1. The actions of two pharmacological agents, barium ions (Ba2+) and picrotoxin (PTX), were examined on components of the electroretinogram (ERG) in frog retina. Depth profiles of light-evoked field potentials were recorded, and current source densities (CSDs) were computed from these. 2. Ba2+ abolished the M-wave, slow PIII, and the c-wave, but only decreased b-wave amplitude down to approximately 65% of control amplitude. 3. Ba2+ abolished a slow current sink in the inner plexiform layer (IPL) and the source at the inner limiting membrane (ILM). This IPL sink/ILM source appears to generate the M-wave. 4. Ba2+ decreased the current sink at the outer plexiform layer (OPL) to approximately 70% of control amplitude, and it increased an IPL source. This Ba(2+)-resistant OPL sink/IPL source appears to generate a significant portion of the b-wave. The Ba(2+)-sensitive portion of the b-wave might be generated by Müller cells. 5. PTX enhanced retinal field potentials, particularly the M-wave in the proximal retina. This enhanced M-wave was shown to originate from an enhanced IPL sink/ILM source. 6. Our results suggest that the M-wave originates from Müller cells, through the spatial buffering of the light-evoked increase in [K+]o of the proximal retina. A portion of the b-wave may also originate from Müller cells, but a stronger direct contribution from depolarizing bipolar cells is suggested.

The technique of current source-density analysis was applied to several components of the light-evoked field potentials (electroretinogram) from the retina of the superfused eyecup of rabbit. The depth distributions of the major current sources and sinks were: b-wave--sink at outer plexiform layer, source at inner plexiform layer; M-wave--sink at inner plexiform layer, source at retinal surface; and slow PIII--source near outer plexiform layer, sink at retinal surface. These distributions, along with the sensitivities of these responses to certain pharmacological agents, support earlier studies that Müller cells generate the M-wave and slow PIII, but that depolarizing bipolar cells directly generate the b-wave.

The technique of current-source density (CSD) analysis for extracellular potentials is reviewed, along with some methodological features that are important for performing CSD analysis of the electroretinogram. In addition, three formulas for computing CSD's are examined on model circuits of resistors and current generators. Finally, CSD results from frog retina that bear on the origins of the b, d, and M waves, along with slow PIII, are presented. It is concluded that the b and d waves are generated primarily and directly by bipolar cells, whereas the M wave and the slow PIII are generated by Müller (glial) cells through the K+ spatial buffer mechanism.

1. Dark flashes were used to evoke spatiotemporal profiles of the d wave of the electroretinogram in frog retina. The current source density technique was used to compute the extracellular current sinks and sources underlying the d wave. 2. The largest d wave current was a sink at the outer plexiform layer (OPL sink) and a source at the inner plexiform layer (IPL source). Several properties of this sink/source, including its relative insensitivity to Ba2+, suggest that it arises from hyperpolarizing bipolar cells. 3. There was a slower IPL sink, along with a source at the inner limiting membrane (ILM source). Its properties, along with its sensitivity to Ba2+ and enhancement by picrotoxin, suggest it arises from the K+ spatial buffer currents of Müller cells. This IPL sink/ILM source underlies the M wave. 4. A small photoreceptor sink/source also contributes to the d wave. This is most significant during the very beginning of the d wave. 5. Our results indicate that in frog retina and at the stimulus conditions used in this paper, the d wave originates primarily and directly from the hyperpolarizing bipolar cells. Photoreceptors likely contribute a small amount, especially during the initial portion of the d wave. There is possibly a small Müller cell contribution.

1. The technique of current source density (CSD) analysis was used to obtain depth profiles of light-evoked source/sink distributions in the retina of the frog. 2. The effects of a number of methodological considerations on the CSD profiles were explored. Adoption of best technique leads to minimal variability and noise in the CSD profiles. Several situations that give rise to artifactual source/sinks were identified. 3. At the time of the b-wave peak, there is a large sink near the outer plexiform layer (OPL), a source at the inner limiting membrane (ILM), and a complex response at the inner plexiform layer (IPL). 4. The IPL response at light onset consists of 1) an initial, sharp sink, followed by 2) a slower source with a time course similar to the b-wave, and last 3) an even slower sink with a time course similar to the M-wave. 5. The first component of the IPL response is the current sink for the proximal negative response (PNR) elicited by a diffuse light stimulus. The PNR has current sources both proximal and distal to this layer. 6. The OPL sink underlies the b-wave. The second component of the IPL response is a source for the b-wave. Part of the ILM source also might be a b-wave source. 7. A large part of the ILM source is the current source for the M-wave. The third component of the IPL response is the M-wave sink, but it is small because it subtractively interacts with the IPL b-wave source.