Vanderbilt Institute of Chemical Biology



Discovery at the VICB







Modulation of Memory Acquired Under Stress


By: Carol A. Rouzer, VICB Communications
Published:  July 18, 2017



Co-activation of the β-adrenergic receptor and metabotropic gluatamate receptor-3 on astrocytes suppresses formation of memories under stressful conditions.


Synaptic plasticity is a property of neurons that enables them to increase or decrease the strength of signal transmission in response to changes in synaptic activity. This property plays an important role in cognition, memory, and learning. Key players in the modulation of synaptic plasticity – particularly in the development of long-term memories associated with novel or stressful stimuli – are neurons located in the brain region known as the locus coeruleus.  These neurons project their axons to the hippocampus and cortex where they use noradrenaline (NA) to activate β-adrenergic receptors (β-ARs)  located on both neuronal and glial cells. In the case of CA1 pyramidal neurons (Figure 1) in the hippocampus, this NA-mediated neurotransmission facilitates the development of long-term potentiation (LTP), a form of synaptic plasticity characterized by a sustained increase in synaptic transmission following stimulation. However, evidence indicates that NA-mediated activation of β-ARs in hippocampal astrocytes also modulates LTP in CA1 cells, and that glutamate, acting at group II metabotropic glutamate receptors (mGluRs) plays a role. Now, Vanderbilt Institute of Chemical Biology members Jeff Conn and Craig Lindsley, along with their colleagues, use novel mGluR-selective allosteric modulators and knockout mice to explore the regulation of LTP in CA1 pyramidal cells [A. G. Walker, et al., (2017) Neuropsychopharmacology, published online June 30, 2017, DOI:10.1038/nnp.2017.136].



FIGURE 1. Outline of a typical CA3 and CA1 neuron, revealing the pyramidal-shaped cell body and extensive dendritic trees. Image reproduced by permission from Macmillan Publishers Ltd, from N. Spruston, (2008) Nat. Rev. Neurosci., 9, 206. Copyright 2008.


From prior research, the investigators knew that simultaneous activation of β-ARs and group II mGluRs in astrocytes results in a much larger increase in cAMP than is observed when β-ARs are activated alone. However, this earlier work did not define which of the two group II mGluRs, mGlu2 or mGlu3, was responsible for this phenomenon. To answer this question, the researchers began by monitoring the production of cAMP in rat hippocampal and mouse cortical brain slices in response to the β-AR agonist isoproterenol (ISO) in the presence or absence of LY379268 (LY), a nonspecific agonist of mGlu2 and mGlu3. They confirmed that addition of LY along with ISO to the slices resulted in a greater production of cAMP than was observed with ISO alone. By itself, LY had no effect on cAMP production. To differentiate between the two mGluRs, the researchers next used a negative allosteric modulator (NAM) of mGlu3. NAMs act by inhibiting the effects of an agonist, in this case LY, and the results indicated that, indeed, addition of the mGlu3 NAM blocked LY's ability to increase the cAMP response to ISO. To investigate a potential role for mGlu2, the investigators employed a positive allosteric modulator (PAM) of that receptor. PAMs act by enhancing the effects of an agonist. In this case, however, the mGlu2 PAM had no effect on LY's ability to increase the cAMP response to ISO. These results suggested that mGlu3, but not mGlu2 was responsible for the effects of LY on ISO-stimulated cAMP production. Further support for this conclusion came from experiments using mGlu2 and mGlu3 knockout mice. In these experiments cortical slices from wild-type littermates and mGlu2 knockout mice all demonstrated LY-dependent increases in cAMP produced in response to ISO. However, cortical slices from mGlu3 knockout mice were unresponsive to LY in this setting. These observations confirmed a role for mGlu3 in modulating cAMP production in response to β-AR activation in hippocampal or cortical brain slices.


After having defined the group II mGluR responsible for cAMP modulation, the researchers next explored the functional relevance of this phenomenon. For this work, they focused on hippocampal CA1 pyramidal neurons, which are known to play a role in long-term memory. These neurons receive input from CA3 pyramidal neurons that are located in a different region of the hippocampus. Axonal projections, known as Schaffer collaterals (SCs) extend from CA3 to CA1 neurons, where they terminate in excitatory synapses that release glutamate. In this case, the glutamate acts on ionotropic receptors on the post-synaptic membranes of CA1 cell dendrites. Using rat hippocampal slices, the researchers delivered different levels of stimulation via an electrode placed in the region of the SCs and then assessed the development of LTP by monitoring the slope of the field excitatory post-synaptic potential (fEPSP) that could be detected by a recording electrode placed in the region of the CA1 neurons. They found that two distinct patterns of initial stimulation (Figure 2) resulted in either weak or strong LTP in response to subsequent stimuli. They also found that addition of ISO increased the modest LTP developed following the weak initial stimulus but had no effect on the robust LTP developed after exposure to the strong initial stimulus. Thus, β-AR activation augmented a suboptimal LTP response. Addition of LY with ISO blocked this augmentation of the weaker LTP, and inclusion of the mGlu3 NAM reversed the effects of LY. These findings suggested that activation of mGlu3 along with the β-AR served to block the effects of β-AR-mediated enhancement of LTP in the hippocampus.




FIGURE 2. Patterns of electrical stimulation used to induce LTP in hippocampal CA1 neurons. A weak initial stimulus (weak TBS) consisted of a single series of four 100 hz bursts spaced 230 ms apart. This produced a suboptimal LTP. The strong initial stimulus (strong TBS) comprised four series of 100 hz bursts spaced 100 ms apart. This produced a maximal LTP. Image reproduced by permission from Macmillan Publishers Ltd, from A. G. Walker, et al., (2017) Neuropsychopharmacology, published online June 30, 2017, DOI:10.1038/nnp.2017.136]. Copyright 2017.



Prior studies had suggested that group II mGluR-mediated effects on β-AR-dependent signaling occurred in astrocytes rather than neurons. These studies had shown that increased cAMP formation resulting from simultaneous mGluR and β-AR activation led to high levels of adenosine that subsequently activated A1 adenosine receptors on neighboring neurons. To further explore this mechanism, the investigators showed that LY had no effect on ISO-mediated LTP enhancement in hippocampal slices if the weak initial stimulation was administered in the presence of an A1 receptor antagonist. Similarly, a metabolic toxin that blocks astrocyte function abolished the effects of LY in this model system. These findings demonstrated that both astrocytes and the A1 receptor play a role in modulation of β-AR-mediated LTP enhancement by group II mGluR activation. Further support for a role of astrocytes in this phenomenon came from fluorescence microscopy experiments that demonstrated the co-expression of β-AR and mGlu3 in hippocampal astrocytes (Figure 3).




FIGURE 3. Fluorescence microscopy reveals co-expression of β-AR and mGlu3 in astrocytes. Both antibody (IHC)- and RNA (RNAscope)-based probes were used. Fluorescent labels were green (β-AR), blue (mGlu3) or red (GFAP, an astrocyte marker). Arrowheads indicate locations where both receptors were observed in a single astrocyte. Image reproduced by permission from Macmillan Publishers Ltd, from A. G. Walker, et al., (2017) Neuropsychopharmacology, published online June 30, 2017, DOI:10.1038/nnp.2017.136]. Copyright 2017.



To determine if mGlu3-dependent modulation of LTP is important in vivo, the researchers employed a rat model of contextual fear conditioning. This model was chosen because, as noted above, noradrenergic signaling is believed to be important in modulating synaptic plasticity under conditions of fear and stress. In these experiments, the investigators placed rats in specialized cages that delivered a shock to their feet under highly controlled conditions. This first step was the training phase. Following 24 h, in the reactivation phase, a second exposure to the cage (without foot shock) reinforced the fear. Then 24 h later, the rats were placed in the cage once again without foot shock, and the researchers assessed their level of fear by the percentage of time spent motionless (freezing) (Figure 4). Administration of drugs at the time of reactivation enabled the researchers to evaluate the effects of β-AR and/or mGlu3 modulation on the conditioned fear response. The results demonstrated that administration of either propranolol (a β-AR antagonist) or LY resulted in a diminished fear response in the rats. Co-administration of the mGlu3 NAM with LY resulted in blockade of LY's effects. An agonist of the A1 receptor also reduced the fear response, while an A1 antagonist blocked LY's effects. These findings support the hypothesis that β-AR-dependent signaling enhances LTP in vivo under the context of a fear-inducing stimulus. This enhancement can be blocked by mGlu3 stimulation through a process that requires A1 receptor signaling.




FIGURE 4. Protocol for in vivo rat model of contextual fear conditioning. In the training phase, rats are exposed to a combination of foot shocks and sound in a plexiglass box. On the next day, they are re-exposed to the box in the presence or absence of drug treatments (reactivation phase). Finally, on the third day, they are exposed to the box, and the percent of the time spent motionless (freezing) is measured to assess the level of fear. Image reproduced by permission from Macmillan Publishers Ltd, from A. G. Walker, et al., (2017) Neuropsychopharmacology, published online June 30, 2017, DOI:10.1038/nnp.2017.136]. Copyright 2017.



The researchers proposed a model to explain their observations (Figure 5). According to the model, noradrenergic neurons project terminals to both CA1 pyramidal cells and astrocytes. In the CA1 cells, β-AR receptor activation by these terminals enhances LTP induced by glutamate-dependent excitatory signaling from SCs. In the astrocytes, β-AR activation leads to increases in cAMP. Simultaneous activation of mGlu3 by glutamate released by the SCs augments cAMP production in the astrocytes, leading to adenosine formation and A1 receptor activation of the SC. The outcome of A1 receptor activation is dampening of the signals coming from the SC and a subsequent reduction in LTP in the CA1 cell.



FIGURE 5. Model for mGlu3-mediated modulation of LTP. Noradrenergic nerve terminals (purple) excite β-ARs on CA1 pyramidal cells (blue) to enhance LTP resulting from excitatory glutamate coming from the Shaffer collateral (orange). Stimulation of β-AR signaling on the astrocyte (green) results in activation of adenylate cyclase (AC) and production of cAMP. Glutamate from the Schaffer collateral also stimulates mGlu3 on astrocytes leading to augmented cAMP production. Hydrolysis of cAMP produces adenosne, which is released and activates A1 receptors on the Schaffer collateral resulting in reduced signal transmission and suppression of LTP in the CA1 cell. Image reproduced by permission from Macmillan Publishers Ltd, from A. G. Walker, et al., (2017) Neuropsychopharmacology, published online June 30, 2017, DOI:10.1038/nnp.2017.136]. Copyright 2017.



These findings have important clinical implications. For example, post-traumatic stress disorder results from aberrant processing of fear-related stimuli in the hippocampus. Modulation of LTP by mGlu3 activation may, therefore, be a novel approach to treatment or prevention of this disorder. Similarly mGlu3 activation may help protect neurons from damage due to excessive excitatory stimulation in times of stress. Further work to develop direct agonists or PAMs of mGlu3 will facilitate exploration of these possibilities.



View Neuropsychopharmacology article: Co-Activation of Metabotropic Glutamate Receptor 3 and Beta-Adrenergic Receptors Modulates Cyclic-AMP, Long-Term Potentiation, and Disrupts Memory Reconsolidation







The Vanderbilt Institute of Chemical Biology, 896 Preston Building, Nashville, TN 37232-6304, phone 866.303 VICB (8422), fax 615 936 3884
Vanderbilt University is committed to principles of equal opportunity and affirmative action. Copyright © 2014 by Vanderbilt University Medical Center