Mechanosensitive MicroRNAs in Health and Disease

Myung-Jin Oh, Tzu-Pin Shentu, Daksh Chauhan, and Yun Fang

Department of Medicine, University of Chicago, 5841 S. Maryland Ave., MC 6026, Office M628 Chicago, Illinois 60637 This email address is being protected from spam bots, you need Javascript enabled to view it

Introduction

The increasing amount of emerging genetic data have shed light on the functional importance of small noncoding RNAs in mechanotransduction and, ultimately, the progression of a wide range of human diseases [2-4]. Small noncoding RNAs are a subgroup of RNA transcripts that are not translated into proteins. While several noncoding RNAs exist, they are typically classified by size and function (microRNAs [miRNAs], long noncoding RNAs, circular RNAs, Piwi interacting RNAs, enhancer RNAs, etc.) [5, 6]. This chapter is devoted to microRNAs (miRNAs) and their biological roles in cellular mechanotransduction, by which cells exert,

Modern Mechanobiology: Convergence of Biomechanics, Development, and Genomics Edited by Juhyun Lee, Sharon Gerecht, Hanjoong Jo, and Tzung Hsiai Copyright © 2021 Jenny Stanford Publishing Pte. Ltd.

ISBN 978-981-4800-58-7 (Hardcover), 978-0-429-29483-9 (eBook) www.jennystanford.com sense, and convert mechanical stimuli to biochemical responses. Noncoding RNAs indeed contribute to a major fraction of the mammalian transcriptional output, and many of these noncoding RNAs carry out regulatory functions by directing varied stages of gene expression, such as epigenetic modification, messenger RNA (mRNA) stability, and translational control [7, 8]. For instance, miRNAs are highly conserved small RNAs of 19-26 nucleotides that post-transcriptionally inhibit their target genes via mRNA instability and translational suppression. Disruption of the tightly regulated spatial and temporal miRNA expression is implicated in many human diseases, such as cancer, metabolic syndromes, and cardiovascular diseases [9]. Mechanistically, miRNA plays important roles in gene expression, cellular proliferation/differentiation, organ development, tissue homeostasis, and central biological processes in which mechanical cues are instrumental [10]. This chapter summarizes the regulation of miRNAs in vascular cells exposed to two major biomechanical cues, hemodynamic shear stress and cyclic stretch. We also discuss the putative role of miRNAs in regulating the responses of cells to microenvironments.

MicroRNA in Hemodynamics Sensing

Cells reside in a 3D microenvironment, in which they not only contribute to but also exert and respond to mechanical cues of varying magnitudes, directions, and frequencies [11]. This is especially important in the vasculature, where endothelial cells are constantly exposed to hemodynamics due to the flowing blood right from the time the heart starts beating [12-16]. Large amounts of data have supported the critical role of hemodynamics, particularly the shear forces mainly received by the endothelium, in regulating vascular homeostasis and pathology. Vascular health and vessel remodeling are dynamically regulated by the shear forces acting on the vessels via the endothelial interface and the resulting mechanosensing mechanisms. Meanwhile, the flow-induced mechanotransduction mechanism in the endothelium has been implicated in vascular diseases such as aneurysms, poststenotic dilations, acute lung injury, arteriovenous malformations, aortic valve diseases, and atherosclerosis. The interaction between blood flow and local vessel geometry creates complex spatiotemporal shear stresses on the vessel wall. It has now been well established that endothelial cells are mechanosensitive to different forms of blood flow. There are two types of blood flow patterns that have been established related to atherosclerosis, a disease that classically has been defined as the thickening and hardening of arteries [17]. One of these types of blood flow, which we will term as "disturbed flow”

(DF), that typically occurs at vascular sites of atypical geometries such as branches and bifurcations is characterized as complicated patterns of multidirectional hemodynamics at variable frequencies leading to fluid disturbance featuring oscillation, flow reversal, and low time-averaged shear stress (Fig. 7.1). In contrast to DF, blood flow patterns in the straight part of the blood vessels are considered to have a unidirectional flow (UF) and higher time-average shear stress. Both in vitro and in vivo studies have demonstrated that unidirectional laminar flow with high wall shear stress promotes the quiescent endothelial phenotype while DF in the arterial regions of atypical vascular geometry prone to atherosclerosis causatively activates endothelial cells [12-16].

Figure 7.1 Depiction of a bifurcated human carotid artery [1].

While these two types of blood flow are found in the human body, the third and less relevant type of flow pattern used in several studies is one in which endothelial cells are not subjected to any flow pattern, often called a static condition. Studies using microarray data and sequencing results have found multiple miRNAs that can be categorized as proatherogenic or antiatherogenic and are either upregulated or downregulated in UF and DF regions, respectively [18]. While it is generally true that miRNAs that are upregulated in UF regions are downregulated in DF regions and vice versa, a few miRNAs have been shown to have a dual role, depending on the type of shear stress and the type of study involved.

Due to the fact that flow patterns can alter cell signaling in multiple directions on the basis of their flow patterns, we will first discuss the proatherogenic miRNAs that are generally upregulated in DF regions/downregulated in UF regions, then move to miRNAs that are downregulated in DF regions/upregulated in UF regions, and finally cover miRNAs that are modulated by flow but are dependent on the type and experimental conditions, causing some controversy. While the miRNAs highlighted are in no way the complete list, we will try and cover miRNAs that have relevant findings, novel discovery, or clinical outcomes.

• miR-17-92 cluster: The M1R17HG gene encodes a single transcript that folds into six stem loops and comprises six miRNAs—17, 18a, 19a, 19b, 20a, and 92a—based by how close each miRNA was in relation to the others at the same locus. This cluster is also known as "oncomiR-1,” given its well-established role in cancer biology and in regulating cell cycle, proliferation, and apoptosis [19]. Later studies demonstrated an important role of the miR-17-92 cluster in vascular homeostasis regulated by biomechanical cues. First, many miRNAs in this cluster, particularly miR-92a, an endothelial cell-enriched miRNA, were shown to be regulated by hemodynamic forces [20, 21]. In vivo and in vitro studies demonstrated that miR-92a expression is significantly upregulated in the endothelium exposed to atherosclerosis- promoting DF and downregulated under UF. Mechanistic investigations showed that miR-92a is a major noncoding RNA that promotes endothelial activation by directly inhibiting transcription factors Kriippel-like factor-2 (KLF2) and KLF4 [20, 21], phospholipid phosphatase 3 (also known as PLPP3

and PPAP2B) [22], Sirtuin 1 [23], and integrin subunits a5 [23, 24].

• • miR-34a: This particular miRNA is found in atherosclerotic plaques and is critical for endothelial cell senescence [25,
• 26]. Fan et al. demonstrated that this miRNA was upregulated in oscillatory shear stress and downregulated by UF with high time-averaged shear stress [27]. They furthermore showed that miR-34a alters endothelial cell inflammation by increasing vascular cell adhesion molecule-1 (VCAM-1). VCAM-1 is an important inflammatory molecule that is highly upregulated in atherosclerotic plaques [28]. This, therefore, becomes clinically relevant in miR-34a, potentially altering plaque formation. However, this has currently not been studied. Other studies have also linked miR-34a proliferation in retinal pigment epithelium [29].
• • miR-663: Microarray studies looking at miRNAs on endothelial cells in DF flow regions found miR-663 to be one of the most upregulated miRNAs [30]. While in endothelial cells there have been links to inflammation [30], miR-663 has been found to be important in vascular smooth muscle cells (SMCs) as well [31]. Interestingly, miR-663 when overexpressed in SMCs caused decreased neointimal formation, an important cause of atherogenesis [31]. Furthermore, miR-663 has been tied to cell proliferation and tumor growth in nasopharyngeal carcinoma cells and pancreatic cancer [32, 33]. This suggests that miR-663 can have multiple roles in clinical outcomes of disease beyond that of mechanosensitive regulation.

Next, we will discuss the antiatherogenic miRNAs that are upregulated in UF regions and downregulated in DF regions.

• • miR-19a and miR-23b: While miR19a is part of the miR-17-92 cluster, endothelial miR-19a appears to have opposing effects in terms of upregulation by UF compared to downregulation like the rest of the miR-17-92 cluster [13, 34]. These studies tested human umbilical vein endothelial cells (HUVECs) compared to HUVECs under static conditions (cells not treated to any flow) [35]. Another miRNA that was highly upregulated in studies testing laminar shear was miR23b [36]. In vascular endothelium, miR-19a was shown to promote cell cycle arrest at Gl/S transition by inhibiting cyclinDl [35] while miR-23b mediates UF-induced Gl/Go arrest by suppressing transcription factor E2F1 and Rb hypophosphorylation, consistent with the reported roles of miR-19a and miR-23b in cancer and cell proliferation [36].
• • miR-Юа; Several reports have found miR-lOa to be decreased in DF regions and increased in UF regions both in vitro and in vivo [18, 37] and linked reduced miR-lOa expression to increased endothelial inflammation. First, Fang et al. reported that miR-Юа serves as an important negative regulator of nuclear factor kappa В (NF-кВ) activation by directly suppressing mitogen-activated kinase kinase kinase 7 (MAP3K7), also known as TAK1, and p-transducin repeat- containing gene (pTRC) [18], two key regulators of IkBoc degradation. Second, Lee et al. showed that endothelial miR- 10a reduces VCAM-1 expression via GATA6 suppression [37]. There is growing literature that miR-Юа alterations have different roles depending on the cell type [38,39]. For instance, miR-Юа represses proliferation and induces apoptosis in the ovarian granulosa cells [38]; however in acute myeloid leukemia cell studies decreased miR-Юа caused increased apoptosis, suggesting miR-Юа can cause opposing functions on different cellular pathways [39,40].

Finally, a look at the dual-modulated miRNAs that have been found increased in both UF and DF depending on the reported conditions. This sets up a controversial set of miRNAs as it is unknown how these particular miRNAs function in terms of disease progression.

• • miR-21: When compared to static conditions, endothelial miR- 21 has been reported to be upregulated by both laminar UF and oscillatory DF in a time-dependent manner [41, 42]. The proatherogeneic role of endothelial miR-21 was suggested by its function of suppressing translation, but not transcription, of peroxisome proliferators-activated receptor-a (PPARa) by З'-UTR targeting [41]. In contrast, the antiatherogenic property of miR-21 was suggested by experiments detecting decreased apoptosis and activated nitric oxide production in endothelium of miR-21 overexpression [42]. Of more clinical relevance, miR-21 is found to have roles in breast, colon, and hematological cancers [2, 43]. Similar to cancer, abdominal aortic aneurism (AAA) has also been tied to cell proliferation and apoptosis [43]. In AAA, miR-21 was once again upregulated.
• • miR-126: This is one of the most abundant miRNA clusters expressed in vascular endothelium. Schober et al. reported that DF causatively reduces miR-126-5p expression in vivo, leading to increased Notchl inhibitor delta-like 1 homolog and promoting atherosclerosis [44]. In contrast, Mondadori dos Santos etal. showed thatUF confers the anti-inflammatory endothelial phenotype by increasing the expression of miR- 126, which suppresses stromal cell-derived factor-1 SDF-1/ CXCL12 and VCAM-1 [45]. Nevertheless, administration of miR-126-5p mimics was shown to significantly lessen high fat-induced atherosclerosis in apolipoprotein E knockout (ApoE•/•) mice [44].