Inhibition of mammillary body neurons by direct activation of Group II metabotropic glutamate receptors

Abstract

The mammillary body is an important neural component of limbic circuitry implicated in learning and memory. Excitatory and inhibitory inputs, primarily mediated by glutamate and gamma-amino butyric acid (GABA), respectively, converge and integrate in this region, before sending information to the thalamus. One potentially overlooked mechanism for inhibition of mammillary body neurons is through direct activation of Group II metabotropic glutamate receptors (mGluRs). Here, whole-cell patch clamp recordings of in vitro slice preparations containing the mammillary body nuclei of the mouse were employed to record responses to bath application of pharmacological agents to isolate the direct effect of activating Group II mGluRs. Application of the Group II mGluR specific agonist, APDC, resulted in a hyperpolarization of the membrane potential in mammillary body neurons, likely resulting from the opening of a potassium conductance. These data suggest that glutamatergic inputs to the mammillary body may be attenuated via Group II mGluRs and implicates a functional role for these receptors in memory-related circuits and broadly throughout the central nervous system.

Introduction

The constituent nuclei of the limbic system underlie many learning, memory and emotive functions in the brain and link forebrain centers for integrating affective information with concurrent sensorimotor processing streams^[1,2]. Among these limbic nuclei, the mammillary body is an essential structure for transmitting information to higher forebrain centers via the mammillothalamocortical pathway and is implicated in the neural circuitry for the recollection of memories^[2–5]. Consequently, understanding the neural processes occurring in the mammillary body nuclei is essential for deciphering its functional role in memory^[2].

In this regard, glutamatergic synaptic inputs to the mammillary body have been considered to have only an excitatory influence through activation of ionotropic glutamate receptors (iGluRs)^[6]. However, the actions of glutamate may also have postsynaptic inhibitory effects through the activation of Group II metabotropic glutamate receptors (mGluRs)^[7–9], as demonstrated in the sensory cortical areas^[10,11]. Indeed, postsynaptic roles for Group II mGluRs have only recently become more appreciated, since their canonical localization is mainly presynaptic, acting as autoreceptors^[9,12]. Recently, we reported the expression of Group II mGluRs in several of the constituent nuclei of the limbic system^[13–15], including most notably, the mammillary body^[16]. Since these receptors have been recently implicated in the direct hyperpolarization of neurons in other brain structures^{[7,8,10,14,17]}, it was reasonable to test whether these receptors had similar functions in mammillary body neurons, using whole-cell patch clamp recordings from in vitro slice preparations containing these nuclei in the mouse.

Whole-cell responses were recorded from mammillary body neurons in the presence of bath application of pharmacological agents to isolate Group II mGluRs^[10,11], consistent with the previously described receptor expression in the mammillary body and other limbic-related structures^[16]. These data demonstrate that Group II mGluR activation directly inhibits mammillary body neurons, suggesting an unappreciated direct mechanism mediated by the neurotransmitter glutamate that hyperpolarizes neuronal membrane potentials in the mammillary body and that may serve as a potential molecular factor linking disparate elements of limbic circuitry.

Material and Methods

The Institutional Animal Care and Use Committee of the Louisiana State University approved all procedures. Acute brain slices for in vitro physiology were prepared from juvenile C57BL/6J mice (P10-P13). Animals were deeply anesthetized via inhalation of isoflurane in an enclosed chamber. The brains were then quickly dissected, removed and submerged in oxygenated, artificial cerebral spinal fluid (ACSF). A clean razor blade was used to block the brain coronally at the rostral tip to set the sectioning angle for preserving the mammillary body nuclei. The rostrally-blocked surface was then mounted on a cutting stage with instant adhesive glue and sectioned with a vibratome at a thickness of 500 µm. Recovered slices were incubated on a meshed platform submerged in ACSF at 32°C, then transferred to a recording chamber on a modified microscope stage, perfused with ACSF. The mammillary body could be recovered in 1–2 slices per animal and was readily identifiable under DIC optics on an Olympus BX-51 upright microscope (Olympus America, Center Valley, PA) (Fig. 2A).

Figure 2.

Mammillary body neuronal responses to bath application of Group II mGluR agonists and antagonists recorded using whole-cell patch clamp in current clamp mode. A. DIC image of the recording site (white star) in the mammillary body. B. Summary statistics of the magnitude of hyperpolarizing voltage changes from the bath application of APDC, a cocktail of antagonists (see Methods), or the Group II mGluR antagonist, LY341495. C-E. Recorded traces in current clamp mode following bath application of APDC alone (C), with APDC and an antagonist cocktail (D) or with APDC and the Group II mGluR antagonist (E).

Whole-cell patch clamp recordings were made using the Multiclamp 700B amplifier, digitized with a Digidata 1440 board, then captured and analyzed using pCLAMP software (Molecular Devices, Sunnyvale, CA). Glass pipettes were pulled on a Flaming/Brown P-97 micropipette puller (Sutter Instrument, Novato, CA) to a tip resistance of 4–6 MΩ, then filled with an intracellular solution containing in mM: 135 K-gluconate, 7 NaCl, 10 HEPES, 1–2 Na₂ATP, 0.3 GTP, and 2 MgCl₂, at a pH of 7.3 and osmolality of 290 mOsm obtained with distilled water. Recordings were uncorrected for liquid junction potentials (~10 mV). Intracellular current injections in current clamp mode were performed prior to recording in order to characterize intrinsic membrane properties. These procedures have been described in detail^[10,11,18].

Pharmacological reagents (TOCRIS, Ellisville, MO) were prepared as stock solutions in ddH₂0 or DMSO and then diluted to their final concentrations prior to bath application^[10,11,18]. Agonists and antagonists for mGluRs were applied at the following concentrations: APDC as an agonist for Group II mGluRs (100 µM), LY341495 as an antagonist for Group II mGluRs (100 nM), LY367385 as an antagonist for mGluR1 (50 µM), MPEP as an antagonist for mGluR5 (30 µM). To block GABA receptors: SR 95531 (20 µM) for GABAA and CGP 46381 (40 µM) for GABAB were used. To block iGluRs: DNQX (50 µM) for AMPA and MK-801 (40 µM) for NMDA were used. The cocktail for isolating Group II mGluR responses was prepared from the antagonists listed above in a low Ca2+ (0.02 mM) / high Mg (6 mM) ACSF solution to reduce synaptic activity and with TTX (1 µM) to block action potential generation. The final bath concentration of pharmacological agents was estimated to be one-fourth of the initial concentration.

To test neuronal input resistance during recordings in current clamp mode, hyperpolarizing current pulses (10 pA, 50 ms, 0.2 Hz) were applied to the neurons. During application of the drug, the membrane potential was stepped briefly to the pre-agonist level, in order to eliminate potential voltage effects on assessment of the membrane resistance. To test membrane conductance changes in voltage clamp mode, slow ramped voltage commands (−60 to −120 mV, 3-s duration) were applied to the neuron and 5 consecutive current traces were averaged for each condition for analysis.

Results

Whole-cell responses from neurons in the mammillary body were recorded in response to bath applied Group II mGluR agonist, APDC (medial mammillary body: n=11; lateral mammillary body: n=6). In each neuron, hyperpolarizing and depolarizing current injections were used to characterize intrinsic membrane properties (Fig. 1B). In all recorded neurons, depolarizing current injections elicited repetitive spikes that adapted over time, i.e. increasing inter-spike intervals during the applied steps (Fig. 1A, E). In addition, pronounced sag currents (I_h) and after-hyperpolarization spikes were observed in response to hyperpolarizing current injections (Fig. 1C, D).

Figure 1.

Intrinsic membrane properties of mammillary body neurons recorded using whole-cell patch clamp in current clamp mode. A. Adapting spiking response to depolarizing current injections. B. Traces of current injections for the intrinsic responses. C. Response of neuron to hyperpolarizing current injections. D. Plot of current-voltage relationships at time points indicated by the blue diamond and red square. E. Plot of inter-spike intervals (ISIs) from varying depolarizing current injections demonstrating adapting spike intervals.

The Group II mGluR agonist, APDC, was then tested for potential effects on the membrane response of recorded mammillary body neurons (Fig. 2A). Bath applied APDC resulted in a pronounced hyperpolarization of the membrane potential (12.8 ± 1.3 mV; n=7) (Fig. 2B,C). This hyperpolarization persisted in the presence of a cocktail composed of antagonists to ionotropic glutamate receptors, ionotropic and metabotropic gamma-amino butyric acid (GABA) receptors, and Group I mGluRs (see Methods) (Fig, 2D). Even in the presence of these antagonists, recorded neurons (n=3) exhibited a pronounced hyperpolarization of the membrane potential (11.3 ± 1.4 mV) that did not differ statistically from that of the agonist alone (p>0.05, t-test) (Fig. 1B). However, in the presence of the Group II mGluR antagonist, LY341495, the hyperpolarizing response was profoundly reduced (1.5 ± 0.3 mV; n=3) (Fig. 1E) and significantly different from either that of the agonist alone (p<0.0001, t-test) or agonist with cocktail (p<0.001, t-test) (Fig. 1B).

Conductance changes that accompanied the APDC-mediated hyperpolarizing response were then characterized. During the application of APDC, hyperpolarizing current injections were applied to test the membrane resistance (Fig. 3). Membrane resistance decreased during the bath application of APDC (67.9 ± 5.5 %; n=3) (Fig. 3Aii) relative to that before the drug was applied (Fig. 3Ai). Following wash out of APDC from the bath, the membrane resistance returned to near control levels (92.5 ± 8.9 %; n=3) (Fig. 3Aiii). These results suggest that the APDC-mediated hyperpolarization results from an increase in membrane conductance, likely due to the opening of ion channels.

Figure 3.

Group II mGluR activation results in an increased hyperpolarizing conductance. A. Response of neuron to bath application of APDC recorded in current clamp mode. Downward deflections during recording are current injections to test membrane resistance at the time points indicated (i, ii, iii). Top trace depicts the time-course of membrane response to APDC application applied at period indicated by solid bar. The membrane potential was manually stepped to the pre-drug level at (ii) to compensate for any potential voltage effects on membrane resistance. Lower traces are expanded from the time points (i, ii, iii). B. Reversal potential assessed by voltage ramps applied during recordings in voltage clamp mode: top trace, calculated I-V curves prior to (blue line) and during (red line) application of APDC, and bottom trace, subtraction I-V plot (green line) from above to determine the APDC-induced reversal potential.

Finally, voltage ramps (−60 to −120 mv; n=4) were applied to assess the reversal potential before and after the application of APDC (Fig. 3B: top panel), and the drug-induced reversal potential was determined to be −82.3 ± 3.2 mV (Fig. 3B: bottom panel), uncorrected for liquid junction potentials of ~10 mV. Furthermore, altering the concentration of K⁺ in the ACSF to 6 mM resulted in a depolarizing shift to the assessed reversal potential (−70.1 ± 3.8 mV). These data suggest that the APDC-mediated hyperpolarization is mainly due to the opening of a potassium conductance.

Discussion

These data demonstrate that neurons in the mammillary body hyperpolarize in response to the bath application of the Group II mGluR agonist, APDC, which was not blocked by antagonists to iGluRs, GABARs or Group I mGluRs, but was blocked by bath application of a Group II mGluR antagonist. This Group II mGluR-induced hyperpolarization resulted in the opening of a hyperpolarizing membrane conductance, with a reversal potential of approximately −82 mV, uncorrected for liquid junction potentials of ~10 mV, suggestive of an underlying potassium conductance mediating this hyperpolarizing response, similar to that observed in other brain regions^{[7,8,10,14,15,17,19–23]}.

These results complement prior findings of highly expressed Group II metabotropic glutamate receptors throughout the brain of the developing mouse^[10,11,16], particularly in nuclei of the limbic system, e.g. the mammillary body, amygdala, and entorhinal cortical areas^[16], as observed by several investigators^{[13–15,24,25]}. Expression of Group II mGluRs is localized on cell bodies in those regions^[14–16,19], consistent with related physiological results suggesting a postsynaptic activation of Group II mGluRs^[14,15]. The expression and activation of Group II mGluRs in the mammillary body, and throughout the nuclei of the limbic system, suggest a possible molecular factor unifying the operations of limbic circuitry, especially with regards to the neural encoding and retrieval of memory^[2–5].

These results also suggest a novel mechanism for direct inhibition of mammillary body neurons via glutamatergic activation of postsynaptic Group II mGluRs. In general, Group II mGluRs are considered to be presynaptically localized throughout the nervous system^[9,12]. Thus, in the mammillary body, the neurotransmitter glutamate could have both excitatory and inhibitory actions at the same synapse, depending on the degree of synaptic activity, i.e. increased activity likely recruits Group II mGluRs, perhaps expanding the dynamic range of the synapse and acting as a potential gain control mechanism^[10].

In addition, Group II mGluRs may have both developmental and age-dependent profiles in some regions of the brain^[25–28]. Thus, these receptors may also serve as a molecular guidance cue for linking disparate elements of related functional circuitry or as a contributing factor to the cognitive changes experienced with age^[27,28]. It remains to be ascertained how such changes to expression and activation of Group II mGluRs in normal and diseased states contribute to their many varied functions.

Conclusion

Group II mGluRs are activated in the mammillary body of the mouse in response to the pharmacological agonist, APDC, resulting in a hyperpolarization of the membrane potential, likely through opening of a downstream potassium conductance. Thus, glutamate can potentially act as both an excitatory and inhibitory neurotransmitter in the mammillary body, which may underlie neural processes associated with memory in this structure.