|Home | About | Journals | Submit | Contact Us | Français|
Nitric oxide (NO) is a potent regulator of vascular tone and hemorheology. The signaling function of NO was largely unappreciated until approximately 30 years ago, when the endothelium-derived relaxing factor (EDRF) was identified as NO. Since then, NO from the endothelium has been considered the major source of NO in the vasculature and a contributor to the paracrine regulation of blood hemodynamics. Because NO is highly reactive, and its half-life in vivo is only a few seconds (even less in the bloodstream), any NO bioactivity derived from the intraluminal region has traditionally been considered insignificant. However, the availability and significance of NO signaling molecules derived from intraluminal sources, particularly erythrocytes, have gained attention in recent years. Multiple potential sources of NO bioactivity have been identified in the blood, but unresolved questions remain concerning these proposed sources and how the NO released via these pathways actually interacts with intravascular and extravascular targets. Here we review the hypotheses that have been put forward concerning blood-borne NO and its contribution to hemorheological properties and the regulation of vascular tone, with an emphasis on the quantitative aspects of these processes.
Nitric oxide (NO) is a pivotal signaling molecule that regulates blood flow and hemorheology, stimulating vasoactivity, inhibiting smooth muscle cell proliferation, inhibiting platelet adhesion, and regulating cell adhesion, vascular permeability, and erythrocyte deformability . NO can be enzymatically synthesized in a reaction catalyzed by nitric oxide synthase (NOS). Three isoforms of NOS have been identified: neuronal NOS (nNOS or NOS1), inducible NOS (iNOS or NOS2), and endothelial NOS (eNOS or NOS3). Since the discovery of NO as an important signaling molecule more than 30 years ago, the NO molecules synthesized by endothelial NOS present in the vascular endothelium have been considered the dominant source in the vasculature [34, 50].
NOS3 is localized in the endothelial lining of the vascular wall, adjacent to the smooth muscle of the vessel. This enzyme catalyzes the degradation of L-arginine to L-citrulline, with the release of NO molecules . Once formed, NO diffuses out of the endothelial cells and enters the extravascular and intraluminal spaces . In the vascular wall, NO reacts with soluble guanylate cyclase (sGC), which catalyzes the formation of 3’,5’-cyclic guanosine monophosphate (cGMP) to induce vasodilation. In the extravascular region, NO can inhibit mitochondrial respiration by binding to the heme group of cytochrome c oxidase, further regulating oxygen delivery to tissue . Inside the lumen, NO can reach its cellular targets and regulates certain rheologic factors, such as erythrocyte deformability and platelet-platelet adhesion [9, 11] at least partially through the transient interaction with the heme group of guanylate cyclase expressed on erythrocytes and platelets [4, 48, 56]. Other proposed regulation mechanisms may involve the interaction between NO and ion transport channels on erythrocytes and platelets [21, 55]. Thus, in the vasculature, NO can be consumed as a result of its reactivity with a variety of species in the vasculature, including O2, reactive oxygen species, reactive nitrogen species, and cytochrome c oxidase, while carrying out its signaling role.
The life-time of NO in the extravascular tissue is relatively longer, on the scale of sub-seconds to a few seconds [43, 65]. However, inside the lumen of the blood vessel, the NO that is formed is rapidly scavenged by hemoglobin contained in erythrocytes and converted into relatively inert nitrite and nitrate , and its lifetime can be as short as a few milliseconds [36, 65]. The surviving NO in the lumen has the capacity to regulate hemorheological properties. However, for many years the blood was viewed solely as a sink for NO molecules. Targets of NO signaling in the lumen were thought to be regulated by locally produced NO from the endothelium. One interesting possibility is that NO could be “carried” through the circulation by sources in the blood itself, with the NO bioactivity being released when needed and thereby providing an endocrine regulation of hemorheology. Also, if a significant amount of NO from these sources could eventually reach the smooth muscle and react with sGC, vascular tone could be regulated by the blood-borne NO. However, in vitro experiments and theoretical modeling suggested that most of the free NO in the bloodstream is rapidly converted to relatively inert nitrite and nitrate [28, 39, 42, 43] because of the fast reactivity of NO with hemoglobin contained in erythrocytes and the abundance of such hemoglobin in the blood. These findings appeared to close the door on the possibility that NO bioactivity could be stored in the blood and that blood-borne NO could regulate vascular tone and other biological functions.
The NO concentration around blood vessels has been measured, and the values reported have been in the range of 200–1000 nM [54, 67, 72], with the measured NO generally considered to be endothelial in origin. However, the microelectrodes used in these experiments measured the local NO level (discussed in Ref. ) that is derived from all possible NO sources in the vasculature. Furthermore, theoretical biochemical pathway analyses [15, 16] have shown that NOS3-derived NO cannot account for the large concentration of NO measured in the perivascular region, suggesting a discrepancy between the predicted and measured NO concentrations. This discrepancy, coupled with real-time NO concentration mapping evidence, indicated that a significant amount of NO might be produced from the intraluminal sources and might reach the vascular wall [14, 38]. Indeed, several hypotheses have been proposed since the 1990s to account for the possible release of NO from the blood (reviewed below).
Despite the theoretical prediction that NO cannot escape the scavenging of erythrocyte-contained hemoglobin, experiments have provided strong evidence of NO bioactivity after its release from erythrocytes. For example, inhalation of NO gas can reduce pulmonary hyptertension through NO-related mechanisms . In the 1990s, a novel endocrine signaling pathway was proposed  in which NO bioactivity was hypothesized to be preserved, rather than destroyed, by blood proteins, with the NO molecules being released under certain conditions, particularly hypoxia. Furthermore, nitrite, which in the past has been considered a stable species, has also been proposed to be reduced to NO radicals in low-pH environment or by hemoglobin in erythrocytes (and possibly by other heme-containing globins in other cells beyond the intraluminal region) [33, 60, 78]. Recent studies have also shown evidence of NO release from other sources, such as iron-nitrosyl-hemoglobin (HbNO) and via the activity of NOS3 in erythrocytes. The likely pathways that lead to NO release are summarized in Fig. 1. In addition to in vitro and ex vivo experiments that have focused on the specific pathways for the release of NO, in vivo measurements in animals have been conducted to detect NO in the blood. Kashiwagi et al.  observed strong fluorescent signals from intraluminal sources from their in vivo experiments using 4,5-diaminofluorescein diacetate in rat mesentery, suggesting the existence of intraluminal NO within the mesenteric vessels.
The potential release of intraluminal NO poses several conceptual challenges: First, it is not clear how the NO molecules that are formed can escape being scavenged by hemoglobin. The large quantity of erythrocyte-contained hemoglobin acts as a natural sink for the highly reactive NO molecules. Second, we still lack direct and specific measurements of the amount of NO derived from each source, especially from the intraluminal region. Finally, we have only an incomplete understanding of the quantity of NO released from the blood and how this blood-derived NO compares to endothelial NO in terms of physiological activity. The NO synthesized by NOS3 in endothelial cells diffuses to other parts of the vascular wall and to the extravascular region, and its periarteriolar level is reported as several hundred nanomolar . How much NO can reach that region from the blood and there regulate a variety of physiological functions?
In the next section, we will review the current hypotheses regarding the release of bioactive NO from the blood and discuss the likely mechanisms by which NO molecules might survive the scavenging activity of hemoglobin.
In considering NO transport within the lumen of the blood vessel and its signaling activity, particularly in the case of NO derived from erythrocytes, the first question that arises is how NO molecules can be transported after their formation. Is NO transported as the free nitrogen monoxide itself? Are there more stable intermediate species that are involved in NO production from its original source? Must NO be bound to other species in order to carry out its regulatory function? A thorough understanding of these processes will help us to achieve a complete picture of the physiology of NO signaling and develop NO-targeted therapeutics, such as vasodilating agents.
While it is unclear how the NO that is formed is transported out of erythrocytes, there are several hypotheses to account for the ability of NO to escape from the scavenging effect of erythrocyte-contained hemoglobin . The first potential explanation is that NO freely diffuses inside the lumen once it is formed. As discussed below, computational modeling has shown that a free diffusion mechanism would allow only a negligible amount of NO to reach the vascular wall and carry out its signaling functions [12, 17, 36, 45]. Another hypothesis states that once NO is formed inside the erythrocytes, NO can bind to erythrocyte membrane-bound proteins that can shuttle NO molecules out of the cell, avoiding immediate scavenging by the abundant intracellular hemoglobin . Yet another possibility is that an intermediate species with a longer lifetime in vivo is produced prior to the formation of NO itself. Such an intermediate molecule would not be reactive with hemoglobin and would be converted to NO in the extracellular environment [8, 59].
In addition to these potential transport mechanisms, the unique characteristics of the erythrocyte membrane seem to help prevent rapid scavenging of luminal NO by intracellular hemoglobin: NO is apparently restricted to some extent in its ability to diffuse freely through the erythrocyte membrane. As an uncharged small molecule, NO was initially thought to be able to diffuse freely through the lipid bilayer of the erythrocyte membrane, but experiments have now indicated that although the reaction between free hemoglobin and NO is rapid and limited only by the diffusion rate [26, 28], the reaction between NO and hemoglobin encapsulated in an erythrocyte is significantly slower [10, 44]. Thus, if NO synthesized in erythrocytes is shuttled out of the cells through a facilitated mechanism, this controlled permeability to NO would help decrease the scavenging of NO by intracellular hemoglobin by slowing the movement of some of the NO molecules back into the erythrocyte. Furthermore, computational modeling has predicted that most of the endothelial NO that is responsible for regulating vascular tone would be rapidly converted to relatively inactive nitrite or nitrate by erythrocytic hemoglobin soon after the NO was formed, as a result of the enzymatic activity of NOS3 in the endothelium . These mechanisms, taken together, suggest that the erythrocyte membrane may provide a barrier that reduces NO scavenging by intracellular hemoglobin. Other experiments, in conjunction with computational modeling [5, 70, 71], have indicated that the membrane of erythrocytes indeed exhibits an intrinsic resistance to NO diffusion, which helps preserve the bioactivity of NO from both the vascular wall and intraluminal sources.
It should be noted that this potential unidirectional high resistance of the erythrocyte membrane to NO is under debate. Some studies [45, 68] have suggested that the membrane permeability may be higher than previously reported, because the slow rate of NO binding to intraerythrocytic hemoglobin that was reported in previous studies [27, 70] might be due to other hematocrit-dependent mechanisms (i.e., extracellular unstirred solution layer ). Thus, the nature of this membrane resistance to diffusion of uncharged NO requires further investigation if we are to gain a more complete understanding of endocrine NO-mediated signaling.
Next, we review the possible sources of NO inside the erythrocyte.
Under both normoxic and hypoxic conditions, NO can rapidly react with erythrocytic hemoglobin by binding to the iron group in the heme pocket, with a kinetic rate that is limited only by the rate of diffusion [28, 44]. This rapid binding has meant that erythrocytes have long been regarded as a potent NO sink. In addition to reacting with the iron group in the heme pocket, nitrosonium cation (NO molecule after being oxidized) binds to the thiol group on certain proteins through S-nitrosation [62, 63]. Thus, the formation of S-nitrosohemoglobin (SNOHb) is another potential reaction between hemoglobin and NO.
The SNOHb hypothesis proposes that NO binds to T-state (deoxygenated) hemoglobin to form Fe2+NO; the bound NO group is then transferred during hemoglobin oxygenation from the ferrous iron group to the cysteine residue at position 93 in the β chain (ß -93 cysteine) of the hemoglobin molecule in the R-state, forming the relatively stable species, SNOHb [57, 63]. Then, under hypoxic conditions, SNOHb undergoes allosteric changes that result in the transfer of the NO group to the thiols of the anion exchange protein present on the membrane. The NO group also could be transferred to glutathione via transnitrosation. This bioactivity of NO could then be exported from the erythrocytes and regulate blood flow [2, 62]. According to the SNOHb hypothesis, in addition to playing a role in decreasing NO bioactivity, hemoglobin is also able to sense a change in the ambient O2 and release NO under hypoxic conditions, thereby regulating O2 delivery and hemorheological properties through an endocrine mechanism. It is not yet clear whether the nitrogen oxides released from erythrocytes are present in the form of nitrogen monoxide (free NO) [58, 74] or other NO equivalents . It is likely that NO bioactivity is preserved in the form of S-nitrosothiols; a variety of physiological regulations are dependent on the intracellular uptake of these compounds .
Because of the complex chemical reactions involved in the binding of NO to the thiol group of hemoglobin, it is difficult to distinguish any change in NO derived from this source from that of NO released from other sources. Thus far, there have been no direct experimental measurements in vivo of the amount of NO delivered to smooth muscle by this mechanism (or by other possible intraerythrocytic mechanisms). We have recently constructed a computational model that simulates NO production and transport within the framework of SNOHb hypothesis, with the assumption that free NO is finally released . The model predicted that the amount of NO delivered by intraerythrocytic SNOHb to the vascular wall is in the neighborhood of 6 pM, according to the SNOHb hypothesis; this NO availability would be as low as 0.25 pM if the SNOHb concentration were lower (~50 nM), as has been reported in some studies. It should be noted that if there is no protected NO transport mechanism (i.e., if only free diffusion occurs), the value of the NO released from this source would be three orders of magnitude lower than the 6-pM value, according to our computational simulation. Moreover, computational modeling showed that the amount of NO present in the smooth muscle depends strongly on facilitated membrane transport, membrane resistance to NO diffusion, and the physiological level and half-life of SNOHb.
Although intraerythrocytic SNOHb could be an important NO source in the intraluminal region, the role of SNOHb in hypoxic vasodilation is currently under debate [25, 32, 35, 57]. A complete understanding of how much NO comes from intraerythrocytic SNOHb requires further investigation. S-nitrosothiol (SNO) itself can also induce vasodilation, independent of NO and regardless of the NO release rate [49, 52, 53, 64]. NO bioactivity may be present in the form of other small molecules, such as S-nitrosoglutathione (GSNO).
Nitrite has been considered one of the stable metabolic end-products of NO. Recent experimental evidence suggests that nitrite acts as a reservoir that preserves NO bioactivity and that it can be enzymatically reduced to NO by hemoglobin under hypoxic conditions . The maximal nitrite reductase activity occurs around the P50 of hemoglobin because of the balance between the enzymatic activities of R-state (oxygenated) and T-state (deoxygenated) hemoglobin. R-state hemoglobin is thermodynamically favorable for nitrite reduction, with a lower redox potential for the hemes, while T-state hemoglobin provides the more ligand-free hemes as the reaction site with nitrite [29, 33]. The chemistry involved in this reaction can be expressed as
The reaction shown above provides an interesting pathway that could regulate blood flow and hemorheology under conditions of hypoxia. As previously mentioned, hemoglobin is abundant in erythrocytes; on the other hand, experiments have shown that erythrocytes are the major intravascular storage sites of nitrite in human blood , although there are considerable levels of nitrite in the plasma as well (~100 nM). Thus, erythrocytes are assumed to export NO molecules, particularly at the pre-capillary arterioles where the P50 of hemoglobin can most likely be achieved, provided NO bioactivity can be transported through protected mechanisms. Physiological experiments from a number of labs have shown that the addition of nitrite can cause downstream vasodilation and nitrite does indeed act as a vasodilator under hypoxic conditions [6, 20, 24, 41]. In addition to the nitrite reduction by hemoglobin, certain proteins in the extracellular region, such as xanthine oxidoreductase and mitochondrial proteins, can reduce nitrite to NO to regulate tissue oxygenation and other physiological functions [7, 22, 46].
However, there have been no direct experimental measurements in vivo of the amount of NO delivered through nitrite reduction, in part because of the complex chemistry involved in this mechanism. Computational modeling has been applied to simulate the reduction process and predict the amount of NO released via this pathway. Jeffers et al.  reported that the NO concentration in smooth muscle is 0.08 pM after the intraerythrocytic nitrite is reduced to NO, which is then transported through free diffusion. We also simulated the NO transport that occurs after nitrite reduction in erythrocytes, considering the three possible transport mechanisms discussed above : free diffusion, a membrane-associated mechanism, and a mechanism involving an intermediate product that only liberates NO as it reaches the vascular wall (Fig. 2). Our calculation for the intraerythrocytic sources, which predicted a level of ~0.04 pM NO (Figs. 2A and 2B), was in agreement with the value of Jeffers et al. Our model also predicted that NO is present in smooth muscle at 43 pM (Figs 2C and 2D) if the NO that is formed can be released from erythrocytes through a facilitated mechanism. This value could be increased to around 260 pM (Figs. 2E and 2F) in an extreme case in which an intermediate reaction exists, which would carry the NO and liberate it before it reached the vascular wall. From these results, we can see that, as opposed to free diffusion, a transport mechanism that protects NO bioactivity and facilitates its export out of erythrocytes is potentially important for deliverinFg significant amounts of NO from erythrocytes.
It is also clear that cell-free hemoglobin, which can be produced by hemolysis in sickle cell anemia or as a result of the administration of hemoglobin-based oxygen carriers, can either scavenge NO or react with nitrite to produce NO. Experimental evidence has indicated that hypertension and vasoconstriction are associated with a significant presence of cell-free hemoglobin in the blood [20, 66]. Furthermore, computational modeling has shown that the reaction of cell-free hemoglobin with nitrite generates an insignificant net amount of free NO (~0.02 pM) that reaches the vascular wall because of the lack of a possible protected transport mechanism (thus, there is a strong self-capture effect) . Therefore, erythrocytes, rather than the plasma in which a certain amount of cell-free hemoglobin is present, should be considered the major site for nitrite reduction to NO.
It is also interesting to consider that hemoglobin could potentially synthesize SNOHb by utilizing nitrite as a substrate, as an adjunct to the binding of free NO to the heme group via the cysteine in the hemoglobin β chain [3, 58]. This finding provides a link between the two hypotheses. Thus, it seems that there is a complex interaction among the hemoglobin iron group, the hemoglobin thiol group, free nitric oxide, and nitrite in erythrocytes.
As was discussed above, NOS3 in the endothelial cells has been considered the major source of the bioavailable NO in the vasculature. In addition to its presence in endothelial cells, functional NOS3 has been detected in erythrocytes, offering a possibility that hemoglobin-containing erythrocytes are indeed NO generators. In 1998, Chen and Mehta  reported the presence of the L-arginine-nitric oxide pathway in human erythrocytes, suggesting that NO-producing enzymes are indeed present in erythrocytes. However, the enzymes were later suggested to be inactive . Recently, Kleinbongard et al. , using multiple experimental approaches, have shown that erythrocytes express a functional NOS3 that can potentially serve as an intraluminal NO source. This NOS3 was located on the plasma membrane as well as in the cytoplasm of the erythrocytes. Note that NO from this source has been shown to be able to regulate the erythrocyte deformability and platelet aggregation, which are important factors that determine the blood flow in the microvasculature.
The controversy regarding the existence of NOS3 in erythrocytes is not yet resolved: Several recent studies have confirmed the presence or function of NOS3 in erythrocytes [69, 73], but others have questioned its existence . NO from this source has been thought to regulate the function of erythrocytes and platelets [40, 75]. If the NO synthesized by intraerythrocytic NOS3 can escape scavenging by hemoglobin as does other blood-borne NO, it could serve as another important source regulating vascular tone and hemorheological properties.
Because of the technical difficulties involved in attempting to obtain quantitative measurements of intraluminal NO, there are no published data regarding how much NO from this source is available to the intraluminal region or the vascular wall. Computational modeling  can simulate NO production from NOS3 in erythrocytes and its transport through an arteriole. Computer simulations indicate that, as has been shown for other intraerythrocytic pathways, whether a significant amount of NO from this source can be delivered to vascular smooth muscle is strongly dependent on the existence of a protected mechanism for NO delivery. If the expression level of NOS3 in erythrocytes is similar to that in endothelial cells, a combined NOS biochemical pathway analysis model [15, 16] and NO transport model  would predict a level of ~13 pM NO in the vascular smooth muscle from this source, given the existence of such a protected NO transport mechanism .
Free NO binds to deoxygenated hemoglobin to form iron-nitrosyl-hemoglobin (HbNO), which is present in erythrocytes at a relatively high concentration. This species has been put forward as a possible NO source within the lumen, in particular as an intermediate product during nitrite reduction . However, the dissociation rate of NO from the heme of hemoglobin is slow, in the range of 1 × 10 3 to 1 × 10 5 s 1 . This slow dissociation rate would make it possible for the capture of the released NO erythrocytes to occur quickly enough to quench any NO bioactivity from this source. Also, the vasodilation produced by the infusion of nitrogen oxide into the blood would require a faster dissociation rate if the HbNO were to have any regulatory effect on hemorheology.
Recent experiments have indicated that the presence of oxidants in the blood can expedite the release of NO from HbNO [30, 61]. The NO that is generated by this accelerated mechanism may either be transported out of the erythrocyte or be converted back into nitrite ions that can go through the nitrite reduction pathway in response to hypoxia. How much free NO can be released from HbNO under physiological conditions is not yet clear and requires further investigation.
Erythrocytes apparently play a dual role in regulating NO availability: they scavenge NO and generate methemoglobin, HbNO, nitrite, and nitrate; and they can serve as a site for NO-related reactions involving SNOHb, nitrite reduction, enzymatic cleavage by NOS3 expressed on erythrocytes, and/or iron-nitrosyl-hemoglobin. These pathways have been identified on the basis of experimental evidence and theoretical modeling; however, the amount of NO that each of these pathways contributes to the regulation of vascular tone and hemorheological properties remains to be elucidated in future theoretical and experimental research. Both the NO carried by and delivered from erythrocytes and the NO present in the vessel lumen may contribute significantly to the regulation of vascular tone and hemorheological properties. It is likely that NO derived from the intraluminal pathways, as well as that from previously recognized endothelial sources, is of great importance for vascular physiology. A full understanding, particularly at the quantitative level, of the relative contributions of the intraluminal NO pathways and other enzymatic sources of NO is critical to the future development of therapeutics targeting NO signaling in the vasculature.
This study is supported by NIH grants R01 HL018292 and R01 HL079087.
*This article is based on a paper given by Dr. K. Chen in Symposium 29 at the 13th International Congress of Biorheology and the 6th International Conference of Clinical Hemorheology, Penn State University, PA, USA, July 9-14, 2008