Restraint Stress Intensifies Interstitial K(+) Accumulation during Severe Hypoxia

Abstract

Chronic stress affects neuronal networks by inducing dendritic retraction, modifying neuronal excitability and plasticity, and modulating glial cells. To elucidate the functional consequences of chronic stress for the hippocampal network, we submitted adult rats to daily restraint stress for 3 weeks (6 h/day). In acute hippocampal tissue slices of stressed rats, basal synaptic function and short-term plasticity at Schaffer collateral/CA1 neuron synapses were unchanged while long-term potentiation was markedly impaired. The spatiotemporal propagation pattern of hypoxia-induced spreading depression episodes was indistinguishable among control and stress slices. However, the duration of the extracellular direct current potential shift was shortened after stress. Moreover, K(+) fluxes early during hypoxia were more intense, and the postsynaptic recoveries of interstitial K(+) levels and synaptic function were slower. Morphometric analysis of immunohistochemically stained sections suggested hippocampal shrinkage in stressed rats, and the number of cells that are immunoreactive for glial fibrillary acidic protein was increased in the CA1 subfield indicating activation of astrocytes. Western blots showed a marked downregulation of the inwardly rectifying K(+) channel Kir4.1 in stressed rats. Yet, resting membrane potentials, input resistance, and K(+)-induced inward currents in CA1 astrocytes were indistinguishable from controls. These data indicate an intensified interstitial K(+) accumulation during hypoxia in the hippocampus of chronically stressed rats which seems to arise from a reduced interstitial volume fraction rather than impaired glial K(+) buffering. One may speculate that chronic stress aggravates hypoxia-induced pathophysiological processes in the hippocampal network and that this has implications for the ischemic brain.

Keywords: anoxia; hippocampus; spatial buffering; spreading depression; synaptic function.

Figure 1

Chronic restraint stress leaves basal synaptic function intact but impairs LTP. (A) Control and stressed rats were kept under an inverse light–dark cycle. For a total duration of 21 days they were exposed to restraint stress during their activity periods. (B) As indicated by the input–output curves, basal synaptic transmission at the Schaffer collateral/CA1 pyramidal neuron synapse was not affected by chronic stress. The averages of 16 control and 15 stress slices are plotted, error bars represent standard deviations. (C) PPF, recorded in the same set of slices, was not affected by chronic stress. (D) Sample traces of fEPSPs recorded in control and stress slices. (E) In stress slices, stable LTP could not be induced. The LTP-inducing stimulus was delivered at time 0 (arrow mark) and evoked an indistinguishable PTP in both groups. In stress slices, however, the amplitude of the fEPSPs then declined to pre-stimulus baseline conditions in the course of the next 60 min (n = 5 control, n = 7 stress slices). For clarity, error bars were omitted for every second data point. The bar plot summarizes the averaged fEPSP amplitudes right after the 100 Hz stimulus (PTP) as well as in the time window of 50–60 min later (LTP). The number of slices analyzed is indicated at the bottom of the bars; asterisks indicate statistically significant changes (**P < 0.01).

Figure 2

Hypoxia-induced spreading depression episodes are not markedly affected in stressed rats. (A) Recordings of the HSD-associated negative deflection of the extracellular DC potential and the closely correlated changes in [K⁺]_o as monitored in st. radiatum of the CA1 subfield. Each slice underwent three successive hypoxic episodes. O₂ was resubmitted 25 s after the onset of the rapid DC potential deflection. Hypoxic episodes were separated by at least 20 min to ensure complete recovery. Note that stress slices showed a more rapid and intense increase in [K⁺]_o during early hypoxia before HSD onset and a slower posthypoxic recovery of [K⁺]_o. (B) Definition of the analyzed characteristic parameters of the DC potential deflection and the associated changes in [K⁺]_o (ΔV amplitude of the DC shift, Δt time to onset since O₂ withdrawal, t_1/2 duration measured at the half amplitude level).

Figure 3

Statistical summary of the electrophysiological signs of HSD during three consecutive episodes of hypoxia. (A) Summary of the characteristic DC potential parameters in control slices, stress slices and control slices treated with FAc (5 mM, >1.5 h). The DC potential parameters time to onset and duration refer to the left hand time axis, the amplitude refers to the right hand voltage axis. (B) Summary of [K⁺]_o at the distinct time points defined in Figure 2B. The lower two diagrams show the slopes of the rise in [K⁺]_o early during hypoxia before HSD onset and of the posthypoxic recovery, respectively. Asterisks indicate significantly different changes as compared to the respective HSD episode in control slices (*P < 0.05, **P < 0.01, ***P < 0.001).

Figure 4

Intrinsic optical signal analyses define the spatiotemporal pattern of HSD propagation. (A) Subtraction images of the changes in tissue reflectance visualizing the propagation of HSD within the hippocampal subfield. Numbers report the time since occurrence of the first optical changes (t = 0) and the last image shows the maximum spatial spread of HSD. The reflectance changes are coded in a gray-scale covering a range of ±20% as referred to prehypoxic baseline. (B) Time course of the reflectance changes in control and stress slices. Neither the kinetics nor the intensity of the scattering increase reveal marked changes among the two groups. Plotted are the averages of 16 control and 15 stress slices for the first, second, and third hypoxic episode; for clarity, error bars were omitted. (C) Quantification of the reflectance increase shows only a trend of a somewhat reduced intensity at the height of HSD for stress slices, which reaches the level of significance only for the third hypoxic episode. (D) During posthypoxic recovery, control slices showed a moderate secondary reflectance increase [see also arrow marks in (B)] which was absent in stress slices. For clarity, error bars are shown for every 5th data point only; the range of significant changes is marked. (E) The propagation velocity of the HSD wave front does not differ in control and stress slices during the three hypoxic episodes. (F) Normalized tissue invasion reports the maximum spread of HSD within the hippocampal formation. Differences were not observed among control and stress slices.

Figure 5

Hypoxia-induced failure of synaptic function and posthypoxic recovery. Synaptic function is lost within 1–2 min upon O₂ withdrawal (filled triangle mark) in control and stress slices. Upon reoxygenation (open triangle mark), recovery of synaptic function is slower in stress slices. The time points at which the fEPSPs regained 50% of their normoxic amplitudes are marked by the black arrows. Furthermore, the final degree of recovery tended to be less complete in stress (n = 28) than in control slices (n = 24; see gray arrows). Since the semi-automated analysis of fEPSPs amplitudes detected noise peaks within the traces, zero-amplitudes were not quite reached during hypoxia. Nevertheless, in the raw traces, detectable fEPSPs were absent upon stimulation.

Figure 6

Quantification of GFAP and Kir4.1 expression. (A) Immunolabeling of the astrocytic marker GFAP revealed an increased GFAP immunoreactivity in the hippocampal CA1 subfield in sections from stressed rats. Relative optical density of the sections was determined in st. oriens and st. radiatum (bar plots on the right). In both layers, GFAP immunoreactivity was more dense in stressed as compared to control rats. The number of sections analyzed is reported (so st. oriens, sp st. pyramidale, sr st. radiatum). (B) Immunolabeling also revealed a downregulation of Kir4.1 in stressed rats that was obvious in all layers of the CA1 subfield. (C) Western blots confirmed the decreased Kir4.1 immunoreactivity, yielding a decreased expression of Kir4.1 as compared to β-actin content in stressed rats (n = 8 hippocampi each group).

Figure 7

Electrophysiological analyses of hippocampal glial cells. (A) St. radiatum astrocytes were identified by labeling with the astrocyte-specific fluorescent marker sulforhodamine101. (B) The resting membrane potential (V_m) and input resistance (R_i) of astrocytes did not differ among control and stress slices. (C) Whole-cell current responses of astrocytes to voltage steps from −80 mV holding potential to various test potentials ranging from −160 to +50 mV in 10 mV increments (see insert). (D) Summarized IV-curves of astrocytes from control and stress slices.

Figure 8

Quantification of K⁺ fluxes in astrocytes. (A) Voltage-clamp recordings of st. radiatum astrocytes (V_hold = −80 mV) that were repetitively subjected to defined voltage step protocols (see Figure 7C). Application of 50 mM K⁺ elicited an inward current (I_K, see arrows) of similar amplitude in astrocytes from control and stress slices. (B) Statistical comparison of the 50 mM K⁺-mediated inward current (I_K) and potentiation of that current by hyperpolarizing voltage steps from −80 to −160 mV [ΔI₋₁₆₀; see arrow marks in (A)].