While the activation of eNOS in response to Ca²⁺-elevating, receptor-dependent agonists, such as acetylcholine and bradykinin is relatively transient, the activation of eNOS following the application of fluid shear stress differs in that it can be maintained for several hours (i.e., as long as shear stress is applied). This shear stress-induced NO production is insensitive to the removal of extracellular Ca²⁺ and is not inhibited by the calmodulin antagonist, calmidazolium which abrogates the agonist-induced vasodilatation to acetylcholine. This apparently Ca²⁺-independent eNOS activation is inhibited by the tyrosine kinase inhibitors erbstatin A and herbimycin A suggesting that the tyrosine phosphorylation of eNOS or an associated regulatory protein is crucial for its Ca²⁺-independent activation. This Ca²⁺-independent activation of eNOS is also sensitive to the phosphatidylinositol 3-kinase inhibitor, wortmannin.

To determine the effects of shear stress on the phosphorylation of eNOS we incubated porcine aortic endothelial cells with ³²P for 6 hrs prior to cell stimulation, then immunoprecipitated eNOS and subjected the hydrolysed protein to two dimensional phosphoamino acid analysis. In unstimulated ³²P-labelled endothelial cells eNOS was phosphorylated on serine, threonine and tyrosine residues. Fluid shear stress (2-5 min) rapidly increased the phosphorylation of eNOS on serine and tyrosine residues (P-Tyr by 267 ± 44%; P-Ser by 130 ± 20%). Phosphorylation on P-Tyr and P-Ser returned to basal levels within 30 min and decreased below basal upon prolonged exposure to shear stress (up to 4 hrs). eNOS phosphorylation on tyrosine, but not serine residues, was prevented by herbimycin A. Removal of extracellular Ca²⁺ on the other hand prevented the shear stress-induced phosphorylation of eNOS.

Thus the initial transient phosphorylation of eNOS in response to fluid shear stress appears to be a Ca²⁺-dependent phenomenon. As an enhanced tyrosine phosphorylation of eNOS was therefore unlikely to regulate the Ca²⁺-independent activation of eNOS we investigated other shear stress-induced changes in eNOS. The application of fluid shear stress to cultured endothelial cells results in the rapid Ca²⁺-independent of the serine kinase Akt. Preliminary data suggest that the Akt-induced serine phosphorylation may be responsible for its maintained Ca²⁺-independent activation as the latter is sensitive to wortmannin.

One of the most marked effect of shear stress on the eNOS protein was its redistribution from a detergent (Triton X-100)-soluble to an insoluble cell fraction. A change in eNOS detergent solubility was first evident 15 to 30 min after initiating fluid shear stress and was sensitive to the tyrosine kinase inhibitors erbstatin A and herbimycin A as well as the chaperone-binding agent geldanamycin. A pharmacologically identical (Ca²⁺-independent, tyrosine kinase inhibitor-sensitive) activation of eNOS and alteration in its detergent solubility can be induced by protein tyrosine phosphatase inhibitors underlining the importance of an increase in the tyrosine phosphorylation of endothelial proteins in the sustained activation of eNOS. However given that the maintained activation of eNOS was not associated with the tyrosine hyperphosphorylation of eNOS, the tyrosine phosphorylation/activation of an eNOS-associated regulatory protein, rather than eNOS itself, appears to be crucial for its Ca²⁺-independent activation.

A change in the detergent solubility of a protein is frequently indicative of the formation of a protein-protein or protein-lipid complex for example, p91-phox, p22-phox, p47-phox and p67-phox in activated neutrophils; focal adhesion kinase (pp125^FAK) and Crk-associated substrate (p130^Cas) in fibroblasts. In both native and cultured endothelial cells a number of proteins are specifically co-precipitated with eNOS, most notably proteins corresponding to ~200, 90, 69-72 and 53 kDa. The Ca²⁺-elevating receptor-dependent agonist bradykinin, which does not alter the detergent solubility of eNOS, did not change the pattern of eNOS associated proteins. In contrast, the tyrosine phosphatase inhibitor, phenylarsine oxide which Ca²⁺-independently activates eNOS and renders it Triton-insoluble, decreased the recovery of 210 and 69-72 kDa proteins from the detergent-soluble fraction.

Proteins of approximately 103 and 87-91 kDa were recovered from the soluble fraction of cells treated with phenylarsine oxide but not with solvent, bradykinin or the combination of herbimycin A and phenylarsine oxide. These observations suggest that eNOS exists as part of a multi-molecular complex and that the Ca²⁺-independent activation induced by phenylarsine oxide seems to be linked to changes in the constituents of this complex and results in the alteration of its detergent solubility. Moreover since these changes were sensitive to herbimycin A, this process appears to involve activation of tyrosine kinases.

Apart from fluid shear stress hemodynamically relevant cell-cell-generated forces affect endothelial NO production. One example is the isometric contraction in which the development of contractile force within the smooth muscle cell layer counteracts the distending transmural pressure. Under such conditions there is a relative displacement of opposing cell layers within the vascular wall (e.g., smooth muscle cells vs. elastic lamina and endothelial cells) despite the fact that no net movement occurs. Although the displacement induced may be subtle, the close physical arrangement of endothelial focal adhesion contacts and the smooth muscle would tend to suggest that the forces developed at the abluminal surface of endothelial cells may be greater than those generated by shear stress on the luminal surface.

While these forces cannot be expressed as a simple physical term, isometric contraction of endothelium-intact arterial segments has been demonstrated to elevate NO production. In rings preconstricted under isometric conditions with PGF_2ALPHA, up to 40% of the maximal phenylephrine-induced contraction, the NOS inhibitor, NGnitro-L-arginine, elicited an additional contraction which was dependent on the level of preconstriction. A less pronounced effect of N^Gnitro-L-arginine was observed in rings preconstricted to over 50% of the maximally inducible tone. These observations indicate that stretch elicited by isometric contraction, within a certain range, activates eNOS in endothelial cells. More direct evidence for the release of NO by isometric contraction was obtained in bioassay experiments where the superfusate from isometrically contracted rings increased cyclic GMP levels in detector segments.

This NO production and the supplementary NGnitro-L-arginine-induced increase in vascular tone were inhibited by the non-selective kinase inhibitor staurosporine and the tyrosine kinase inhibitors erbstatin A and herbimycin A, whereas the calmodulin antagonist calmidazolium and the selective protein kinase C inhibitor Ro 31-8220 were without effect. Coincident with the NO formation was an increase in endothelial tyrosine phosphorylation which also correlated with the preconstriction level. Thus isometric contraction, tyrosine phosphatase inhibitors and fluid shear stress appear to activate the Ca²⁺/calmodulin-independent formation of NO via a similar tyrosine kinase-linked signalling pathway.

Activation of eNOS has up till now generally been associated with an increase in [Ca²⁺]_i however recent overwhelming evidence has demonstrated activation of eNOS by a Ca²⁺/calmodulin-independent pathway. This latter pathway, which is abrogated by the tyrosine kinase inhibitor erbstatin A but not by genistein, is associated with a change in the detergent solubility of eNOS and appears to be mainly activated by mechanical stimuli such as shear stress and isometric stretch. Receptor-dependent and -independent agonists however activate eNOS via the classical Ca²⁺/calmodulin pathway. Tyrosine kinases appear to be involved in both the Ca²⁺-dependent and -independent activation of eNOS, however the characteristic spectrum of sensitivity to the tyrosine kinase inhibitors used suggests that two separate signalling pathways exist.

REFERENCES

1. Ayajiki, K., Kindermann, M., Hecker, M., Fleming, I. & Busse, R. (1996) Intracellular pH and tyrosine phosphorylation but not calcium determine shear stress-induced nitric oxide production in native endothelial cells. Circ.Res. 78, 750-758.
2. Busse, R. & Fleming, I. (1998) Pulsatile stretch and shear: physical stimuli determining the production of endothelium-derived relaxing factors. J.Vasc.Res. 35, 73-84.
3. Fleming, I., Bauersachs, J. & Busse, R. (1997) Calcium-dependent and -independent activation of the endothelial NO synthase. J.Vasc.Res. 34, 165-174.
4. Fleming, I., Bauersachs, J., Fisslthaler, B. & Busse, R. (1998) Calcium-independent activation of the endothelial nitric oxide synthase in response to tyrosine phosphatase inhibitors and fluid shear stress. Circ.Res. 81, 686-695.
5. Kuchan, M.J. & Frangos, J.A. (1994) Role of calcium and calmodulin in flow-induced nitric oxide production in endothelial cells. Am.J.Physiol. 266, C628-C636