Flow pattern-dependent endothelial cell responses through transcriptional regulation

ABSTRACT

Blood flow provides endothelial cells (ECs) lining the inside of blood vessels with mechanical stimuli as well as humoral stimuli. Fluid shear stress, the frictional force between flowing blood and ECs, is recognized as an essential mechanical cue for vascular growth, remodeling, and homeostasis. ECs differentially respond to distinct flow patterns. High laminar shear flow leads to inhibition of cell cycle progression and stabilizes vessels, whereas low shear flow or disturbed flow leads to increased turnover of ECs and inflammatory responses of ECs prone to atherogenic. These differences of EC responses dependent on flow pattern are mainly ascribed to distinct patterns of gene expression. In this review, we highlight flow pattern-dependent transcriptional regulation in ECs by focusing on KLF2 and NFκB, major transcription factors responding to laminar flow and disturbed flow, respectively. Moreover, we introduce roles of a new flow-responsive transcriptional co-regulator, YAP, in blood vessel maintenance and discuss how these transcriptional regulators are spatiotemporally regulated by flow and then regulate EC functions in normal and pathological conditions.

KEYWORDS:

• shear stress
• blood flow
• transcription factor
• YAP
• atherosclerosis

Overview of endothelial cell (EC) responses to shear stress

Endothelial cells (ECs) lining blood vessels are continuously exposed to blood flow. While blood flow affects ECs by conveying molecules including growth factors, cytokines, and hormones, it acts on ECs by providing mechanical forces. ECs sense mainly 2 types of mechanical stimuli from blood flow: one is shear stress, the frictional force tangential to ECs; the other is mechanical strain, the force perpendicular to the direction of flow, related to rhythmic heart beating.^1-4 Flow-mediated mechanical stimuli regulate various aspects of vascular functions including blood vessel formation, maintenance, and remodeling by acting on ECs and smooth muscle cells (SMCs). Vascular remodeling, which includes alterations of the distribution and diameter of blood vessels, allows them to meet the demand of tissues so that they receive appropriate perfusion with maximal efficiency.^5,6 Although our knowledge about mechanical strain- or hydrostatic pressure-mediated mechanotransduction in ECs is limited, extensive studies provide much information about shear stress-induced multiple mechanotransduction signals.^1,2

Shear stress is a weak force that is less than 1/100 of mechanical strain.⁷ Therefore, highly sensitive sensing mechanisms is necessary to sense shear stress. Because flow is directly in touch with apical surfaces of ECs, the structures located on the apical side, including primary cilia, glycocalyx, ion channels, and lipid bilayer membranes, have been proposed as direct shear stress sensors.^8-12,13 A mechanosensory complex at cell-cell junctions, consisting of PECAM-1, VE-cadherin, and VEGFR2,¹⁴ and a mechanosensitive cationic channel, Piezo1,¹⁵ have been characterized as sensor molecules responding to flow. In addition, caveolae and vimentin intermediate filaments are also proposed as shear stress sensors,^16,17 although they are not mainly located on the apical surfaces of ECs. It still remains a central topic how sear stress is sensed by ECs in vivo. These sensing mechanisms might act individually depending on the situations. Alternatively, multiple sensing mechanisms might function simultaneously in a coordinated manner.

ECs differentially respond to different patterns of flow. In straight portions of a blood vessel, ECs are exposed to uniform laminar shear stress (Fig. 1A). In contrast, the flow pattern is variously altered dependently on vessel curvatures, stenosis, and branching (Fig. 1A). In those regions, the flow becoming reversed and separated gives low shear stress to ECs.^1,18-20 High laminar shear stress promotes EC survival and renders ECs quiescent with low proliferation rate (Fig. 1B). Laminar shear stress suppresses cell cycle progression by upregulating cyclin-dependent kinase inhibitor 1 (p21) and reduced phosphorylation of retinoblastoma protein (Rb).^21,22 On the contrary, disturbed flow enhances apoptosis and proliferation of ECs, thereby leading to increased EC turnover (Fig. 1B).^23-26 Furthermore, disturbed flow enhances vascular permeability and promotes inflammatory gene expression that induces accumulation of monocytes.^27,28 It also promotes endothelial senescence via a p53 dependent signaling pathway.²⁹

Figure 1. Different endothelial responses to different patterns of flow. (A) ECs lining straight blood vessels are exposed to laminar shear stress, whereas those linging branching or curved blood vessels are exposed to disturbed flow. (B) High laminar flow induces quiesent phenotypes in ECs including low proliferation rate, reduced cell death, and anti-inflammatory gene expression (left panels), while low or disturbed flow induces atheroprone phenotypes in ECs including high proliferation rate, increased cell death, and inflammatory gene expression (right panels). High laminar flow induces the expression of KLF2 and KLF4 via the MEK5/ERK5/MEF2 signaling pathway. KLF2 promotes the activation of Nrf2 and both KLF2 and Nrf2 promote quiescent phenotypes of ECs. Low or disturbed flow upregulates the expression of NFκB through JNK and its downstream ATF2. Flow-driven NFκB activation is mediated by integrin-dependent signaling upon binding to fibronectin. AP-1-, EGR1-, and YAP-dependent gene expression is also induced by disturbed flow. RhoA and F-actin are responsible for determinging the localization of YAP in sheared ECs, although it is not determined whether disturbed flow regulates the canonical Hippo pathway.

These disturbed flow-dependent regulation of ECs is involved in the onset and progression of atherosclerosis. High steady laminar flow or pulsatile flow is atheroprotective, while low or disturbed flow creates an atheroprone environment for blood vessels.^1,18-20 Indeed, atherosclerotic lesions develop primarily at bifurcations of medium-sized arteries in which disturbed flow is observed.³⁰ However, it is unknown how ECs can distinguish the different flow patterns. Distinct shear stress sensors might be used according to flow patterns.

Transcriptional regulators responsible for EC responses to flow

Shear stress induces mechanotransduction signals inside of ECs, although the surface mechanosensory systems of ECs are yet unclear. The mechanotransduction signals are also differentially regulated by different flow patterns.^1,18,20 It is clear that shear stress affects EC functions mainly through global changes in gene expression mediated by flow-regulated transcriptional factors (TFs). Kruppel-like factor 2 (KLF2) is a major flow-regulated transcription factor, because KLF2 regulates ∼46% of highly flow-responsive genes in ECs.^31,32 Atheroprotective laminar flow, but not disturbed flow, upregulates KLF2 and KLF4, another member of KLF family TFs.^32-35 KLF2 promotes the activation of antioxident TF, Nrf2. Both KLF2 and Nrf2 induce quiescence of ECs characterized by the reduction of inflammatory responses (Fig. 1B).^36-38 Nuclear factor-κb (NFκB) is a representative TF that is enhanced by disturbed flow.^39,40 Whereas expression of NFκB is driven by disturbed flow in a sustained manner, longer laminar shear stress reduces the activity of NFκB below the base line. Similar to NFκB, activator protein-1 (AP-1) and early growth response-1 (EGR1) are also induced by disturbed flow (Fig. 1B).⁴⁰ Through the coordinated expression of these TFs, disturbed flow upregulates pro-atherosclerotic genes.^1,18 In lymphatic endothelial cells (LECs), disturbed flow induces FOXC2, a forkhead TF important for lymphatic vessel development and lymphatic valve formation.⁴¹

Recently, we and other group have identified Yes-associated protein (YAP), a transcriptional cofactor downstream of the Hippo signaling pathway, as a novel mechanotransducer in ECs.^42-44 Endothelial YAP responds to flow-mediated mechanical stimuli both in vitro and in vivo and is important for vessel maintenance and atherosclerosis.

KLF2 – a transcription factor responsible for atheroprotective flow-induced regulation of vascular tones

KLF2 is a member of the zinc finger family of TFs and is expressed in ECs in vivo.^45,46 High laminar shear stress and unidirectional pulsatile shear stress, both of which are atheroprotective flow, induce the expression of KLF2 and its paralog KLF4 in a sustained manner via the MEK5/ERK5/MEF2 signaling pathway.^32-35 KLF2 expression in ECs is dependent on blood flow in vivo, because endothelial KLF2 expression is lost without blood flow.^32,47 KLF2 expression is increased in the region exposed to high shear stress in developing chick embryos.⁴⁸ In mouse embryo, endothelial KLF2 expression is higher in arteries than in veins.⁴⁹ Consistently, higher KLF2 expression is induced by unidirectional pulsatile flow mimicking arterial flow than by steady laminar flow in cultured ECs.³⁵ In human, KLF2 expression is relatively low at the outer wall of the bifurcation of the aorta and iliac arteries and at the branch of the common carotid artery from the aortic arch, where blood flow is disturbed and atheroprone.^35,50 Similarly in cultured ECs, atheroprone flow does not enhance KLF2 expression.³² Thus, KLF2 responds to atheroprotective flow.

In response to atheroprotective flow, KLF2 functions as a transcriptional integrator for multiple vascular protective functions.^31,32,51 KLF2 controls vascular tones by inducing the expression of endothelial nitric oxide (NO) synthase (eNOS) and caveolin1.^32,50 Under laminar shear stress, KLF2 reduces vessel permeability, thrombosis, and endothelial metabolism.^32,52-54 In addition, KLF2 represses inflammatory responses induced by IL-1β or TNF-α^53,55 Hemizygous deficiency of KLF2 increases atherosclerosis in apolipoprotein E (apoE) deficient mice.⁵⁶ However, the contribution of endothelial KLF2 expression to this phenotype remains to be defined. KLF4 also represses inflammatory responses in ECs.^57,58 Endothelial specific loss-of-function of KLF4 results in increased atherosclerosis in apoE deficient mice,⁵⁸ demonstrating atheroprotective functions of endothelial KLF4.

In addition to the vascular protective role of KLF2, an angiogenic role of KLF2 is reported in zebrafish. Klf2a is important for flow-stimulated angiogenesis during aortic arch development.⁵⁹ Klf2a induces expression of an EC-specific microRNA, miR-126, to activate VEGFR2 signaling. KLF2 is also responsible for ectopic angiogenesis in cerebral cavernous malformation (CCM) mutant zebrafish.⁶⁰ The question whether KLF2 is a general regulator of flow-stimulated angiogenesis should be solved in the future. In zebrafish, Klf2a is also important for the maintenance of haematopoietic stem cells (HSCs), another flow-regulated process, because knockdown of Klf2a resulted in the loss of HSC.⁴⁷ These phenotypes are shown in the morphants of klf2a using morpholino antisense oligonucleotides (MOs). Analyses using klf2 mutant fish would be more beneficial to understand its physiologic role. klf2 mutant zebrafish (klf2a and klf2b) exhibit reduced diameter of the axial blood vessels and increased blood flow velocity,⁶¹ indicating an essential function of KLF2 in the vascular tone regulation. These mutant fish also show defects in flow-induced EC polarization.⁶¹ Consistently, constitutive and EC-specific KLF2 knockout mice die at embryonic day 10.5 owning to lack of vascular tone, bleeding, and cardiac dysfunction,^46,49 although vessel patterning and density appear normal in these knockout mice.^46,49,62 KLF2 deficient mice exhibit disorganized SMC due to the failure of SMC migration.⁶² Thus, endothelial KLF2 might affect smooth muscle tone in vivo.

Gene expression studies using cultured ECs identify numerous KLF2 target genes that could affect vascular smooth muscle tone.^32,50 Hergenreider et al. reveal that KLF2 upregulates miR-143/145 controling SMC phenotype.⁶³ In HUVECs, shear stress induces miR-143- and miR-145-containing extracellular vesicles through KLF2. These miRNAs can be transferred to SMCs and regulate SMC differentiation, suggesting that endothelial KLF2 regulates communication between ECs and SMCs under shear stress.

In mice, KLF2 is detected in the endocardium overlying the endocardial cushions and the common outflow tract that are exposed particularly to high shear stress.⁴⁹ Also, in zebrafish, Klf2a is highly expressed in the endocardium within the atrioventricular canal (AVC) in a manner dependent on oscillatory flow.^64,65 klf2a mutant zebrafish exhibit variety of abnormal phenotypes in atrioventricular valve formation, including valve deficiency or thickened valve leaflets.⁶⁴ In endocardium, KLF2a regulates fibronectin synthesis in the AVC for proper valve formation, suggesting the essential responses of endocardial endothelial cells to high shear stress.

NFκB – a transcription factor involved in inflammatory responses to atheroprpne flow

NFκB is a ubiquitous TF that participates in inflammatory, immune, and developmental processes.⁶⁷ NFκB family members, p65 (RelA), RelB, c-Rel, p105/p50 (NFκB1), p100/p52 (NFκB2), exist as homo- or heterodimers. Among them, p50/p65 is predominantly expressed in ECs.⁶⁸ NFκB is inactive when localized in the cytoplasm, where it is retained with inhibitors of NFκB (IκB) proteins.⁶⁹ After stimulation, IκB proteins are phosphorylated by IκB kinase (IKK) complex, ubiquitinated, and subjected to subsequent degradation in 26S proteasome. NFκB then translocates into the nucleus, thereby promoting the expression of target genes.

In cultured ECs, disturbed flow, and steady or pulsatile low shear stress enhance nuclear NFκB (p65) and its transcriptional activity in a sustained manner.^39,40 In contrast, higher laminar shear stress induces NFκB activation in a rapid and transient manner (∼30 min).³⁹ Notably, the extent of NFκB activation in response to laminar shear stress is determined by flow direction relative to the cell axis.⁷⁰ Whereas laminar flow in a direction which highly diverges from the cell axis induces NFκB activation, laminar flow parallel to the cell axis does not. Therefore, ECs that have already been pre-flowed and aligned to the flow direction do not respond to laminar flow parallel to the cell axis. Besides flow direction, the frequency components of shear stress waveforms are important for effective NFκB activation.⁷¹ The IKK-IκB pathway mediates nuclear translocation of NFκB induced by shear stress. Consistently, IKK activation and subsequent IκB degradation are induced by shear stress.⁷²

Nuclear NFκB binds to a shear stress responsive element (SSRE) found in the promoter of atherogenic adhesion molecules, including ICAM-1, VCAM-1, and E-selectin,⁷³ as well as PDGF, a main regulator of SMC development.⁷⁴ Because expression of ICAM-1 and VCAM-1 is increased in ECs of atherosclerotic-prone regions and atherosclerotic lesions,^75,76 NFκB may participate in early events in atherosclerotic progression. Consistently, NFκB is detected in the nucleus of ECs in the atherosclerotic lesions.⁷⁷

In arterial ECs, NFκB (p65) is highly expressed in atheroprone regions where blood flow is disturbed.^27,78 In these ECs, disturbed flow enhances NFκB expression through JNK and its downstream transcription factor ATF2.⁷⁸ Unexpectedly, in these NFκB-expressing ECs, nuclear translocation of NFκB is detected only in a small population (< 15%) probably because of higher IκB expression in these regions.^27,79 Thus, in the ECs experiencing disturbed flow in vivo, NFκB is not fully activated but is primed for the activation to respond to activation stimuli including atherogenic diet. Thus, disturbed shear stress might not be sufficient for full activation of NFκB in vivo. Additional extracellular factors including inflammatory cytokines, eg, TNF-α, and subendothelial ECM may be required for triggering NFκB activation. For example, increased fibronectin deposition is associated with atherosclerotic progression in apoE deficient mice.^80,81 Flow triggers NFκB activation in ECs on fibronectin but not on collagen or laminin.^80,82 Integrin αvβ3 and α5β1 are activated by shear stress on fibronectin but not on collagen^83-85 and mediate shear stress-induced NFκB activation in cultured ECs.^86,87 Thus, flow regulates endothelial NFκB activity in concert with surrounding microenvironment and integrin signaling. NFκB mediates an increase in fibronectin expression under disturbed flow, thereby creating a positive feedback loop to maintain inflammatory responses.⁸²

Several lines of evidence using an in vitro model and a mouse retinal neovascularization model suggest an angiogenic role of NFκB.^88-90 In zebrafish embryo, NFκB signaling functions in maintaining vessel integrity and stabilization.⁹¹ Therefore, flow-regulated NFκB might also functions in normal vascular development besides atherosclerosis.

YAP – a novel transcriptional co-activator responding to blood flow in vivo

YAP, originally known as a final effector molecule of the Hippo signaling pathway,⁹² has been recently identified as a mediator for mechanical stimuli.⁹³ YAP is a transcription cofactor that shuttles between the cytoplasm and nucleus where it associates with transcriptional factors. Among the TFs YAP potentially binds to including RUNX1, RUNX2, Smad, p73, and Nkx2.5, TEA domain (TEAD) family members are regarded as the main partners of YAP.^93-95

The Hippo signaling pathway has emerged as a key negative regulator of tissue growth in Drosophila and mammalian systems.⁹⁶ The Hippo signaling pathway negatively regulates YAP by inducing nuclear export of YAP. In the Hippo signaling cascade, MST1/2 kinases phosphorylate and activate LATS1/2 kinases, which in turn directly phosphorylate YAP, thereby exporting YAP from the nucleus.^92,97 Hippo-mediated YAP phosphorylation is a fundamental input for YAP regulation but not the only one, because nuclear localization of YAP can be often regulated by upstream components independent of the core Hippo signaling cascade or YAP phosphorylation by LATS1/2.^42,98-101

The mechanical stimuli, which are generated by extracellular matrix rigidity, cell strain, and cell density, control nuclear translocation of YAP through the modulation of filamentous-actin (F-actin) structures.^{98,99,102-105} Recently, we and other groups have shown that endothelial YAP responds to shear stress. In cultured ECs, YAP and its paralogue, transcriptional coactivator with PDZ-binding motif (TAZ, also known as WWTR1), translocate into the nucleus and promotes TEAD-dependent transcription by disturbed or oscillatory flow in a sustained manner (Fig. 2).^43,44,106 On the contrary, unidirectional laminar shear stress transiently induces nuclear translocation of YAP in acute phase (10–60 min).⁴² However, because YAP is excluded from the nucleus under longer laminar shear stress (∼6 h), its transcriptional activity is suppressed.⁴⁴ Thus, the localization of YAP might be controlled by different flow patterns. Similar to NFκB, the response of YAP to flow might vary according to flow direction relative to the cell axis. YAP phosphorylation is reduced by disturbed flow but is not altered by acute laminar shear stress.^42-44 It is not determined whether shear stress regulates the canonical Hippo signaling pathway. F-actin appears to be a common regulator of YAP under different patterns of flow. RhoA, a small GTPase involved in assembly and bundling of F-actin, regulates YAP in sheared ECs (Fig. 1B).^43,44 Consistently, thick cortical actin bundles along cell-cell boundaries are commonly assembled by both transient laminar shear stress and longer (48 h) oscillatory shear stress in ECs, whereby nuclear localization of YAP is enhanced.^42,106,107 Angiomotin (AMOT) links actin organization to YAP localization. AMOT sequesters YAP in the cytoplasm via a direct interaction in the static state.^42,108-110 Upon acute laminar shear stress, part of AMOT pool preferentially binds to enhanced cortical actin bundles and thereby releases YAP. Subsequently, released YAP enters into the nucleus to promote TEAD-dependent transcription.⁴² The similar mechanism might also control nuclear localization of YAP under disturbed flow.

Figure 2. Flow-regulated YAP-dependent transcriptional activity, and its targets, and physiologic/pathological functions. While acute laminar shear stress induces nuclear translocation of YAP and enhances TEAD-dependent transcription, longer laminar shear stress reverses them. In contrast, disturbed flow induces nuclear translocation of YAP in a sustained manner. In response to disturbed flow, YAP enhances a variety of genes involved in cell cycle and inflammatory responses. Then, YAP regulates EC proliferation, vascular maintenance, and atherogenesis.

We developed the transgenic (Tg) zebrafish systems that could indirectly monitor spatio-temporal localization and transcriptional activity of Yap1 (zebrafish YAP) in living zebrafish and revealed that endothelial Yap1 responded to blood flow in vivo.^42,111,112 In accordance with YAP regulation by different flow patterns in cultured ECs, Yap1-dependent transcription is sustained in the vessels including the dorsal longitudinal anastomotic vessel (DLAV), where blood flow is disturbed, while it is transient in the dorsal aorta (DA), where blood flow is unidirectional.⁴² Also in mice, in outer curvature of aortic arch and thoracic aorta, where blood flow is unidirectional, YAP is predominantly localized in the cytoplasm, while in the inner curvature of aortic arch, where blood flow is disturbed, YAP is mainly localized in the nucleus.⁴⁴

In cultured ECs, YAP regulates cell cycle progression.⁴³ YAP promotes hyperphosphorylation of RB protein and controls the expression of cell cycle regulatory genes in ECs exposed to shear stress (Fig. 2). In zebrafish, Yap1 is important for the maintenance of some blood vessels including the DLAV and the central artery (CtA), since yap1 mutant larvae exhibit enhanced lumen stenosis in these vessels (Fig. 2).⁴² Future studies are needed to elucidate the mechanisms by which Yap1 stabilizes blood vessels in response to blood flow. In addition, nuclear translocation of YAP and TAZ leads to disturbed flow-induced inflammatory responses by increasing JNK (Fig. 2).^43,44 Consistently, flow-induced nuclear translocation of YAP/TAZ is related to atherosclerosis. CTGF and CYR61, well-characterized YAP/TAZ target genes, are highly expressed in human atherosclerotic lesions.^113-115 In apoE deficient mice, increased atherosclerotic plaques are induced by EC-specific overexpression of YAP, whereas the overall size of atherosclerotic plaques is reduced by EC-specific YAP knockdown or by blockade of YAP/TAZ translation in the partial ligation (PL)-operated left carotid artery (LCAs).^43,44 Thus, inhibition of YAP will become a promising therapeutic strategy against atherosclerosis.

We also report flow-independent nuclear translocation of YAP in vessel sprouts of zebrafish brain vessels.⁴² Nuclear localization of YAP is similarly observed in the angiogenic front of the mouse retina where flow is absent.¹¹⁶ In the developing mouse retina, YAP functions as an important regulator for angiogenesis through the regulation of Angiopoetin-2 expression. Thus, YAP could be regulated in both flow-dependent and flow-independent manners in ECs. It would also be a next challenge to elucidate what extracellular stimuli regulate YAP functions during angiogenesis in vivo.

Crosstalks among shear stress-regulated transcriptional factors

NFκB and KLF2 signaling mutually play antagonistic roles in ECs.¹¹⁷ NFκB reduces KLF2 expression by inhibiting MEF2 transcriptional activity.¹¹⁸ This inhibitory effect is due to the cooperative action of NFκB (p65) and histone deacetylases (HDAC4 and HDAC5). On the other hand, KLF2 and KLF4 inhibit NFκB function and expression of adhesion molecules.^55,57 Similarly, MEF2 that can stimulate KLF2 expression also inhibits the cytokine-induced NFκB activation.¹¹⁹ KLF2 alters neither nuclear translocation nor DNA-binding property of NFκB. Instead, KLF2 binds to CBP/p300, a coactivator required for optimal NFκB activity. By competing with NFκB for binding to CBP/p300, KLF2 reduces NFκB-dependent transcriptional activity.⁵⁵ Furthermore, KLF2 inhibits nuclear activity of ATF2, one of the heterodimeric components of AP-1, in response to laminar shear stress.¹²⁰ Thus, shear stress-regulated TFs interact with each other in ECs. Therefore, it is assumed that ECs differentially respond to different flow patterns through these crosstalks between TFs. The Hippo-YAP signaling coordinately functions with various other key signaling pathways including TGFβ/SMAD, Delta/Notch, JAK/STAT, and Wnt/β-catenin signaling.⁹⁵ Therefore, YAP might associate with the flow-regulated TFs in a complex transcriptional network. Because disturbed flow-driven YAP/TAZ activation in ECs promotes the expression of inflammation-related genes by enhancing JNK activity,⁴⁴ there might be a functional relationship between YAP/TAZ and NFκB signaling in sheared ECs.

Concluding remarks

Transcriptional regulation is a fundamental process in EC responding to blood flow-mediated mechanical stimuli. Multiple expression profiling studies have uncovered shear stress-responsive genes regulated by various flow patterns both in cultured ECs and in vivo.^{32,35,121-124} Expression of atherosclerosis-related genes is closely related to disturbed flow in vivo.^79,125,126 However, it is not fully understood why atherosclerosis-related genes should be enhanced in ECs where flow is disturbed. This might be due to the maladaptation of ECs to disturbed blood flow.²⁰ Alternatively, the genes upregulated by disturbed flow might have some positive functions in normal blood vessels. For example, the inflammatory NFκB signaling is involved in the HSC emergence in the DA and the vessel integrity during normal development of zebrafish.^91,127,128 YAP also has dual functions, vessel maintenance and atherosclerosis, in the regions where blood flow is disturbed.^42-44 It would be interesting to investigate non-pathological function of disturbed flow in vivo. In addition, shear stress is not a sole mechanical stimulus mediated by blood flow in living animals. Therefore, it is important to understand how shear stress and other mechanical factors including mechanical strain cooperatively regulate gene expression.

Disclosure of potential conflicts of interest

No potential conflicts of interest were disclosed

Acknowledgments

We thank all laboratory members for their help, advice, and encouragement.

Funding

The studies in our laboratory were supported by JSPS KAKENHI (No. 16 H02618 to N.M.; No. 17K08560 to H.N.); AMED-CREST (No. 13414779 to N.M.)