amygdala "network-18" and "network-19" affect the language, mood and reward circuit of BD.

The screening method of the bipolar affective disorder biomarker is based on clinical and questionnaire data, 100 depression bipolar disorder patients and 49 health control groups are selected, the intestinal microbiota is analyzed through metagenome bomb sequencing, the serum metabolome is analyzed through non-targeted mass spectrometry, and the brain dysfunction is expressed through the functional connection of the whole brain resting state.

A kit prepared according to the bipolar affective disorder biomarker as a detection target or detection target, wherein the bipolar affective disorder biomarker comprises four intestinal microorganisms, Akkermansia muciniphila, Citrobacter spp.

The bipolar disorder biomarkers further comprise pantothenic acid, riboflavin, folic acid, pyridinol, kynurenic acid, GABA, and SCFA.

The kit is used for diagnosis of bipolar affective disorder or screening of a drug for treating bipolar affective disorder.

A biological preparation prepared from said bipolar affective biomarker for use in balancing or restoring bipolar affective biomarker in a patient with bipolar affective disorder.

The invention has the beneficial technical effects that:

the invention further provides a comprehensive view of the intestinal-cerebral axis in acute bipolar disorder depression by analyzing three aspects of intestinal microorganisms, serum metabolites and brain function connection conditions of bipolar disorder patients and healthy people, thereby screening biomarkers highly related to bipolar disorder and providing guidance for pathological mechanism research, clinical recognition, development trend prediction and accurate treatment of bipolar disorder.

The multigroup study provided by the invention has higher value for revealing the pathophysiology basis of the bipolar affective disorder and early diagnosis of diseases. First, bipolar disorder is a systemic disease, and the intestinal flora and serum metabolism are changed, and the correlation between the change of the intestinal flora and the etiology of bipolar disorder cannot be well revealed in the previous research because of the problems of limited sample size, inconsistent methodology and the like. Second, the availability, operability, and safety of stool samples, serum samples, and brain function MRI all ensure patient compliance. And the detection of the fecal sample is realized based on a macro-gene sequencing technology, and the detection of the serum sample is realized based on a non-targeted mass spectrometry technology, so that the obtained marker has higher sensitivity and specificity. Finally, the three-component assay of the present invention can also be used for clinical identification and precise treatment of bipolar affective patients.

Drawings

FIGS. 1a and 1b illustrate the analysis of the difference in microbial function between a bipolar disorder patient and a healthy control patient according to an embodiment of the present invention.

FIG. 2a is a graph showing the difference in relative abundance of bacterial populations of a bipolar disorder patient and a healthy control patient at the species level, according to one embodiment of the present invention.

FIG. 2b shows major species with significant differences in abundance between bipolar disorder patients and healthy controls, including significant decreasesYersiniaceae(5 kinds of the plants which are selected,Yersinia aleksiciae（GENOME143346），Yersinia frederiksenii (GENOME143407, GENOME143408), Yersiniaceae spp (GENOME231599), and Serratia spp. (GENOME232192)）， Akkermansiaceae(4 kinds of the chemical substances,Akkermansia muciniphila (GENOME113322, GENOME143211) and Akkermansia spp. (GENOME115272, GENOME142411)），Enterobacteriaceae(9 kinds of the plants with different molecular weights,Citrobacter freundii(GENOME103766)，Citrobacter werkmanii(GENOME231453)，Citrobacter spp. (GENOME000595)，Enterobacteriaceae spp. (GENOME177471)，Enterobacter cancerogenus(GENOME095996)，Enterobacter cloacae(GENOME143740、GENOME143746)，Enterobacter spp. (GENOME000057)andEnterobacter mori(GENOME000047)) ); 9 types of remarkable risesStreptococcaceae（Streptococcus mitis(GENOME058339、GENOME231431、GENOME231966)，Streptococcus oralis(GENOME176717)，Streptococcus pseudopneumoniae(GENOME207884)AndStreptococcus spp. (GENOME000595、GENOME207787、GENOME244205、GENOME207866) ）。

FIG. 3 is a graph showing the analysis of the changes in serum metabolites of a bipolar disorder patient and a healthy control patient according to an embodiment of the present invention. The graph shows that the bipolar affective patients have extensive metabolic changes, the disease states of the bipolar affective patients are main factors of the serum metabolic changes, and the academic type, the educational age and the sex also have statistically significant influence on the serum metabolites of the BD patients. The steroid hormone synthesis and tryptophan metabolic pathways in bipolar disorder patients were reduced compared to healthy controls.

FIG. 4 is an analysis of serum metabolites of bipolar disorder patients and healthy controls according to one embodiment of the present invention. The upper panel is rich in metabolites for healthy controls and the lower panel is rich in metabolites for BD patients.

FIG. 5 is a schematic representation of a "neuronal marker" metabolic pathway associated with gut flora, according to one embodiment of the present invention. By degrading dietary derived aromatic amino acids and pyruvate, gut microbes produce serum "neuronal markers". Enter blood through active and passive transport intestinal tract cells, and finally affect the brain through ways such as immunity, vage, endocrine and the like.

FIG. 6 is a graph of the relative abundance of key enzymes involved in the biosynthesis of "neuronal markers" in a species and their correlation with corresponding serum metabolites, according to one embodiment of the present invention. These abundances are significantly higher in BD patients and in BD-enriched microbial species, except vitamin B6 (pyridoxine).

FIG. 7 is a graph of the correlation between each "neuroactive metabolite" and the species abundance of the metabolic pathway/synthase encoding gene comprising that particular "neuroactive metabolite" based on a random forest model, identifying 154 species, according to one embodiment of the present invention.

FIG. 8 is a predictive model in which species associated with the production of "neuroactive metabolites" are closely correlated with quantitative measures of bipolar disorder severity (MADRS, HAMD, HAMA, and YAMARS), according to one embodiment of the present invention.

Fig. 9 is a graph of the performance of a random forest model evaluated for classification according to one embodiment of the present invention, for BD patients/healthy control classified subjects with an area under the working characteristic curve of 0.81.

FIG. 10 is a graphical representation of gut specific microbial function as a function of concentration of bipolar disorder associated metabolites based on Spearman's correlation analysis in accordance with one embodiment of the present invention. Functional analysis of the species showed that there was a significant difference in the functions of acetyl-coa metabolism, polyamine biosynthesis, cofactor and vitamin biosynthesis, aromatic amino acid metabolism between 15.4% BD-enriched species and 17.3% HC-enriched species, which probably reflected the interference of BD with the availability of these functions. These functions vary with the concentration of the corresponding BD associated serum metabolites.

Fig. 11a, 11b show the rs-fMRI brain features of patients with bipolar disorder, Principal Component Analysis (PCA), and the ROI-to-ROI connectivity matrix of BD was found to be significantly different from the healthy control group, according to an embodiment of the present invention.

FIGS. 12a, 12b are functional connectivity profiles of a critical cortical region, hippocampal and amygdala functional connectivity patterns (network-18 and network-19) showing a decrease in FC intensity between the limbic region of BD patient species and the DMN/DAN/language network (network-8, 9, 10, 12), in accordance with an embodiment of the present invention; FC enhancement of auditory (network-1, 7), language (network-3, 5, 8) and sensorimotor (network-16) regions, with abnormally high and low connectivity between the sensory (both auditory and speech) and subcortical regions.

FIGS. 13a, 13b, 13c, 13d are graphs showing the analysis of the correlation between the serometabolome, rs-fMRI brain function, and the gut microbiome, according to one embodiment of the present invention. The graph shows a significant correlation between serum metabolites/gut microbes and rs-fMRI (p < 0.05).

Detailed Description

The terms used herein have meanings commonly understood by those of ordinary skill in the relevant art. However, for a better understanding of the present invention, some definitions and related terms are explained as follows:

"bipolar disorder" refers to a group of chronic mental diseases with unknown etiology and repeated attacks, wherein the attack age is mainly focused on the late adolescence or early adulthood, and the clinical features of the diseases are depression, mania or hypomania alternately or mixedly, and the diseases are manifested as sensory perception, emotional processing, cognition and other aspects.

Examples

Collecting and processing samples: the present invention co-collects venous blood, fecal samples and nmr data from subjects including bipolar affective patients (n = 100) and healthy controls (n = 49). In the present invention, the bipolar disorder patients who are enrolled are hospitalized and/or outpatient patients who meet the criteria of diagnostic and statistical manual for mental disorders (DSM-IV), fourth edition BD-I, BD-II and BD (NOS), not specifically mentioned. The following grouping criteria are also met: 1) HAMD-24 score not less than 14 points; 2) no drug or no drug for at least 3 months; 3) there were no other comorbid mental diseases or obvious suicidal thoughts (suicidal ideation score in MADRS ≦ 2). Healthy subjects were recruited from the local community without any family history of psychiatric disorders and diseases. Age, gender and body mass index were matched between the two groups. Exclusion criteria for all subjects included: 1) chronic, severe cardiovascular and cerebrovascular diseases; 2) chronic or acute inflammatory, autoimmune diseases; 3) history of drug abuse, such as alcohol and tobacco; 4) a subject who is pregnant or lactating; 5) antibiotics, prebiotics or probiotics are taken orally 4 weeks before screening; 6) no informed consent was provided. Blood samples are collected in a fasting state at 6-7 points in the morning, venous blood (5 ml) of the elbow of a patient is collected by a vacuum tube containing heparin, and the blood samples are immediately centrifuged for 10 minutes (3000 rm, 4 ℃). Collected plasma samples were stored in aliquots (1.5 ml) in a-80 ℃ refrigerator. Stool samples from all subjects collected were kept at-80 ℃ for half an hour.

The intestinal microbiology analysis comprises the following specific steps:

collecting a fecal sample: collecting fresh feces of all subjects in morning, subpackaging the samples in the feces collecting tube by an operator, taking about 0.2g of each sample, and storing the feces samples in a refrigerator at-80 ℃ for later use within 30 minutes after the feces samples are collected.

Fecal DNA extraction DNA was extracted from thawed fecal samples using an OMEGA-soil DNA kit (OMEGA Bio-Tek, USA). RNase detection without dnase was used to eliminate RNA contamination. DNA quality was checked using a NanoDrop 2000 UV-visible spectrophotometer and 1% agarose gel electrophoresis.

Library construction and sequencing: DNA libraries were constructed using Illumina. A paired-end (PE) library with an insert size of 350 bp was constructed for each sample, followed by high-throughput sequencing using NEXTFLEX Rapid DNA-Seq (Bioscientific, Austin, TX, USA) with a sequencing strategy of 150 bp paired-end. Removing low-quality or human genomic DNA and finally obtaining high-quality sequencing fragments (reads). The relative abundance of the microbial species can be calculated from the high quality sequencing fragment reads as follows: 1) aligning the high quality sequencing fragments to a reference marker gene; 2) counting the number of the inserted fragments according to the comparison result; 3) the number of inserts was normalized to the length of the marker gene to obtain the corresponding abundance.

Species count, α diversity and β diversity: the number of non-zero species in each sample is calculated. Alpha diversity (diversity within a sample group) was estimated on the basis of the gene profile of each sample according to the shannon index and the method using R (3.5.0) vegan package. The vegdist function in vegan was used to calculate the Bray Curtis distance to estimate the β -diversity (diversity between sample groups).

The results show that the species count (p < 0.01) and bacterial shannon diversity (p = 0.062) were significantly lower in BD patients than in the control group, indicating that the stability of the intestinal microbiome was relatively weak in BD patients. The beta diversity of BD gut microbes was higher (p < 0.01), indicating that the community between BD individuals was more heterogeneous than the healthy control group. dbRDA showed that the taxonomic composition and functional potential of the dysdiad microbiome differed significantly from the healthy controls (fig. 1a, fig. 1 b), the present invention identified 600 BD-associated species with an increase in 136 flora and a decrease in 464 flora in BD patients (fig. 2 a).

The main species with significant abundance differences between BD and healthy controls included significant decreasesYersiniaceae (5 kinds of the plants which are selected,Yersinia aleksiciae（GENOME143346），Yersinia frederiksenii (GENOME143407、GENOME143408)，Yersiniaceae spp. (GENOME231599)and Serratia spp. (GENOME 232192)), Akkermanspiaceae (4 species, Akkermansia muciniphila (GENOME113322、GENOME143211)AndAkkermansia spp. (GENOME115272、GENOME142411)）， enterobacteriacea (9 species, Citrobacter freundii (GENOME103766), Citrobacter werkmanii(GENOME231453)，Citrobacter spp. (GENOME000595)，Enterobacteriaceae spp. (GENOME177471)，Enterobacter cancerogenus(GENOME095996)，Enterobacter cloacae(GENOME143740、GENOME143746)，Enterobacter spp. (GENOME000057)AndEnterobacter mori(GENOME000047)））；and a significant riseStreptococcaceae （9 kinds of， Streptococcus mitis(GENOME058339、GENOME231431、GENOME231966)，Streptococcus oralis(GENOME176717)，Streptococcus pseudopneumoniae(GENOME207884)AndStreptococcus spp. (GENOME000595、GENOME207787、GENOME244205、GENOME207866) ). (FIGS. 2a, 2 b).

The serum metabonomics analysis comprises the following specific steps:

sample preparation: bio-ovogene (Suzhou, China) used LC/MS for non-targeted metabonomics analysis, i.e., serum samples were mixed with 80% methanol, spun, centrifuged at 12000 rpm for 10 min at 4 ℃ to obtain NIST standard curve calibration solution supernatants. Samples for quality control and LC-MS detection were prepared under the same centrifugation conditions as described previously.

LC/MS serum metabolic analysis and metabolite identification: the raw data was converted to mzXML format using proteo wizard (v3.0.8789). The peak identification, screening and comparison are carried out by adopting R-package XCMS (R-v3.1.3). After quality control, annotation of metabolites data verifies that databases including LipidMaps (http:// www.lipidmaps.org), massbank (http:// www.massbank.jp /), Human Metabolome Database (HMDB) (http:// www.hmdb.ca), Metlin (http:// Metlin. script. edu), mzclound (https:// www.mzcloud.org), and metabolite databases were used, and were constructed by biovogene (Suzhou, China), to avoid missing any important metabolites.

Metabolite pathway recognition: the biological pathways of key metabolites that show significant differences between BD patients and Healthy Controls (HCs) were annotated. Biological pathway analysis was performed by Metabolite Set Enrichment Analysis (MSEA) using the MetaboAnalyst tool kit 1.

12127 metabolic features were determined in 131 serum samples according to the present invention. Of these, 8,918 features (73.54%) were altered in BD (false discovery rate [ FDR ] < 0.05), suggesting that bipolar affective patients have extensive metabolic alterations. Annotation of metabolic profiles by the database yielded a total of 265 annotated serum metabolites, including plasma metabolites with important functions in humans. Patient behavioral data, including diagnostic questionnaires and demographic information, were also collected in view of the complexity of BD. There is no evidence that behavioral measures are associated with serum metabolites (PERMANOVA test, p > 0.05). However, the severity of BD symptoms (measured in MADRS, HAMD and HAMA) and T cell subsets were significantly associated with serum metabolites (fig. 3, PERMANOVA test, p < 0.01). In addition, the present inventors found that the scholarly type, age and sex of education had a moderate but statistically significant effect on serum metabolites in BD patients (figure 3, PERMANOVA test, p < 0.05). BD disease status accounts for nearly 12% of variance, is a major factor in changes in serum metabolism, and can be distinguished from BD patients in healthy controls (fig. 3).

Further study of distance-based data redundancy analysis (dbRDA), the present inventors found that serum metabolites of BD patients were in sharp contrast to healthy controls (fig. 3, PERMANOVA test, p < 0.01), with 138 of 265 metabolites significantly correlated with BD (fig. 4). In addition, more than half (73.2%) of the metabolites were also observed to be reduced during BD (fig. 4). These 138 serum metabolites are involved in 64 metabolic pathways, including the citric acid cycle, fatty acid biosynthesis, glutathione metabolism, arginine and proline metabolism (fig. 3). The attenuated metabolite pathways in BD are steroid hormone synthesis, tryptophan metabolism, phenylalanine metabolism, pyrimidine metabolism, pantothenic acid and coenzyme a synthesis, and butyrate metabolism.

To further explore the link between gut microbiota and metabolic composition, intergroup Co-inertial analysis (CIA) was performed on the abundance of gut species and serum metabolites. The results indicate a close relationship between the gut microbiome and serum metabolites (RV = 0.265, p)<0.05), wherein 86.9% of the serum metabolites are associated with at least one intestinal microorganism. In particular vitamins known to be involved in brain function (pantothenic acid, riboflavin, folic acid and pyridoxine), short chain fatty acids (3-methylthiopropionic acid and 2-hydroxybutyric acid), kynurenine and gamma-aminobutyric acid, and certain intestinal microorganisms, such asAkkermansia muciniphila，Faecalibacterium prausnitzii，Enterobacter cloacaeAndYersinia aleksiciae(FIG. 5). To confirm the potential role of gut microorganisms in the generation of "neuronal markers", microbial metabolic pathways/genes encoding key enzymes in the major synthetic pathways of these substances were intensively studied (fig. 5). Finally, 1840 species were co-identified, includingThe entire metabolic pathway or key enzymes encoding "neuroactive metabolites" (table 1). In addition to pyridoxine, the abundance of these genes/modules was significantly higher in BD patients and BD-enriched microbial species (fig. 6). The serum concentration of a "neuronal marker" is significantly correlated with the abundance of genes encoding metabolic pathways/cognate synthetases in a particular species. In particularAkkermansia muciniphila, Citrobacter spp., Eubacterium spp.AndYersiniaceae spp.was significantly associated with various serum metabolites (figure 6).

To further screen potential biomarkers of risk of developing bipolar disorder, a random forest model was used to establish correlations between each "neuroactive metabolite" and species abundance of metabolic pathway/synthetase encoding genes comprising that particular "neuroactive metabolite". The random forest model maximally improved the ability to predict the concentration of "neuroactive metabolites" in serum, and 154 species were identified (fig. 7). The model averages 22% of the variance in the concentration of the target metabolite in serum. This indicates that the corresponding species promoted the production of "neuroactive metabolites" to a large extent. Consistent with the "neuroactive metabolite" changes in BD patients, most species (28.2%) were less in BD (fig. 7). Importantly, the species associated with the production of "neuroactive metabolites" were closely related to quantitative measures of bipolar disorder severity (MADRS, HAMD, HAMA and YAMARS) (fig. 8). It can therefore be assumed that the gut microbiome may influence the pathophysiology of BD by the above-mentioned "neuroactive metabolites" (in particular vitamins).

To date, the stochastic model suggested that BD patients/healthy control classified subjects had an area under the working characteristic curve of 0.81 (fig. 9). In this model, Akkermansia muciniphila, Citrobacter freundii, Enterobacter cloacae and Yersinia frederiksenii are the major differential flora. These findings indicate that microorganisms involved in the production of "neural substances" are likely candidates for the diagnosis of BD. Furthermore, functional analysis of a single species indicates that the species rich in BD/HC coding functions are associated with substances that regulate neurotransmitter secretion (referred to as "neuroactive metabolites"), such as acetyl-coa metabolism, polyamine biosynthesis, cofactor and vitamin synthesis, aromatic amino acid metabolism. There were significant differences in these functions between the 15.4% BD-rich species and the 17.3% HC-rich species, which probably reflected interference of BD with acetyl-coa, polyamines, aromatic amino acids, cofactors and vitamin availability. BD/HC-enriched species encode a variety of amino acid, carbohydrate and methane metabolisms. These functions were altered with the concentration of the corresponding BD-associated serum metabolites (fig. 10). Therefore, the enrichment of the metabonomics of BD patients is related to the intestinal microorganism-mediated AAA (intestinal Amino Acids) biosynthesis, SCFA (Short chain fatty acid) biosynthesis, choline-related functions, cofactors and vitamin biosynthesis. These data reveal that microbial dysbiosis of the gut flora and enriched metabolomics of BD are associated with gut flora biosynthetic function.

The brain functional group analysis comprises the following specific steps:

acquiring and processing nuclear magnetic resonance data: 44 BD patients and 37 healthy controls received 20 minutes of MRI treatment and were scanned structurally and functionally on a 3.0 Tesla GE Signal HDxt scanner (GE Healthcare, Waukesha, Wisconsin, USA) equipped with an 8-channel phased array head coil. Throughout the procedure, all subjects were instructed to remain still and awake with both eyes open. The cushion is used to limit head movement and the earplugs are used to reduce noise. The resting state functional image (rs-fMRI) obtained the following parameters using an echo-planar imaging protocol: TR (repetition time, pulse repetition interval time) = 1800ms, TE (echo time, repetition time) = 30 ms, flip angle = 90 °, voxel = 3.75 × 3.75 × 4 mm³Field of view = 240 × 240mm²28 axial slices per unit volume, 180 time points per cubic millimeter. High-resolution 3D T1 weighted magnetization gradient echo fast acquisition (MPRAGE) structural image is only used for anatomical reference, and parameters TR = 7.05ms, TE = 2.85ms, and the structure is invertedCorner = 8 °, voxel =1 mm³Nondirectional, field of view = 240 × 240mm²。

And (3) processing data: structural and functional images were preprocessed with CONN-20.b toolbox2 (http:// www.nitrc.org/projects/CONN, default preprocessing pipeline), the corresponding T1 image was realigned based on SPM123. functional scans and resampled in the phase encode direction, susceptibility distortion correction was performed to adjust for head motion and possible distortion due to field inhomogeneity (realignment and warping). The functional slices (interleaved and bottom-up) are time-shifted and resampled for slice time correction. Since resting state function images are prone to head motion artifacts, a more conservative approach has been taken to detect outlier scans due to excessive head motion. Outliers were obtained when the composite subject's motor threshold exceeded 0.5 mm or observed global BOLD signal changes exceeded 3 standard deviations. All anatomical and functional images were normalized to standard mni (montreal Neurological institute) space and then segmented into gray matter, white matter and cerebrospinal fluid (CSF) classes (segmentation and normalization). Reference structure data of the mapping function data (interpolated at isotropic 2 mm voxels; resolution consistent with the MNI mean mask) were directly normalized. The resulting function data was smoothed using a 6mm Full Width Half Maximum (FWHM) gaussian kernel to improve BOLD (blood oxygen level dependent) signal-to-noise ratio (SNR). To further mitigate the effects of motion-related and physiological noise, the necessary denoising procedure is applied. The anatomical composition correction strategy of CONN calculates the confounding effect of white matter and CSF noise components that linearly regress global signals using acompar 5 Band-pass filtering 0.008Hz, 0.09Hz for the pre-processed functional time series.

Regions of Interest (Regions of Interest, ROIs): the invention divides the whole brain (excluding cerebellum) into 136 different regions of interest (ROIs) and performs ROI-based resting state functional connectivity (rsFC) analysis. 91 bilateral subcortical ROIs and 15 bilateral subcortical ROIs are defined from the FSL Harvard-Oxford atlas maximum likelihood cortical atlas. Based on the CONN default clustering and ranking algorithm, another 30 network-based ROIs were also included in the rsFC analysis, which defined the ICA analysis of the human connected group project dataset from the CONN.

Neuroimaging analysis: ROI-based rsFC was investigated at the subject level using CONN-20.b toolbox 7. The average BOLD time series of all predefined ROIs are analyzed pairwise to calculate the fischer-tropsch binary correlation coefficient between each pair of ROIs. One paired ROI-ROI connectivity (RRC) matrix may describe the entire connected network for each subject. RRC matrices were extracted from CONN at the subject level and studied using serum metabolomics and gut microbiology data. The present invention examines the inter-sample RRC matrix (BD vs HC) and performs functional network connection analysis using multivariate parameters GLM8, where population-related functional connections will be organized into significant network clusters with cluster-level FDR correction p < 0.05.

To verify that the gut microbiota may drive bipolar disorder, at least in part, by "neuroactive metabolites," the present invention collected resting state functional MRI (rs-fMRI) images of 81 subjects. Based on the CONN data-driven hierarchical clustering algorithm of ROI-to-ROI spatial proximity and functional similarity measurement, 9180 pairwise connections of 136 ROIs are divided into 210 clusters. Principal Component Analysis (PCA) was first performed and the ROI-to-ROI connectivity matrix for BD was found to be significantly different from the healthy control group (FIG. 11a, PERMANOVA test, p)<0.05). 69 of the 210 clusters were significant (Table 2, GLM, FDR)<0.05) containing 1401 significant individual connections (post hoc)tInspection of p<0.05) that form 20 "networks" (fig. 11b, table 2).

Table 2 functional network connectivity analysis results

The results of the present invention show a general decrease in functional connectivity between the limbic zones of the BD group compared to the HC group, mainly including hippocampal formation and the amygdala ("network-18" and "network-19") and cortical regions. The cortical areas highlighted are the temporal gyrus (classified as default mode network DMN; "network-9" and "network-10"), the temporal gyrus (temporal-occipital part, classified as dorsal attention network DAN; "network-12"), and the prefrontal gyrus (IFG, classified as language network, "network-8"). In fact, more than half (61.3%) of the prominent connectivity impairment/low connectivity clusters (HC > BD) surround the edge system ("network-18" and "network-19"), suggesting that cognitive dysfunction and mood regulation disorders may be present in BD. However, strong functional connections were also found in the subcortical region of BD, particularly in the limbic system ("network-18" and "network-19"), thalamus and striatum (including caudate and putamen ("network-20"), which may be associated with potentially elevated mood and neural communication in the reward circuit (fig. 12 a).

Another finding of the present invention regarding functional connectivity of bipolar disorder is that the sensory-motor network has a complex role. In one aspect, the super-connectivity of BD patients includes auditory ("network-1"), language (STG [ "network-3", "network-5" ], IFG [ "network-8" ]), and sensory-motor (PreCG/PostCG, the "network-16" region of FIGS. 11b and 12 b). On the other hand, the connection between the partial sensory area and the subcortical area of BD patients is greatly weakened. It seems to suggest that BD patients may experience abnormal sensory information processing and emotional assessment of internal sensory activity. It can be said that the FC pattern between the auditory area ("network-1" and part "network-7") and the subcortical region is less deterministic (fig. 12 b). Hippocampal and amygdala auditory areas FC increased and thalamic auditory areas FC decreased, suggesting that patients with bipolar disorder may experience a psychiatric event.

Next, the present invention evaluated the degree of correlation of FC, serum metabolomics, and gut microbiome using Spearman correlation analysis. The relevance of serum metabolism and FC (Pearson r = 0.395, p = 0.039) is greater than that of gut microbes and FC (Pearson r = 0.368, p = 0.073) (fig. 13a, 13 b), as neural signals can alter the sensorimotor and secretory functions of the gut through complex neurohumoral regulation, while derivatives of gut microbes can modulate brain function through visceral and endocrine circulatory afferent signals. In addition, the invention also adopts PERMANOVA detection to evaluate the influence of the rs-fMRI cluster spectrum on the biphasic disorder related serum metabolites and the intestinal flora species. The results showed that 86.96% (120/138) of the BD associated serum metabolites were significantly associated with at least one individual linkage (PERMANOVA test, p < 0.05). In particular folic acid, which is associated with the regulation of brain development, mood and cognition, the results showed significant association with most of the clusters (85.51%, 59/69, fig. 13 b), especially certain brain regions and networks, including hippocampus and amygdala ("network-18" and "network-19"), thalamus and striatum ("network-20"), language region ("network-3", "network-4", "network-5", "network-10") and sensorimotor region ("network-16") (fig. 13 c). In addition, other "neuroactive metabolites," such as kynurenic acid, vitamin B6, gamma-aminobutyric acid, and riboflavin were significantly associated with the thalamus and striatum ("network-20"), auditory zone ("network-7"), language zone ("network-3", "network-4", "network-5"), dorsal attention network ("network-12"), hippocampal formation, and functional connectivity of amygdala ("network-18" and "network-19"), suggesting that disorders in serum of "neuroactive metabolites" may affect the language, mood, and circuitry of BD (fig. 13 c).

Likewise, 78.33% (470/600) of gut microorganisms were significantly associated with at least one individual linkage (PERMANOVA test, p)<0.05). In particular, microorganisms associated with serum "neuroactive metabolites" are also associated with specific brain networksAre related to each other.Akkermansia spp. (mainly isAkkermansia muciniphila ), Citrobacter freundii, Yersinia spp.(Yersinia frederiksenii And Yersinia aleksiciae), andenterobacter spp. (Enterobacter cloacae and Enterobacter kobei)There is significant correlation with the language region ("network-3", "network-4", "network-10"), thalamus and striatum ("network-20"), sensory motor region ("network-16"), hippocampal formation and amygdala ("network-18" and "network-19"). (FIG. 13 d).

Unless otherwise indicated, the techniques used in the examples are conventional and well known to those skilled in the art, and may be performed according to the third edition of the molecular cloning, laboratory Manual, or related products, and the reagents and products used are commercially available. Various procedures and methods not described in detail are conventional methods well known in the art, and the sources, trade names, and components of the reagents used are indicated at the time of first appearance, and the same reagents used thereafter are the same as those indicated at the first appearance, unless otherwise specified.

In the present invention, the sequencing of NEXTFLEX Rapid DNA-Seq and the non-targeted mass spectrometry are well known in the art and can be adjusted by the skilled person according to the specific situation.

In the present invention, the use methods of the random forest model and the ROC curve are well known in the art, and those skilled in the art can set and adjust parameters according to specific situations.

In the invention, training sets of 'neuroactive metabolites' of bipolar affective disorder subjects and non-bipolar affective disorder subjects are constructed, and the 'neuroactive metabolites' of the sample to be tested are evaluated on the basis.

One skilled in the art will appreciate that when the sample size is further expanded, the normal content value interval (absolute value) of each "neuroactive metabolite" in the sample can be derived using sample detection and calculation methods well known in the art. The absolute value of the content of the detected neuroactive metabolites can be compared with the normal content value, and a statistical method can be combined to obtain the risk evaluation and diagnosis of the bipolar affective disorder and the efficiency for monitoring the treatment effect of the bipolar affective disorder patient.

These "neuroactive metabolites" are present in the serum of the human body. The correlation analysis of the neuroactive metabolites, the intestinal flora and the brain function connection condition of the testee is carried out by the method, which indicates that the intestinal flora influences the brain function connection condition of the BD patient by mediating the metabolism of the neuroactive metabolites.

The results show that the three-group chemical analysis of the intestinal flora biomarker, the serum 'nerve activity metabolite' and the brain function connection condition, disclosed by the invention, can well provide comprehensive landscapes and the screening and the application of the biomarker on the basis of the pathophysiology of acute-phase bipolar depression, and provide guidance for the pathological mechanism research, clinical recognition, development trend prediction and accurate treatment of bipolar disorder.

Thus, the present invention proposes applications including, but not limited to:

the intestinal flora-based bipolar affective disorder biomarker combination is used as a detection target or an application of a detection target in preparation of a detection kit.

The application of the serum-based bipolar affective disorder biomarker combination as a detection target or a detection target in preparing a detection kit.

The application of the combination of the bipolar affective disorder biomarker based on the intestinal flora and the brain function connection condition of serum 'nerve activity metabolite' as a target point in screening the prediction risk of the bipolar affective disorder.

Although embodiments of the present invention have been shown and described above, it will be understood that the above embodiments are exemplary and not to be construed as limiting the present invention, and that changes, modifications, substitutions and alterations can be made to the above embodiments by those of ordinary skill in the art within the scope of the present invention.

Claims

1. Bipolar affective disorder biomarkers characterized by four gut microbiota, Akkermansia muciniphila, Citrobacter spp, Yersiniaceae spp and Enterobacter spp, in which the abundance is significantly reduced in bipolar affective disorder patients, with gut signaling to the brain via neuroactive metabolites.

2. A bipolar affective biomarker according to claim 1, wherein said four gut microbiota derived neuroactive metabolites in serum, including pantothenic acid, riboflavin, folic acid, pyridinol, kynurenic acid, GABA and SCFA, are dysregulated in bipolar affective patients;

disorders significantly associated with the language zones network-3, network-4, "network-10, thalamic and striatal network-20, sensorimotor zone network-16, hippocampal formation and functional connectivity of amygdala network-18 and network-19 affect the language, mood and reward circuit of BD.

3. A method of screening the biomarkers of bipolar affective disorder according to claim 1, wherein based on clinical and questionnaire data, 100 patients with depressive bipolar disorder and 49 healthy control groups were selected, gut microbiota was analyzed by metagenomic bomb sequencing, and serum metabolome was analyzed by non-targeted mass spectrometry to demonstrate brain dysfunction in functional connectivity in whole brain resting state.

4. A kit prepared according to claim 1, wherein the bipolar affective disorder biomarker comprises four gut microbes, Akkermansia muciniphila, Citrobacter spp.

5. The kit of claim 4, wherein the bipolar disorder biomarker further comprises pantothenic acid, riboflavin, folic acid, pyridinol, kynurenic acid, GABA, and SCFA.

6. The kit according to claim 4, for diagnostic use or for screening of a medicament for the treatment of bipolar affective disorder.

7. A biological preparation prepared according to the bipolar affective biomarker of claim 1 or 2, for use in balancing or restoring the bipolar affective biomarker in a patient with bipolar affective disorder.