Metaplasticity of hypothalamic synapses following in vivo challenge

Abstract

Neural networks that regulate an organism's internal environment must sense perturbations, respond appropriately, and then reset. These adaptations should be reflected as changes in the efficacy of the synapses that drive the final output of these homeostatic networks. Here we show that hemorrhage, an in vivo challenge to fluid homeostasis, induces LTD at glutamate synapses onto hypothalamic magnocellular neurosecretory cells (MNCs). LTD requires the activation of postsynaptic alpha2-adrenoceptors and the production of endocannabinoids that act in a retrograde fashion to inhibit glutamate release. In addition, both hemorrhage and noradrenaline downregulate presynaptic group III mGluRs. This loss of mGluR function allows high-frequency activity to potentiate these synapses from their depressed state. These findings demonstrate that noradrenaline controls a form of metaplasticity that may underlie the resetting of homeostatic networks following a successful response to an acute physiological challenge.

Figure 1. Hemorrhage Lowers Pr at Glutamate Synapses

(A) eEPSCs in slices obtained from hemorrhaged rats (blue) displayed an increased PPR when compared with controls (black). Current traces are averages of 30 consecutive responses from each condition. For paired-pulse recordings, synaptic stimuli were delivered 50 ms apart. Scale bars: 10 ms, 20 pA. (B) Summary graph illustrating that hemorrhaged rats (blue; n = 15) had a reduced Pr compared to nonhemorrhaged naive rats (black; n = 57), indicated by significant increases in PPR, failure rate, and CV. *p < 0.05, **p < 0.01. (C–E) Example of an injection site from animals microinjected with prazosin, yohimbine, or vehicle into the PVN prior to hemorrhage. The image shown is from case 5 of the yohimbine-injected group (see Figure S1). (C) Low-power view (4X) of a transmitted light image of a coronal slice of the PVN. (D) The injection site as indicated by the red fluorescence of the Fluospheres. (E) Superimposed transmitted light and epifluorescence images. 3V, third ventricle. (F–H) Bar chart summaries of PPR, failure rate, and CV of eEPSCs obtained from control naive slices (control; n = 57) and slices from animals with yohimbine (10 nmol; n = 6), prazosin (10 nmol; n = 8), or vehicle (n = 4) microinjected into the PVN prior to hemorrhage.

Figure 2. Noradrenaline Induces Presynaptic LTD

(A) Representative single eEPSC recording illustrating LTD of evoked glutamate release induced by NA (LTD_NA). Traces show average paired-pulse eEPSC responses in control (black) and 20 min after wash of NA (red) at the times indicated. Scale bars: 10 ms, 20 pA. (B) Data normalized and averaged from multiple neurons shows time course and extent of LTD_NA (n = 17). (C) LTD_NA was accompanied by an increase in PPR, failure rate, and CV. *p < 0.05; n = 7.

Figure 3. LTD_NA Requires Postsynaptic α2-Adre-noceptor Activation and Endocannabinoids

(A) Average normalized eEPSC amplitude of experiments where NA was applied in the presence of the α2-adrenoceptor antagonist yohimbine (20 μM; n = 7). Inset scale bars: 10 ms, 50 pA. (B) Average eEPSC amplitude of experiments where clonidine (50 μM; n = 6) was applied alone (open circles) and where application of clonidine was chased with yohimbine (n = 7; closed circles) at the time indicated. Inset scale bars: 10 ms, 20 pA. (C) Effects of clonidine on eEPSCs recorded with an intracellular solution containing GDP-βs (1 mM; n = 8). Inset scale bars: 10 ms, 50 pA. (D) Average normalized eEPSC amplitude of experiments where the CB1 receptor antagonist AM251 (1 μM) was applied at least 5 min prior to NA (n = 8). Inset scale bars: 10 ms, 200 pA. (E) Summary of NA experiments where 10 mM BAPTA was included in the patch pipette. BAPTA was allowed to perfuse into the neuron for at least 15 min before NA was applied (n = 12). Inset scale bars: 10 ms, 100 pA. (F) Summary data showing the percent change in eEPSC amplitude after 20 min wash of NA (100 μM; n = 17), phenylephrine (PE; 50 μM; n = 5), NA in prazosin (praz + NA; 10 μM; n = 5), NA in yohimbine (yoh + NA; 20 μM; n = 7), NA chased with yohimbine (NA + yoh chaser; n = 5), clonidine (clon; 50 μM; n = 6), clonidine followed by a yohimbine chaser (clon + yoh chaser; n = 7), clonidine in cells patched with GDP-βs included in the intracellular solution (GDP-βs + clon; n = 8), NA in the presence of AM251 (AM251 + NA), and NA in cells patched with an intracellular solution containing BAPTA (BAPTA + NA). *p < 0.05; **p < 0.01.

Figure 4. Hemorrhage and In Vitro NA Inactivates Presynaptic Group III mGluRs

(A) Representative experiment showing that the group III mGluR agonist (L-AP4, 25 μM) depresses eEPSCs obtained from MNCs in the PVN from a naive rat. Inset shows average eEPSCs before (black) and after (green) L-AP4. Scale bars: 10 ms, 50 pA. (B) L-AP4 depressed eEPSCs presynaptically as indicated by increases in PPR, failure rate, and CV (n = 9). (C) Representative experiment illustrating that L-AP4 failed to depress eEPSCs in rats hemorrhaged in vivo 30 min prior to slicing. Inset shows sample eEPSCs from a hemorrhaged rat in control (blue) and after L-AP4 (green). Scale bars: 5 ms, 50 pA. (D) In slices obtained from rats that were microinjected with yohimbine into the PVN in vivo prior to hemorrhaging, L-AP4 was not effective in depressing evoked glutamate release. Inset shows eEPSCs in control (hem + yoh; microinjected with yohimbine and then hemorrhaged) and after application of L-AP4 (green). Scale bars: 20 ms, 50 pA. (E) After induction of LTD_NA subsequent application of L-AP4 failed to depress glutamate synapses further. Inset shows sample eEPSCs after LTD_NA induction (red) and subsequent application of L-AP4 (green). Scale bars: 10 ms, 20 pA. (F) Representative recording showing the effect of L-AP4 application after clonidine. Group III mGluRs are not inactivated following induction of LTD with clonidine. Inset shows average eEPSCs after wash of clonidine (orange) and following subsequent application of L-AP4 (green). Scale bars: 20 ms, 20 pA. (G and H) Summary bar graphs illustrating the effect of L-AP4 on the amplitude of evoked glutamate release (G) and PPR (H) in slices obtained from control naive rats (con; black), after induction of LTD_NA (NA; red; n = 5), after clonidine LTD (clon; orange; n = 8), after in vivo hemorrhage (hem; blue; n = 5), and after micro-injection of yohimbine into the PVN followed by hemorrhage (hem + yoh; n = 5). *p < 0.05.

Figure 5. Hemorrhage Unmasks Activity-Dependent Plasticity

(A) Sample experiment showing HFS-induced (3 × 100 Hz for 1 s, 10 s interval) LTP in a rat hemorrhaged in vivo. Horizontal black lines indicate the average eEPSC amplitude before and after HFS. Inset shows eEPSCs from a hemorrhaged rat before (blue) and after (purple) HFS. Scale bars: 10 ms, 20 pA. (B) Summary of experiments illustrating the response to HFS in hemorrhaged rats (n = 15). (C) LTP in hemorrhaged rats was accompanied by a reduced PPR, failure rate, and CV. *p < 0.05; **p < 0.01. (D) Summary of experiments showing that HFS applied to naive slices failed to increase synaptic strength (n = 11). Inset shows sample eEPSCs in control and after HFS. Scale bars: 20 ms, 50 pA. (E) Normalized group data showing the effects of HFS on eEPSCs in slices from animals microinjected with prazosin in vivo prior to hemorrhage (n = 5). Inset shows sample eEPSCs before (hem + praz) and after HFS. Scale bars: 20 ms, 20 pA. (F) Bar graph summarizing the group data of the changes in eEPSC amplitude and PPR after HFS in naive slices, following hemorrhage, and in animals with prazosin microinjected into the PVN prior to hemorrhage. *p < 0.05; **p < 0.01.

Figure 6. Noradrenaline-Mediated LTD Unmasks Metaplasticity

(A) Representative experiment showing HFS-induced LTP following LTD_NA induction. Sample traces are paired-pulse eEPSCs in control (black), during LTD_NA (red), and after HFS (purple). Scale bars: 10 ms, 50 pA. (B) Summary of experiments showing the effect of HFS on eEPSC amplitude after induction of LTD_NA (closed black). The average time course of LTD_NA in the absence of HFS is replotted from Figure 2B for comparison (open gray). (C) HFS restored PPR, failure rate, and CV to pre-LTD_NA values. **p < 0.01.

Figure 7. Functional Reactivation of Group III mGluR Signaling after HFS

(A) Average time course showing that mGluR function recovers following post-LTD_NA HFS-induced LTP (n = 6). Inset shows sample eEPSCs in control, after LTD_NA induction (red), after HFS (purple), and after subsequent application of L-AP4 (green). Scale bars: 5 ms, 20 pA. (B and C) Summary graphs illustrating the percent change in eEPSC amplitude and PPR in response to L-AP4 when it’s perfused on naive slices (green), on slices after the induction of LTD_NA (red; replotted from Figure 4), or on slices after LTD_NA induction with subsequent HFS (purple). Each bar represents an individual experiment. *p < 0.05.