These studies establish that P-Rex2, a GEF that specifically activates Rac small GTPases, is expressed in a cell- and synapse-specific manner in the retina. P-Rex2 localized to the synaptic terminals of photoreceptor and bipolar cells, as indicated by extensive double labeling with VGluT1 and other bipolar cell terminal markers. In contrast, P-Rex2 was not expressed in horizontal or amacrine cell processes or terminals or in the dendrites of bipolar or ganglion cells, as indicated by the absence of colocalization with an array of cell- and synapse-specific markers, including calbindin, synaptotagmin 2, PKC, GAD-65, GlyT1, and MAP-1 [21, 23, 25, 28, 31, 32, 34]. Thus, P-Rex2 in the retina is expressed specifically in the specialized glutamatergic ribbon synaptic terminals of photoreceptors and bipolar cells that transmit information vertically through the retina, and not the conventional synapses of amacrine cells or the processes of horizontal cells that mediate lateral processing in the IPL and OPL, respectively.
P-Rex2 localized to both rod and cone photoreceptor terminals in the OPL and to both rod and cone bipolar cell terminals in the IPL, indicating that P-Rex2 expression is not restricted specifically to either rod- or cone-driven synaptic pathways. Similarly, P-Rex2 expression is not exclusively associated with either ON or OFF pathways in the retina as P-Rex2 localized to bipolar cell terminals distributed across the ON and OFF sublayers of the IPL. The distribution of P-Rex2-positive bipolar cell terminals throughout the depth of the IPL further indicates that multiple bipolar cell types contained P-Rex 2 in their terminals. Double labeling for P-Rex2 in conjunction with PKC and synaptotagmin 2 positively identified P-Rex2 expression in the terminals of rod bipolar cells and Type 2 and Type 6 cone bipolar cells [21, 28, 31]. These studies also showed that the terminals of many additional bipolar cells also contained P-Rex2. On the basis of depth of stratification and the size characteristics of these terminals, it is clear that P-Rex2 also must be present in the terminals of several additional cone bipolar cell types and potentially could be present in the terminals of all bipolar cell types.
The distribution of P-Rex2 does not precisely match the reported distribution of its effector, Rac1. Rac 1 has been localized to photoreceptor outer segments and undergoes light-induced activation . Activation of Rac1 in the outer segment appears to have a key role in light-induced photoreceptor cell death [40, 41]. P-Rex2 seems unlikely to modulate light induced activation of Rac1 in the outer segment, as our data show little evidence for enrichment of P-Rex2 in photoreceptor outer segments. In contrast, our data suggest that P-Rex2 may be more important in the regulation of Rac1 activity specifically in photoreceptor terminals. Consistent with this notion, Rac1 has been localized to the distal portion of the mouse OPL . Although Rac1 is known to be important for proper development and polarization of Drosophila photoreceptors , conditional knockout of Rac1 from mouse rods does not appear to greatly disrupt the structure or function of mouse rods , although the structural organization and plasticity of photoreceptor terminals in vertebrate photoreceptors lacking Rac1 has not been examined in detail. The expression of Rac1 by cells in the inner retina and Rac1 labeling in the IPL has been reported previously [40, 43], but little is known regarding the cell-specific distribution or activation of Rac1 in the inner retina. The finding that P-Rex2 is selectively localized to bipolar cell terminals suggests that P-Rex2 provides specific regulation of Rac1 activity in those terminals.
The P-Rexes regulate actin cytoskeleton remodeling by activating Rac GTPases. P-Rex activation requires coincident signals via PI3K and G-protein receptor activation [18, 19, 44] and is a key mechanism for the regulation of membrane dynamics and remodeling of cytoskeleton in response to external cues [11, 15, 18, 19, 44, 45]. Diminished P-Rex function in neurons leads to aberrations in growth cone structure, membrane ruffling, neurite outgrowth, and neuritic architecture, resulting in functional deficits and impaired synaptic plasticity [11–13, 20]. It is likely that P-Rex2 serves a similar function in the terminals of photoreceptors and bipolar cells. One attractive possibility is that P-Rex2 may mediate adaptive remodeling of the terminal in response to simultaneous activation of G-protein and PI3K mediated pathways in the terminal. The terminals of photoreceptors and bipolar cells and their synaptic partners undergo significant anatomical remodeling in response to changes in illumination, including the extension and retraction of processes from the terminal itself and rearrangements associated with post-synaptic processes [46–54]. Plasticity of this nature is best known in the retinas of non-mammalian species [48–53], but adaptive structural changes also occur in mammalian photoreceptor and bipolar cell terminals [46, 47, 54]. This structural remodeling is dependent at least in part on the actin cytoskeleton as treatment with cytochalaisin D inhibits remodeling [50, 52], which would be consistent with a role for P-Rex2 in adaptive remodeling.
Another potential function for P-Rex2 is coordination of adaptive remodeling of the synaptic machinery within photoreceptor and bipolar cell terminals, presumably via activation of Rac1 which is known to be present in photoreceptors and other retinal cells [39–41, 43, 55]. For example, synaptic ribbons, and active zones in rod photoreceptor terminals undergo adaptive light-dependent (i.e., activity-dependent) remodeling [56–58]. Ribbon and active zone material is removed in the first few hours after light onset resulting in shortening or disappearance of some synaptic ribbons and active zones, and detachment of other ribbons from the terminal plasma membrane. This remodeling is then reversed in darkness. Synaptic vesicle density can also change with light- or dark-adaptation . The mechanism(s) mediating the movement and remodeling of ribbon and active zone material is currently unknown, but P-Rex2-mediated activation of Rac1 leading to local remodeling of actin within the terminal is a plausible contributor.
P-Rex2 also potentially might modulate functional plasticity at photoreceptor and bipolar cell terminals via the regulation of receptors at the surface of the terminal. Double knockout of P-Rex1 and 2 interferes specifically with late phase consolidation of long-term potentiation at the parallel fiber to Purkinje cell synapse in the cerebellum, most likely due to the failure of the synapse to consolidate changes in AMPA receptor density in the absence of P-Rex . It is unlikely that P-Rex2 would modulate functional plasticity at photoreceptor or bipolar cell terminals by direct modulation of transmitter release, as synaptic vesicle exocytosis is not actin-dependent. However, P-Rex2 activation potentially could affect recycling and trafficking of synaptic vesicles in the reserve pool, which are tethered to the actin cytoskeleton by the synaptic vesicle protein synapsin in conventional synapses [60–62]. However, it is not clear whether P-Rex2 might modulate synaptic vesicle interactions with the actin cytoskeleton in photoreceptor and bipolar terminals, which lack synapsins .
While the signals that regulate P-Rex2 activity in the terminals of photoreceptors and bipolar cells are not known, P-Rex2 activation requires coincident signaling via PI3K and G-protein-coupled receptor mechanisms [18, 19, 44]. Furthermore, the available evidence suggests that activated P-Rexes translocate to spatially restricted domains of the plasma membrane in order to activate Rac GTPases [19, 45]. Thus, P-Rex2 activation would appear to be ideally suited to tight spatial and temporal regulation of Rac GTPase activation in photoreceptor and bipolar cell terminals in response to very specific combinations of external signals. Although the precise signals that activate P-Rex2 in photoreceptor and bipolar cell terminals are not known, defects in phosphoinositide signaling disrupt the structure, maintenance and function of photoreceptor terminals [64, 65] and photoreceptors and bipolar cells possess a variety of G-protein coupled receptors that might contribute to activation of P-Rex2. A challenge for the future will be to define the precise functional role of P-Rex2 in ribbon synaptic terminals and the signals regulating its function.