Intracranial hemorrhage management in the multi-omics era

Abstract

Intracranial hemorrhage (ICH) is a devastating disorder. Neuroprotective strategies that prevent tissue injury and improve functional outcomes have been identified in multiple animal models of ICH. However, these potential interventions in clinical trials produced generally disappointing results. With progress in omics, studies of omics data, including genomics, transcriptomics, epigenetics, proteomics, metabolomics, and the gut microbiome, may help promote precision medicine. In this review, we focused on introducing the applications of all omics in ICH and shed light on all of the considerable advantages to systematically analyze the necessity and importance of multiple omics technology in ICH.

1. Introduction

Spontaneous intracranial hemorrhage (ICH) accounts for 15% of all strokes and causes higher mortality and disability than ischemic stroke worldwide [1]. Survival and recovery from ICH are associated with the hemorrhage site, mass size, intracranial pressure from the hematoma, and subsequent secondary brain injuries (SBI) such as neurotoxicity, inflammation, and perihematomal edema (PHE), as shown in Fig. 1. Significant progress has been made in animal models on therapies and mechanisms underlying brain injury in ICH. However, few effective drug therapies are currently available to limit direct mechanical brain damage and SBI in ICH in clinical trials [2]. Therefore, how to effectively develop precise therapeutic targets has been a critical topic in ICH.

Fig. 1.

Pathophysiology after ICH. The Hematoma causes early primary brain injury after ICH. Subsequently perihematomal edema (PHE), disruption of inflammation and immunity, clot component-induced neurotoxicity, and blood brain barrier (BBB) damage cause the secondary brain injury and result in neurofunctional deficits.

In the early 21st century, DNA sequencing technology improved immensely, especially next-generation sequencing. At the same time, the omics analysis related to genomics, transcriptomics, epigenetics, proteomics, metabolomics, and gut microbiome has increased rapidly. Biomedical sciences have progressed omics century due to rapid development in molecular biology technology and bioinformatics. The multi-omics technologies help achieve the better understanding of the molecular interactions, disease mechanism and processes, thus representing a promising approach for ICH study.

This review first introduces the general application of omics in ICH, as shown in Fig. 2. Hopefully, this review will help readers better understand the application of multiple omics in ICH.

Fig. 2.

Multi-omics in ICH. ICH, intracranial hemorrhage.

2. Pathophysiology of ICH

Hypertension is the most common cause of spontaneous ICH. Other etiologies that cause ICH include cerebral amyloid angiopathy (CAA), aneurysms, brain tumors, cerebral venous thrombosis, vascular malformations, hemorrhagic transformation of ischemic stroke (IS), and use of anticoagulants. Once ICH occurs, there are two types of damage: primary brain injury and SBI [3], as shown in Fig. 1.

2.1. Primary brain injury

The initial bleeding causes a primary brain injury in ICH. The hematoma damages neurons and glia, compresses nearby brain regions to ischemia, increases intracranial pressure, and even causes brain herniation. Patients usually undergo hematoma expansion within 24 h [4]. Hematoma expansion increases the progression of disease and is an essential cause of ICH-related death [5]. The peroxisome proliferator-activated receptor-γ (PPAR γ) induces the expression of CD36 in macrophages and microglia and is considered one of the critical factors in improving the resolution of hematoma and improving neurofunctional outcomes in ICH [6].

2.2. Secondary brain injury (SBI)

The primary injury will cause a series of secondary brain injuries (SBI), such as PHE, inflammation and immunity, the release of clot components, and damage to the blood brain barrier (BBB). These factors cause further damage to the surrounding parenchyma from days through weeks, which is an essential time window for therapeutic intervention [7].

PHE progresses rapidly during the first 24–72 h. Subsequently, the slow progressive edema increases in the second week [8]. The progression of PHE experiences three different phases [9]. The first phase appears within 1–4 h of hematoma formation and is associated with clot retraction and hydrostatic pressure without disruption of the BBB. In the second phase, PHE progresses between 4 and 72 h after ICH and is related to thrombin release. On the one hand, thrombin mediates conversion of fibrinogen to fibrin and leads to the formation of clots. On the other hand, thrombin mediates inflammatory responses and BBB disruption. In the late phase, PHE is related to erythrocyte lysis after 72 h. Erythrocyte lysis leads to the accumulation of hemoglobin, which stimulates radical oxygen production and causes neuronal death. There are conflicting reports involving the association of PHE and ICH results.

Inflammation and immunity are important host defense responses against brain injury in ICH. Increasing evidence has shown that inflammation and immunity play an essential role in ICH-induced SBI [10]. When ICH-induced brain injury occurs, many neural cells die, and subsequently secondary inflammatory and immune responses occur. Excessive microglial activation, which releases many cytokines and chemokines, and reverses recruitment of peripheral immune cells into the injured brain regions. At the same time, danger-associated molecular patterns (DAMPs), which are released from dead neurons, induce circulating inflammatory cells into the brain to aggravate brain injury. These processes are called focal brain inflammation. Emerging evidence shows that immunity and inflammation also occur in areas remote from the primary injury site in ICH, which is named global brain inflammation [11]. Furthermore, growing evidence shows that ICH is characterized by systemic inflammatory and immune responses with numerous extra-central nervous system (CNS) manifestations such as brain-spleen interaction and the change of cellular mediators of the circulating lymphocyte and monocyte subpopulations [12].

Growing evidence shows that clot component-induced neurotoxicity plays a vital role in ICH-induced SBI. Erythrocyte lysis occurs very early in the hematoma and releases hemoglobin (Hb) [13]. When Hb is released to the extracellular space, heme, the Hb degradation product, is gathered in the brain parenchyma and releases iron. The latter is oxidized from ferrous (²⁺) to ferric (³⁺) and accumulates in the brain regions. ICH also activates the coagulation cascade and causes release of thrombin and subsequent accumulation in the brain [14]. Thrombin plays a critical role in the upregulation of pro-inflammatory cytokines, BBB disruption, and PHE. The Hb molecule and ferric iron result in inflammatory reactions, BBB disruption, PHE, oxidative stress, neuronal death and contribute to delayed neurodegeneration and brain atrophy in ICH [15,16].

Endothelial cells, pericytes, and astrocytes adhere to the ECM to maintain the integrity of the BBB [17]. When ICH occurs, inflammation, immunity, and neurotoxicity are involved in ICH-induced BBB dysfunction in PHE [18].

3. Omics in ICH management

3.1. Genomics in ICH

Importantly, genome-wide association studies (GWAS) and next-generation sequencing have revealed many variations associated with an increased risk of stroke [19]. ICH is a complex disease mainly caused by the interactions of genetic and environmental risk factors [20]. Unraveling the genomics of ICH sheds new light on the pathophysiology of ICH and opens new avenues for pharmaceutical therapy [21].

Genetic variation plays a key role in the risk and prognosis of ICH [22]. Collaborative meta-analyses of GWAS from six studies showed that had identified rs11179580 in chromosomal region 12q21.1 was associated with lobar ICH, and rs2984613 in chromosomal region 1q22 was associated with non-labor ICH in European [23]. APOE is a polymorphic gene with three alleles ε2/ε3/ε4. Studies of Candidate genes have shown that APOE was one of the risk genes for ICH. Population-based researches have proved ε2/ε4 carriers were at increased risk of ICH [24]. ε2/ε4 alleles were found to account for nearly 30% of lobar ICH risk [22]. CAA is degenerative vasculopathy with accumulation of amyloid beta deposition in the walls of the cerebral arteries. CAA-related hemorrhages (CAAH) account for 5%–20% of all ICH [25]. In addition, a study found that patients with CAAH were more likely to be APOE ε2carriers [26]. Several studies indicated that variations of COL4A1 and COL4A2 genes were associated with elevated risks in ICH. The another study also identified two rare mutations rs138269346 and rs201716258 in association with sporadic ICH [27]. Cerebral autosomal dominant arteriopathy with subcortical infarcts and leukoencephalopathy (CADASIL), which was caused by NOTCH3 gene mutations, is a single gene disorder. Choi et al. found that 25% of the CADASIL had ICH [28]. Owing to the fact that genomics does not change with age, genomics plays a key prole in ICH research, and provides new insights into ICH pathogenesis.

3.2. Transcriptomics and epigenetic modifications in ICH

Transcriptomics is the study aimed at capturing both coding and non-coding RNA and further influencing gene expression heterogeneity. Transcriptomics analysis is critical because it refers to the complete set of gene transcripts [29]. Messenger RNA (mRNA) is coding RNA and will translated into protein [30]. Transcriptomics plays an important role in cerebro-cardiovascular disease [31,32]. Cheng et al. evaluated microRNAs (miRNA) and mRNA targets in the whole blood of patients and found that there are changed expression of 29 miRNAs (22 down and 7 up) and 250 target mRNAs (136 up and 114 down) after ICH compared to controls. Furthermore, analysis of the pathways of integrated mRNA-miRNA networks showed that the regulated miRNAs-mRNAs targets regulated many inflammatory/immune/cell death pathways such as toll-like receptor, natural killer cells, apoptosis, etc [33]. miRNAs are the most abundant RNA type and are a well-studied non-coding RNA that regulate gene expression through mediating mRNA degradation [34]. Increasing numbers of researches have demonstrated that miRNAs play pivotal roles in pathophysiological mechanisms, diagnosis, and therapy aspects of ICH ³³. Compared to the control group, ICH clinical trials have shown that the expressions of circulating miR-145, miR-223, and miR-155 were increased in ICH, but circulating miR-181b and Hsa-miR-21-5p was decreased [35,36]. In animal experiments, miR-126-3p and miR-27a-3p attenuated ICH-induced BBB disruption [37,38]. miR-7, miR-590-5p, miR-21, miR-146a, Let-7a, miR-124, miR-132, and miR-223 reduced ICH-induced inflammation [[39], [40], [41], [42], [43], [44], [45]]. The study reported that miR-223 downregulates NLRP3 to decrease microglial activation and neuron injury after ICH [45]. Long non-coding RNAs (lncRNAs) and circular RNAs (cirRNAs) are also potential therapeutic targets for brain injury following ICH. A study showed that lncRNA-FENDRR regulated the expression of vascular endothelial growth factor A (VEGFA) through interaction with miRNA-126 in ICH mice [46]. Jeong-Min Kim et al. studied the expression pattern of lncRNAs and mRNA in two different ICH rat models on the1st, 3rd, and 7th days and found that H19 was the most up-regulated lncRNA. Additionally, this study also illustrated the dynamic expression of lncRNAs after ICH [47]. A lot of studies also revealed expression changes of circRNA profiles in ICH. Dou et al. found a significant alteration in the expression of cirRNAs in rat brain tissue after ICH, and 93 cirRNAs were up-regulated and 20 cirRNAs were down-regulated at 6 h, 12 h, and 24 h [48]. Bai et al. found hsa circ 0001240, hsa circ 0001947 and hsa circ 0001386, three peripheral blood circRNAs, are potential diagnostic marker for ICH [49]. In conclusion, transcriptomic modifications provide important insights for the deeper understanding of ICH prevention, treatment, and prognosis.

Epigenetic modifications modulate mainly gene expression without changing the DNA sequence. Some studies have demonstrated that epigenetic modifications modulate different physiological processes of stroke [50]. DNA methylation is an essential part of the epigenetic mechanism. Increasing evidence showed that DNA methylation is critical in ICH. Zhang et al. found that genome-wide DNA methylation was altered in whole blood of ICH patients. Additionally, methylated genes were involved mainly in inflammatory pathways [51]. 5-hydroxymethylcytosine (5-hmC), derived from 5-methylcytosine, is an intermediate in the DNA demethylation process and an alternative epigenetic marker [52]. The recent study showed global 5hmC levels on promoters of the Akt2, Pdpk1, and Vegf genes were significantly decreased from 24 h to 72 h in mouse brains after ICH [53]. These findings demonstrate the value and feasibility of transcriptomics and epigenetic modifications in ICH research. However, none of these differentially expressed biomarkers has yet been successfully applied to clinical practice.

3.3. Proteomics in ICH

Proteomics is an innovative approach that can simultaneously measure the entire set of proteins in a given biologic system [54]. With the rapid development of proteomics technology, quantitative proteomics has been introduced to explore the pathophysiological mechanisms, new biomarkers, and candidate therapeutic interventions in animal models and clinical trials of stroke [55,56]. Deng et al. found that 96 proteins identified in brain tissue at 24 h following ICH had significantly different expressions than the control group [57]. Liu et al. reported that rhubarb exhibited a protective effect in ICH through the regulation of oxidative stress, calcium-binding protein, angiogenesis, and energy metabolism with relative and absolute quantification (iTRAQ)-based LC-MS/MS proteomics technology [58]. Liu et al. also screened 201 differentially expressed proteins using iTRAQ‐based LC-MS/MS technology and found dysregulation of energy metabolism may be an important cause of ICH-induced SBI [59]. In another study, researchers screened 207 proteins expressed significantly altered after ICH and were associated with 13 main signaling pathways by employing a quantitative proteomic approach combined with bioinformatics [60]. Furthermore, Chen et al. found 884 proteins of perihematomal tissue were dysregulated in the ICH patients, and fibronectin 1 was central within the protein-protein interaction networks [61]. In the current study, Cheng et al. confirmed that Nrf2/OPTN-mediated mitophagy alleviated SBI based on iTRAQ proteomics technology [62]. These findings support these proteomics datasets will be valuable for future studies to explore diagnostic biomarkers and therapeutic strategies for ICH.

3.4. Metabolomics in ICH

Metabolomics is a crucial new omics that focuses on identifying novel metabolite biomarkers.

It plays a key role in exploring the pathophysiological mechanisms and novel treatment targets in stroke [63,64] In a prospective cohort study, researchers found that ten metabolites acted as new diagnostic biomarkers for the thrombotic stroke by high-resolution metabolomics (HRM) [65]. A clinical trial of subarachnoid hemorrhage (SAH) study showed that elevated plasma taurine level identified by HRM at admission predicted a better 90-day recovery [66]. In an intraventricular hemorrhage (IVH) study of preterm neonates, researchers reported urine metabolites were promising biomarkers for IVH progression [67]. To search for biomarkers to differentiate ICH from acute ischemic stroke (AIS), Zhang et al. confirmed 11 potential metabolites to build the artificial neural network (ANN) model to distinguish ICH from AIS [68]. In addition, Zhang et al. found that metabolite 20-OH-LTB4 and its key enzyme CYP4F2 may be potential ICH biomarkers by a UHPLC/MS-based targeted metabolomic method [69]. Mader et al. described changes in l-arginine metabolism in cerebrospinal fluid (CSF) and blood after ICH, and found marked differences in the concentrations of l-arginine between ICH group and control. Furthermore, the change in the concentration of the same metabolite has a different temporal pattern. Lastly, the authors found that the concentration of l-arginine in CSF with early reduction was an independent risk factor for poor prognosis [70]. Furthermore, to explore the mechanisms of rhubarb and wine-processed rhubarb treatment in ICH, researchers found wine-processed rhubarb had a better effect by regulating amino acid metabolism by integration and analysis of proteomics and metabolomics in the ICH rat model [71]. Therefore, the metabolomic analysis may provide substantial values to better understand pathogenesis and discover new biomarkers for diagnosis, prognosis, and treatment of ICH.

3.5. Gut microbiome in ICH

Increasing evidence shows that dynamic changes of gut microbiota play essential roles in maintaining the intestinal microbiota and brain physiology. When central nervous system disorders (CNS) occur, on the one hand, it results in alteration of the microbiome by neurotransmitters or neuromodulators. On the other hand, the gut microbiota releases metabolites and inflammatory cytokines to enter the CNS further to promote the pathogenesis of neurological disorders [72]. In an ICH study, Luo et al. found that gut dysbiosis worsened the poor prognosis of ICH by 16S rRNA gene sequencing [73]. Trimethylamine N-oxide (TMAO), a gut microbiota-derived metabolite, plays a critical role in the pathogenesis of cerebrovascular disease. A recent study showed that elevated levels of TMAO were identified as an independent risk factor for 3-month function outcome after ICH [74]. However, the influence of TMAO on ICH is not yet understood. Therefore, it is necessary to clarify the mechanism by which TMAO is involved in the treatment of ICH. At the same time, it is unclear whether the changes of intestinal microbes and products after different subtypes of ICH are consistent. So, more studies are needed to study the association between the gut microbiome and ICH.

4. Multi-omics in ICH

The analysis of only each type of omics data is limited. We now realize that the integration of various omics data is more effective in elucidating potential causes of pathologies and identifying an appropriate treatment. In a study, the authors demonstrated that obesity induced by a high-fat diet (HFD) impairs CD8 T cell function and accelerates tumor growth by integration and analysis of various genomics, proteomics and metabolomics omics data [75]. In another study, the authors first used proteomics to find that the metabolic enzyme GPD1 was at low levels in bladder cancer tissues, and then used transcriptomics and metabolomics to reveal GPD1 promotes tumor cell apoptosis via the lysoPC-PAFR-TRPV2 axis [76]. In the study of ischemic stroke, Xian et al. found the combination of acupuncture and NaoMaiTong could more effectively improve neurobehavioral deficits than a single treatment via the potential mechanism of connection of Turicibacter and phytoestrogens metabolite isoflavones by integrating sequencing of the 16S rRNA gene and LC/MS-based metabolomics in rat models of ischemic stroke [77]. In the ICH study, the authors first applied proteomics analysis to find the differentially expressed proteins between the brain tissue of the animal model after ICH and normal brain tissue, then used transcriptomics and metabolomics to reveal the specific mechanism of the target protein in ICH and uncover the cell signaling pathways that target protein might participate in. Xie et al. discovered sex differences in gene and protein expression by combining transcriptomic gene expression and proteomics in brain tissue of mice after ICH [78]. So, we can intensively study. The pathogenesis of ICH and the role of neuroprotective drugs in ICH by combinations of different omics⁶⁴. The greatest challenge for the integration and analysis of multi-omics data is to build efficient criteria for disease classification, improve prognostic prediction performance, and ultimately provide precision treatment for patients [79]. Machine learning (ML) methods have helped to discover new biological biomarkers in omics by integrating and analyzing the various omics data [80]. Milan et al. focused on studies of integration of multi-omics datasets through ML analysis and presented five different integration strategies [81]. However, these ML methods are still limited for the integration of multi-omics analysis in ICH. In the future, with the development of ML, ML algorithms and deep neural networks may be considered to integrate various omics data to help elucidate the biological processes of ICH.

5. Future perspectives

Omic analysis can contribute to the development of accurate tools for early diagnosis and prediction, to the better selection of drug treatments, a better understanding of the pathological mechanism of ICH. Evidence from current omics studies supports the previous findings that inflammation, immunity, autophagy, and oxidative stress dysfunctions all participate in the progress in ICH. However, there are some challenges to face. First, it will become essential to integrate omics and clinical data and develop a novel algorithm's strategies [82]. The integration and analysis of omics and non-omics (OnO) data can propose better predictive models for the prevention, diagnosis, progression, and prognosis of disease, and therefore provides the help to precision treatment in the future [83]. However, few studies performed integration of OnO data in ICH. Second, it cannot distinguish whether the alteration of molecules networks by omics approach analysis is the cause or consequence of the disease. In this sense, the mechanism of the candidate molecules in ICH needs to be performed with a more significant number of basic experiments. Lastly, with the progress in omics, single-cell and spatiotemporal omics will be the direction of development. The development of omics in ICH is still far behind other fields, and a standard and specialized database needs to be established. Currently, there is no proper drug treatment for ICH; however, omics techniques may help solve this problem.

Author contribution statement

All authors listed have significantly contributed to the development and the writing of this article.

Funding statement

This work was supported by National Key Research and Development Projects [2022YFC3602400], Key Technologies Research and Development Program [2022YFC3602401], Innovative Research Group Project of the National Natural Science Foundation of China [Grant No. 82271369]. This work was supported bythe Project Program of the National Clinical Research Center for Geriatric Disorders of Xiangya Hospital in China [2020LNJJ16].

Data availability statement

No data was used for the research described in the article.

Additional information

No additional information is available for this paper.

Declaration of competing interest

The authors declare that there are no conflicts of interest.