Recent Insights Concerning Autophagy and Endothelial Cell Nitric Oxide Generation

Abstract

Although endothelial cell (EC) dysfunction contributes to the etiology of more diseases than any other tissue in the body, EC metabolism is an understudied therapeutic target. Evidence regarding the important role of autophagy in maintaining EC homeostasis is accumulating rapidly. Here we focus on advances over the past two years regarding how EC autophagy mediates EC nitric oxide generation in the context of aging and cardiovascular complications including coronary artery disease, aneurysm, and stroke. In addition, insight concerning the efficacy of maneuvers designed to boost EC autophagy in an effort to combat cardiovascular complications associated with repressed EC autophagy is discussed.

Introduction

Autophagy is a highly conserved trafficking system whereby intracellular cargo is delivered to lysosomes for degradation by acid hydrolases, producing substrates that can be recycled for use in new biosynthetic reactions, or diverted to metabolic pathways that generate ATP [1, 2]. Defective autophagy leads to the accumulation of misfolded proteins or damaged organelles, thereby driving pathologies such as neurodegeneration. Challenging the existing paradigms about how autophagy controls cellular homeostasis, recent advances indicate endothelial cell (EC) autophagy influences EC synthesis and release of vasoactive molecules which contributes to vascular smooth muscle cell tone. Of the factors that cause smooth muscle cells to relax, myriad studies indicate decreased synthesis of nitric oxide (NO) by EC NO synthase (eNOS) and / or increased destruction of NO contributes importantly to vascular dysfunction associated with aging and cardiovascular diseases, but precise molecular mechanisms are unclear. Here we highlight new insight into how EC autophagy impacts EC NO generation in the context of aging and cardiovascular disease (Figure 1).

Figure 1. Recent research.

The number of publications identified by keywords “autophagy and endothelial cell” has grown over the past decade (A). This review focused upon publications selected from 41 that were identified by keywords “autophagy and endothelial cell and nitric oxide” from 2020, 2021, and between January and July 2022 (B).

Autophagy and EC NO generation and aging

LaRocca et al. first reported that steady-state autophagy is compromised at baseline in primary arterial ECs from older vs. adult male subjects that subsequently displayed impaired brachial artery vasodilation upon acetylcholine infusion [3]. Because additional risk factors associated with aging have potential to impair eNOS activity, Bharath et al. discerned whether a lack of EC autophagy per se was causal. Indeed, activating phosphorylation of eNOS at Ser1177 (p-eNOS^S1177) and NO generation were prevented, whereas reactive oxygen species (ROS) production and inflammatory responses were amplified, in bovine aortic endothelial cells (BAECs) exposed to shear stress after siRNA-mediated knockdown of Atg3, a protein necessary for formation of autophagosomes [4]. These findings were confirmed by others [5, 6], and evidence supporting a molecular mechanism was revealed [7]. Impaired EC autophagy suppresses glucose transporter 1 expression and EC glycolysis, resulting in deficient ATP synthesis and release, and defective purinergic 2Y₁-receptor (P2Y₁-R) mediated signaling to eNOS via protein kinase C δ.

Autophagy is a dynamic process and it is instructive to differentiate between trafficking of the autophagosome to the lysosome e.g., autophagic flux, and lysosomal degradation of the autophagosome e.g., lysosomal function. Bafilomycin A1 (BAF A1) is a V-ATPase inhibitor that blocks the terminal phases of autophagy by impairing autophagosome-lysosome fusion, thereby preventing degradation of autophagolysosome contents. Increased accumulation of LC3-II and p62 in the presence vs. the absence of BAF A1 indicates increased autophagosome formation, and is an estimate of autophagic flux.

Cho et al., determined the impact of aging on autophagic flux [8*]. Compared to the respective vehicle treatment, shear-stress increased LC3-II and p62 to a greater extent in ECs isolated from adult but not older mice after treatment with BAF A1. This strongly suggests that aging represses EC autophagic flux. In addition to these findings, and congruent with an earlier study wherein autophagy was genetically repressed, shear-stress increased glycolysis, ATP production, p-eNOS^S1177, and NO generation in ECs from adult but not older mice [8*].

Translating findings from ECs exposed to shear-stress in vitro, to ECs exposed to active hyperemia, the same authors demonstrated that autophagy and p-eNOS^S1177 activation were robust in arterial ECs from adult but not older mice after 60-min treadmill-running [8*]. Providing translational relevance from mice to humans, Cho et al. substantiated that steady state autophagy is compromised in primary arterial ECs from older vs. adult male volunteers that exhibit impaired peripheral arterial function [8*, 9]. Importantly, even though active hyperemia evoked by rhythmic handgrip exercise elevated radial artery shear rate similarly from baseline between groups, autophagy initiation, p-eNOS^S1177 activation, and NO generation occurred in radial artery ECs obtained from adult but not older subjects.

The in vitro and in vivo findings from ECs isolated from older mice, coupled with results from primary ECs obtained from older humans, suggest an inability to upregulate EC autophagy might contribute to the age-associated repression of peripheral arterial function. Accordingly, compromised intraluminal flow-mediated vasodilation displayed by femoral arteries from older mice was recapitulated in vessels from adult animals by : (i) NO synthase inhibition; (ii) acute autophagy impairment using 3-methyladenine (3-MA); (iii) inducible disruption of Atg3 specifically in ECs; (iv) P2Y₁-R blockade; and (v) germline depletion of P2Y₁-Rs. Importantly, P2Y₁-R activation using 2-methylthio-ADP improved vasodilatory capacity in arteries from : (i) adult mice treated with 3-MA; (ii) adult Atg3^EC−/− mice; and (iii) older animals with repressed EC autophagy [8*]. Collectively, NO-mediated arterial dysfunction evoked by EC autophagy compromise is improved by activating P2Y₁-Rs. G protein-coupled P2Y receptors might represent a target to manipulate endothelial dysfunction that is secondary to or associated with defective EC autophagy in the context of aging. [10]

Autophagy and EC NO generation and coronary artery disease

The Beyer / Gutterman group established that shear-induced EC release of NO and endothelium-derived hyperpolarizing factors (EDHFs) stimulate flow-mediated vasodilation of adipose arterioles from healthy adult volunteers. In older individuals [11] and coronary artery disease (CAD) patients [12], however, a shift toward mitochondrial generated hydrogen peroxide (H₂O₂) governs this response. Based on findings that : (i) EC autophagy promotes EC NO formation and limits ROS generation; [4–8*, 13] (ii) telomerase reverse transcriptase (TERT; the catalytic subunit of telomerase) inhibits mitochondrial-derived ROS and induces a switch in the mechanism of dilation from H₂O₂ to NO in adipose arterioles from CAD patients [14–16] and (iii) overexpression of TERT in mouse embryonic fibroblasts increases LC3-I to LC3-II conversion, whereas inhibiting TERT reduces LC3-I to LC3-II [17], Hughes et al. tested the hypothesis that telomerase activity influences autophagy to an extent that impacts NO-mediated, flow-induced dilation in the microvasculature [18**].

As predicted, human adipose arterioles displayed increased autophagic flux and NO-mediated vasodilation that switched to H₂O₂-mediated vasodilation upon lysosomal inhibition or Atg3 siRNA. Further, arterioles obtained from CAD patients were refractory to shear-induced autophagy, but vasodilatory mechanisms could be reverted from mitochondrial generated H₂O₂ to EC mediated NO in response to autophagy activation using the histone deacetylase inhibitor trichostatin-A (TSA) or the disaccharide trehalose. Implicating autophagy as “downstream” from telomerase, neither gain nor loss of autophagy impacted TERT activity in either group. Underscoring the importance of telomerase, TERT inhibition of non-CAD arterioles limited shear-induced autophagic flux that could be rejuvenated by concurrent treatment with TSA. Congruent with these findings, TERT activation of CAD arterioles restored autophagic flux that could be subsequently abrogated by autophagy inhibition using BAF A1. In support of the original hypothesis, autophagy activation maintained NO as the primary mechanism of vasodilation in non-CAD arterioles despite TERT inhibition, whereas increasing TERT expression in the presence of autophagy inhibition maintained H₂O₂ as the vasodilatory mechanism of CAD arterioles. The authors concluded that autophagy acts downstream of telomerase to determine NO-mediated, flow-induced dilation in human adipose arterioles [18**].

The ability for aspirin to reduce cardiovascular risk and attenuate EC dysfunction is well-known. Using human coronary arteries ECs (HCAECs), Chen et al. showed that aspirin dose-dependently : (i) increased steady-state autophagy (e.g., ↑ LC3-II / LC3-I, ↓ p-62) and autophagic flux (tandem mRFP-GFP-tagged LC3); and (ii) heightened p-eNOS^S1177 and NO generation, to an extent that reduced protein expression of p-p38, p-NF-kB, sICAM-1, and sVCAM-1, which are otherwise elevated by treatment with oxidized low-density lipoprotein, angiotensin-II, or high glucose. Remarkably, the ability of aspirin to attenuate cell injury in response to each intervention was lost in HCAECs transfected with beclin-1 vs. scrambled siRNA. Determining the translational relevance of these interesting findings using a relevant preclinical model is warranted [19].

Autophagy and EC NO generation and aneurysm

Impaired NO bioavailability and heightened oxidative stress directly contribute to vascular smooth muscle cell growth, proliferation, contraction, and differentiation, which are major determinants of aneurysm development [20]. Implicating a contribution from defective autophagy, Atg5 deficiency increases aortic dissection [21] whereas Atg7 depletion promotes aortic aneurysm induced by high cholesterol [22]. Iraci et al. determined the translational relevance of these in vitro findings by comparing aortic tissue and blood samples from patients undergoing surgery to repair thoracic aortic aneurysm (TAA) to patients referred for aortic valve replacement or repair (control) [23*]. Protein expression of LC3-II, Atg5, and Atg7 was repressed, whereas p62 was elevated, in aortic lysates from TAA vs. controls, but the source of the defect i.e., EC or vascular smooth muscle, was not discerned. Because mRNA abundance for these markers was similar between groups, the authors concluded that post-translational modifications contributed to lowering vascular autophagy in TAA patients but mechanisms were not evaluated. Of note, aortic lysates from aneurysm patients with repressed vascular autophagy exhibited lower circulating and tissue nitrite (estimates of NO bioavailability), coupled with NADPH oxidase 2 activation and elevated H₂O₂, (estimates of oxidative stress). The authors acknowledge the observational nature of their study and relatively low sample size, but nevertheless this report provides translational evidence that contexts involving repressed vascular autophagy couple with reduced NO bioavailability and amplified oxidative stress in the setting of aneurysm.

Autophagy and EC NO generation and stroke

Acute ischemic stroke (AIS) limits delivery of oxygen and glucose to cerebral artery ECs. Upon detecting a nutrient-stress, the activity of mammalian target of rapamycin (mTOR) complex 1 (mTORC1) decreases, stimulating EC autophagy to support basal metabolism [24, 25]. In addition to canonical roles for EC autophagy in this context [26], evidence exists that EC autophagy enables NO-mediated cerebral arterial vasodilation. In this regard, low-dose mTORC1 inhibition using rapamycin might increase collateral perfusion during and reperfusion following AIS, to an extent that reduces infarct volume and improves neurological outcomes in normotensive [27**, 28] and hypertensive rats [27**]. Underscoring the importance of ECs, rapamycin-induced dilation of leptomeningeal anastomoses (i.e., collaterals) from normotensive and hypertensive rats was abolished by NO synthase inhibition ex vivo [27**]. The authors speculated that benefits of rapamycin on cerebral perfusion might be secondary to induction of autophagy and activation of p-eNOS^S1177 on penetrating arterioles [27**].

mTORC1 inhibition might improve stroke outcome via alternative NO-mediated mechanisms. For example, mTORC1 diminution reduces inflammation secondary to cyclooxygenase repression in the context of oxygen-glucose deprivation or AIS, thus shifting the balance toward epoxyeicosatrienoic acid (EET) production. EET generation might be protective in two ways. First, 14,15-EET activates sirtuin (SIRT) 1-mediated mitophagy in brain microvascular ECs [29]. Second, 14,15-EET induces NO generation by human microvascular ECs [30]. Either mechanism has potential to improve stroke outcome and warrants further pursuit.

The Sciarretta laboratory demonstrated that directly re-activating EC autophagy via pharmacological [31**] or nutraceutical [32–34*] maneuvers is protective in the context of hemorrhagic stroke [31**]. ECs isolated from brains of stroke-prone hypertensive rats (SP-SHRs) prior to stroke-onset exhibit depressed autophagy and greater cell death, that associates with downregulation of the mitochondrial complex 1 subunit NDUFC2, mitochondrial dysfunction, and NAD⁺ depletion upon exposure to salt. Remarkably, treating primary ECs from SP-SHRs in vitro with the peptide autophagy agonist Tat-D11 [35] improved their viability, and administering Tat-D11 to SP-SHRs in vivo reduced stroke occurrence [31**]. Of interest, this study demonstrated translational relevance. Endothelial progenitor cells derived from individuals harboring a variant of NDUFC2 that associates with increased incidence of juvenile ischemic stroke [31**] exhibit depressed steady-state autophagy and ineffective autophagic responses to salt exposure. Underscoring the importance of autophagy, Tat-D11 treatment reduced indexes of cell senescence in mutant progenitors exposed to salt.

Mitochondrial complex I maintains the NAD⁺ / NADH ratio, and since NAD⁺ supplementation stimulates autophagy [36], Forte et al. sought to determine whether nicotinamide mononucleotide (NMN) treatment could be restorative in vascular smooth muscle cells (i.e., A10 cells). Indeed, NMN treatment improves NAD⁺ levels that associate with heightened autophagy and mitophagy, and this intervention reduced stroke occurrence in SP-SHRs. Although not studied directly, NMN-induced upregulation of EC autophagy and EC NO generation could have contributed partly to the observed benefits. NMN treatment increases NAD⁺ levels in ECs [37], optimizes NAD⁺ / NADH in aortic tissue [38], improves NO-dependent vasodilation in carotid arteries from older mice [39], possibly by sirtuin (e.g., SIRT1) -mediated upregulation of eNOS [40], and direct administration of nicotinamide to human umbilical vein endothelial cells (HUVECs) treated concurrently with H₂O₂ improves NO bioavailability and cell viability [34*]. In the latter study, trehalose exerted benefits similar to nicotinamide in HUVECs exposed to oxidative stress. The same group showed increased acetylcholine-evoked mesenteric artery vasorelaxation in SP-SHRs that display reduced stroke occurrence [32], providing proof of concept that this approach might improve stroke outcome by preserving cerebral arterial function and subsequent perfusion. Taken together, amplifying EC autophagy has potential to benefit stroke outcomes in the context of AIS and / or hypertensive stroke by mediating EC NO generation.

Directions for future study

First, our review is “NO-centric” by design and we acknowledge that additional work is justified concerning the impact of EC autophagy on other vasoactive molecules. Underscoring this point, McCarthy et al. tested whether rejuvenating whole-body autophagy resolves dysregulated vasomotion displayed by mesenteric resistance arteries (MRAs) from spontaneously hypertensive rats (SHRs). Arteries from SHRs administered trehalose exhibited: (i) decreased p62 protein abundance; (ii) reduced cyclooxygenase 1 and cyclooxygenase 2 expression; and (iii) lower dihydroethidium fluorescence vs. vehicle-treated SHRs, that altogether associated with diminished vasocontraction to high-dose acetylcholine in a manner could be recapitulated by indomethacin or tempol. Notably, although boosting autophagy did not impact NOS activity in this study, the potassium channel inhibitor tetraethylammonium impaired acetylcholine-evoked vasorelaxation to a greater extent in MRAs from trehalose vs. vehicle-treated SHRs. These findings support that whole-body autophagy amplification via trehalose in the context of systemic hypertension restrains acetylcholine-stimulated generation of endothelium-derived vasoconstrictor prostaglandins and reactive oxygen species, while increasing endothelium-dependent hyperpolarization, in MRAs from SHRs. Future studies investigating the impact of EC autophagy on EDRFs other than NO together with EDCFs are warranted. Second, to translate in vitro [19, 31**, 34*] and ex vivo [3, 18**, 27**, 28, 32] “restoration of EC autophagy” interventions to pre-clinical models, studies using EC specific amplification procedures are required. Underscoring this need, trehalose ingestion increased stiffness and fibrosis in arteries from Wistar rats that exhibit intact autophagy [41]. Third, although physiological maneuvers, even if initiated late-in-life, can restore autophagic flux in cardiomyocytes that associates with improved myocardial function [42], the extent to which this and other (e.g., time-restricted feeding) lifestyle interventions improves EC autophagic flux and NO generation is unknown. Finally, whether sex differences exist concerning EC autophagy and EC NO generation have not been explored. Progress toward these and other issues will certainly be made in the coming years.

Figure 2. EC autophagy impacts EC NO generation.

Evidence exists that EC autophagy is repressed in the context of aging and cardiovascular complications including coronary artery disease, aneurysm, and stroke. Amplifying whole body or arterial EC autophagy might improve outcomes in each context by improving nitric oxide (NO) bioavailability. The numbers in brackets refer to the specific citation. EDHF, endothelium-derived hyperpolarizing factor; EDCF, endothelium-derived constricting factor; TSA, trichostatin A; NMN: nicotinamide mononucleotide.

Table 1.

Autophagy activators.

Activator	Target cells or tissues	Disease model	Treatment	Verification	References
Trehalose	Aorta	Aging	2% trehalose in drink water for 4 weeks in old mice	Arterial Beclin 1, WIPI-1, LC3-II, LAMP-2A, and p62 protein levels	[3]
Trehalose	Artery	CAD	10 mmol/L trehalose for overnight in adipose and atrial resistance arterioles from CAD patients	No verification	[18**]
Trehalose	Rat brain	Stroke	2% trehalose in drinking water for 4 weeks in SHRSP† received a JD‡	LC3-II and p62 protein expression levels	[32]
Trehalose	Mesenteric artery	Hypertension	2% trehalose in drink water for 4 weeks in SHRΨ	p62 protein expression	[41]
TSA	Artery	CAD	100 mmol/L TSA for 30 min or overnight in adipose and atrial resistance arterioles from CAD patients	Lysotracker Red DND-99 Fluorescence in artery	[18**]
Aspirin	HCAECs#	CAD	2.5 mM Aspirin for 16 h in HCAECs	LC3-II and p62 protein expression levels and mRFP-GFP-tagged LC3	[19]
Tat-Beclin1	CECsΩ	Stroke	15 mg/kg/d of Tat-Beclin1 D11 in SHRSP receiving JD for 3 weeks	LC3-II protein expression	[31**]
Rapamycin	Rat brain	Stroke	250 ug/kg rapamycin after 30 min of MCAO in SHR	No verification	[27]
Rapamycin	Mouse/Rat brain	Stroke	10 mg/kg rapamycin after MCAO in SD rats or C57BL/6 mice	No verification	[28]
Nutraceutical	HUVECs	Stroke	Combination of trehalose (10 uM), cathechin (10 uM), epicatechin (10 uM), spermidine (5 uM) in HUVECs	No verification	[34*]

^†SHRSP: Stroke-prone spontaneously hypertensive rat

^‡JD: Japanese-style diet

^#HCAECs: Human Coronary Artery Endothelial Cells

^ΩCECs: Cerebral Endothelial Cells

^ΨSHR: Spontaneous hypertensive rat; Nutraceutical

^ΔNutraceutical : 10 uM trehalose, 10 uM catechin, 10 uM epicatechin, and 5 uM spermidine for Mix 1. 10 uM trehalose, 5 uM spermidine, 5 uM nicotinamide for mix 2