Multi-omic approaches for biomarker discovery in Moyamoya disease

Moyamoya disease (MMD) is a rare cerebrovascular condition characterized by a chronic and progressive narrowing of the terminal portion of the bilateral internal carotid arteries, causing the formation of an abnormal vascular network. These compensatory brain vessels often prove insufficient, leading MMD patients to severe ischemic or hemorrhagic clinical manifestations. Surgical treatment, mainly based on direct and indirect revascularization, represents the preferred procedure for MMD patients until now, for improving cerebral hemodynamics and decreasing the pathological collateral network development. The specific mechanism underlying both the progressive arterial wall thickening and the spontaneous angiogenesis of the defective moyamoya vessels remains poorly understood. Moreover, the lack of reliable animal or cellular pre-clinical models and the heterogeneous data on MMD pathophysiology have hampered the clinical validation of powerful biomarkers, as well as the development of tailored therapeutic options.

The previous investigations aimed at biomarker discovery mainly addressed patients’ peripheral blood samples, which may not reflect the real pathological changes of MMD cerebral vessels. On the other hand, prior interesting studies involving cerebral artery specimens (e.g. middle cerebral artery [MCA]) were performed through RNA microarray techniques, which have several limitations as compared to high-throughput RNA sequencing (RNA-seq). As an example, the long noncoding RNA profile of MMD patients’ MCA provided data regarding antibacterial response, T-cell receptor signalling pathway and cytokine production.¹ Interestingly, Xu et al. carried out an RNA-seq analysis of MMD patients’ MCA, as compared to atherosclerosis-associated intracranial artery stenosis/occlusion. The authors identified several differential expressed genes mainly involved in extracellular matrix organization and mitochondrial function, thus highlighting novel insights into disease pathogenesis.²

Since the challenging sampling of cerebral artery specimens for transcriptomic studies, other ultrasensitive techniques were recently carried out for molecular profiling of circulating biomarkers from cerebrospinal fluid (CSF) or blood. The study by Ota et al. through a next-generation sequencing (NGS) approach demonstrated that specific changes occurred in the expression levels of extracellular vesicle-derived microRNAs (miRNAs), extracted from intracranial CSF of MMD patients when compared to controls.³ The authors suggested that MMD has a specific regulatory mechanism for angiogenesis, different from that found in other ischemic disorders. Proteomic approaches towards MMD patients’ circulating fluids have already been reported. Tandem mass tag (TMT)-labelled proteome analysis was performed on serum-derived exosomes, extracted from pure ischemic or hemorrhagic MMD patients and healthy controls.⁴ Bioinformatic analysis revealed deregulated cell growth/maintenance and indicated disturbed actin dynamics in MMD versus control subjects, as well as immunity dysfunction in hemorrhagic MMD. Similarly, serum proteins were identified using TMT labelling combined with liquid chromatography-tandem mass spectrometry (LC-MS/MS).⁵ Among the differentially expressed proteins, apolipoprotein E (APOE) was selected and evaluated to determine its feasibility as a candidate diagnostic marker of MMD. Another study recently carried out by He and colleagues investigated the potential impact of abnormally expressed serum proteins in MMD pathogenesis.⁶ The authors explored a large number of MMD serum samples by using a data-independent acquisition MS proteomic approach and found differences among MMD subgroups versus control subjects. Interestingly, the quantitative analysis revealed that Filamin A (FLNA) was significantly up-regulated in all MMD subgroups when compared to control subjects. Such evidence is particularly relevant if considering the results obtained through a comprehensive, label-free, quantitative MS-based proteomic characterization of dura mater samples from MMD patients.⁷ Indeed, exactly FLNA emerged among the most abundant detected proteins and was indicated as a second hit gene in MMD pathogenesis, by promoting MMD-like vascular formation. It is a large actin-binding protein, important for cell motility due to providing scaffolds for multiple cytoskeletal proteins.

Despite numerous findings collected in the proteomics field, very little is known about the overall lipidomics of MMD, or changes in the lipid composition in patients’ circulating fluids. Nevertheless, it has been recently reported that the main susceptibility genetic factor identified in MMD patients (i.e. Ring Finger Protein 213) is involved in lipid metabolism.⁸ A previously performed untargeted lipidomic analysis showed a cumulative depletion of lipid assets in plasma of MMD patients, as compared to healthy donors.⁹ Specifically, a decrease in membrane complex glycosphingolipids circulating in MMD plasma, with respect to healthy donors, was observed, likely suggestive of cerebral cellular recruitment.

More recently, He and colleagues reported a comprehensive and detailed multi-omics analysis of MMD molecular profiles, including genomics (whole-exome sequencing), transcriptomics (RNA-seq) and metabolomics (ultra-high-performance LC-high-resolution MS, UHPLC-HRMS) approaches (Figure 1).¹⁰ Notably, the untargeted metabolomics suggested lysophosphatidylcholine (i.e. LPC 16:1-2) as potential diagnostic blood (i.e. serum) biomarker, able to identify MMD patients from healthy subjects and to differentiate ischemic MMD patients from the hemorrhagic ones. LPC is a bioactive lipid known to cause inflammation, endothelial dysfunction and proliferation/migration of vascular smooth muscle cells, displaying effects on angiogenesis function and vascular wall integrity of cerebral vessels. Interestingly, RNA-seq analysis of MMD neurosurgical specimens (e.g. superficial temporal artery; MCA) revealed that the gene expression of LPC-related enzymes (i.e. phospholipase A1 Member A; phospholipase A2 group IIA) was consistently decreased, thus corroborating the lipidomic findings. Such evidence was also confirmed by conventional immunoassays aimed at determining the expression levels of these LPC-related enzymes in MMD patients’ peripheral blood. The study conclusions confirmed that multi-omic signatures are associated with MMD pathophysiological features and suggested a candidate blood biomarker for MMD patients’ stratification.

Data shown in all these research studies strongly underlined the crucial role that omics approaches can exert in attempting to better clarify MMD pathogenesis (Figure 2). Indeed, omics techniques coupled and sustained by experimental in vitro models could pave the way for the search for specific MMD biological markers, hopefully leading to improved disease diagnosis and screening.

In this regard, multi-omic approaches may turn out to be crucial in patients’ risk stratification and precision medicine, allowing for the selection and customization of medical treatment based on individual signatures. In addition, combining multiple omics strategies may also shed light on potential novel targets for MMD therapeutic intervention, as shown above for FLNA and LPC.^{6, 7, 10}

Laura Gatti: Conceptualization; original draft writing and editing. Gemma Gorla: Artworks. Laura Gatti, Anna Bersano, Gemma Gorla, Giuliana Pollaci, Tatiana Carrozzini, and Antonella Potenza: Critical revision, editing and approval of the final version of the article.

The authors declare no conflict of interest.

Not Applicable.