Role of neurotransmitter receptors in mediating light-evoked responses in retinal interplexiform cells

Abstract

Interplexiform (IP) cells are a long-neglected group of retinal neurons the function of which is yet to be determined. Anatomical study indicates that IP cells are located in the inner nuclear layer, juxtaposed with the third-order neurons. However, the synaptic transmission of IP cells in the inner retina is poorly understood. Using whole cell patch-clamp and pharmacological techniques, we extensively studied synaptic receptors in IP cells. The IP cells in amphibian retinal slices were identified by electrical and morphological properties with voltage-clamp recording and Lucifer yellow dialysis. We find that light-evoked excitatory postsynaptic currents (L-EPSCs) are mediated by AMPA and N-methyl-d-aspartate receptors in IP cells. Although both receptors contributed to the amplitude and kinetics of L-EPSCs, AMPA receptor desensitization substantially shaped L-EPSCs in the neurons, similar to those found in the third-order neurons. The light-evoked inhibitory postsynaptic currents (L-IPSCs) in IP cells were primarily mediated by strychnine-sensitive glycine receptors with a small component of GABA(C) receptors. GABA(C) receptor rho2 subunits were detected in IP cells with single-cell RT-PCR assays. Expression of GABA(C) receptors is one of the special features for IP cells, distinct from most of the third-order neurons that depend on GABA(A) and glycine receptors to relay the inhibitory signals. However, GABA(A) receptors in IP cells acted like nonsynaptic receptors that were activated by exogenous GABA application. Furthermore, L-IPSCs in IP cells were inhibited by the serial inhibitions between amacrine cells in the inner retina. In addition, application of neurotransmitters on the axon terminals of IP cells had no significant current generated in the cells, indicating that the synaptic inputs of IP cells are mainly from the inner retina. This study demonstrates the important role that light signals are encoded by both experiment of inhibitory receptors in IP cells.

Fig. 1.

The morphology and electrophysiology of interplexiform (IP) cells in whole cell recording experiments. A: morphology of an IP cell in a retinal slice was revealed in Lucifer yellow dialysis during whole cell recording. B: the voltage-dependent channel currents were recorded from an IP cell activated by various voltage steps, from −100 to +60 mV. Large sodium currents were recorded from the cell when the membrane potential held at −40 mV or more positive voltage. The current-voltage (I-V) relationship was measured and plotted (see right). C: light-evoked action potentials were recorded from an IP cell in a dark-adapted retinal slice in current-clamp mode.

Fig. 2.

Pharmacological study of glutamate receptor subtypes in IP cells. A: light-evoked excitatory currents (L-EPSCs) were recorded from an IP cell voltage-clamped at −65 mV. B: the L-EPSCs are partially blocked by 1-(4-aminophenyl)-4-methyl-7,8-methylendioxy-5H-2,3-benzo-diazepine (GYKI), an AMPA receptor antagonist, and are fully blocked by GYKI with d,l-2-amino-7-phosphonoheptanoic acid (AP-7), a N-methyl-d-aspartate (NMDA) receptor antagonist. The effect of the antagonists can be partially washed out in an IP cell.

Fig. 3.

Differentiation of AMPA and NMDA receptor-mediated L-EPSCs. L-EPSCs were recorded from IP cells with 100 μM picrotoxin and 2 μM strychnine in the bath solution. A: application of cyclothiazide to block AMPA receptor desensitization increases the L-EPSCs. B: the statistical analysis of AMPA receptor desensitization-affected peak current (*P < 0.005, n = 4) and charge influx of L-EPSCs at the onset of a light stimulus from 8 cells (**P < 0.001, n = 4). C: AP-7, a NMDA receptor antagonist, reduces L-EPSCs. D: statistically, AP-7 significantly reduces the amount of peak current (*P < 0.005, n = 5) and charge influx) (**P < 0.005, n = 5), measured at the onset of a light stimulus. Error bar: SD.

Fig. 4.

Evidence of glycine receptors and GABA_C receptors that mediate light-evoked inhibitory postsynaptic currents (L-IPSCs) in IP cells. The onset of L-IPSCs were recorded from IP cells in dark-adapted retinal slices (A) Strychnine blocks most L-IPSCs and the remaining currents are blocked by imidazole-4-acetic acid (I4AA, n = 4). B: the effect of high concentrations of strychnine and SR95531, a potent GABA_A receptor antagonist, on L-IPSCs. The strychnine blocks the same amount L-IPSCs with and without SR95531; picrotoxin blocks the rest of the L-IPSCs. C: average reductions of L-IPSCs at the onset of a light by strychnine with and without I4AA, calculated from 8 IP cells (*P < 0.001, **P < 0.0001, n = 8). Error bar: SE.

Fig. 5.

Expression of GABA_C receptor ρ2 subunit in IP cells detected in single-cell RT-PCR. A: the cDNA gel indicates the RT-PCR products from various samples including total retinal mRNA, single-cell mRNA, and negative control. B: alignments of GABA_C receptor ρ2 subunit protein sequences from different species. The protein sequence of salamander ρ2 subunit shows 93.7, 94.3, and 93.0% homology with that of human, mouse, and morone (fish), respectively. The black background highlights the amino acid sequences with high similarity.

Fig. 6.

Pharmacological study of L-IPSCs in IP cells. The cells were voltage-clamped at 0 mV. Blockage of network inhibition by a GABA_A receptor antagonist, bicuculline, largely increases L-IPSCs that are significantly blocked by strychnine (n = 5). B: picrotoxin blocks both GABA_A and GABA_C receptors in network, which significantly enhances L-IPSCs in an IP cell; with picrotoxin, the L-IPSCs are fully blocked by strychnine (n = 6). The small inward (downward) currents in the bottom trace are glutamate residue currents. C and D: the effects of bicuculline and picrotoxin on the L-EPSCs were recorded from amacrine cells held at −70 mV. Picrotoxin, but not bicuculline, substantially increases the L-EPSCs in amacrine cells.

Fig. 7.

GABA-elicited currents in IP cells are blocked by GABA_A and GABA_C receptor antagonists. IP cells were voltage clamped at 0 mV, and GABA was very briefly applied at the inner retina in slice preparation. A: puff GABA-elicited currents are reduced by bicuculline; the currents are further reduced by bicuculline with 10 μM I4AA, and the currents are fully blocked by 100 μM picrotoxin. B: the average suppressions of GABA-elicited currents by bicuculline (*P < 0.005, n = 8), I4AA (**P < 0.005, n = 6), I4AA or 1,2,5,6-tetrahydropyridine-4-yl-methylphosphinic acid (TPMPA; NS: no significant difference, n = 6), and picrotoxin (***P < 0.001, n = 6). Error bar: SE.

Fig. 8.

Study of the effects of potassium, GABA, and glycine at the axon processes of IP cells. A: the diagram of the experimental procedure. B: puff 100 mM potassium at the outer plexiform layer (OPL) elicits excitatory currents in an IP cell; the currents can be completely blocked by application of Cd²⁺ in bath solution. C and D: in the presence of Cd²⁺, puff 100 μM GABA or 100 μM glycine on the OPL has generated no response in IP cells.