Changes in ultrastructure and voltage-dependent currents at the glia limitans of the frog optic nerve following retinal ablation

Application of CNTF or FGF-2 increases the number of M2-like macrophages after optic nerve injury in adult Rana pipiens

We have previously shown that a single application of the growth factors ciliary neurotrophic factor (CNTF) or fibroblast growth factor 2 (FGF-2) to the crushed optic nerve of the frog, Rana pipiens, increases the numbers and elongation rate of regenerating retinal ganglion cell axons. Here we investigate the effects of these factors on the numbers and types of macrophages that invade the regeneration zone. In control PBS-treated nerves, many macrophages are present 100 μm distal to the crush site at 1 week after injury; their numbers halve by 2 weeks. A single application of CNTF at the time of injury triples the numbers of macrophages at 1 week, with this increase compared to control being maintained at 2 weeks. Application of FGF-2 is equally effective at 1 week, but the macrophage numbers have fallen to control levels at 2 weeks. Immunostaining with a pan-macrophage marker, ED1, and a marker for M2-like macrophages, Arg-1, showed that the proportion of the putative M2 phenotype remained at approximately 80% with all treatments. Electron microscopy of the macrophages at 1 week shows strong phagocytic activity with all treatments, with many vacuoles containing axon fragments and membrane debris. At 2 weeks with PBS or FGF-2 treatment the remaining macrophages are less phagocytically active, containing mainly lipid inclusions. With CNTF treatment, at 2 weeks many of the more numerous macrophages are still phagocytosing axonal debris, although they also contain lipid inclusions. We conclude that the increase in macrophage influx seen after growth factor application is beneficial for the regenerating axons, probably due to more extensive removal of degenerating distal axons, but also perhaps to secretion of growth-promoting substances.

Ultrastructural studies of dorsal root axons regenerating through adult frog optic and sciatic nerves

Optic nerves of adult fish and amphibia can successfully regenerate, in part because their glial cells, unlike those of mammals, provide an environment permissive to regrowth. We altered the environment of regenerating dorsal root axons in the frog, Rana pipiens, by grafting segments of optic nerve to test the permissiveness of CNS glial cells to other sensory neurons. We compared these preparations to grafts of segments of sciatic nerve. After allowing various times for survival, light and electron microscopy were used to evaluate the grafts. An agglomeration of astrocytes, tightly joined by desmosomes, initially formed in the center of the optic nerve grafts. Around this grew regenerating dorsal root axons, accompanied by Schwann cells. At early stages, some axons formed dilated terminal structures, which were not seen in peripheral nerve grafts. The appearance of blood vessels within the graft and the dispersion of cells allowed larger numbers of axons to grow through the graft. By eight weeks, 48% of dorsal root sensory axons had grown through optic nerve grafts, compared to 84% for sciatic nerve. These results suggest that astrocytes from optic nerve are not inhibitory to, and provide a suitable substrate for, regrowing sensory neurons.

Potassium currents in cultured glia of the frog optic nerve

The processes that participate in clearing increases in [K+]o produced by active neurons include KCl uptake, Na pump stimulation, and spatial buffering. The latter process requires glial cells to carry: 1) inward K+ currents in regions where K+ is elevated at a glial membrane potential more negative than EK; and 2) outward K+ currents at normal K+ and glial membrane potential more positive than EK (Orkand et al: J Neurophysiol 29:788, 1966). Techniques for isolation and culturing glial cells brought new possibilities for studying ionic channels involved in spatial buffering. However, they raised the question of the extent to which the properties of ionic channels are changed due to the process of culturing when glial cells are exposed to an artificial environment and deprived of direct interaction with neurons. We studied potassium currents in glial cells from the frog optic nerve that were cultured for 1-8 days. At 24-48 h, cells exhibited an inwardly rectifying Cs+ blocked current (IK(IN)) that increased in amplitude and shifted its threshold of activation to EK when [K+]o was increased from 3 to 6 or 10 mM. IK(IN), diminished after 3 days in culture and virtually disappeared after 5 days. At 24-48 h, a potassium delayed rectifier current (IKD) was relatively small but became large at 3 days, and was practically the only current present after 5 days. IKD was activated at -8.5 +/- 0.58 mV(SE, n = 48) and 58 +/- 2.2% (SE, n = 48) blocked by 20 mM tetraethylammonium. The results of this study support the idea that the inward rectifying potassium channels (Kir) are responsible for carrying K+ into glial cells whenever [K+]o increases. However, the delayed rectifier potassium channels (KD) cannot provide the pathway for outward K+ current during spatial buffering, and another mechanism must be involved in this process. Our study provides further evidence that culture conditions can greatly influence functional expression of ionic channels in glial cells.

Unilateral optic nerve transection decreases 2-[125I]-iodomelatonin binding in retinorecipient areas and visual pathways of chick brain

In chick brain, specific 2-[125I]-iodomelatonin-binding was localized primarily in the visual system, i.e., retinorecipient and relay nuclei and fiber tracts of the tectofugal, thalamofugal, circadian and accessory visual pathways. Unilateral transection of the optic nerve (ONX) significantly reduced the binding of 2-[125I]-iodomelatonin (75 pM) in many, but not all, primary retinal targets and visual pathways at 7 and 14 days, but not 1 day, postlesion. As measured using quantitative autoradiography, 2-[125I]-iodomelatonin binding was decreased by 90% in both the central portion of the lesioned optic tract and one of its targets, the nucleus of the basal optic root (nBOR). Other retinorecipient areas exhibiting substantial decreases (60%) in 2-[125I]-iodomelatonin-binding included the optic tectum, lateroventral and dorsolateral geniculate nuclei and tectal gray area contralateral to the lesion. These findings are consistent with the hypothesis that melatonin receptors are located presynaptically on incoming optic nerve terminals in many retinorecipient areas. This localization may account for most of the binding sites in nBOR. In other primary visual areas, however, melatonin receptors also appear to be located on postsynaptic cells and/or non-retinal afferents. ONX had no significant effect on 2-[125I] -iodomelatonin binding in two retinorecipient areas, the visual suprachiasmatic nucleus and the dorsolateral anterior thalamus, which are part of the circadian/oculomotor and thalamofugal pathways, respectively. An unexpected consequence of ONX was that 2-[125I]- iodomelatonin binding was decreased in certain secondary (nucleus rotundus, isthmi nuclei) and tertiary level (ectostriatum) nuclei along the prominent tectofugal visual pathway. Binding in the tectorecipient nucleus triangularis was not significantly altered, however. Analysis of secondary level relay nuclei in the oculomotor pathway revealed that binding after ONX was decreased in the ipsilateral Edinger-Westphal nucleus but not in the oculomotor nuclei. Selective transsynaptic changes in 2-[125I]-iodomelatonin binding after lesion of the visual input most likely reflect activity-dependent regulation and functional plasticity of central melatonin receptors.

Voltage-dependent ion channels in glial cells

Glial cells, although non-excitable, express a wealth of voltage-activated ion channels that are typically characteristic of excitable cells. Since these channels are also observed in acutely isolated cells and in brain slices, they have to be considered functional in the intact brain. Numerous studies over the past 10 years have yielded detailed characterizations of glial channels permitting comparison of their properties to those of their neuronal counterparts. While for the most part such comparisons have demonstrated a high degree of similarity, they also provide evidence for the expression of some uniquely glial ion channels. An increasing number of studies indicate that the expression of "glial" channels is influenced by the cells' microenvironment. For example, the presence of neurons can induce or inhibit (depending on the preparation and type of channel studied) the expression of glial ion channels. Like ion channels in excitable cells, glial channels can be functionally regulated by activation of second-messenger pathways, allowing for short-term modulation of their membrane properties. Although the extent to which most of the characterized ion channels are involved in glial function is presently unclear, a growing body of data suggests that certain channels play an active role in glial function. Thus inwardly rectifying K+ channels in concert with delayed rectifying K+ channels are thought to be involved in the removal and redistribution of excess K+ in the brain, a process referred to as "spatial buffering". Glial K+ channels may also be crucial in modulating glial proliferation. Cl- channels and stretch-activated cation channels are believed to be involved in volume regulation. Na+ channels appear to be important in fueling the glial Na+/K(+)-pump, and Ca2+ channels are likely involved in numerous cellular events in which intracellular Ca2+ is a critical second messenger.

Na+ channels of Müller (glial) cells isolated from retinae of various mammalian species including man

Within the last few years, the expression of voltage-dependent, TTX-sensitive Na+ channels has been demonstrated in several types of neuroglial cells such as astrocytes and Schwann cells. Recently, we reported the occurrence of such Na+ currents in retinal Müller (glial) cells from dog and cat. This paper deals with the description of the properties of Na+ currents in Müller cells isolated from retinae of several mammalian species, as well as from human retinae. These Na+ currents were eliminated by TTX (1 microM), and by exposure to sodium-free extracellular solution; typically, they were demonstrable only after blocking most of the K+ conductance by Ba2+ (1 mM). Voltage-dependent activation and inactivation characteristics and time constants of the Na+ currents were similar to those of currents carried by neuronal Na+ channels. The estimated number of sodium channels per cell was low (about 1,500 channels per 7,500 microns 2), and the K+ conductance exceeded the peak Na+ conductance by an average factor of 5. Thus, the cells were incapable of generating action-potential-like responses under current clamp. Modelling estimations show that triggering of glial Na+ currents under physiological conditions, if any, can at best occur by emhaptic transmission at perinodal sites of optic axons. It is speculated that glial Na+ channels might be involved in neuroglial signalling events.

Facilitation of sodium currents in frog neuroglia by nerve impulses: dependence on external calcium

Facilitation of voltage-gated sodium currents in glial membranes by nerve impulses has been studied by using both the whole cell and loose-patch clamp techniques in the isolated intact optic nerve of the frog. During facilitation there is a shift in the voltage dependence of glial Na+ channels such that a given depolarization produces a larger inward Na+ current. Decreasing external calcium from 4 times normal to 0.2 times normal produced a similar shift in the current-voltage relation. Increasing the external calcium concentration to 4-5 times normal blocks facilitation. In reduced calcium, 0.1-0.2 times normal, the peak of facilitation was unaffected, but its decay was slowed. The addition of 1 mM nickel and 2 mM cobalt or 2 mM cadmium, to prevent depletion of extracellular calcium that might result from voltage-dependent entry of calcium into the axons, did not block the facilitation. The results suggest that, even though facilitation is blocked by high extracellular calcium, a decrease in extracellular calcium produced by axon impulses is not the cause of the facilitation.