"After prolonged (60 min) incubation.

"After prolonged (60 min) incubation.

In teleost, avian, and amphibian retinas, 3H-labeled GABA appears to accumulate exclusively in neurons of the inner nuclear and ganglion cell layers (Lam and Steinman, 1971; Marshall and Voaden, 1974a; Voaden et al., 1974; Pourcho et al., 1984). On the other hand, 3H-GABA is taken up primarily by the Müller cells in elasmobranch (Lam, 1975; Ripps and Wit-kovsky, 1985) and most mammalian retinas (Fig. 4.1). However, in rabbit and

Figure 4.1. Light microscope autoradiographs of 3H-GABA uptake in the retinas of several vertebrate species. Activity appears confined to the Müller cell fibers in rat (B), guinea pig (C), and skate (D) retinas. In rabbit (A), amacrine cells as well as Müller cells are labeled (Marshall and Voaden, 1974b, 1975; Ripps and Witkovsky 1985). ([B] copyright 1974Academic Press, Inc., reprinted with permission. [A], [C] copyright 1975 Elsevier Science, [D] copyright 1985 Elsevier Science, reprinted with permission.)

Figure 4.1. Light microscope autoradiographs of 3H-GABA uptake in the retinas of several vertebrate species. Activity appears confined to the Müller cell fibers in rat (B), guinea pig (C), and skate (D) retinas. In rabbit (A), amacrine cells as well as Müller cells are labeled (Marshall and Voaden, 1974b, 1975; Ripps and Witkovsky 1985). ([B] copyright 1974Academic Press, Inc., reprinted with permission. [A], [C] copyright 1975 Elsevier Science, [D] copyright 1985 Elsevier Science, reprinted with permission.)

mouse retinas, amacrine cells as well as Müller cells accumulate GABA (Marshall and Voaden, 1974b, 1975; Ehinger, 1977; Neal and Iverson, 1972; Blanks and Roffler-Tarlov, 1982) ; with prolonged incubation, mouse ganglion cells also take up radiolabeled GABA (Blanks and Roffler-Tarlov, 1982). Thus, GABA transport into Müller cells and into GABAergic nerve terminals can be expected to limit the spread of GABA from its release sites and thereby regulate GABA concentration in the synaptic clefts and extracellular space.

The identification of cells engaged in neurotransmitter uptake is sometimes complicated by inherent technical limitations. Because Müller cell endfeet occupy almost the entire surface area of the inner retina, caution needs to be exercised in interpreting results obtained when radiolabeled agents are applied to the vitreal surface of the intact retina. This is a particularly serious problem when a transmitter is delivered in vivo by intravitreal injection (cf. Ishida and Fain, 1981). The avid uptake of amino acids into Müller cells can mask uptake into neurons that would otherwise have accumulated these substances. This limitation will become more apparent when comparing some of the conclusions drawn from autoradiography with results obtained by electrophysiological methods (Section 4.2).

A second source of uncertainty is that the radiolabeled transmitter may be metabolized and lost to the incubation medium. This problem can often be minimized by pretreatment with compounds that inhibit degradation of the transmitter, and result in greater retention of radioactivity. Thus, the use of agents, e.g., hydroxylamine or amino-oxyacetic acid (AOAA) , that block the activity of the degradative enzyme GABA-T, has revealed that the magnitude of transmitter uptake may be grossly underestimated because of degradation (Martin, 1976; Neal and Starr, 1973; Ehinger, 1977).

Although most of the GABA taken up by Müller cells is probably rapidly metabolized (see Chapter 3), it is important to recognize that carrier-mediated transport processes subserving the cellular uptake of amino acids can operate also in the reverse direction to discharge these substances (cf. Levi and Raiteri, 1993). In goldfish, for example, [3H]GABA efflux can be evoked from horizontal cells, both in vitro and in situ (Ayoub and Lam, 1984; Yazulla, 1985), and a similar phenomenon has been reported in mammalian Müller cells. When rat retina is preloaded with radioactive GABA, the neurotransmitter appears to be taken up mainly by Müller cells (Neal and Iverson, 1972). It can then be released by electrical stimulation as well as by the application of solutions containing high K+, veratridine, or the GABA mimetic ethylenediamine (cf. Voaden and Starr, 1972; Sarthy, 1983). However, the assumption that Müller cells alone take up exogenous GABA in rat retina is questionable, and other cell types may also be responsible for its discharge. Interestingly, both Ca2+-dependent and Ca2+-independent mechanisms have been implicated (Sarthy, 1983), the latter presumably mediated by reversal of the uptake mechanism. More definitive information on the voltage- and ionic-dependence of "reverse transport" has been obtained with electrophysiological methods, and results obtained with this approach leave little doubt that Müller cells are capable of transmitter release (cf. Szatkowski et al., 1990). It has yet to be determined whether GABA is released under normal conditions, and in sufficient concentration to affect neuronal excitability.

The uptake of glutamate, the major excitatory transmitter in the retina, has also been well studied. Fewer species have been examined for glutamate uptake using radioactive tracers, but several studies demonstrated that glutamate is accumulated by Müller cells and photoreceptors in the mammalian retina by a Na+-dependent, high-affinity uptake mechanism (White and Neal, 1976; Ehinger, 1977; Sarthy et al., 1985). However, the use of electrophysiological methods in the analysis of membrane transport can often provide greater insight into the ionic dependence of the process. A more complete description of the glutamate uptake system is provided in Section 4.2.

4.1.2. GABA and Glutamate Transporters

Considerable progress has been made in elucidating the molecular structure and diversity of neurotransmitter transporters. Briefly, the GABA transporters belong to a large gene family that includes transporters for dopamine, serotonin, and several other neuroactive substances (Guastella et al., 1990,1992; Clark et al., 1992; Kanner, 1994). Several glutamate transporters have also been cloned and characterized (Storck et al., 1992; Kanai and Hediger, 1992; Pines et al., 1992; Fairman et al., 1995; Arriza et al., 1997; Eliasof et al., 1998). Their molecular structure indicates they constitute a distinct family, unrelated to the GABA family of transport proteins. Schematic drawings representative of the two families of neurotransmitter transporters illustrate their predicted membrane topology (Fig. 4.2), and a number of excellent reviews provide detailed descriptions of the structural features, putative regulatory mechanisms, and ionic requirements of these transport proteins (cf. Uhl, 1992; Amara and Kuhar, 1993; Kanner, 1994; Worrall and Williams, 1994; Malandro and Kilberg, 1996).

Within each family of transporters, there are several subtypes, which are often distributed differentially among neurons and glial cells. Regions of divergence in the amino acid sequences of the various transport proteins probably determine their specificity for a particular transmitter, whereas conserved regions may account for a feature common to most high-affinity plasma membrane transporters, namely, their sodium dependence. Using the electrochemical gradient of sodium ions, these transporters can accumulate neurotransmitter against significant concentration gradients.

(A), and 8 for the glutamate transporter (B). Filled circles indicate potential glycosylation sites (Worral and Williams, 1994). (Copyright 1994 The Biochemical Society, reprinted with permission.)

4.1.3. Localization of GABA Transporters

Studies of neuronal and glial preparations showed that GABA uptake systems for these different cell types exhibit different pharmacological properties (Amara and Kuhar, 1993). It was assumed, therefore, that there were at least two GABA transporters (GATs) expressed in the CNS: one specific to neurons and another specific to glial cells. In agreement with this expectation, several GABA transport proteins have been cloned and characterized (Guastella et al., 1990; Borden et al., 1992). In the mammalian retina, three GABA transporters (GAT-1, GAT-2, and GAT-3), which differ in peptide sequence, structure, and pharmacologic properties, have been reported (Ruiz et al., 1994; Brecha and Weigmann, 1994; Honda et al., 1995; Johnson et al., 1996). Immunocytochemical and in situ hybridization studies show that GAT-1 and GAT-3 are expressed in the neural retina, whereas GAT-2 is confined to RPE and the ciliary epithelia. However, the expression of GAT-1 and GAT-3 by inner retinal neurons and Müller cells illustrates the complexity of GABA uptake pharmacology in the retina. GAT-1 is highly expressed by amacrines, displaced amacrines, and some ganglion cells, but is found only at a low level in Müller cells (Fig. 4.3) (Ruiz et al., 1994; Honda et al., 1995;


Figure 4.3. Localization of GABA transporter, GAT-1, in the mouse retina. (A) and (B) are bright- and dark-field micrographs, respectively, showing GAT-1 mRNA localization by in situ hybridization using 35S-labeled probes. Silver grains are localized to cell bodies in the inner nuclear and ganglion cell layers (Ruiz et al., 1994). (Copyright 1994 Association for Research in Vision and Ophthalmology, reprinted with permission.)

Brecha and Wegmann, 1994). In contrast, GAT-3 is found predominantly in Müller cells, as well as in some amacrines in the proximal INL. GAT-3 immunoreactivity is found throughout the Müller cell, including regions where there is no evidence for GABAergic synapses, e.g., the outer nuclear layer. This result is similar to GAT-3 localization in astrocytic processes that are not in the vicinity of GABAergic synapses (Minelli et al., 1996).

Perhaps the most intriguing outcome of the localization studies is the finding that nonmammalian retinal horizontal cells, which are known to avidly take up GABA, do not express either GAT-1 or GAT-3, suggesting that another plasma membrane transporter mediates GABA uptake by these cells (Yang et al., 1997). Moreover, a particular retinal cell could express more than one GAT, and many neurons that synthesize GABA may not express a GABA transporter. It appears, therefore, that GABA uptake may not be an ideal marker for identifylng GABAergic neurons.

4.1.4. Localization of Glutamate Transporters

Based on DNA sequence, pharmacology, and channel properties, five types of glutamate transporter (also called excitatory amino acid transporter, EAAT)—EAAT1 (GLAST), EAAT2 (GLT-1), EAAT3 (EAAC1), EAAT4, and EAAT5—have been identified (Seal and Amara, 1999). EAAC1 is localized mostly to neurons whereas GLT-1 and GLAST appear to be the predominant glutamate transporters in astrocytes (Seal and Amara, 1999). EAAT4 is found only in the cerebellum whereas EAAT5 appears to be predominantly in the retina (Fairman et al., 1995; Arriza et al., 1997; Eliasof et al., 1998). Interestingly, EAAT4 and EAAT5 appear to function both as transporters and as ion channels permeable to CI- (see Section 4.2.2). Immuno-cytochemical studies suggest that EAAC1 is present in many neurons in the INL and the ganglion cell layer (GCL) (Rauen et al., 1996; Schultz and Stell, 1996) whereas EAAT5 is localized to both neurons and Müller cells (Arriza et al., 1997; Eliasof et al., 1998). Surprisingly, GLT-1 is localized mainly to cones and bipolar cells and is absent from both Müller cells and astrocytes (Rauen and Kanner, 1994). GLAST, however, has been localized exclusively to Müller cells (Fig. 4.4) (Rauen et al., 1996; Derouiche and Rauen, 1995; Lehre et al., 1997; Pow and Barnett, 1999; Wang et al., 1999). These results are in accord with earlier biochemical and electrophysiological data which suggested that Müller cells express a Na+-dependent, high-affinity glutamate transporter (Neal and White, 1971; Sarthy et al., 1986; Brew and Attwell, 1987; Schwartz and Tachibana, 1990).

Recently, GLT-1 and GLAST knockout mice have been generated (Ta-naka et al., 1997; Harada et al., 1998). Although the ERG in mice lacking GLT-1 appear to be normal, the b-wave and oscillatory potentials are depressed in GLAST-knockout mice. In addition, lack of GLAST leads to

Was this article helpful?

0 0

Post a comment