Graded responses and spiking properties of identified first-order visual interneurons of the fly compound eye.

R O Uusitalo,M Juusola,M Weckström

DOI: https://doi.org/10.1152/jn.1995.73.5.1782

1995-01-01

Journal of Neurophysiology

Abstract:1. We studied the graded and spiking properties of the "non-spiking" first-order visual interneurons of the fly compound eye in situ with the use of intracellular recordings. Iontophoretical QX-314 injections, Lucifer yellow marking, and (discontinuous) current-clamp method together with transfer function analysis were used to characterize the neural signal processing mechanisms in these neurons. 2. A light-OFF Spike was seen in one identified anatomic subtype (L(3), n = 6) of the three first-order visual interneurons (L(1), L(2), and L(3), or LMCs) when recorded from synaptic region (i.e., in the 1st visual ganglion, lamina ganglionaris) in dark-adapted conditions. Hyperpolarization of the membrane potential by current caused the identified L(1) (n = 4), as well as L(3) (n = 6), to produce an OFF spike, a number of action potentials, and some subthreshold depolarizations after the light-ON response. In L(2) the OFF spike or action potentials could not be elicited. 3. To produce action potentials in L(1) and L(3), it was found to be necessary to hyperpolarize the cells similar to 35-45 mV (n = 43) below the resting potential (RP) in the synaptic zone. Recordings from the axons of these cells revealed that near the second neuropil (chiasma) the threshold of these spikes was near to (similar to 10 mV below, n = 16) or even at the RP when an ON spike was also produced (n = 4). 4. The recorded spikes were up to 54 mV in amplitude, appeared with a maximum frequency of up to 120 impulses/s, and had a duration of similar to 8 ms. In L(1) and L(3) the spikes were elicited either after a light pulse (L(3)) Or after a negative current step that was superimposed on a hyperpolarizing steady-state current (L(3) and L(1)). A positive current step (similarly superimposed on a hyperpolarizing steady-state current) also triggered the spikes during the step. 5. Iontophoretic injection of a potent intracellularly effective blocker of voltage-gated sodium channels, QX-314, irreversibly eradicated the spikes and subthreshold depolarizations (n = 5). In addition, further injections elongated the light-on responses and decreased or even abolished the Light-OFF response. 6. Negative prepulses followed by positive current steps were applied from the RP, to test the activation-inactivation properties of the channels responsible for the OFF spike. During this experiment the increase of the negative prepulse (removal of the inactivation) increased the amplitude of the OFF spike from 7 to 21 mV, whereas the increase of the positive test pulse (activation test) led to the increase of the spike from 5 to 51 mV. 7. These results indicate that voltage-gated sodium channels that are normally highly inactivated are responsible for the OFF-spike generation in L(3) and L(1), and that probably a component of the light-OFF response is mediated via the same conductance as well. The fact that L(2) did not show any spiking properties suggests that this subtype is responsible of feeding visual information to a different functional subsystem than L(1) or L(3) 8. Frequency domain analysis suggested that the putative sodium channels in LMC axons considerably increase the gain of the signals at high frequencies and produce a resonance. Thus in these neurons the OFF spike is probably used to compensate the gain loss and to improve the signal-to-noise ratio (SNR) during the passive propagation through the long and thin axon. All three interneurons therefore regulate the gain in different manner and could be considered as parallel pathways with differently modulated responses.