Activation of group I metabotropic glutamate receptors enhances persistent sodium current and rhythmic bursting in main olfactory bulb external tufted cells

Abstract

Rhythmically bursting olfactory bulb external tufted (ET) cells are thought to play a key role in synchronizing glomerular network activity to respiratory-driven sensory input. Whereas spontaneous bursting in these cells is intrinsically generated by interplay of several voltage-dependent currents, bursting strength and frequency can be modified by local intrinsic and centrifugal synaptic input. Activation of metabotropic glutamate receptors (mGluRs) engages a calcium-dependent cation current (I_CAN) that increases rhythmic bursting, but mGluRs may also modulate intrinsic mechanisms involved in bursting. Here, we used patch-clamp electrophysiology in rat olfactory bulb slices to investigate whether mGluRs modulate two key intrinsic currents involved in ET cell burst initiation: persistent sodium (I_NaP) and hyperpolarization-activated cation (I_h) currents. Using a BAPTA-based internal solution to block I_CAN, we found that the mGluR1/5 agonist DHPG enhanced I_NaP but did not alter I_h. I_NaP enhancement consisted of increased current at membrane potentials between −60 and −50 mV and a hyperpolarizing shift in activation threshold. Both effects would be predicted to shorten the interburst interval. In agreement, DHPG modestly depolarized (∼3.5 mV) ET cells and increased burst frequency without effect on other major burst parameters. This increase was inversely proportional to the basal burst rate such that slower ET cells exhibited the largest increases. This may enable ET cells with slow intrinsic burst rates to pace with faster sniff rates. Taken with other findings, these results indicate that multiple neurotransmitter mechanisms are engaged to fine-tune rhythmic ET cell bursting to context- and state-dependent changes in sniffing frequency.

in the olfactory bulb, glutamatergic external tufted (ET) cells play important functional roles in the intra- and interglomerular networks. ET cells rhythmically burst from approximately 0.5 to 10 Hz, matching the theta frequency range characteristic of rodent sniffing rates (Hayar et al. 2004a; Liu and Shipley 2008b; Wachowiak and Shipley 2006). ET cell bursts entrain olfactory nerve input and in turn provide monosynaptic dendrodendritic input to periglomerular and short axon cell inhibitory interneurons, most of which are themselves devoid of direct olfactory nerve input (Hayar et al. 2004b; Shao et al. 2009). Activation of periglomerular cells by ET cell bursts provides a major source of feedback and feedforward inhibition that shapes the temporal structure of intraglomerular activity with rhythmic respiration (Shao et al. 2009, 2012; Wachowiak and Shipley 2006). ET cell input to short axon cells, by contrast, drives an interglomerular inhibitory circuit (Aungst et al. 2003; Liu et al. 2013; Whitesell et al. 2013). Dendritic glutamate release from ET cells is thought to generate long-lasting depolarizations that amplify olfactory nerve input onto, and synchronize spike output among, mitral/tufted cells associated with the same glomerulus (Carlson et al. 2000; De Saint Jan et al. 2009; Gire and Schoppa 2009; Hayar et al. 2005; Najac et al. 2011; Shao et al. 2012).

The bursting properties of ET cells thus reinforce or strengthen the rhythmic nature of olfactory input onto the glomerular network. ET cell bursting arises spontaneously and is dependent on a number of intrinsic voltage-dependent channels. Thus ET cell burst generation and frequency are unperturbed when ionotropic glutamate and GABA receptors are blocked (Hayar and Ennis 2007; Hayar et al. 2004a; Liu and Shipley 2008b). However, ET cell bursting can be potently regulated by several classes of modulatory neurotransmitter receptors. Serotonin, for example, increases ET cell burst frequency by enhancing a nonselective cation current in these cells (Liu et al. 2012). More recently, dopamine was found to increase ET cell burst frequency by positive modulation of the hyperpolarization-activated cation current (I_h) that plays a key role in intrinsic bursting (Liu et al. 2013).

ET cells express several modulatory metabotropic glutamate receptor (mGluR) subtypes, in particular, high levels of group I mGluRs (i.e., mGluR1 and mGluR5; Heinbockel et al. 2004; Romano et al. 1995; Sahara et al. 2001; Shigemoto et al. 1992). These receptors can be engaged directly by olfactory nerve input and also by glutamate spillover from the apical dendrites of ET and mitral/tufted cells (De Saint Jan and Westbrook 2007; Dong et al. 2009; Ennis et al. 2006; Johnston and Delaney 2010; Schoppa and Westbrook 2001; Yuan and Knöpfel 2006). mGluR1/5 activation triggers a calcium-dependent nonselective cation current (I_CAN) that robustly increases both the strength and frequency of ET cell bursting. However, mGluR-evoked effects on bursting were incompletely reduced by nonselective cation channel antagonists. This suggests that mGluR activation may enhance ET cell excitability by modulation of intrinsic currents involved in burst initiation. The goal of this study therefore was to investigate whether mGluRs modulate intrinsic ET cell currents involved in spontaneous burst initiation using patch-clamp electrophysiology in rat olfactory bulb slices.

MATERIALS AND METHODS

Slice preparation.

Sprague-Dawley rats (18–30 days old of either sex) were decapitated in accordance with an approved Institutional Animal Care and Use Committee protocol and National Institutes of Health guidelines. The olfactory bulbs were removed as previously described (Dong et al. 2009) and immersed in oxygenated sucrose artificial cerebrospinal fluid (ACSF) composed of (in mM): 62 NaCl, 26 NaHCO₃, 1.25 NaH₂PO₄, 3 KCl, 4 MgSO₄, 0.1 CaCl₂, 15 glucose, and 120 sucrose; pH 7.3, 310 mOsm. Horizontal slices were cut with a Vibratome 3000 (Vibratome, St. Louis, MO) at thickness of 400 μm. After recovery at 33°C for 20 min, slices were then incubated until used at room temperature (22°C) in normal ACSF equilibrated with carbogen (95% O₂-5% CO₂) and composed of (in mM): 126 NaCl, 26 NaHCO₃, 3 KCl, 1.25 NaH₂PO₄, 2 MgCl₂, 2 CaCl_2, and 20 glucose; pH 7.3, 310 mOsm. For voltage-clamp experiments, cadmium and nickel were added to the ACSF, and NaH₂PO₄ was excluded. For recording, a single slice was placed in a recording chamber and continuously perfused with carbogen-saturated ACSF at the rate of 1.5 ml/min.

Electrophysiology.

All recordings were performed at 30°C. Neurons were visualized using an upright microscope (BX50WI; Olympus, Tokyo, Japan) equipped with epifluorescence and near-infrared differential interference contrast optics. Whole cell voltage-clamp recordings were made with a MultiClamp 700B amplifier (Molecular Devices, Sunnyvale, CA). The pipette resistance was 6–8 MΩ. Only neurons with access resistance <30 MΩ were included in this study. Unless otherwise noted, voltage-clamp recordings from ET cells were made with pipettes containing (in mM): 125 cesium methanesulfonate (CsMeSO₃), 1 NaCl, 10 phosphocreatine di(tris) salt, 3 MgATP, 0.3 Na₂GTP, 0.5 EGTA, 10 HEPES; pH 7.3, 290 mOsm. The intracellular solution for current-clamp recording was composed of (in mM): 124 potassium gluconate, 1 NaCl, 10 phosphocreatine di(tris) salt, 3 MgATP, 0.3 Na₂GTP, 0.5 EGTA, and 10 HEPES; pH 7.3, 290 mOsm. Lucifer yellow (0.02%) was added to the intracellular solution in all experiments for in situ labeling.

Analog signals were low-pass filtered at 2 kHz (MultiClamp 700B) and digitized at 5 kHz using a Digidata 1440A interface and pClamp 10 software (Molecular Devices). Current-voltage (I–V) relationships of persistent sodium currents (I_NaP) were studied with a voltage-ramp protocol (−80 to −40 mV, 26 mV/s, 1.5-s duration) from a holding potential (HP) of −60 mV with 6-cyano-7-nitroquinoxaline-2,3-dione (CNQX; 10 μM), dl-2-amino-5-phosphonovaleric acid (APV; 50 μM), gabazine (SR-95531; 10 μM), tetraethylammonium chloride (TEA; 10 mM), cadmium (200 μM), and nickel (1 mM) in the ACSF. The I_CAN elicited by 3,4-dihydroxyphenylglycine (DHPG) requires internal store calcium release and persists in the presence of the voltage-gated calcium channel blockers cadmium and nickel (Dong et al. 2009). In some experiments, I–V curves for I_NaP were converted to conductance according to the formula g = I/(V_t − V_r), where V_t = membrane test potential and V_r = the estimated reversal potential for sodium ions. To estimate the voltage at which half of the current was activated (V_a), the curves of conductance vs. voltage (g–V) were fitted with the Boltzmann equation: 1/{1 + exp[(V_a − V_t)/k]}. Membrane resistance was also calculated from currents elicited by negative voltage pulses (−20 mV, 100-ms duration) from a −60-mV HP. In current-clamp, ET cell resting or minimum membrane potential and the depolarizing envelope amplitude and duration were defined and measured as previously described (Dong et al. 2009; Liu and Shipley 2008b). Spike detection was performed offline using Mini Analysis software (Synaptosoft, Decatur, GA), and custom-designed algorithms were used to measure parameters of ET cell spike bursts as previously described (Dong et al. 2009): firing frequency, burst frequency (defined as the number of bursts per second), spikes per burst, and burst duration (defined as the time interval between the 1st and the last spike in a burst). A spike burst was defined as a series of 2 or more consecutive spikes that had interspike time intervals of <75 ms as previously reported (Hayar et al. 2004a). Group data I–V curves were plotted with a sigmoidal function (Origin 8.5; MicroCal, Northampton, MA). Data, expressed as means ± SE, were statistically analyzed using ANOVA followed by Newman-Keuls post hoc comparisons and with t-tests.

ET cells were identified by their characteristic electrophysiological and morphological characteristics (Hayar and Ennis 2007; Hayar et al. 2004a,b, 2005; Liu and Shipley 2008b). Electrophysiologically, they exhibited spontaneous rhythmic bursting in the cell-attached or current-clamp recordings. Rhythmic bursting was typically confirmed in extracellular recordings before switching to whole cell voltage-clamp mode. Additionally, ET cells exhibited randomly occurring (i.e., nonbursting) spontaneous excitatory postsynaptic currents in voltage-clamp recordings (HP = −60 mV; Hayar et al. 2004a,b, 2005). Morphologically, all recorded ET cells were characterized by soma location within the glomerular layer and by a primary dendrite with a tuftlike arborization that ramifies within a single glomerulus. They also lacked secondary dendrites. Cells that lacked these characteristics were discarded.

Drugs and solutions.

In most experiments, drugs and solutions were applied to the perfusion solution with a three-way valve system. Recording media, gabazine, CNQX, APV, TEA, cadmium chloride, nickel(II) chloride hexahydrate, and BAPTA were obtained from Sigma-Aldrich (St. Louis, MO). DHPG, L-CCG-I, LY 367385, 2-methyl-6-(phenylethynyl)-pyridine (MPEP), l-2-amino-4-phosphonobutyric acid (AP4), riluzole, and TTX were purchased from Tocris Bioscience (Ellisville, MO).

RESULTS

Activation of group I mGluRs enhance I_NaP.

Data are reported from voltage- or current-clamp recordings obtained from 54 ET cells in 51 slices from 31 rats. Previous findings indicate that activation of mGluR1/5 by DHPG elicits a robust I_CAN at voltages corresponding to the resting or interburst membrane potential of ET cells (Dong et al. 2009). Block of this current reduced but did not abolish DHPG-evoked increases in ET cell bursting. We therefore investigated potential effects of DHPG on intrinsic ET cell membrane currents, focusing on currents that are essential to burst initiation and/or control burst frequency, namely I_NaP and I_h. Both are active at ET cell resting membrane potential (Hayar et al. 2004a; Liu and Shipley 2008b) and are therefore potential candidates for DHPG effects on bursting. To investigate whether DHPG modulates I_NaP, we applied a slow voltage-ramp (see materials and methods) in the presence of CNQX (10 μM), APV (50 μM), gabazine (10 μM), TEA (10 mM), cadmium (200 μM), and nickel (1 mM) in the ACSF to block ionotropic glutamate and GABA receptors and voltage-gated calcium and potassium channels. As previously reported (Hayar et al. 2004a; Liu and Shipley 2008b), slow depolarizing ramps elicited a TTX-sensitive inward current, I_NaP, developing at −65 to −60 mV and increasing with depolarization to approximately −40 mV (Fig. 1A). As shown in Fig. 1, DHPG (30 μM) elicited an inward current shift at all voltages and also appeared to increase I_NaP superimposed on this current. Application of TTX (1 μM) blocked I_NaP, leaving a residual linear current identical to the previously described TTX-insensitive I_CAN evoked by DHPG (Dong et al. 2009). The I_CAN was associated with a significant decrease in membrane resistance measured with negative voltage steps as previously reported (−50 ± 8.8 MΩ, n = 7; P = 0.002, paired t-test). Subtraction of the TTX-sensitive component (Fig. 1, B–D) revealed a significant impact of DHPG on I_NaP (n = 7; P < 0.05, 2-way ANOVA). Compared with control values, I_NaP amplitude was increased in the range of −65 to −50 mV with a peak increase of −15.6 ± 5.8 pA at −55 mV (P < 0.05, ANOVA followed by Newman-Keuls test; Fig. 1C). Boltzmann equation fits of g-V curves (Fig. 1D) showed that DHPG shifted the activation threshold of I_NaP to more negative voltages with half-maximal activation values of −50.2 ± 0.9 and −55.8 ± 1.3 mV in control and DHPG (n = 7; P = 0.001, paired t-test). Comparable results were obtained by determining significant increases in I_NaP from −80 mV, which identified values of −50 and −60 mV in control and DHPG (n = 7; P < 0.05, ANOVA followed by Newman-Keuls tests). At potentials positive to approximately −40 mV, the magnitude of I_NaP in the presence of DHPG was equivalent or slightly less than in control. Thus DHPG shifts the overall window of I_NaP toward hyperpolarizing potentials.

Fig. 1.

3,4-Dihydroxyphenylglycine (DHPG) elicits calcium-dependent cation current (I_CAN) and enhances persistent sodium current (I_NaP). A: inset indicates order of drug application. Voltage-clamp recordings from an example external tufted (ET) cell in control (trace 1) show that a slow depolarizing voltage-ramp (shown at bottom) elicits a TTX-sensitive I_NaP that developed at approximately −60 mV. Im, membrane current. DHPG (30 μM, trace 2) elicited a linear TTX-insensitive current, I_CAN, across all membrane potentials (Vm) and also increased the magnitude of I_NaP. Note that DHPG shifts the activation threshold of I_NaP toward more negative potentials. Subsequent addition of 1 μM TTX (trace 3) eliminated I_NaP but did not affect I_CAN. B: current-voltage (I–V) profiles of I_NaP from data in A in control and DHPG calculated by subtracting the currents in each condition in the presence and absence of TTX. Note that DHPG produced a leftward shift to decrease the I_NaP activation threshold. C: group subtraction data as in B from 7 ET cells show I_NaP in control and DHPG conditions. D: group data conductance vs. voltage plot showing the leftward shift in I_NaP activation threshold; dotted lines denote half-maximal activation values in control and DHPG.

Low concentrations of TTX (e.g., 10–20 nM) and riluzole have been reported as effective blockers of sustained or persistent sodium currents, including I_NaP (Cramer et al. 2007; D'Ascenzo et al. 2009; Pace et al. 2007). TTX at 20 nM eliminated I_NaP (n = 8) and its enhancement by DHPG (n = 4) but did not alter I_CAN (Fig. 2). Riluzole (30 μM) had similar effects (n = 2, data not shown). Group I mGluRs include mGluR1 and mGluR5. Both subtypes are expressed by neurons in the glomerular layer and by ET cells specifically (Dong et al. 2009; Fotuhi et al. 1993; Heinbockel et al. 2004; Romano et al. 1995; Sahara et al. 2001; Shigemoto et al. 1992; Van den Pol et al. 1995). Activation of I_CAN and the enhancement of I_NaP were prevented when DHPG (30 μM) was applied in the presence of the mGluR1 and mGluR5 antagonists LY 367385 (100 μM) and MPEP (50 μM), respectively (Fig. 2, C and D, and Fig. 3). We did not attempt to determine the relative contribution of mGluR1 and mGluR5.

Fig. 2.

Low concentrations of TTX and group I metabotropic glutamate receptor (mGluR) antagonists block DHPG-evoked enhancement of I_NaP. A and B: individual cell (A) and group (B) data subtraction plot show that 20 nM TTX blocks I_NaP in control and the presence of DHPG but does not alter I_CAN (red plot in B). C and D: in the presence of 2-methyl-6-(phenylethynyl)-pyridine (MPEP; 50 μM) and LY 367385 (LY; 100 μM), DHPG did not alter I_NaP or evoke I_CAN.

Fig. 3.

Bar graph comparing the magnitude of I_NaP at −55 mV in different experimental conditions. L-AP4, l-2-amino-4-phosphonobutyric acid. *P < 0.05 compared with respective control values, paired t-tests. #P < 0.05, unpaired t-test.

ET cells express group II and III mGluRs (Dong et al. 2009; Hayar and Ennis 2007), and we next investigated whether agonists for these mGluR classes alter I_CAN or I_NaP. We found that the group II mGluR agonist L-CCG-I (10 μM) enhanced I_NaP in a manner similar to DHPG (n = 6; Fig. 3). Unlike DHPG, L-CCG-I did not evoke an I_CAN, consistent with a lack of effect on ET cell input resistance measured with negative voltage steps (−9.3 ± 17.4 MΩ change, n = 6; P > 0.24, paired t-test). The group III mGluR agonist AP4 (10 μM) did not elicit I_CAN, alter I_NaP (Fig. 3; n = 3; P = 0.45), or alter input resistance (−21.3 ± 46 MΩ change; P > 0.17).

To investigate DHPG modulation of I_NaP further, recordings were made with BAPTA-based pipettes to block the DHPG-evoked I_CAN in ET cells (Dong et al. 2009). BAPTA alone tended to increase slightly the basal I_NaP magnitude at hyperpolarized potentials (−80 to −55 mV) and to increase I_NaP at more depolarized potentials (e.g., −50 to −40 mV; Fig. 4C). As shown in Fig. 4, with BAPTA pipettes, the DHPG-evoked I_CAN was absent, whereas I_NaP was enhanced. At −55 mV, DHPG increased I_NaP amplitude by −13.7 ± 6.2 pA (n = 7; Figs. 3 and 4C), a value that did not differ from that obtained with standard pipettes (−15.6 ± 5.8 pA; P > 0.6, unpaired t-test). DHPG did not alter membrane resistance measured with negative voltage steps (−6.6 ± 18.2 MΩ change, n = 7; P > 0.9, paired t-test), further consistent with the absence of I_CAN.

Fig. 4.

DHPG enhances I_NaP when I_CAN is blocked. A: voltage-ramps elicit the TTX-sensitive I_NaP in this example ET cell using a BAPTA-based recording pipette (see main text). Note that DHPG (30 μM) enhances I_NaP, whereas the DHPG-evoked I_CAN is absent in this condition (compare with Figs. 1A and 2A). B and C: single cell (B) and group (C; n = 7) data subtraction plots show the DHPG-evoked enhancement of I_NaP. C includes the I–V profile of I_NaP obtained with the standard recording solution (Control-No BAPTA). Note that I_NaP magnitude is larger at depolarized potentials (−45 to −40 mV) compared with that obtained with a BAPTA-based internal solution.

ET cells exhibit an I_h that is active at resting membrane potential and cooperates with I_NaP to initiate ET cell bursts (Ennis and Hayar 2008; Liu and Shipley 2008b). However, we found no evidence for modulation of I_h by DHPG (n = 6; P > 0.15, 2-way repeated-measures ANOVA). I_h was eliminated by bath application of the selective I_h blocker ZD 7288 (200 μM, n = 4; P < 0.001, 2-way ANOVA).

Does DHPG modulation of I_NaP enhance rhythmic bursting?

Activation of group I mGluRs by DHPG enhances ET cell rhythmic bursting (Dong et al. 2009). This is mediated in part by activation of I_CAN. Does the enhancement of I_NaP contribute to rhythmic bursting following mGluR activation? To investigate this question, we initially tested whether blockade of I_NaP by 10 nM TTX or 30 μM riluzole could attenuate DHPGs effects on bursting in current-clamp recordings. Consistent with previous studies (Liu and Shipley 2008b), we found that 10 nM TTX (n = 3) and riluzole (n = 2) invariably and completely eliminated spontaneous bursting; bursting could not be restored by depolarizing current injection (data not shown). As an alternative approach, we tested the effects of DHPG in the presence of APV-CNQX-gabazine using BAPTA-based pipettes, a condition that abolishes DHPG-evoked I_CAN. Compared with recordings with standard pipettes in similar conditions [n = 11 ET cells, data from Dong et al. (2009)], BAPTA did not significantly alter ET cell resting membrane potential or membrane resistance (n = 11; P > 0.5 and P > 0.7, unpaired t-tests). DHPG (30 μM) depolarized the membrane potential by 3.6 ± 0.8 mV (n = 11; P < 0.006, paired t-test), a value significantly less than that obtained with standard recording pipettes in similar conditions [7.3 ± 1.1 mV, n = 11 ET cells from Dong et al. (2009); P = 0.012, unpaired t-test]. The mean 3.6-mV depolarization is similar to the 2.8-mV value predicted by the 13.7-pA mean inward current evoked by DHPG at −55 mV and the average ET cell input resistance (201 ± 11 MΩ, n = 11). DHPG did not alter the depolarizing envelope amplitude or duration (P > 0.18 and P > 0.4, respectively, paired t-tests).

As shown in Fig. 5, DHPG significantly increased the firing rate and burst frequency by approximately 252 and 243%, respectively, but had no effect on the number of spikes per burst or burst duration. DHPG did not alter the intraburst frequency (40.4 ± 8.6 vs. 37.7 ± 8.9 Hz, n = 11; P = 0.32, paired t-test). Thus the increased firing rate is due to a faster burst rate rather than an increase in the number of spikes per burst or the intraburst frequency. We asked whether the effect of DHPG depended on the basal burst properties, finding an inverse relationship between the basal burst rate and the increase evoked by DHPG (Fig. 5; n = 11; r² = 0.47, P = 0.02). Thus DHPG evoked the largest increase in ET cells with the slowest basal burst rates. By contrast, the increase in burst frequency was unrelated to the basal number of spikes per burst (r² = 0.005, P > 0.8) or the magnitude of the DHPG-evoked depolarization (r² = 0.002, P > 0.8). It should be noted that the burst frequencies recorded with BAPTA were slower than in previous extracellular recordings studies or via whole cell current-clamp recordings with a standard internal solution (approximately 0.5–11 Hz; Hayar and Ennis 2007; Hayar et al. 2004a; Liu and Shipley 2008b). This suggests that BAPTA reduces ET cell burst rate, perhaps by modulating calcium-dependent intracellular processes.

Fig. 5.

DHPG enhances rhythmic bursting in presence of BAPTA. A1: current-clamp recording with a BAPTA-based recording solution shows that DHPG increases ET cell burst frequency in the presence of dl-2-amino-5-phosphonovaleric acid (APV; 50 μM), 6-cyano-7-nitroquinoxaline-2,3-dione (CNQX; 10 μM), and gabazine (10 μM). A2: faster time-scale records taken from the 3 points indicated in A1. Note increase in burst frequency and decrease in number of spikes per burst with DHPG. A3: in another ET cell, DHPG increases burst frequency and the number of spikes per burst. B1–B4: scatterplots showing the effects of DHPG on burst parameters in 11 ET cells. Filled circles denote means ± SE. *P < 0.05, paired t-tests. Note increase in firing frequency and burst frequency. NumSpike/Burst, number of spikes per burst. C: scatterplot of the relation between the basal (intrinsic) burst frequency and the increase in frequency elicited by DHPG. Note the larger enhancement of burst frequency in cells with slower basal burst rates. D: graph shows that the DHPG-evoked depolarization is significantly reduced in ET cells recorded with a BAPTA-based internal solution. *P = 0.012, unpaired t-test.

I_NaP activation drives the ET cell membrane potential toward the threshold of the depolarizing envelope underlying spike bursts (Liu and Shipley 2008b). DHPG facilitation of I_NaP is consistent with the observed increase in burst rate. However, the modest depolarization by DHPG could also underlie the increased burst rate (Hayar et al. 2004a). To address this, we prevented spontaneous bursting and compensated for the DHPG-evoked depolarization with negative current injection and then applied a slow current ramp (1.5-s duration, 20 pA) to evoke a burst. As shown in Fig. 6, DHPG reduced the burst latency from 0.69 ± 0.09 to 0.46 ± 0.08 s (n = 6; P < 0.02, paired t-test). DHPG also increased the slope between the ramp onset and first burst as well as the slope of the depolarizing envelope.

Fig. 6.

DHPG hastens evoked ET cell bursting. A: a slow current ramp (shown at bottom) was applied to evoke a burst in ET cells recorded with BAPTA-based internal solution in the presence of APV-CNQX-gabazine. DHPG (10 μM) shortens the delay to the 1st burst and increases the slope from ramp onset to the burst depolarizing envelope onset as well as the rising phase of the depolarizing envelope. B1–B3: bar graphs of group data from 6 ET cells showing changes in burst latency (B1), slope to 1st burst (B2), and rising slope of depolarizing envelope. *P = 0.017 (B1), P = 0.046 (B2), P = 0.047 (B3), paired t-tests.

DISCUSSION

The present results reveal that group I mGluR activation enhances ET cell excitability via positive modulation of I_NaP, a key intrinsic current involved in rhythmic burst generation in these cells. The facilitation of I_NaP elicited a modest depolarization and an increase in the rate of rhythmic bursting without effect on other major parameters of ET cell bursting. The increase in burst frequency evoked by DHPG was most pronounced in cells with slower basal burst rates, suggesting that mGluR activation may preferentially enable ET cells with slow intrinsic bursting to pace with faster sniffing frequencies.

ET cell burst initiation and frequency are largely determined by I_h, I_NaP, and low voltage-activated (LVA) calcium currents (Hayar et al. 2004a; Liu and Shipley 2008b). In each burst cycle, postburst hyperpolarizations engage I_h, which drives the membrane potential to activate I_NaP and subsequently LVA currents. The present results demonstrate that a major effect of DHPG is facilitation of I_NaP, consisting of a hyperpolarizing shift in the activation threshold together with an increased current magnitude in the range of −60 to −50 mV. Comparable effects of DHPG and group I mGluR agonists on I_NaP have also been reported in neocortical pyramidal cells and nucleus accumbens neurons (Carlier et al. 2006; D'Ascenzo et al. 2009). The impact of mGluR activation on I_NaP would be predicted to allow I_h to engage a more vigorous I_NaP at more hyperpolarized voltages. As observed here, burst frequency would therefore increase by facilitating I_NaP activation at an earlier stage in the interburst cycle. Since depolarization increases ET cell burst rate (Hayar et al. 2004a), the depolarization elicited by DHPG alone would also be expected to increase burst rate. However, when this depolarization was compensated by negative current injection, the latency of evoked bursts decreased. The rate of rise for evoked bursts and the increased slope of the rising phase of the depolarizing envelope are also consistent with facilitation of I_NaP. As previously reported (Dong et al. 2009) and observed here, DHPG also elicited a BAPTA-sensitive I_CAN that robustly depolarizes ET cells and also increases burst frequency, burst duration, and number of spikes per burst. Taken together with these results (Dong et al. 2009), the present findings suggest that mGluR enhancement of I_CAN and I_NaP cooperatively increase burst rate, whereas I_CAN facilitation also increases burst strength, i.e., number of spikes per burst and burst duration.

Our results indicate that the other key current involved in burst initiation, I_h, is unaffected by DHPG and therefore unlikely to contribute to the observed modulation of burst frequency. I_h modulation would also be expected to alter burst duration and the number of spikes per burst (Liu and Shipley 2008b), which were not altered by DHPG in the presence of BAPTA. We cannot exclude possible effects of DHPG on LVA currents. However, modulation of LVA currents alters the number of spikes per burst as well as burst duration (Liu and Shipley 2008b), effects incompatible with the lack of effect of DHPG on these burst properties. It is conceivable that DHPG may modify other currents, such as the large conductance BK-type calcium-dependent potassium current that shapes ET cell burst duration and spikes per burst (Liu and Shipley 2008b), but these properties were not altered by DHPG. Additionally, the BAPTA concentration used in this study is known to block I_BK (Schwindt et al. 1992; Velumian and Carlen 1999), and the observed effect of DHPG on bursting is not consistent with modulation of this current.

Electrical and chemical synapses made by ET cells play a robust role in coordinating intraglomerular network activity and amplifying output from mitral/tufted cells (De Saint Jan et al. 2009; Gire et al. 2012; Gire and Schoppa 2009; Hayar et al. 2004b, 2005; Najac et al. 2011; Shao et al. 2009, 2012). The intrinsic bursting properties of these cells are thought to reinforce the rhythmic, sniff-driven temporal pattern of sensory input onto the glomerular network. As a population, ET cells spontaneously burst at frequencies encompassing the approximately 1- to 12-Hz theta range of rodent sniffing (for review, see Wachowiak 2011; Welker 1964). Although the spontaneous burst rate in individual ET cells is remarkably stable, bursting can synchronize to olfactory nerve input delivered at rates faster than the intrinsic burst frequency (Hayar et al. 2004a). Olfactory nerve input therefore can override intrinsic burst frequency, perhaps allowing these cells to tune to the prevailing rhythmic sniffing rate. However, burst entrainment to olfactory nerve input begins to fail in the upper theta range (approximately >5 Hz), characteristic of active or investigational sniffing. Context- and state-dependent increases in sniffing (Kay and Laurent 1999; Wachowiak 2011), however, may engage mechanisms to offset burst failures. The activity of brain-stem serotonergic cell groups that innervate the glomerular layer including ET cells increases with arousal (Jacobs and Azmitia 1992; McLean and Shipley 1987). Serotonin engages a nonselective cation current that increases ET cell burst frequency (Liu et al. 2012). Notably, similar to the present results, this occurred preferentially in cells with slower intrinsic burst rates. Dopamine released from glomerular interneurons was recently found to increase ET cell burst frequency by enhancing I_h (Liu et al. 2013). mGluR activation in the olfactory bulb is preferentially engaged by high-frequency olfactory nerve input (De Saint Jan and Westbrook 2007; Ennis et al. 2006; Johnston and Delaney 2010; Yuan and Knöpfel 2006). I_NaP plays a key role in boosting the strength of subthreshold excitatory input to ET cells, particularly at high frequencies (Liu and Shipley 2008a). Thus the mGluR enhancement of I_NaP and burst frequency reported here may operate at the upper theta range of sniff-driven sensory input to offset burst failure. Taken together with previous findings, our results suggest that modulation of ET cell excitability by neurotransmitters intrinsic to the glomerular circuitry as well as extrinsic centrifugal inputs allow fine-tuning of ET cell bursting to state- and context-dependent changes in sniff rates.

GRANTS

This work was supported by National Institute of Deafness and Other Communications Disorders Grant DC-003195.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the author(s).

AUTHOR CONTRIBUTIONS

H.D. and M.E. conception and design of research; H.D. performed experiments; H.D. analyzed data; H.D. and M.E. interpreted results of experiments; H.D. prepared figures; H.D. drafted manuscript; H.D. and M.E. edited and revised manuscript; H.D. and M.E. approved final version of manuscript.