OMICs in Stroke: Insight Into Stroke Through Epigenomics, Transcriptomics, Proteomics, Lipidomics, and Metabolomics

Key Points

OMICs-based technologies are identifying novel molecules and pathways associated with stroke.
Markers identified through an OMICs-based approach may have utility as biomarkers to aid in the diagnosis of stroke, prediction of stroke risk, and assessment of stroke complications and stroke outcome.
OMICs approaches include epigenetics, transcriptomics, proteomics, metabolomics, and lipidomics.
Stroke is a complex disease. Broad assessment at the whole genome, proteome, metabolome, and lipidome level provides unique insight into the complexity of the biology present in stroke.
Ongoing studies over the next decade will reveal novel precision markers to aid physicians in the diagnosis, treatment decisions, and risk classification of patients with stroke.

Introduction

Our knowledge regarding the genetics of stroke is ever expanding, particularly with the advent of genome-wide association studies. However, the functional implications of many of the identified genes and loci in stroke remain unclear. This will change with the completion of the Human Genome Project coupled with technologies, including rapid sequencing, mass spectrometers, and nuclear magnetic resonance (NMR) spectroscopy. These tools are making it possible to examine the entire expressed genome (transcriptomics), gene regulatory elements (epigenomics), the entire proteome (proteomics), and downstream effectors including metabolites (metabolomics), lipids (lipidomics), and circulating blood cells.

While still in the early days, over the next decade OMICs-based studies will provide novel insight to the molecular biology occurring in patients with stroke. This likely will result in new therapeutic targets to treat stroke and stroke biomarkers to aid in diagnosis and risk assessment. Most tissues or fluids can be studied by OMICs-based technologies, although in stroke most have focused on blood. Since platelets, white blood cells, and the factors involved in clotting are all included within blood, it seems reasonable to apply the OMICs approaches to blood of a patient with ischemic stroke (IS). This logic may apply to intracerebral hemorrhage (ICH) as well if there is a tendency toward poor clotting and bleeding. Cerebrospinal fluid (CSF) and brain are other samples that have been analyzed. In this chapter we will demonstrate the types of questions being addressed by OMICs studies in stroke, including epigenomics, transcriptomics, proteomics, metabolomics, and lipidomics. While space precludes a discussion of all studies, the focus is on stroke patients highlighting potential areas where OMICs-based studies could impact stroke clinical care.

Epigenomics

Epigenetics is the study of factors besides DNA sequence variation that influence the process of gene expression. These can range from DNA methylation, histone modifications, transcription factors, microRNAs, long noncoding RNAs, and others. This is a relatively new field of stroke research ( Table 50.1 ).

TABLE 50.1

Epigenomic Markers Associated With Stroke.

Molecule Type	Marker	Role	Reference
DNA methylation	ASB10 methylation	Stroke risk	Davis (2018)
	TTC37 methylation	Stroke risk	Davis (2018)
	Long interspersed nucleotide element-1 methylation	Stroke risk	Krupinski (2018)
	Sortilin promotor hypomethylation	Moyamoya risk	Sung (2018)
	PTEN methylation	Cavernous malformations	Zhu (2009)
Histone modification	HDAC9	Large vessel stroke	Wang (2019), Shroff (2019)
	Histone deacetylase	Stroke recovery	Elder (2013), Kassis (2017), Schweizer (2013)
microRNA	Plasma microRNA	Stroke diagnosis	Eyileten (2018)
	Whole blood microRNA	Stroke diagnosis	Jickling (2014), Tan (2009), Spepramaniam (2014), Huang (2016), Jickling (2016), Gilles (2018)
Long noncoding RNA	Whole blood	Stroke diagnosis, moyamoya	Dykstra-Aiello (2016), Deng (2018), Wang (2017)
	INK4 locus (ANRIL)	Atherosclerosis risk	Holdt (2018)
	ZFAS1	Large vessel stroke	Wang (2018)

HDAC9 , Histone Deacetylation 9.

DNA Methylation

Experimental studies suggest focal ischemia increases DNA methylation that generally suppresses gene expression, and that this appears to be harmful since DNA methylation inhibitors can improve stroke outcome. ^, Increasing evidence supports altered methylation in human cerebral ischemia. Hypermethylation at two 5’-C-phosphate-G-3’ (CpG) sites located in ASB10 and TTC37 genes have been associated with fewer strokes. Changes of methylation have been associated with occurrence of atherosclerosis, hypertension, serum lipid profiles, smoking, response to antiplatelet agents, risk of stroke recurrence, and functional outcome following stroke. Patients with lower long interspersed nucleotide element-1 (LINE-1) methylation have a higher risk of ischemic heart disease and stroke. One surprise has been the lack of global methylation differences between the major causes of stroke—cardioembolic, large vessel, and lacunar. Hypomethylation of the sortilin promoter may be a biomarker for moyamoya disease, and hypermethylation of the phosphatase and tensin homolog (PTEN) promoter has been associated with cerebral cavernous malformations (CCMs).

Histone Modifications

Histone acetylation is a powerful modulator of gene expression. Certain histone deacetylation 9 (HDAC9) polymorphisms are a risk factor for large vessel stroke that produce large effects on gene expression of peripheral blood cells. Treatment with valproate, a histone deacetylase inhibitor, decreases the risk of stroke following a transient ischemic attack (TIA) or previous stroke. Histone deacetylases and other epigenetic alterations have been implicated in stroke recovery, ^, hypertension, atherosclerosis, vascular wall repair, neuroinflammation, ischemic tolerance, and neuroprotection.

MicroRNA (Short Noncoding RNA, miRNA), Plasma, Blood/Leukocytes/Platelets

There have been multiple studies of miRNA expression in IS, 19 examining plasma or serum and 6 examining whole blood. These studies have used different platforms (polymerase chain reaction, microarray, RNA sequencing), different timing following stroke, different inclusion and exclusion criteria, and different bioinformatic methods resulting in a varied set of differentially expressed miRNA in the different studies.

Studies of plasma and serum are of interest because the miRNA identified would likely be associated with exosomes or Argonaut, are extracellular. Moreover, they could be derived from ischemic brain, but also from nonischemic heart, lung, kidney, leukocytes, or other tissues.

Whole blood miRNA studies are quite different from those of serum or plasma, since whole blood represents predominantly intracellular miRNA from leukocytes, platelets, and other blood cells, which are an order of magnitude greater than extracellular miRNA. One whole blood study of acute stroke found miR-122, miR-148a, let-7i, miR-19a, miR-320d, and miR-4429 were decreased and miR-363 and miR-487b were increased compared to vascular risk factor controls. These miRNA were predicted to regulate toll-like receptor signaling, NF-κ β signaling, leukocyte extravasation signaling, and the prothrombin activation pathway. Other whole blood studies of acute stroke have found increased expression of let7 family members and miR-19a. Let-7i and other family members regulate the immune response, ^, and miR-19a regulates the inflammatory response, blood coagulation, and platelet activation.

Noncoding RNA (lncRNA)

Fewer studies have assessed lncRNA at the whole genome level. In the first whole genome lncRNA study 299 lncRNAs were differentially expressed between stroke and control males, whereas 97 lncRNAs were differentially expressed between stroke and control females ( n = 266 subjects). The lncRNA changed expression over time, and some were mapped close to previously identified stroke risk genes, including lipoprotein, lipoprotein(a)-like 2, ABO (transferase A, α1-3-N-acetylgalactos-aminyltransferase; transferase B, α1-3-galactosyltransferase) blood group, prostaglandin 12 synthase, and α-adducins. It was also found that specific lncRNA correlated with cause of stroke and other lncRNA correlated with the National Institutes of Health stroke scale.

A study of peripheral blood mononuclear cells found 70 upregulated and 128 downregulated lncRNA in IS patients compared to controls. qRT-PCR validation demonstrated that three lncRNAs (linc-DHFRL1-4, SNHG15, and linc-FAM98A-3) were significantly upregulated in IS patients compared with healthy controls and TIA patients. The lncRNAs outperformed brain-derived neurotrophic factor and neuron specific enolase (NSE) with an area under the ROC curve of 0.84. Another study found differentially expressed lncRNA in blood of patients with moyamoya disease compared to healthy controls that were involved in Toll-like receptor (TLR), chemokine, and mitogen-activated protein kinase signaling.

There are many reports of single lncRNA in IS. Of interest, the Chr9p21 risk locus has emerged as a top signal in genome-wide association studies of atherosclerotic cardiovascular disease, including stroke. The risk SNPs are near a non-protein-coding DNA, containing the gene body of the long noncoding RNA (lncRNA) antisense noncoding RNA in the INK4 locus (ANRIL). ANRIL splicing results in circular noncoding RNA that regulates atherogenesis. The long noncoding RNA ZFAS1 is down-regulated in IS compared to controls, and most down-regulated in large vessel compared to other causes of stroke.

Transcriptomics

Table 50.2 lists transcriptomic markers associated with stroke.

TABLE 50.2

Transcriptomic Markers Associated With Stroke.

Molecule Type	Marker	Role	Reference
Stroke diagnosis	22 gene panel	Stroke vs. control	Moore (2005)
	18 gene panel	Stroke vs. control	Tang (2006)
	9 gene panel	Stroke vs. control	Barr (2010)
	97 gene panel	Stroke vs. control	Stamova (2010)
	489/63 gene panel	Ischemic vs. hemorrhagic stroke	Stamova (2018)
TIA	34 gene panel	TIA vs. control	Zhan (2011)
	26 gene panel	TIA vs. mimics	Jickling (2012)
Stroke etiology	23 gene panel	Cardioembolic vs. large vessel	Xu (2008)
	40 gene panel	Cardioembolic vs. large vessel	Jickling (2010)
	37 gene panel	Atrial fibrillation vs. nonatrial fibrillation	Jickling (2010)
	41 gene panel	Lacunar vs. nonlacunar	Jickling (2011)
	40 + 41 gene panels	Cryptogenic stroke	Jickling (2012)
Sex differences	Male/female differences in gene expression in stroke	Stroke biology	Stamova (2014), Tian (2012)
Hemorrhagic transformation	6 gene panel	Risk of tPA related hemorrhagic transformation	Jickling (2013)
Smoking	63 genes	Inflammation associated with smoking in stroke	Cheng (2019)

TIA , Transient ischemic attack.

Stroke Diagnosis

Ischemic Stroke Versus Controls

Moore et al. (22 genes), Barr et al. (9 genes), and Tang et al. (17 genes) have compared mRNA expression at the whole genome level for IS versus control patients. There was some overlap between each study in spite of different platforms or cells for each study. Stamova et al., using the same methods as Tang et al., found that the 17 Tang et al. genes had a 94% sensitivity and 90% specificity in predicting IS versus controls in a new cohort. In addition, a 97 probe set panel correctly distinguished IS at 3 and 24 hours (86%) from controls (84%) and from myocardial infarction (75%). Though the measurement of mRNA is generally quite slow, new technology is being developed that might make it possible to measure mRNA in minutes at the bedside.

Ischemic Stroke Versus Hemorrhagic (Intracerebral Hemorrhage) Stroke

A recent whole genome study found 489 transcripts were differentially expressed between ICH and controls, and 63 between IS and controls. ICH had differentially expressed T-cell receptor and CD36 genes, and inducible nitric oxide synthase, TLR, macrophage, and T-helper pathways. IS had more noncoding RNA. ICH and IS both had angiogenesis, CTLA4 in T lymphocytes, CD28 in T-helper cells, nuclear factor of activated T-cells regulation of immune response, and glucocorticoid receptor signaling pathways. T-cell receptor and T-cell genes were sufficient to differentiate IS and ICH. Theoretically, a stroke blood test that differentiates IS and ICH seems possible.

Transient Ischemic Attack

In a study of 26 TIA and 26 controls there were 449 differentially expressed genes associated with systemic inflammation, platelet activation, and prothrombin activation. Hierarchical cluster analysis of the identified genes suggested the presence of two patterns of RNA expression in patients with TIA, one group possibly being TIA mimics. Prediction analysis identified a set of 34 genes that discriminated TIA from controls with 100% sensitivity and 100% specificity. In a follow-up study TIA ( n = 26) were compared to IS ( n = 94) and controls ( n = 44). Seventy-four genes expressed in TIA were common to those in IS with function pathways relating to activation of innate and adaptive immune systems. A prediction model using 26 of the 74 ischemia genes distinguished TIA and stroke subjects from control subjects with 89% sensitivity and specificity. In a validation cohort, 17 of 17 TIA diffusion-weighted imaging-positive/minor strokes were predicted to be ischemic, and 10 of 13 nonischemic transient neurologic events were predicted to be nonischemic. In transient neurologic events of unclear etiology, 71% were predicted to be ischemic. Transcriptomics may be useful for predicting which TIAs progress to stroke, and which are most likely TIA mimics and need less of a workup and may not need treatment.

Stroke Etiology

Cardioembolic Versus Large Vessel Versus Small Vessel

In the initial transcriptome study of IS cause, expression profiles in the blood of cardioembolic stroke patients were shown to be distinctive from those of large-vessel atherosclerotic stroke patients. Seventy-seven genes differed at least 1.5-fold between them, and a minimum number of 23 genes differentiated the two types of stroke with at least 95.2% specificity and sensitivity. Genes regulated in large-vessel atherosclerotic stroke are expressed in platelets and monocytes and modulated hemostasis. Genes regulated in cardioembolic stroke were expressed in neutrophils and modulate immune responses to infectious/ inflammatory stimuli.

In a follow-up study a 40-gene profile differentiated cardioembolic stroke from large vessel stroke with >95% sensitivity and specificity. A separate 37-gene profile differentiated cardioembolic stroke due to atrial fibrillation from nonatrial fibrillation causes with >90% sensitivity and specificity. The identified genes elucidate differences in inflammation between stroke subtypes. When applied to patients with cryptogenic stroke, 17% are predicted to be large-vessel and 41% to be cardioembolic stroke. Of the cryptogenic strokes predicted to be cardioembolic, 27% were predicted to have atrial fibrillation.

In a third study of 184 stroke patients a 41-gene profile discriminated lacunar from nonlacunar stroke with >90% sensitivity and specificity. Of the 32 small deep infarcts of unclear cause, 15 were predicted to be lacunar and 17 were predicted to be nonlacunar. The identified profile represents differences in immune response between lacunar and nonlacunar stroke.

Cryptogenic

In the next study RNA was isolated from peripheral blood of 131 cryptogenic strokes and compared with profiles derived from 149 strokes of known cause. Cryptogenic strokes were divided into cortical and subcortical. Cryptogenic strokes were predicted to be 58% cardioembolic, 18% arterial, 12% lacunar, and 12% unclear etiology. Cryptogenic stroke of predicted cardioembolic etiology had more prior myocardial infarction and higher CHA(2)DS(2)-VASc scores compared with stroke of predicted arterial etiology. Predicted lacunar strokes had higher systolic and diastolic blood pressures and lower National Institutes of Health Stroke Scale compared with predicted arterial and cardioembolic strokes. Cryptogenic strokes of unclear predicted etiology were less likely to have a prior TIA or IS.

Lacunar Stroke and Cerebral White Matter Hyperintensities

Gene profiles for the known causes of stroke have the potential for predicting the causes of cryptogenic strokes. One surprising finding for lacunar stroke was that the lacunar profile ^, did not overlap the profile for white matter hyperintensities. Function analyses suggested that white matter hyperintensities (WMH)-specific genes were associated with oxidative stress, inflammation, detoxification, and hormone signaling, and included genes associated with oligodendrocyte proliferation, axon repair, long-term potentiation, and neurotransmission. In contrast, lacunar stroke was associated with growth of myeloid cells and leukocytes, monocyte and leukocyte activation and recruitment, immune response, cardiovascular process of blood vessel, endothelial adhesion, and angiogenesis of endothelial cells. Thus, though both WMH and lacunar stroke are believed to be associated with “small vessel” disease, the peripheral immune, clotting, and blood-brain barrier responses to each are quite distinct suggesting unique pathophysiology and causes.

Stroke Biology and Complications

You're Reading a Preview

Become a Clinical Tree membership for Full access and enjoy Unlimited articles

Become membership

If you are a member. Log in here