Neuronal and microvascular alterations induced by the cholinergic toxin AF64A in the rat retina

Depletion of cholinergic amacrine cells by a novel immunotoxin does not perturb the formation of segregated on and off cone bipolar cell projections

Cone bipolar cells are the first retinal neurons that respond in a differential manner to light onset and offset. In the mature retina, the terminal arbors of On and Off cone bipolar cells terminate in different sublaminas of the inner plexiform layer (IPL) where they form synapses with the dendrites of On and Off retinal ganglion cells and with the stratified processes of cholinergic amacrine cells. Here we first show that cholinergic processes within the On and Off sublaminas of the IPL are present early in development, being evident in the rat on the day of birth, approximately 10 d before the formation of segregated cone bipolar cell axons. This temporal sequence, as well as our previous finding that the segregation of On and Off cone bipolar cell inputs occurs in the absence of retinal ganglion cells, suggested that cholinergic amacrine cells could provide a scaffold for the subsequent in-growth of bipolar cell axons. To test this hypothesis directly, a new cholinergic cell immunotoxin was constructed by conjugating saporin, the ribosome-inactivating protein toxin, to an antibody against the vesicular acetylcholine transporter. A single intraocular injection of the immunotoxin caused a rapid, complete, and selective loss of cholinergic amacrine cells from the developing rat retina. On and Off cone bipolar cells were visualized using an antibody against recoverin, the calcium-binding protein that labels the soma and processes of these interneurons. After complete depletion of cholinergic amacrine cells, cone bipolar cell axon terminals still formed their two characteristic strata within the IPL. These findings demonstrate that the presence of cholinergic amacrine cells is not required for the segregation of recoverin-positive On and Off cone bipolar cell projections.

Cholinergic amacrine cells are not required for the progression and atropine-mediated suppression of form-deprivation myopia

Muscarinic cholinergic pathways have been implicated in the visual control of ocular growth. However, the source(s) of acetylcholine and the tissue(s) which regulate ocular growth via muscarinic acetylcholine receptors (mAChRs) remain unknown. We sought to determine whether retinal sources of acetylcholine and mAChRs contribute to visually guided ocular growth in the chick. Cholinergic amacrine cells were ablated by intraocular injections of either ethylcholine mustard aziridinium ion (ECMA; a selective cholinotoxin) or quisqualic acid (QA; an excitotoxin that destroys many amacrine cells, including those that release acetylcholine). Disruption of cholinergic pathways was assessed immunocytochemically with antibodies to the acetylcholine-synthesizing enzyme choline acetyltransferase (ChAT) and three different isoforms of mAChR, and by biochemical assay for ChAT activity. ECMA (25 nmol) destroyed two of the four subtypes of cholinergic amacrine cells and attenuated retinal ChAT activity, but left retinal mAChR-immunoreactivity intact. QA (200 nmol) destroyed the majority of all four subtypes of cholinergic amacrine cells, and ablated most mAChR-immunoreactivity and ChAT activity in the retina. ECMA and QA had no apparent effect on mAChRs or cholinergic fibres in the choroid, only marginally reduced choroidal ChAT activity, and had little effect on ChAT activity in the anterior segment. Toxin-treated eyes remained emmetropic and responded to form-deprivation by growing excessively and becoming myopic. Furthermore, daily intravitreal injection of 40 microg atropine for 6 days into form-deprived toxin-treated eyes completely prevented ocular elongation and myopia. We conclude that neither cholinergic amacrine cells nor mAChRs in the retina are required for visual regulation of ocular growth, and that atropine may exert its growth-suppressing influence by acting upon extraretinal mAChRs, possibly in the choroid, retinal pigmented epithelium, or sclera.

The AF64A model of cholinergic hypofunction: an update

Based on numerous reports in the literature since 1980, one can now conclude that ethylcholine aziridinium (AF64A) is selective for the cholinergic system in vivo, and that the effect is both dose- and site-dependent. Thus, AF64A treatment, under the correct conditions of dose and time will result in selective reductions in levels of ACh, AChE, ChAT, HAChT, and K(+)- and ouabain-stimulated release of ACh. While other neurotransmitters may also be affected in brains of AF64A treated rats, the effect is only transient and is most probably secondary to the initial cholinergic deficit-induced by AF64A, reflecting an adaptive reaction of these neurotransmitter systems, which are normally integrated with cholinergic interconnections, to the cholinergic deficiency induced by AF64A. This paper provides a historical perspective for the development of AF64A as a selective cholinotoxin, and surveys its potential mechanisms of action at the neurochemical and molecular levels. Moreover, the availability of an animal model such as the AF64A-treated rat, in which the cholinergic system has been compromised selectively for an extended period of time, has allowed investigators to study a wide variety of questions that relate to factors controlling cholinergic function in vivo. Several key illustrations are presented at the end of this paper.

Direct cytotoxicity of ethylcholine mustard aziridinium in cerebral microvascular endothelial cells

The choline analogue ethylcholine mustard aziridinium (AF64A) is a potent and irreversible inhibitor of choline uptake in brain synaptosomes and is used as a neurotoxin to produce animal models of cholinergic hypofunction. However, previous studies have shown that intraocular administration of AF64A in rats not only reduced the number of cholinergic neurons in the retina, but also induced ultrastructural alterations in the microvasculature. The purpose of this study was to investigate whether AF64A has a direct cytotoxic effect on endothelial cells. As revealed by the measurement of lactate dehydrogenase activity in the culture medium, AF64A produced similar concentration-dependent cellular damage in cultures of bovine cerebral endothelial cells and in the human cholinergic neuroblastoma cell line SK-N-MC, but not in bovine cerebral smooth muscle cells. The toxic effect of AF64A correlated well with the affinity of the choline transport system detected in each cell type. The effect of the toxin on endothelial cells was mediated by its interaction with the endothelial cell choline carrier, as demonstrated by the following observations: (a) AF64A inhibited [3H]choline uptake in a concentration-dependent manner in both cultured and freshly isolated cerebral endothelial cells, and (b) the addition of choline or hemicholinium-3 to the culture medium prevented the AF64A-induced toxicity in endothelial cell cultures.

Possible involvement of cholinergic and glycinergic amacrine cells in the inhibition exerted by the ON retinal channel on the OFF retinal channel

In the frog retina, the inhibition exerted by the ON channel on the OFF channel was evidenced by the increase in transient ganglion cell OFF responses, when the ON channel was blocked by 2-amino-4-phosphonobutyrate (APB). Intraocular administration of the neurotoxic choline analog ethylcholine mustard arizidinium ion (ECMA) also provoked an increase in the number of spikes of transient ganglion cell OFF responses, without suppressing the ON responses. APB, when administrated after ECMA, abolished the ON responses, but did not modify the OFF responses already increased by ECMA. Neurons located in the inner part of the inner nuclear layer were histologically altered by the toxin, and choline acetyltransferase activity was significantly depressed in ECMA-treated retinas. A double immunostaining experiment showed that amacrine cells containing glycine bear muscarinic binding sites. These results confirm the participation of cholinergic neurons in the inhibition exerted by the ON retinal channel on the OFF retinal channel, and suggest the involvement of a cholinergic/glycinergic loop of amacrine cells in this mechanism.