Disynaptic amplification of metabotropic glutamate receptor 1 responses in the olfactory bulb

Abstract

Sensory systems often respond to rapid stimuli with high frequency and fidelity, as perhaps best exemplified in the auditory system. Fast synaptic responses are fundamental requirements to achieve this task. The importance of speed is less clear in the olfactory system. Moreover, olfactory bulb output mitral cells respond to a single stimulation of the sensory afferents with unusually long EPSPs, lasting several seconds. We examined the temporal characteristics, developmental regulation, and the mechanism generating these responses in mouse olfactory bulb slices. The slow EPSP appeared at postnatal days 10-11 and was mediated by metabotropic glutamate receptor 1 (mGluR1) and NMDA receptors. mGluR1 contribution was unexpected because its activation usually requires strong, high-frequency stimulation of inputs. However, dendritic release of glutamate from the intraglomerular network caused spillover-mediated recurrent activation of metabotropic glutamate receptors. We suggest that persistent responses in mitral cells amplify the incoming sensory information and, along with asynchronous inputs, drive odor-evoked slow temporal activity in the bulb.

Figure 1.

Single ON stimulation evoked long-lasting synaptic responses in mitral cells. a, Mitral cell EPSPs evoked in an olfactory bulb slice by a single stimulation of the ON layer (1 shock, 100 μs) at different intensities. Each trace is an average of five episodes. b, For the same cell as shown in a, mitral cell responses were all or none at a stimulation intensity that produced ∼50% responses and 50% failures (9 V stimulation in this example). The average of 49 EPSPs (gray trace) is superimposed. c, In voltage-clamp recordings, the EPSCs (top, −70 mV) reversed polarity at a positive holding potential (bottom, +50 mV). d, The duration of the mitral cell EPSCs increased markedly after postnatal day (PN) 10. The intensity of the stimulation was set to evoke responses with peak amplitudes of 300–700 pA. Bottom, The integral of the EPSC was normalized to the peak amplitude and plotted as a function of age.

Figure 2.

ON stimulation evokes ongoing release of glutamate from mitral cell dendrites. a, Subsets of PG interneurons receive different inputs in the glomerulus (scheme) (see also Hayar et al., 2004). Some PG cells are directly contacted by ON terminals and respond to ON stimulation with a short-latency monosynaptic EPSC (right, top trace). Others are connected to mitral cells (MC) and respond with a long-latency burst of EPSCs (bottom trace). b, At a longer time scale, the latter responded to a stimulation of the ON (8 V) with a prolonged barrage of NBQX-sensitive EPSCs at V_h = −70 mV (bottom trace) and a barrage of NMDA receptor-mediated EPSCs at V_h = +50 mV (top trace).

Figure 3.

mGluR1 activation contributes to the synaptic depolarization of mitral cells. a, Top trace, The mGluR1-specific antagonist CPCCOEt (100 μm) reduced the mitral cell EPSP (second trace). The remaining slow component was blocked by the NMDA receptor antagonist AP-5 (150 μm, third trace). The fast EPSP that persisted in the presence of CPCCOEt and AP-5 was blocked by the AMPA receptor antagonist NBQX (20 μm, bottom trace). b, Inhibition by CPCCOEt or MCPG did not correlate with the amplitude of the mitral cell EPSP. c, In mGluR^−/− mice, the prolonged component of the ON-evoked mitral cell EPSP was entirely blocked by AP-5 (150 μm). NBQX (20 μm) inhibited the fast EPSP. d, Left, The integral of the EPSP was larger in wild type (WT) than in mGluR1^−/− mice (KO). Right, The amplitude at t = 1 s normalized to the peak amplitude was smaller in knock-out mice (KO) than in wild type (WT). Responses with similar peak amplitudes (WT: 15.5 ± 0.5 mV, n = 16; KO: 14.4 ± 0.5 mV, n = 18; p > 0.05) were selected for these comparisons.

Figure 4.

Synaptic activation of mGluR1 in the presence of ionotropic GluR blockers. a, Left, The mitral cell EPSP (inset) was nearly abolished by NBQX (20 μm) and AP-5 (150 μm). Trains of stimuli (stim.; 3, 5, or 10 stimuli at 100 Hz) increased the small residual EPSP, which was attributable to activation of mGluR1 (bottom traces). Right, Integral of the mGluR1-mediated EPSP evoked by single or multiple stimulations reached a maximum size after three stimuli (number of cells in parentheses). b, Inhibition of glutamate transporters with TBOA (100 μm) prolonged the duration of the mGluR1-mediated EPSP evoked by a train of stimuli (integral, 235 ± 35% of control; n = 4) without affecting its amplitude (109 ± 5.1% of control; n = 4).

Figure 5.

mGluR1 activation in the presence of the NMDA receptor antagonist AP-5. a, Left, AP-5 (150 μm) inhibited >80% of the mitral cell EPSP evoked by ON stimulation. Right, The remaining slow component (top trace) was selectively blocked by the mGluR1-specific antagonist LY367385 (25–50 μm; bottom trace). b, Left, NBQX (20 μm) blocked not only the fast component but also the mGluR1 component as recorded in the presence of AP-5. Right, NBQX inhibition of the slow EPSP for six cells in the presence of AP-5.

Figure 6.

The intraglomerular network generates glutamatergic inputs onto mitral cell dendrites. a, Left, Spontaneous (spont.) depolarizations in a pair of mitral cells projecting to the same glomerulus had a similar time course to the evoked EPSPs. Note also that the depolarization were synchronous in the two cells. Right, Evoked events (thin traces) had faster rise times than spontaneous events (thick traces). b, In voltage clamp, spontaneous and ON-evoked currents also had similar shapes and reversed polarity at positive holding potentials (top trace). c, EPSPs were evoked by a single stimulation of the ON at different intensities in a mitral cell from a connexin36 knock-out mouse. d, Comparison of the integral (left) and decay (right) of EPSPs evoked in wild-type (WT) or Cx36^−/− (KO) mice. As in Figure 3, the ratio of the EPSP amplitude at t = 1 s and peak amplitude were used as a measure of the response duration. Responses with similar amplitude (WT: same as in Fig. 3; KO: 14.7 ± 0.5 mV, n = 17, p > 0.1) were selected for comparison. e, ON-evoked EPSCs in a wild-type mitral cell recorded at V_h = +50 mV in control solution (i.e., 1 mm external Mg²⁺; top traces). Most, but not all, responses exhibited a slow component that prolongs the NMDA receptor-mediated-EPSC. Increasing extracellular Mg²⁺ from 1 mm to 3 mm (bottom traces) selectively blocked the slow current responsible for the prolonged decay of the EPSC. Ten consecutive responses are shown in each condition. f, A series of 15 EPSCs recorded at +50 mV in a connexin36^−/− mouse lacked the slow components observed in wild-type animals. Solutions for recordings in b, e, and f included gabazine (2 μm).

Figure 7.

Selective inhibition of dendritic release reduced the slow component of the mitral cell response. a, Nickel had no effect on the amplitude and paired-pulse ratio of EPSCs evoked by ON stimulation in a PG neuron that received monosynaptic input from ON terminals (scheme). Twelve consecutive traces are shown in control conditions and in the presence of nickel. b, In a PG neuron synaptically connected to mitral cells (MC; scheme), nickel blocked the EPSC evoked by an extracellular stimulation (1 pulse, 100 μs) in the mitral cell layer. Ten consecutives traces are shown in each condition. c, Top, Inhibiton by nickel of the ON-evoked mitral cell EPSP. Nickel reduced the integral of the response (bottom, opened circles) but not the peak of the response (•; n = 6). d, Series of consecutive ON-evoked EPSCs in a mitral cell recorded in voltage clamp at V_h = +50 mV in control condition (top traces) or in the presence of nickel (bottom traces). The nickel concentration was 300 μm in all experiments.