Disruption of circadian timing increases synaptic inhibition and reduces cholinergic responsiveness in the dentate gyrus

Abstract

We investigated synaptic mechanisms in the hippocampus that could explain how loss of circadian timing leads to impairments in spatial and recognition memory. Experiments were performed in hippocampal slices from Siberian hamsters (Phodopus sungorus) because, unlike mice and rats, their circadian rhythms are easily eliminated without modifications to their genome and without surgical manipulations, thereby leaving neuronal circuits intact. Recordings of excitatory postsynaptic field potentials and population spikes in area CA1 and dentate gyrus granule cells revealed no effect of circadian arrhythmia on basic functions of synaptic circuitry, including long-term potentiation. However, dentate granule cells from circadian-arrhythmic animals maintained a more depolarized resting membrane potential than cells from circadian-intact animals; a significantly greater proportion of these cells depolarized in response to the cholinergic agonist carbachol (10 μM), and did so by increasing their membrane potential three-fold greater than cells from the control (entrained) group. Dentate granule cells from arrhythmic animals also exhibited higher levels of tonic inhibition, as measured by the frequency of spontaneous inhibitory postsynaptic potentials. Carbachol also decreased stimulus-evoked synaptic excitation in dentate granule cells from both intact and arrhythmic animals as expected, but reduced stimulus-evoked synaptic inhibition only in cells from control hamsters. These findings show that loss of circadian timing is accompanied by greater tonic inhibition, and increased synaptic inhibition in response to muscarinic receptor activation in dentate granule cells. Increased inhibition would likely attenuate excitation in dentate-CA3 microcircuits, which in turn might explain the spatial memory deficits previously observed in circadian-arrhythmic hamsters.

Keywords: Siberian hamster; acetylcholine; carbachol; dentate; hippocampus; sex differences.

FIGURE 1

Representative examples of locomotor activity patterns that were used to designate hamsters as entrained (ENT, a1) or circadian‐arrhythmic (ARR, a2). Vertical black hash marks indicate daily locomotor movement in the home cage and are double‐plotted on a 48‐hr time scale with data from consecutive days plotted from top to bottom. Most cage activity occurs at night for ENT hamsters as indicated by the black and white rectangles that indicate night (8 hr) and day (16 hr), respectively. By contrast, locomotor activity in ARR hamsters is distributed equally across 24 hr, indicating the loss of circadian timing. These observations were confirmed by chi‐square time series analysis showing a robust daily rhythm with a peak at 24 hr with harmonics at 12 and 36 hr in ENT animals (b1), and a lack of periodicity in the ARR animal (b2)

FIGURE 2

Loss of circadian timing did not affect basic excitatory circuit properties in the hippocampus. Recordings from the perforant path of the dentate gyrus of input/output curves (I/O, a1), paired pulse ratio (PPR, a2) and long‐term potentiation (LTP, a3), with the same measures in the Schaffer collateral CA1 path of hippocampus (b1‐3; n = 24 ENT, n = 23 ARR animals). There were no significant differences in these three measures between ENT (red) and ARR (green) either in dentate or CA1 (2‐way ANOVA with repeated measures for stimulus, ISI, and time, p > .05). Inserts show representative traces at different stimulus intensities. Note that PPR and LTP responses are evaluated by their slopes, not amplitudes. Dotted lines in LTP panels indicate baseline values extrapolated past tetanus (arrow). All scale bars are 0.5 mV, 10 ms. Data expressed as means ± SE; for clarity, PPR only shown up to 250 ms, LTP error bars shown every 600 s

FIGURE 3

Effects of muscarinic receptor activation by carbachol in the dentate gyrus. Data are shown prior to (control) and during (carbachol) bath application of 10 mM carbachol. Stimulus‐evoked fEPSPs (ENT and ARR combined, n = 22 animals) were suppressed by carbachol (a1) while simultaneously‐recorded population spikes were unaffected (a2), however, there were no significant differences between ENT (n = 10) and ARR (n = 12) groups in these measures (b1, b2; p > .05). The high variance in the ENT control group (b1) was largely due to one cell, however, it was not a significant outlier (Grubbs' test, p > .05). Paired pules ratio (PPR) of EPSPs was not affected by carbachol and did not differ between cells from ENT (n = 9) and ARR (n = 11) groups (c1; p > .05). By contrast, PPR of the population spike was significantly increased by carbachol in both ENT and ARR groups, although the effect was significantly reduced in cells from ARR animals (c2). ** Indicates stimulus (0.2–0.7 mA in panel c1), ISIs (40–100 ms in panel c2) that differed significantly between ENT and ARR groups under carbachol treatment (p < .01)

FIGURE 4

Synaptic inhibition induced by muscarinic receptor activation by carbachol was attenuated in dentate cells from circadian‐arrhythmic animals. Carbachol application decreased stimulus evoked inhibitory transmission (a1; n = 18 cells, p < .001), but increased spontaneous tonic inhibition (b1; p = .020). IPSPs are depolarizing because the intracellular electrode solution contained a high concentration of chloride (see methods). Representative traces of inhibitory responses are shown in panels (a1) and (b1). Box plots show median values and extend from the 25th to 75th percentiles; whiskers extend to minimum and maximum data points with individual values shown as circles. Stimulus‐evoked IPSPs were significantly suppressed only in cells from ENT animals (a2; n = 8 ENT, n = 7 ARR); amplitudes are shown for individual cells from ENT (a3) and ARR (a4) animals. Frequency distributions of spontaneous IPSP intervals (b2) and amplitudes (b3) during the first 150 s of recording. Statistical comparisons were made with KS tests from cumulative frequency distributions of the data in panels (b2) and (b3). Carbachol increased the number of spontaneous IPSP events in both ENT (p = .001) and ARR (p = .003) groups. Cells from ARR animals had significantly more IPSP events before (control; p = .014) and during (carbachol; p = .003) drug application compared to ENT (b2). There were no significant differences (p > .05) in amplitudes (b3) of spontaneous ISPS events. *p < .05, **p < .001, n.s. (nonsignificant difference)

FIGURE 5

Membrane potentials of dentate granule cells were more depolarized by carbachol in arrhythmic animals. For data points in panel (a2) potentials were averaged across cells within individual animals so that each point is the mean value for a single animal; (a1, b), and (c1–c2) all represent individual cells. For panel (c3), only cells that depolarized under carbachol were analyzed. Carbachol (10 mM), caused significant but modest depolarization of the resting membrane potential (A1; n = 13). However, when data are separated by rhythm status, significant average depolarization only occurred in cells from ARR animals (a2; ENT, n = 6; ARR, n = 7). The resting potential of most ENT cells were not reactive to carbachol (b), while most cells in slices from ARR animals were reactive (b; Fisher's exact test, p = .008). Membrane potentials of all individual neurons (c1, ENT, n = 15; c2, ARR, n = 13). Membrane potentials only of individual neurons that reacted (depolarized) in response to carbachol (c3). Box plot parameters as in Figure 4. *p < .05, **p < .01

FIGURE 6

Sex differences in carbachol effects on dentate granule cells. Data from ARR and ENT animals were combined to show that overall, cells from both males and females depolarized under carbachol application (a1). When these data are parsed out by rhythm status (a2), no carbachol‐induced depolarization was seen in either female or male ENT animals. Significant depolarization by carbachol was detected, however, in cells from both male and female ARR animals (a2). The resting membrane potential was greater in cells from male ARR compared to those from male ENT animals (a2). We found no significant sex differences in the proportion of cells that were reactive to carbachol (p > .05; b). Frequency distributions of spontaneous IPSP intervals (c1) and amplitudes (c2) during the first 150 s of recording. Statistical comparisons were made with KS tests from cumulative frequency distributions of the data in panels (c1) and (c2). There were no sex differences in control conditions (p > .05). Carbachol increased the number of spontaneous IPSP events in females (p = .021), but not in males (p > .05). There were no significant differences (p > .05) in amplitudes (c2) of spontaneous ISPS events. Box plot parameters as in Figure 4. *p < .05