So far only few studies report about GAL

So far only few studies report about GAL in the visual system: GAL has been detected in the cornea of rat (Jones and Marfurt, 1998), mouse (Henken and Martin, 1992) and pig (Stone et al., 1988), and in the uvea of rat (Strömberg et al., 1987), pig (Stone et al., 1988) and cat (Grimes et al., 1994), and was mainly related to the autonomic nervous system supplying the eye, being either of sympathetic or primary afferent origin. On the other hand, reports in rat revealed also an additional parasympathetic origin of GAL (Jones and Marfurt, 1998). Further, GAL has been reported in the choroid of primates including man (May et al., 2004) and haloperidol haldol (Stubinger et al., 2010), being present in intrinsic choroidal neurons. These in turn have been linked to a local control of choroidal blood vessels and choroidal non-vascular smooth muscle cells (Schrodl et al., 2003). However, GAL-receptors in the eye have not been reported so far. GALR1-3 have been detected in basal layers of the corneal epithelium and conjunctiva, thus indicating a possible tear fluid mediated GAL action. While reports of GAL in lacrimal glands exist, it was up to now exclusively detected in glandular neuronal elements (Adeghate, 1996, Adeghate and Singh, 1994), and its presence in human tear fluid needs to be proven. On the other hand, ocular epithelial GAL might be involved in wound healing, and indeed, GAL upregulation has been detected during the formation of skin granulation tissue with concomitant stimulation of angiogenesis (Yamamoto et al., 2011). This might be of interest for corneal inflammatory processes and the clinical application of GAL-antagonists might lead to a desired antiangiogenetic effect. Besides epithelium, stromal keratinocytes could be involved in these pathological processes as well, especially since recent work was able to demonstrate a progalanin release from fibroblasts (Yamamoto et al., 2014). In the anterior uvea, both iris sphincter and dilator muscle display immunoreactivity for GALR1-3. GAL-immunoreactivity has been detected in iris sensory nerve fibers of non-human primates (macaques and baboon; (Firth et al., 2002) and rat (Strömberg et al., 1987). In the human anterior uvea, GAL-immunoreactivity has been reported “in positively stained nerve fibers at the muscle tips” (Selbach et al., 2000). Sources of these nerve fibers are cranial autonomic ganglia: since human superior cervical ganglion lacks galaninergic neurons (Baffi et al., 1992) but galaninergic neurons are detected in human trigeminal ganglion (Del Fiacco and Quartu, 1994), it is likely that GAL is part of the primary afferent signal cascade in humans. In line with these observations is that GAL elicits no iris sphincter contraction in non-human primates in-vitro (Yamaji et al., 2005) and in-vivo (Almegard and Andersson, 1990). On the other hand, a GAL modulation or inhibition of the cholinergic transmission has been demonstrated for the isolated rabbit iris sphincter (Ekblad et al., 1985, Yamaji et al., 2003), however, the receptors responsible are unknown yet. While information about GAL in human ciliary and pterygopalatine ganglion is missing, it has been detected in many of aforementioned ganglia in cat (Grimes et al., 1994) and birds (Stubinger et al., 2010). In any way, as our results show, it seems that GAL action on iris muscles is mediated via GALR1-3, most likely originating from the autonomic nervous system, with GALR3 is solely reserved for the ciliary muscle. Different in this respect are iris vessels, or vessels in the ciliary body: here, the endothelium displays immunoreactivity for GALR1/GALR2 (iris) or GALR3 only (ciliary body), thus indicating a non-neuronal control mechanism that might be blood-stream mediated. These results differ from in-vitro experiments of primary human dermal endothelial cells in which GALR1-3 were absent (Schmidhuber et al., 2007). As for the here detected GALR1 immunoreactivity in the ciliary epithelium, our results confirm earlier in-vitro observations of GALR1 in non-pigmented epithelium derived cells (Ortego and Coca-Prados, 1998). We propose a non-neuronal involvement in aqueous humor formation that might be driven by the concentration of GAL in the aqueous humor (Ortego and Coca-Prados, 1998), the origin of which is unknown. However, this concentration could also influence the activity of corneal endothelium, since all three receptors were detected, as our results show. Earlier experiments in non-human primates revealed an unaltered outflow after intracameral administration of GAL (Ortego and Coca-Prados, 1998).