Nitric oxide: what’s new to NO?

Abstract

Nitric oxide (NO) is one of the critical components of the vasculature, regulating key signaling pathways in health. In macrovessels, NO functions to suppress cell inflammation as well as adhesion. In this way, it inhibits thrombosis and promotes blood flow. It also functions to limit vessel constriction and vessel wall remodeling. In microvessels and particularly capillaries, NO, along with growth factors, is important in promoting new vessel formation, a process termed angiogenesis. With age and cardiovascular disease, animal and human studies confirm that NO is dysregulated at multiple levels including decreased production, decreased tissue half-life, and decreased potency. NO has also been implicated in diseases that are related to neurotransmission and cancer although it is likely that these processes involve NO at higher concentrations and from nonvascular cell sources. Conversely, NO and drugs that directly or indirectly increase NO signaling have found clinical applications in both age-related diseases and in younger individuals. This focused review considers recently reported advances being made in the field of NO signaling regulation at several levels including enzymatic production, receptor function, interacting partners, localization of signaling, matrix-cellular and cell-to-cell cross talk, as well as the possible impact these newly described mechanisms have on health and disease.

cardiovascular disease is one of the important contributors to disease-related deaths worldwide (78). Nitric oxide (NO) was the first acknowledged physiologically relevant gaso-transmitter, and its dysregulation has been implicated in cardiovascular diseases. Appreciation of its multiple roles in maintaining the function of high-pressure closed circulatory systems was a paradigm shift in medicine. While investigation into the intricacies of this molecule have flourished, NO is significantly understudied compared with other simple molecules such as oxygen. This review will focus on recent insights over the past decade into the signaling pathways that regulate the synthesis of NO as well as NO-mediated target activation and the effects these newly described mechanisms have on homeostasis and pathophysiology.

Regulation of Nitric Oxide Signaling

Nitric oxide synthase (NOS) enzymes catalyze the NADPH and tetrahydrobiopterin (BH4)-dependent oxidation of l-arginine to l-citrulline and produce NO as one of the reaction products (105). Intracellular arginine availability is a rate-limiting factor in cellular NO production, and argininosuccinate lyase that converts citrulline back to arginine has been shown to be important to synthesize not only intracellular arginine but also to utilize extracellular arginine for NOS-dependent NO synthesis (28). There are three mammalian NOS isoforms: neuronal (nNOS), endothelial (eNOS), and inducible (iNOS). In the absence of l-arginine or BH4, eNOS is uncoupled and the monomer form of eNOS synthesizes superoxide in preference to NO (20). Monomeric eNOS cannot deliver electrons to the heme center of other eNOS monomers, and instead it transfers the electrons to a location where they interact with oxygen and thus facilitates the formation of superoxide (20). Consequently, eNOS dimerization is essential for proper enzymatic activity and NO production. This process is regulated by heat shock protein 90 (HSP90) as silencing of HSP90 or its inhibition with geldanamycin destabilizes eNOS dimers, leading to degradation. Conversely, phosphorylation does not appear to play a substantial role in this process since phosphorylation status of S1179 and T497 has no effect upon the eNOS dimer/monomer ratio in endothelial cells or in cell-free preparations of purified eNOS protein (20). Recently, it was shown that endothelin-1 could facilitate superoxide production through eNOS in human microvascular endothelial cells when they were subjected to endotoxins (38). The regulation of eNOS signaling is complex and, although it is mainly regulated by phosphorylation of its amino acid residues at multiple sites on its domains, its interaction with various other proteins and intracellular distribution provides a further degree of modulation.

Phosphorylation and Regulation of eNOS Activity

In humans, eNOS can be phosphorylated at several amino acid residues, including S1177 on the COOH-terminal reductase domain; T495 on the calmodulin (CaM)-binding domain; S633, S615, and Y657 on the autoinhibitory element of the FMN-binding domain; and S114 and Y81 on the NH₂-terminal oxygenase domain (95). The NH₂-terminal oxygenase domain binds to zinc, heme, and tetrahydrobiopterin (BH4) and coordinates the formation of eNOS homodimer. Shear stress, insulin, leptin, statins, bradykinin, VEGF, and IGF-I are some of the factors that activate eNOS and enhance NO production by phosphorylating S1177 (112). The kinases that are responsible for S1177 phosphorylation include CaMKII, AMPK, PKA, Akt, Chk1, and PKG (83, 95). Recently, eNOS was shown to be activated through tyrosine phosphorylation of G-protein-coupled receptor (GPCR) kinase-interacting protein-1 (GIT1) by Src where Akt regulated the ability of Src to phosphorylate GIT1 as well as GIT1-eNOS association (72).

Some other signals including mechanical shear stress, VEGF, statins, 8-Br-cAMP, and bradykinin also lead to phosphorylation or dephosphorylation of sites on eNOS such as S615 by Akt, S633 by PKA and Pim1, Y657 by PYK2, and site T495 by PKC (32, 114). While phosphorylation of eNOS at S615/S633 increases eNOS calcium sensitivity and activity, Y657 and T495 phosphorylation are associated with downregulation of eNOS activity (18, 109). On the oxygenase domain, Y81 is phosphorylated by pp60src kinase and activates the enzyme while S114 is phosphorylated by ERK, decreasing the enzymatic activity and NO production (15, 33). It is evident that eNOS phosphorylation has been studied more extensively compared with eNOS dephosphorylation perhaps secondary to inherent challenges in studying protein dephosphorylation. It is also not clear how these competing protein modifications are coordinated.

Interaction of eNOS With Various Proteins

Calcium-calmodulin.

eNOS directly binds to calcium-activated-calmodulin (CaM) through its canonical binding site located between the NH₂-terminal oxygenase domain and its COOH-terminal reductase domain, and this interaction is essential for its activity (19). CaM can displace eNOS from caveolae in the presence of HSP90 and act as positive regulator of eNOS activity (39).

Caveolin-1.

Discovery of cell-membrane structures called caveoli and the spatial localization as well as regulation of NO production within such “hot” spots represented a new era in the understanding of NO signaling. Caveolin-1 (Cav-1) binds to eNOS in the region of caveolin-scaffolding domain and is a negative regulator of eNOS activity and NO production in contrast to CaM (79). Cav-1 has been shown to prevent CaM binding when calcium levels are low, resulting in reduced eNOS activity (85). Deletion of Cav-1 from mice revealed a phenotype consistent with the removal of a negative regulator of eNOS activity, i.e., reduced blood pressure and enhanced endothelium-dependent relaxation (121).

Other protein-protein interactions.

eNOS has been shown to interact with HSP90, NOSIP, NOSTRIN, PIN1, PP2A, PP1, SIRT1, VDAC1/2, GPCR (B2), GRX-1, β-actin, CAT-1, CHIP (carboxyl terminus of HSP70-interacting protein), endoglin, and HDAC3 (2, 17, 20, 24, 40, 55, 65, 66, 70, 76, 93, 94, 117, 122).

Intracellular Distribution and Membrane Targeting of eNOS

eNOS has been detected at both the Golgi complex and the plasma membrane (57). eNOS is targeted to different subcellular compartments by protein fatty acid acylation. First, eNOS is irreversibly myristoylated, which aids in its association with the plasma membrane, and this process allows for subsequent reversible palmitoylation, which occurs in the Golgi and directs eNOS to the plasma membrane (31). The plasma membrane is an ideal location for synthesis of NO as eNOS can be in close proximity to ion channels, kinases, and ligand-activated GPCRs that are required for its proper function. But at the same time, plasma membrane-targeted eNOS is much more vulnerable to extracellular factors such as oxidized low-density lipoprotein, which reduces the eNOS activity in contrast to Golgi-eNOS, which is less degradable (120). This possibly explains the observation that, in mature blood vessel endothelial cells, only a fraction of the total eNOS resides at the plasma membrane, with a majority being found in the relatively protected Golgi complexes. Conversely, several different factors including NOSIP, NOSTIN, and CHIP (carboxyl terminus of HSP70-interacting protein) promote the dissociation of eNOS from the plasma membrane towards intracellular compartments and such actions have been shown to result in decreased eNOS activity (24, 55, 122). Thus, it appears that attachment of fatty acid moieties and subsequent intracellular distribution of eNOS plays an important role in regulating eNOS activity and NO production.

Matricellular Protein Acts From “Outside In” to Inhibit NO

Soluble guanylate cyclase (sGC) is the major intracellular receptor for NO. NO in turn stimulates sGC, and sGC converts GTP into the second messenger cyclic guanosine monophosphate (cGMP) (118). cGMP subsequently inhibits platelet aggregation, leucocyte adhesion, relaxes vascular smooth muscle cells and thus promotes blood flow. However, once formed, NO, as a biogas, can activate its receptor sGC almost instantaneously. To maintain homeostasis, a number of feedback mechanisms, such as phosphodiesterases types 5, 6, and 9 (PDE 5, 6, 9) exist and work to specifically decrease cytoplasmic levels of cGMP (21). But new work suggests there may be further built-in molecular limiters of NO. Thrombospondin-1 (TSP1), one such NO-limiter, was found residing among the matricellular family of proteins. This name was derived from the fact that these proteins are secreted after injury, bind to extracellular matrix without providing structural support, and serve as ligands for cell membrane receptors (13). TSP1 has been found to limit eNOS activity and NO signaling through directly targeting both sGC and, in some cells, through directly inhibiting targets of second messenger cGMP (51, 87) (Fig. 1). These multiple limiting effects of TSP1 on the NO signaling pathway require interaction with the vascular cell membrane receptor CD47 (49) and occur at picomolar or less amounts of TSP1, thus representing one of the most potent activities so far defined for the protein. Indeed, TSP1 has been proposed as a universal inhibitor of sGC (77). Functionally, TSP1, via CD47, inhibits endothelial-dependent arterial relaxation in arterial segments and limits acetylcholine-mediated hypotension in animals (5). Interestingly, isolated coronary arterioles from young female rats were resistant to TSP1-mediated inhibition of vasodilation (81). On the other hand, intravenous TSP1 and a CD47 agonist antibody, which functions like TSP1 to activate the receptor, increase blood pressure in animals. Conversely, TSP1- and CD47-null mice show constitutively increased levels of cGMP and phosphorylated eNOS (5). Animals lacking either the ligand TSP1 or the receptor CD47 have enhanced angiogenesis and increased tissue blood flow in response to a range of stressors including temperature, ischemia, and reperfusion (47). Recent findings indicate that TSP1-CD47 signaling stimulates increased reactive oxygen species (ROS) production in hypoxic endothelial cells (6), vascular smooth muscle, and renal tubule epithelial cells, the latter via activation of NADPH oxidase (22, 116), which may account, in part, for the inhibitory activity of TSP1 on NO, but this remains to be determined. In healthy human pulmonary arteries, TSP1 potentiates vasoconstriction and suppresses NO-mediated vasodilation (91). Similarly, TSP1 and CD47 are upregulated in diseased human arteries and this is associated with a loss in sensitivity to NO (91). Therapeutic agents that disrupt TSP1-CD47 signaling improve sensitivity to NO-mediated vasodilators in diseased human arteries (91), and in rodents and pigs these agents increase blood flow in response to ischemia (50, 90). Recent initiation of clinical trials testing CD47 targeting therapies, albeit not for cardiovascular disease, confirms emerging interest in CD47 in human disease (71).

Fig. 1.

Recent findings provide for additional levels of nitric oxide (NO) regulation. Receptor-mediated signaling such as vascular endothelial growth factor receptor 2 (VEGFR2) and acetylcholine (ACh) as well as protein partners including heat shock protein 90 (HSP90), calmodulin (CaM), and caveolin-1 (Cav-1) regulate endothelial NO synthase (eNOS) activity and thus NO production in the vascular wall. Once generated, NO diffuses from endothelium to various cell types including vascular smooth muscle cells, platelets, and monocytes where it activates its receptor soluble guanylate cyclase (sGC), increases production of cyclic guanosine monophosphate (cGMP), and controls cell function. In addition to paracrine processes, NO controls or alters intracellular functions such as mitochondria biogenesis, telomerase activity, reactive oxygen species (ROS) production, and low-density lipoprotein (LDL)-oxidation. Heme scavenging proteins, such as hemoglobin-α, are expressed in endothelial cells (EC) and serve as a “buffer” to control NO diffusion, a process that is regulated at endothelial (MEJ)-smooth muscle cell junctions by cytochrome b5 reductase 3 (Cb5r3). Extracellular “outside-in” signals, such as thrombospondin-1 (TSP1)-cluster of differentiation 47 (CD47) also have the ability to limit on NO signaling through inhibition of eNOS, sGC, and cGMP-dependent protein kinase (PKG) activity. EPC, endothelial progenitor cells; MR3, muscarinic receptor 3; MEJ, myo-endothelial junction; SMC, smooth muscle cell; GPCR, G protein-coupled receptor. [Printed with permission from Anita Impagliazzo.]

NO “Corridors” in the Vascular Wall

In blood vessels, NO diffuses in a spherical gradient that results in nondirectional dispersion (14, 67, 68, 108). Because of these properties, questions as to how NO precisely targets its receptor sGC in enough concentration to activate it remains unclear. The ability of cells to spatially localize NOSs has provided some hints as to how NO can manage directional signaling (34, 52). Recently, the existence of a NO corridor at the myoendothelial junction (MEJ) has been proposed (101). The MEJ is a cellular extension that typically arises from endothelium, juxtaposes vascular smooth muscle cells (VSMCs), and serves as a connection that facilitates cross talk between endothelium and vascular smooth muscle in small arteries and arterioles (89, 104). Recent work has demonstrated that eNOS also localizes to the MEJ, and in some arteries, is enriched at the MEJ (101). The expression of eNOS at the MEJ likely limits long-distance NO diffusion, positions it close to VSMC sGC, S-nitrosates local proteins, and decreases NO scavenging by ROS (101) (Fig. 1). More recently, the lipid composition at the MEJ has been shown to add another level of control to regulate NO diffusion and signaling (9). In addition to eNOS localization at the MEJ, hemoglobin-α (Hb-α), potent scavenger of NO, is enriched at the MEJ through an unbiased proteomic screen (103). Functionally, Hb-α serves as an “NO sink” by buffering NO diffusion from endothelium to smooth muscle cells through a dioxygenation reaction forming nitrate and methemoglobin-α, thereby further regulating NOS-mediated signaling and control of arterial vascular reactivity (103). For this to be a sustainable mechanism of NO diffusion control, methemoglobin-α heme iron is reduced by cytochrome b5 reductase 3 (86, 103). Additional studies have now demonstrated that disruption of eNOS and Hb-α binding with an Hb-α mimetic peptide enhances NO signaling in vivo and lowers blood pressure, thereby identifying a novel target for treating cardiovascular disease (102).

NO Signaling in the Heart

Nitric oxide synthase (NOS) is also a source of endogenous NO in cardiomyocytes. The three NOS isoforms—eNOS, nNOS, and iNOS—are each expressed in a unique location within the cardiomyocyte. eNOS is found in the caveolae of the plasma membrane bound to caveolin-3. nNOS is found bound to the ryanodine receptor in the sarcoplasmic reticulum (SR) of cardiomyocytes. iNOS can be induced in inflamed cardiomyocytes. Owing to the unique spatial distribution and concentrations of biogas generated by the NOS isoforms, each NOS can affect the heart differently, with experimental work indicating a cardio-protective function in certain situations and a cardio-damaging function in others.

NO is important for the β-adrenergic response of the heart. Upon β-adrenergic stimulation, the amount of NO in the cardiomyocyte increases. NO produced by nNOS stimulates SR calcium release, ultimately aiding in β-adrenergic stimulation (60). Fittingly, nNOS knockout mice show suppressed inotropic response and reduced force-frequency relationship (FFR), likely due to reduced SR calcium stores (37, 60). NO may increase β-adrenergic stimulation through S-nitrosation and subsequent activation of calcium/calmodulin-dependent protein kinase II (23, 41) or S-nitrosation of ryanodine receptor 2, increasing SR calcium release (37). Furthermore, inhibition of nNOS reduces sensitivity of cardiomyocytes to β-adrenergic stimulation and reduces calcium release (27). Thus, nNOS is important for calcium release from the SR. However, peroxynitrite, which is produced from NO, and superoxide inhibit the β-adrenergic response through decreasing phospholamban phosphorylation (63) and uncoupling nNOS (98, 106). Peroxynitrite and superoxide in the cardiomyocyte decreases NO production and increases superoxide formation, further contributing to inactivation of nNOS (106). Furthermore, peroxynitrite may alter, through sarco/endoplasmic reticulum Ca²⁺-ATPase (SERCA), cardiomyocyte action potentials (11), damage mitochondria through increasing lipid peroxidation (64) and generally suppress cardiac muscle function (26). In addition, increased peroxynitrite leads to SERCA-mediated Ca²⁺ sequestration and cardiomyocyte relaxation (8).

Conversely, eNOS is cardio-protective through inhibition of the β-adrenergic inotropic response through regulation of the L-type calcium channel (111). Thus, eNOS-null mice show enhanced contractility and a normal FFR (60) with hypertension and hypertrophy of the heart (4, 15) and are more sensitive to β-adrenergic stimulation. Specifically, phosphorylation of S1177, consistent with eNOS activation, is increased during exercise (30, 62, 88) and epinephrine stimulation (16), likely contributing to the cardio-protective effects of exercise. This increased phosphorylation possibly occurs through a β₃-adrenergic receptor-dependent pathway, because nebivolol, a β₃ adrenergic receptor agonist, increased phospho-eNOS levels (3). β₃-Adrenergic receptor-null mice have reduced p-S1176 (murine phosphorylation site) eNOS abundance (3, 16), suggesting β₃ adrenergic receptor stimulation is important for eNOS activation in the heart. Mice fed a high-fat diet showed a reduction in β₃-adrenergic receptor protein expression and eNOS activity and this did not correlate with the observed cardio-protective effects of exercise (62). Furthermore, the eNOS dimer/monomer ratio was increased while the amount of protein nitration (measure of peroxynitrite formation) was decreased following exercise. Thus, eNOS dimerization and activation may be cardio-protective through decreasing peroxynitrite formation (30). Changes in iNOS expression may also contribute to the cardio-protective effects of exercise as exercise decreased iNOS abundance while inhibition of iNOS in mice on a high-fat diet reduced infarct size (62). Mitochondrial nNOS may also be cardio-protective during exercise through increasing oxygen consumption (46, 58, 92).

While iNOS is associated with an inflammatory response in the heart, it may play a role in the development of reactive hypertrophy. iNOS-null mice are less sensitive to left ventricular pressure overload secondary to transverse aortic constriction (TAC), showing an improved ejection fraction and reduced heart size compared with wild-type mice (119). This is contrary to eNOS- and nNOS-null mice that show hypertrophy of the heart with age, but only the former shows concurrent elevation in blood pressure (4). iNOS may contribute to hypertrophy development through several mechanisms including increased ROS production, or activating proteins such as protein kinase B, mechanistic target of rapamycin, and ERK (119). Furthermore, iNOS expression is likely regulated through the phosphoinositide 3-kinase (PI3K)-nuclear factor of activated T cells (NFAT) pathway, because PI3Kγ-null mice show increased iNOS abundance, and inhibition of NFAT in these mice reduces iNOS levels (80).

NO is also important for mitochondrial respiration in the heart. NO stimulation increases mitochondrial oxygen consumption (MV̇O₂). For example, cardiac myocytes from nNOS-null mice are more sensitive to bradykinin and carbachol, compounds that stimulate NO production, shown by a bigger reduction in MV̇O₂ (73). Furthermore, nNOS-null myocytes have a greater abundance of superoxide and decreased NO, likely due to increased activity of xanthine oxidase (XO) (92) through increased phosphorylation of p38, which activates XO (61). Angiotensin II may also regulate NOS signaling in the mitochondria. For example, angiotensin II increases nNOS expression and causes a decrease in NADPH oxidase and reactive oxygen species (ROS) (56). Inhibition of NO combined with angiotensin II treatment, however, increases mitochondrial respiration and ROS production (43). Thus, angiotensin II may work with mitochondrial NOSs to regulate respiration.

New findings of regional- and cell-specific NO signaling in the heart, while important, have further implications in light of work suggesting that TSP1-CD47 signaling limits heart function. Cyclic guanosine monophosphate and cyclic adenosine monophosphate levels were elevated in whole hearts from young male TSP1^−/− and CD47^−/− mice compared with wild-type (48), whereas TSP1 expression was upregulated in hypoxic right ventricles (91) and pressure-overloaded left ventricles post-TAC in wild-type mice (96). Mice lacking TSP1-CD47 signaling (TSP1^−/− or CD47^−/− mice) were resistant to ventricular dysfunction and remodeling mediated by chronic hypoxia (6, 91) and TAC (96) compared with wild-type animals. Given the role of TSP1-CD47 signaling to inhibit NO in vascular cells and to limit arterial function, it will be important to determine whether these effects extend to cardiomyocytes.

Another important aspect related to cardiomyocyte function is the control of NO diffusion and NO compartmentalization. As mentioned, NOS isoform expression and localization play a major role in localized NO synthesis and function (4). Beyond this, heme-centered proteins, such as myoglobin and cytoglobin, are predicted to function in the fine-tuning of NO diffusion rates and signaling under normal physiological conditions. Prior work demonstrated that loss of myoglobin acts as an intracellular NO scavenger to regulate cardiac contractility (113). Because cytoglobin is expressed in cardiomyocytes and has a high NO scavenging capacity, it is possible that it controls NO diffusion similar to myoglobin (42, 99). At present, the role of Hb-α as a modulator of cardiac NO signaling is unknown. Dedicated studies deciphering mechanisms of cardiomyocyte NO compartmentalization in physiological and pathophysiological settings are warranted.

NO and NO-Related Therapies

NO and NO-related drugs have been employed to therapeutic effect in a range of cardiovascular (74) and central nervous system (7) diseases where loss of NO signaling is believed to play a contributory role. As an inhaled gas NO has been approved in the US and in Europe for the treatment of pulmonary hypertension (PH) and associated respiratory failure of the newborn (1). Inhaled NO may be beneficial in instances when perioperative PH occurs, as has been reported in mitral valve surgery (44) and cardiac transplantation (35). NO-carrying drugs that link NO donors to existing drugs are in development, with a range of NO-donating molecules under preclinical and clinical investigation including organic nitrates (10, 110) and nitrites (25, 97), the so-called NONOates (53, 59), and NO-metal complexes such as sodium nitroprusside (84). Nonsteroidal anti-inflammatory drugs provided a sensible platform for NO donor-carrying drug development, but questions about NO-delivery remain (12). A refined application of this idea has resulted in the linking of NO donors to topical agents to treat glaucoma (29). Other NO-related species such as nitroxyl (HNO) (82) and S-nitrothiols (100) have been effective in cell culture and animals studies but remain to be proven in clinical trials. Downstream of NO delivering-agents, new drugs that stimulate production of cGMP have been developed. Central to the effect of these agents is heme-independent stimulation of sGC (75). Riocigaut, one such agent, was effective in increasing exercise capacity and decreasing shortness of breath and pulmonary vascular resistance in individuals with pulmonary arterial hypertension (36). Recent patent filings suggest that these agents may be applied to the treatment of glaucoma (WO 2015095515 A1).

In contrast to diseases that are mediated through decreased NO signaling, recent studies have linked hyperactive NO signaling, as manifested by increased iNOS expression, to breast cancer progression and patient outcome (107). Blocking iNOS significantly limited breast cancer (45) and melanoma (115) metastasis in mice. However, it is not clear that selective iNOS suppression can be achieved in the clinic. Furthermore, NO has been found to inhibit tumor cell proliferation (69) and iNOS to limit tumor growth in animals (54), indicating that the final results of targeting NO in relation to cancer are yet to be determined.

Conclusion

Although not exhaustive, this review of recent findings highlights that the gaso-transmitter NO remains a source of persistent fruitful study. Scientists from diverse backgrounds continue to reveal unexpected subtleties to NO regulation that will likely have positive implications for human health and disease. A majority of research findings in NO signaling have been confirmed in male rodents and large mammals. Going forward, it will be necessary to assess the role of sex in gaso-transmitter signaling in general, and NO signaling specifically, both for fixed principles as well as the newly reported regulatory mechanisms.

GRANTS

This work was supported by National Institutes of Health (NIH) National Heart, Lung, and Blood Institute Grants R01 HL-108954, 1R01 HL112914, and 1R21EB017184 (to J. S. Isenberg) and R01 HL 133864, R01 HL 128304 (to A. C. Straub) Further support was provided by the Institute for Transfusion Medicine, the Hemophilia Center of Western Pennsylvania, and the Vascular Medicine Institute of the University of Pittsburgh School of Medicine (to J. S. Isenberg and A. C. Straub).

DISCLOSURES

J. S. Isenberg serves as chair of the scientific advisory board of Radiation Control Technologies, Inc. (Jersey City, NJ) and has equity interest in this company and in Tioma Therapeutics (St. Louis, MO). No other conflicts of interest, financial or otherwise, are declared by the author(s).

AUTHOR CONTRIBUTIONS

K.G., H.M.A., A.C.S., and J.S.I. drafted manuscript; K.G., H.M.A., A.C.S., and J.S.I. edited and revised manuscript; K.G., H.M.A., A.C.S., and J.S.I. approved final version of manuscript.