Differential effects of alpha-adrenoceptor agonists on human retinal microvessel diameter.

The effects of locally administered brimonidine, clonidine, and p-aminoclonidine on microvessel caliber were compared in human retinal tissues grafted into the hamster cheek pouch. Clonidine and p-aminoclonidine, but not brimonidine, potently constricted human retinal microvessels over a broad concentration range. All three agonists elicited significant vasoconstriction in naive hamster cheek pouch microvasculature. The alpha (2)-adrenoceptor antagonist, rauwolscine, inhibited p-aminoclonidine-induced constriction in naive hamster cheek pouch microvessels, but not p-aminoclonidine-induced effects in retinal grafts. Selective alpha (1)-adrenoceptor agonists evoked vasoconstriction in retinal grafts only at relatively high concentrations. These differential effects on the retinal microvasculature could not be readily explained solely on the basis of alpha (1)- or alpha (2)-adrenoceptor involvement. Clonidine, p-aminoclonidine and brimonidine are also imidazoline derivatives that interact with putative non-adrenergic imidazoline-sensitive binding sites, the so-called II-imidazoline binding site subtype implicated by some investigators in mediation of peripheral vasoconstriction. As with p-aminoclonidifie, the potent vasoconstriction in human retinal microvasculature elicited by moxonidine, an alpha -adrenergic agonist that has also been reported to exhibit selectivity for putative I-1-imidazoline binding sites, was not inhibited by the selective alpha -adrenoceptor antagonist, rauwolscine, nor by idazoxan, an antagonist characterized as having substantial activity at putative I-2-imidazoline binding sites. These data suggest the possible involvement of an unconventional non-adrenergic imidazoline-sensitive pathway in regulation of microvascular responses in the inner retina, and that drug activity mediated via such an imidazoline-sensitive component could potentially evoke deleterious effects in the retinal microvasculature.