Apnea produces excitotoxic hippocampal synapses and neuronal apoptosis

Experimental Neurology 238 (2012) 107–113

Contents lists available at SciVerse ScienceDirect

Experimental Neurology journal homepage: www.elsevier.com/locate/yexnr

Apnea produces excitotoxic hippocampal synapses and neuronal apoptosis Simon J. Fung a, b,⁎, MingChu Xi a, b, JianHua Zhang a, b, Sharon Sampogna b, Michael H. Chase a, b, c a b c

VA Greater Los Angeles Healthcare System, Los Angeles, CA 90073, USA WebSciences International, Los Angeles, CA 90024, USA Department of Physiology, School of Medicine at UCLA, Los Angeles, CA 90024, USA

a r t i c l e

i n f o

Article history: Received 22 May 2012 Revised 27 July 2012 Accepted 3 August 2012 Available online 19 August 2012 Keywords: Hypoxia Rat model Obstructive sleep apnea CA1 fEPSP Excitotoxicity Neurodegeneration Plasticity Cognitive processes

a b s t r a c t Obstructive sleep apnea (OSA) results in the degeneration of neurons in the hippocampus that eventuates in neurocognitive deficits. We were therefore interested in determining the effects of apnea on monosynaptic excitatory processes in a hippocampal pathway (cornu ammonis 3–cornu ammonis 1, CA3–CA1) that has been shown to mediate the processing of cognitive information. In addition, to substantiate an anatomical basis for the cognitive dysfunction that occurs in OSA patients, we examined the effects of apnea with respect to neurodegenerative changes (apoptosis) in the same hippocampal pathway. In order to determine the effects of apnea, an automated system for the generation and analysis of single and recurrent periods of apnea was developed. Utilizing this system, the field excitatory postsynaptic potential (fEPSP) generated by pyramidal neurons in the CA1 region of the hippocampus was monitored in α-chloralose anesthetized rats following stimulation of glutamatergic afferents in the CA3 region. A stimulus–response (input– output) curve for CA3–CA1 synaptic activity was determined. In addition, a paired-pulse paradigm was employed to evaluate, electrophysiologically, the presynaptic release of glutamate. Changes in the synaptic efficacy were assessed following single episodes of apnea induced by ventilatory arrest (60 to 80 s duration, mean=72 s; mean oxygen desaturation was 53% of normoxia level). Apnea resulted in a significant potentiation of the amplitude (mean = 126%) and slope (mean = 117%) of the baseline CA1 fEPSP. This increase in the fEPSP was accompanied by a significant decrease in the amplitude (71%) and slope (81%) of normalized paired-pulse facilitation (PPF) ratios. Since the potentiation of the fEPSP is inversely proportional to changes in PPF ratio, the potentiated fEPSP accompanied by the reduced PPF reveals that apnea produces an abnormal increase in the preterminal release of glutamate that results in the over-activation (and calcium overloading) of hippocampal CA1 neurons. Thus, we conclude that individual episodes of apnea result in the development of excitotoxic processes in the hippocampal CA3–CA1 pathway that is critically involved in the processing of cognitive information. Morphologically, the deleterious effect of recurrent apnea was substantiated by the finding of apoptosis in CA1 neurons of apneic (but not normoxic) animals. © 2012 Elsevier Inc. All rights reserved.

Introduction Recurrent periods of apnea, which result in hypoxia, produce changes in cellular responsiveness that precede structural alterations (Fung et al., 2007, 2009; Gozal et al., 2001). Both presynaptic

Abbreviations: AKT, protein kinase B; CA1, cornu ammonis region 1; CA3, cornu ammonis region 3; COPD, chronic obstructive pulmonary disease; fEPSP, field excitatory postsynaptic potential; GABA, gamma-butyric acid; i.v., intravenous; LTP, long-term potentiation; MΩ, megohm; NaCl, sodium chloride; NMDA, N-methyl-D-aspartate; o, stratum oralis; OSA, obstructive sleep apnea; p, stratum pyramidale; PBS, phosphatebuffered saline; PI3K, phosphatidylinositol 3-kinase; PPF, paired-pulse facilitation; r, stratum radiatum; SEM, standard error of the mean; SpO2, oxygen saturation level of hemoglobin. ⁎ Corresponding author at: WebSciences International, 1251 Westwood Blvd., Los Angeles, CA 90024, USA. Fax: +1 310 235 2067. E-mail address: [email protected] (S.J. Fung). 0014-4886/$ – see front matter © 2012 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.expneurol.2012.08.006

(enhanced preterminal release of glutamate) (Fung et al., 2007, 2009) and postsynaptic mechanisms (over-activation of N-methyl-D-aspartate (NMDA) receptors) (Forder and Tymianski, 2009) contribute to elevated intracellular calcium levels and the production of free radicals that eventuate in neurodegeneration/apoptosis (for reviews, see Choi, 1988; Lau and Tymianski, 2010). Even single episodes of apnea in animals (Fung et al., 2007, 2009) produce marked cellular effects such as a decrease in the paired-pulse facilitation (PPF) ratio (that signifies an enhanced release of glutamate transmitter) (Alger and Teyler, 1976; Andersen, 1960; Creager et al., 1980) that accompanies an increase in NMDA receptor-mediated fEPSPs (i.e., postsynaptic hyper-excitability) (Fung et al., 2007). In patients with obstructive sleep apnea (OSA), apnea results in cognitive impairments and damage to neuronal elements (Macey et al., 2002; Morrell et al., 2010). Experimentally, apnea leads to neuronal damage (apoptosis) in the hippocampus (Fung et al., 2007, 2009). Similarly, apoptosis of hippocampal neurons occurs in behaving rats

S.J. Fung et al. / Experimental Neurology 238 (2012) 107–113

Apnea Cycle 2

ire

Respiration

e

Materials and methods

Apnea Cycle 1 ir Exp

that have been subjected to periods of intermittent hypoxia (Gozal et al., 2001). In the present study, we examined the hypothesis that single episodes of apnea result in hyper-excitable CA3–CA1 synapses and an attenuation of the PPF in in vivo rats. To substantiate our hypothesis that apnea produces excitotoxic synaptic processes, we developed an Automated Apnea Program in order to determine electrophysiological and anatomical (neurodegenerative/apoptotic) changes in hippocampal CA1 neurons. Preliminary data have been presented in abstract form (Fung et al., 2011).

Insp

108

t1

t2

Baseline SpO2 Pump Off

Animals

Pump On

Nadir SpO2

Adult Sprague Dawley rats (n = 13, male, mean body weight = 303.90 ± 27.16 g) were used in the present study. All experimental procedures were approved by the Institutional Animal Care and Use Committee of the VA Greater Los Angeles Healthcare System (VAGLAHS) and were conducted in accord with the National Institutes of Health Guide for the Care and Use of Laboratory Animals (NIH publications no. 80‐23, revised1996). Animals were housed in a temperature (22± 1 °C) and humidity (50–70%) controlled environment with a 12:12 h light/dark cycle; the animals had ad libitum access to food and water. Appropriate measures were taken to minimize pain, discomfort or stress of the animals and only the minimum number of animals was used necessary to produce reliable scientific data.

Fig. 1. This figure illustrates the parameters of apnea that are controlled by the Automated Apnea Program to regulate the occurrence of individual and recurrent episodes of apnea for a pre-determined period of time. Apnea is induced experimentally by ventilatory arrest (respirator off, t1); re-ventilation resumes at the termination of an episode of apnea (respirator on, t2). See Automated Apnea Program in Materials and methods section for details.

Surgical preparation

To record the monosynaptic, field excitatory postsynaptic potential (fEPSP), a glass microelectrode (2 M NaCl-filled, 1–1.5 MΩ tip resistance) was positioned under stereotaxic guidance in the CA1 stratum radiatum of the hippocampus that contains the dendritic field of pyramidal neurons (3 mm posterior, 2 mm lateral to bregma, 2 mm below the surface of the cortex). A silver/silver chloride ground wire was placed subcutaneously in the nuchal region. The fEPSP was amplified with an AxoClamp 2A preamplifier (Axon Instruments, Inc., Foster City, CA) operating in current clamp mode in conjunction with a low-pass filter (2 kHz). The high-gain (100×) output signal was monitored on an oscilloscope. In addition, the fEPSP signal was digitized at 20 kHz (50 μs/bin) through a Digidata 1200 A-D interface (Axon Instruments, Inc.) in combination with AxoGraph X Scientific Software. Data were stored on a Power Mac G4 computer for offline analysis. A monopolar stainless steel stimulating electrode was placed in the hippocampal CA3 stratum radiatum (4 mm posterior, 3.5 mm lateral to bregma, 2 mm deep from cortex surface) in order to activate, ipsilaterally, glutamatergic Schaffer collaterals and commissural fibers of the CA3 that synapse monosynaptically with CA1 pyramidal neurons (Andersen et al., 1969; Johnston and Amaral, 1998). Cathodal stimulation (single-pulse, 0.2 ms duration, 0.2 Hz) was applied with a Grass S88 stimulator (Grass Instruments, Quincy, MA) via a Grass photoelectric constant current isolation unit. The optimal placement of electrodes for CA3 stimulation and CA1 recording was determined by varying the depth of the electrodes. A stimulus–response (input–output) relationship of the CA3–CA1 synapse was then constructed based upon variations in the stimulus intensity. Stimulation at an intensity eliciting approximately 30–40% of the maximal fEPSP response was adopted as a baseline throughout each recording session prior to and after single episodes of apnea. Consequently, it was possible to accurately quantify apnea-induced changes in the fEPSP at the steep gradient of the input–output curve in the absence of a ceiling effect or the inclusion of population spikes due to threshold discharge of CA1 neurons. To evaluate the effects of apnea on the CA1 fEPSP, paired-pulse evoked fEPSPs (0.2 ms, inter-pulse intervals 30–50 ms, repetition rate 0.2 Hz) were monitored for stability (baseline, pre-apneic control)

Under isoflurane anesthesia (2.5%), rats (n = 13) were tracheotomized and the jugular vein was catheterized for fluid and drug administration. For electrophysiological studies, the animal (n = 9) was placed in a stereotaxic apparatus; two holes (3 mm in diameter) were trephined in the skull overlying the dorsal part of the hippocampus to provide access for the placement of electrodes to stimulate CA3 and record from CA1. After completion of all surgical procedures, α-chloralose (120 mg/kg, i.v.) was substituted for isoflurane. Supplementary doses of α-chloralose (60 mg/kg, i.v.) were administered as needed. Prior to recording (30 min), the animals were immobilized with gallamine triethiodide (20 mg/kg, i.v.) and artificially ventilated. Rectal temperature was maintained at 35 ± 1 °C by means of a heat lamp. Heart rate, the oxyhemoglobin saturation level (SpO2) and respiratory end-tidal carbon dioxide were continuously monitored by a pulse oximeter (SurgiVet V90041). In this study, we noted that hypercapnia with an end-tidal CO2 of 5–6.5% occurred during recurrent apnea that was similar to that reported in our previous paper (Fung et al., 2009). Automated Apnea Program An Automated Apnea Program was developed in order to control the “on” and “off” functions of respiratory pump by use of a × 10 CM15A wireless controller. Apnea was initiated by ventilatory arrest in order to generate conditions of hypoxemia and hypercapnia that are prevalent in humans with OSA (Beebe and Gozal, 2002). To initiate the program, two initial parameters: duration of apnea cycle (t1 + t2) and off time (t1) (Fig. 1) are set by the user. Each episode of apnea consists of “off time (t1)” + “on time (t2)”. Once initiated, the program controls the occurrence of apnea episodes for the duration of the experiment. The “off time (t1)” is adjustable during the experiment in order to maintain a constant nadir level of oxygen. Data, which are summarized and presented in an Excel spreadsheet at the end of the experiment, include the following: (1) the total and mean durations of t1; (2) the total and mean durations of t2; (3) the total and mean number of cycles (t1 + t2) per minute; and

(4) the duration of the experiment. The program was employed to generate experimental paradigms in the present study that consisted of single as well as recurrent episodes of apnea. Single episodes of apnea

S.J. Fung et al. / Experimental Neurology 238 (2012) 107–113

109

before the initiation of the apnea episode. Apnea was induced experimentally by ventilatory arrest. Using the Automated Apnea Program, the severity of apnea was regulated by varying the “off” time (t1, Fig. 1) so that the nadir oxygen saturation was approximately 55% SpO2; recovery to >95% SpO2 occurred upon re-ventilation. At the termination of the apnea episode, recordings of the fEPSP resumed at 1-min and 5-min intervals for the first 10-min post-apnea period and thereafter, respectively. Ten consecutive, raw fEPSP traces (150 ms duration; sampling rate = 0.2 Hz) were routinely averaged for a given time point prior to (baseline control) and following the termination of a single episode of apnea. Before the measurements were obtained, the potential at the foot of the fEPSP (see Fig. 4A for details) was adopted as the baseline reference potential. If the foot of the fEPSP was masked by the decay of the stimulus artifact, the electrical potential preceding the onset of the stimulus artifact of individual fEPSPs was adopted as the extrapolated baseline reference potential. The slopes of the evoked fEPSPs were computed as the first derivative (i.e., volts per second) of the fEPSP. Data were analyzed statistically based upon the two-tailed paired Student's t test (StatView 4.5 statistical software, SAS Institute, Cary, NC) in order to compare the control (pre-apneic) values with the experimental (post-apneic) variables. Baseline fEPSPs that were recorded for 3 min were pooled and averaged to yield a mean baseline (control) fEPSP prior to apnea. In each animal, changes in fEPSPs post-apnea were normalized to their pre-apneic control for both amplitude and slope measurements. The PPF ratios were normalized in a similar manner prior to statistical analysis. Differences were considered statistically significant at p b 0.05; data were expressed as means ± SEM.

paraformaldehyde, 15% picric acid, and 0.25% glutaraldehyde in 0.1 M phosphate-buffered saline (PBS) (pH 7.4). Tissue blocks containing the hippocampus were postfixed overnight in a 2% paraformaldehyde fixative at 4 °C prior to cryoprotection by immersion in 25% sucrose in 0.1 M PBS at 4 °C for 48 h. Serial frozen sections were cut and stored in 0.1 M PBS with cryoprotectant at 4 °C. Immunostaining for singlestranded DNA was employed as a marker for apoptosis. Detailed immunostaining techniques have been published (Fung et al., 2007, 2009; Zhang et al., 2009, 2010). Briefly, hippocampal sections (30 μm thickness) were mounted onto gelatin-coated superfrost slides. The mounted sections were air‐dried and stored at room temperature. Brain sections were incubated for 2 h in a mouse monoclonal antibody against single-stranded DNA (Millipore, Temecula, CA; MAB3299; 1:10 dilution in 0.1 M PBS). The specificity of the antibody has been shown to label damaged neurons (Frankfurt et al., 1996). Sections were then thoroughly rinsed with PBS prior to incubation with a donkey anti-mouse biotinylated secondary antibody (Jackson ImmunoResearch Labortories, West Grove, PA; 1:200 dilution; 20 min). The sections were then incubated with the avidin–biotin complex (Vector Labortories, Burlingame, CA; 1:100 dilution; 15 min). Immuno-reactivity was identified by development with 0.02% 3,3′-diaminobenzidine and 0.015% hydrogen peroxide in 50 mM Tris buffer (pH 7.6) for 15 to 30 min. After rinsing with PBS, the sections were dehydrated and coverslipped with Permount.

Recurrent apnea

Detailed stimulus–response studies were performed to generate input–output curves. A representative plot is illustrated in Fig. 2 (A, amplitude; B, slope). Optimal CA3 stimulation at low intensities (10–15 μA) produced a stable (1–2 mV) fEPSP which was recorded from CA1. The fEPSP was typified by a smooth rising phase that was uncontaminated by any population spike at the peak of the negative wave (Figs. 2–4). A graded change in the amplitude of the fEPSP at various CA3 stimulus intensities for the CA3–CA1 synapse occurred, which is shown by the superimposed fEPSP traces in Fig. 2C. A non-linear stimulus–response curve that consisted of a steep gradient at mid-range of the stimulus strength was obtained for the amplitude (Fig. 2A) and the slope (Fig. 2B) measurements of the fEPSP responses to graded stimulus intensities (X axis, Figs. 2A and B). For all experiments, the CA3–CA1 stimulus-recording parameters were established according to the preceding input–output relationship. All baseline fEPSPs were set at amplitudes of approximately 1–1.2 mV that was 30–40% of the maximal fEPSP pre-apnea. This control fEPSP was chosen in order to be able to document any reliable increase in post-apneic fEPSPs without their being compromised by a ceiling effect (near the plateau, Fig. 2A) of the input–output relationship of the CA3–CA1 synapses.

Four animals were used for anatomical studies to determine the morphological changes due to recurrent apnea. Accordingly, we developed an Automated Apnea Program with well-characterized parameter controls in order to generate and analyze data obtained in conjunction with recurrent apnea protocols. The parameters were set as follows: duration of each episode (cycle) of apnea = 25 s; “off” time (t1) = 13 s; “on” time (t2) = 12 s; total number of episodes of apnea = 288; experimental duration = 2 h. The duration of the “off time” reflects the intensity/severity of hypoxia as indicated by the lowest level of oxygen saturation (SpO2) reached due to ventilatory arrest. This time-dependent oxygen desaturation response varies during the 2-h period of recurrent apnea, both within and across animals. Our protocol was set at a level of approximately 65% SpO2, over the 2-h experimental session, for the group of experimental (apneic) animals. Whenever the apnea-induced SpO2 value reached a lower or higher level, the “off time” was reduced or increased, respectively, in order to re-set the nadir value to the target 65% SpO2 level. Each animal was surgically prepared similar to those used in the electrophysiological studies (see above). Experimental rats (n = 2) underwent 2 h of recurrent apnea (via ventilatory arrest to 65% SpO2 level), which was followed by re-ventilation to 95% SpO2. Sham-operated (control) rats (n = 2) were maintained with normal respiration for the same period of time. The rationale for our protocol of recurrent apnea was based upon the concurrent presence, in OSA patients, of hypoxia and hypercapnia as well as neurodegeneration (Beebe and Gozal, 2002; Macey et al., 2002). Additionally, the target level of SpO2 was selected based upon the fact that it applies directly to that described in certain population of humans with OSA (see In vivo rat model of apnea section). Morphological studies At the end of the sessions of recurrent apnea, the animals were perfused with saline followed by an ice cold fixative mixture of 4%

Results Input–output relationships of CA3–CA1 synapses

Effects of apnea on the CA1 fEPSP Individual animals (n= 9) were subjected to a single episode of apnea for a mean duration of 72.2 ± 2.4 s (range = 60–80 s). During this period, the oxyhemoglobin saturation reached a mean nadir of 52.6± 4.5% (range = 50–59%) SpO2 level. Upon re-ventilation, the SpO2 recovered to 95% in less than 15 s. Apnea elicited an increase in the postsynaptic responsiveness of CA1 neurons (Fig. 3), with a mean peak time of 13.2 min post-apnea (range = 5.0–20.0 min). Specifically, a maximal increase in the mean amplitude (125.7 ±4.6%, range = 113.0–157.9%, p = 0.001) and mean slope (116.6 ±4.7%, range= 103.3–148.1%, p = 0.008) of the first fEPSP (of the paired-pulse evoked fEPSPs) occurred. The potentiated fEPSP diminished to near baseline level at approximately 20–30 min post-apnea.

110

S.J. Fung et al. / Experimental Neurology 238 (2012) 107–113

B. Slope

2.8

0.6

2.4

0.5

2.0

Slope (V/s)

Amplitude (mV)

A. Amplitude

1.6 1.2

0.4 0.3

0.8

0.2

0.4

0.1

0

0

5

10

15

20

25

30

35

0

0

5

Intensity (µA)

10

15

20

25

30

35

Intensity (µA)

C. Superimposed CA1 fEPSPs

1 mV

5 ms

Fig. 2. Input–output activity of hippocampal CA3/CA1 synapses. Stimulus-evoked synaptic responses of CA1 neurons as determined by the amplitude (A) and the maximal slope (B) of the fEPSP. Each data point in A and B is an average of 10 samples of the CA1 fEPSP evoked by stimulation of CA3 at graded intensities from 2.5 to 35 μA. C: superimposed averages of the CA1 fEPSP response to stimulation of CA3 at different intensities, from bottom downwards, at incremental steps of 2.5, 5, 7.5, 10, 12.5, 15, 20, 25, and 30 μA. Stimulus artifacts (arrowhead) are truncated.

A. Control

B. Post-apnea

0.5 mV 5 ms

C. Superimposed

Control

Post-apnea Fig. 3. Apnea increases CA1 fEPSP. Apnea resulted in the potentiation of CA1 synaptic responses. Maximal potentiation of the averaged fEPSP (B, C) occurred at 12 min following the termination of a single episode of apnea (75 s, 50% SpO2); SpO2 recovered to baseline (95%) level at 15 s following the termination of apnea. The control fEPSP (A, C) is shown at 1 min prior to the initiation of apnea. Prominent apnea-induced potentiation of the fEPSP is illustrated by superimposing both the pre- and post-apneic fEPSPs (C). Stimulus artifacts (arrowhead) are truncated.

Paired-pulse facilitation (PPF) of CA1 fEPSPs The paired-pulse stimulus paradigm elicited a facilitated second fEPSP that was larger in amplitude and the slope measurements were greater than those of the first fEPSP. Quantitatively, the PPF ratio was expressed as the quotient of the increase in the second fEPSP (fEPSP2) over the first one (fEPSP1) divided by the first fEPSP of the pair (fEPSP1) (Fig. 4A). For example, the PPF ratios for the baseline (pre-apneic) fEPSP depicted in Fig. 4A reveal an increase of 53.7% and 70.8% in the amplitude and slope of the baseline fEPSPs, respectively. Since the second fEPSP varied inversely according to changes in the first of the paired-pulse evoked fEPSPs, the effect of apnea-induced potentiation was less prominent vis-à-vis the second fEPSPs. To demonstrate the reduction in the PPF post-apnea, the first apneic fEPSP was graphically scaled i.e., normalized to an amplitude equivalent to that recorded under baseline conditions (note that the peaks of both traces of fEPSP1 are superimposed as depicted in Fig. 4B). For the fEPSP2, however, the superimposed traces indicate a clear reduction in PPF post-apnea. At the same time when maximal potentiation of the first fEPSP occurred, the post-apneic PPF ratios dropped significantly to 71.2±5.7% (range=32.7–89.4%, p=0.001) and 81.0±4.6% (range=63.0–99.1, p=0.003) of the pre-apneic control values for the amplitude and slope measurements, respectively. Recurrent apnea produces cell death of CA1 neurons To evaluate the effects of recurrent episodes of apnea on the morphology of CA1 pyramidal neurons, four animals were divided equally

S.J. Fung et al. / Experimental Neurology 238 (2012) 107–113

A fEPSP1 fEPSP2

0.5 mV 10 ms

B

Post-apnea 0.5 mV 10 ms

Control

Fig. 4. (A) Paired-pulse facilitation (PPF) of CA1 fEPSPs. Quantitatively, PPF ratio was calculated as [(fEPSP2 − fEPSP1)/fEPSP1], where fEPSP1 and fEPSP2 are the mean values of the first and second fEPSPs, respectively. Dotted lines indicate the upper reference baseline potential and the bottom negative peak of individual fEPSPs. (B) Apnea reduces the PPF ratio. The amplitude of the post-apneic fEPSP trace was normalized with that of the pre-apneic control, as shown in this figure. The superimposed traces reveal a clear reduction in the PPF of the second fEPSP, post-apnea. The apnea-induced decrease in PPF was 29% of the pre-apneic control (PPF ratios=0.389 and 0.278 for control and post-apnea conditions, respectively). Stimulus artifacts (arrowhead) are truncated.

into control and experimental groups. Individual animals in the experimental group underwent a 2-h period of recurrent apnea. Based upon 160 cycles of apneic episodes (80 from each rat) that were sampled at intervals during this 2-h period, a mean nadir SpO2 of 64.1 ± 0.3% (range = 50–70%) was determined to have occurred. Upon re-ventilation, the SpO2 recovered to 94.8 ± 0.1% (range = 94– 98%) in approximately 12 s. Fig. 5 presents data from a representative animal from each of the control (normoxic, Fig. 5A) and experimental (apneic, Fig. 5B) groups in this study. Single-stranded DNA immunoreactivity was present in CA1 neurons of the hippocampus, bilaterally. Within individual labeled neurons, immunoreactivity (i.e., as a marker of apoptosis) occurred in both the nuclear as well as cytoplasmic compartments (Fig. 5B). Labeled neurons were located in the layer of stratum pyramidale where CA1 pyramidal neurons were localized (Fig. 5B, arrows). A small number of labeled neurons were also present in adjacent layers (e.g., strata oralis; Fig. 5B, arrows). There was no evidence of apoptosis in the CA1 region of the control animals (Fig. 5A). Discussion In vivo rat model of apnea Previously, in vitro studies have provided fundamental data dealing with hypoxia-related changes in synaptic processes. For example, rat brain slices have been subjected to long periods of hypoxia or anoxia (Coelho et al., 2000; Sebastião et al., 2001); however, these procedures

111

contrast sharply with the episodic patterns of apnea/hypopnea that occur during the nocturnal sleep of OSA patients (Beebe and Gozal, 2002). In addition, a modified in vitro design with a cyclic hypoxia protocol has been employed to study the effects of recurrent hypoxia on the synthesis of catecholamines by the PC-12 cell line (Kumar et al., 2003). In vivo studies have employed a hypoxic chamber wherein rodents are subjected to daily intermittent hypoxia (for a period of months) in order to assess the effects of chronic intermittent hypoxia (e.g., Gozal et al., 2001; Veasey et al., 2004). However, these studies do not include the cyclic patterns of hypoxia and the concomitant conditions of hypercapnia that are prevalent in OSA patients (Beebe and Gozal, 2002; Tafil-Klawe et al., 2004). To more accurately replicate the human pathophysiological condition on an accelerated temporal scale that is amenable to experimental testing, an Automated Apnea Program was developed with welldefined parameter settings that consist of a user defined frequency of apnea duration (once every 25 s or 144 episodes/h) and a low level of oxyhemoglobin saturation (55–60% SpO2). The latter targeted SpO2 was justified since it corresponds to the nocturnal nadir oxygen saturation values of 50–55% SpO2 that have been reported in patients with OSA (Gries and Brooks, 1996; Weiss et al., 2000) and chronic obstructive pulmonary disease (COPD) (Flick and Block, 1977; Wynne et al., 1978). Compared to previous rodent models (Gozal et al., 2001; Veasey et al., 2004), in the present rat model, both the hypoxia and hypercapnic components are present that are characteristic of the intermittent blood gas abnormalities that occur in patients with OSA (Beebe and Gozal, 2002; Tafil-Klawe et al., 2004). Previous findings dealing with hippocampal processes suggest that hypercapnia negatively impacts the excitability of CA1 neurons, causing a diminished synaptic release of glutamate (Vannucci et al., 1997), a reduction in the amplitude of the CA1 population spike (Balestrino and Somjen, 1988), and a decrease in sodium channel activity which results in a decrease in membrane excitability of CA1 neurons (Gu et al., 2007). Despite these suppressant actions of hypercapnia, our findings revealed that recurrent apnea elicits an overwhelming, hyper-excitable effect that eventuates in an increase in apoptosis in hippocampal CA1 neurons (Fig. 5B). This finding is in accord with human brain imaging data that show a loss of gray matter in CA1 of OSA patients (Macey et al., 2002; Morrell et al., 2010). Furthermore, with the present rat model, it is feasible to measure functional synaptic changes induced by apnea in conjunction with anatomical alterations which include neuronal apoptosis. There are other advantages to using this rat model of apnea. Chronic behaving rats, which are also the prototypical species that have been employed in behavioral studies of cognition (e.g., Morris maze test; Gozal et al., 2001), have also been successfully used by ourselves (Fenik et al., 2010, 2011) as well as others (Jelev et al., 2001) to monitor hypoglossal motoneuron activity that is involved in controlling the patency of the upper airway during sleep and wakefulness. In this regard, we have shown, in preliminary studies, that recurrent episodes of apnea not only attenuate the long-term potentiation (LTP) of hippocampal CA3–CA1 synapses, which are processes that are important in the cognitive processing of information, in in vivo rats, but also produce neurodegenerative changes in hypoglossal motoneurons of in vivo rats (Fung et al., 2012, in press). Thus, the present rat model of apnea encompasses many of the abnormal changes that are characteristic of OSA. In addition, key pathophysiological signs that are documented in OSA patients are replicated in this rat model including combined hypoxia–hypercapnia, neurodegeneration/apoptosis of CA1 neurons, and cognitive impairments, as determined by attenuation of the LTP. Whereas OSA consists of recurrent episodes of hypoxia, hypercapnia, cellular oxidative stress, cardiopulmonary disturbances, neurodegeneration, cognitive impairments, depression and anxiety, it is unrealistic for any animal model to recapitulate all of these pathological processes, mechanisms and behaviors. Consequently, our rat model was limited insofar as it was not designed to include certain aspects of OSA that arise in humans such as OSA-related signs of

112

S.J. Fung et al. / Experimental Neurology 238 (2012) 107–113

A

o

B

o

p

p

r

r

Fig. 5. Photomicrographs of recurrent apnea-induced excitotoxicity as labeled by single-stranded DNA immunoreactivity in CA1 neurons. Note that positive labeling of apoptotic changes is abundant among CA1 neurons of the apneic animal (B, arrows), compared to the absence of the dark-stained immunoreactivity of single-stranded DNA in the control animal (A). In addition to the neurons in the stratum pyramidale (p), the adjacent stratum (B: oralis, o) also exhibits some apoptotic, labeled neurons subsequent to recurrent apnea. Other abbreviation: r, stratum radiatum. Scales in A and B = 50 μm.

Valsalva maneuvers evoked autonomic responses (Henderson et al., 2003), paradoxical breathing movements, and/or negative intrathoracic pressure during apnea (Abad and Guilleminault, 2004), due to the fact that our protocol necessitated the administration of a muscle relaxant (Flaxedil) in order to maintain a constant mechanical stability during the period of electrophysiological recordings pre- and post-apnea. Additionally, this model was not designed to assess neurological changes vis-à-vis apnea of central origins; in addition, the etiology of central sleep apnea is different from that of OSA (ibid.). Apnea-induced potentiation of the fEPSP is accompanied by attenuation of the PPF In the present study, correlated changes in both synaptic efficacy and the PPF prior to and after single episodes of apnea were determined. Specifically, we found that apnea-induced potentiation of the fEPSP is accompanied by a reduction in the PPF. Mechanistically, PPF could involve pre- as well as postsynaptic excitatory drives that result in changes in neuronal excitability (Harris and Cotman, 1985; Manabe et al., 1993; Wang and Kelly, 1996, 1997). Classically, it has been considered that a reduction in PPF signifies an increase in the preterminal release of excitatory transmitters in the CA3–CA1 pathway (Andersen, 1960; Creager et al., 1980). An inverse relationship of this nature has previously been explained on the bases of the “residual calcium hypothesis” (Katz and Miledi, 1968). Accordingly, a large increase in the initial evoked fEPSP would provide less residual/ unsequestered calcium in the preterminals; therefore, the probability of release of transmitter vesicles that contribute to the second fEPSP response would be lower, resulting in a reduced PPF, which is exactly that which was observed in the present study. The fact that this same process occurred in response to single episodes of apnea plus the presence of apoptotic CA1 neurons after recurrent episodes of apnea, strongly suggests that a sustained increase in the presynaptic release of glutamate contributes to excitotoxicity (apoptosis) in apneic animals.

al., 2005). It is also known that severe insults involving NMDA receptor over-stimulation can occur as early as 1 h after hypoxia/ischemia (Waters and Machaalani, 2004). Specific temporal patterns of effects of intermittent hypoxia on the anti-apoptotic phosphatidylinositol 3-kinase (PI3K)–protein kinase B (AKT) transduction cascade have been reported in hippocampal CA1 neurons of rats (Goldbart et al., 2003). With the hypoxic chamber technique, decreases in activation of the protein kinases occur as early as 1 h post intermittent hypoxia (ibid). Additionally, mitochondrial dysregulation due to hypoxia is known to lead to activation of pro-apoptotic signal-transduction molecules (Green and Kroemer, 2004) that results in hypoxiarelated hippocampal cell death (Hota et al., 2012). The presence of immunoreactive single-stranded DNA within the cytoplasmic compartment of CA1 neurons in apneic animals confirms the role of mitochondrial mechanisms that mediate the neurodegenerative processes that are evidenced in this study. We also demonstrated, electrophysiologically, that even single episodes of apnea result in hyper-excitable CA1 neurons with a correlated reduction in the PPF, which suggests an excessive synaptic release of glutamate, post-apnea. The abnormal, sustained presence of hyperexcitable CA3–CA1 synapses likely triggers cascades of excitotoxic processes in CA1 neurons (Choi, 1988). Our findings of prominent apoptosis in CA1 neurons in apneic animals validate our hypothesis that the hyper-excitability/excitotoxicity effect is particularly potent in conjunction with recurrent episodes of apnea, such as those that occur in OSA. Since the effects of apnea are global, many transmitter systems are likely to be involved in the excitotoxic processes that were observed in this study. We previously demonstrated that apnea-induced apoptotic damage in guinea pig CA1 neurons could be prevented by eszopiclone (Fung et al., 2009), a positive allosteric modulator of gamma-butyric acid subtype A (GABAA) receptors (Smith et al., 2001). This raises the possibility that apnea also impacts adversely GABAergic interneurons that inhibit CA1 neurons (Klausberger and Somogyi, 2008), resulting in an unimpeded increase in their membrane excitability, post-apnea, which leads to increased apnea-induced apoptotic damage in CA1 neurons in our rat model.

Recurrent apnea-induced apoptotic changes of CA1 neurons Conclusions Utilizing our Automated Apnea Program, recurrent apnea that was conducted for a period of 2 h resulted in the apoptotic degeneration of CA1 neurons. Due to the severity of the experimental outcome (apoptosis) in an accelerated time frame (2 h), our recurrent apnea model mimics the pathophysiological conditions that are present in patients with OSA. Fragmentation of the nucleosomal DNA and caspace activation (hallmarks of apoptosis) have been reported on a similar timescale following ischemia in CA1 slice cultures induced by the oxygen–glucose deprivation method (de Pina-Benabou et

In conclusion, we have demonstrated the usefulness of an Automated Apnea Program and a new animal model of recurrent apnea that was designed to emulate the physiological and anatomical processes that occur in individuals with OSA. This model provides a foundation for examining the cellular, mechanistic changes that occur vis-à-vis the functional and behavioral consequences of apnea/hypoxia in the hippocampus (and other structures), including the development of permanent neurological (learning and memory) impairments and mood

S.J. Fung et al. / Experimental Neurology 238 (2012) 107–113

disorders related to the OSA syndrome (Engleman et al., 2000; Sateia, 2003). Utilizing the Automated Apnea Program and our animal model of recurrent apnea, we determined that single and recurrent episodes of apnea result in the development of excitotoxic synaptic processes in the hippocampus as well as the apoptosis of hippocampal neurons, which we suggest underlie the cognitive impairments that arise in conjunction with OSA. Acknowledgments This work was supported by grants 1|01BX000819 and TGS-109219. The authors gratefully acknowledge the expert technical assistance of Vince Lim and John Lopatto. We are indebted to Johnathan Lim for his successful development of the Automated Recurrent Apnea Program which was adopted to regulate the apnea protocols in this study. References Abad, V.C., Guilleminault, C., 2004. Neurological perspective on obstructive and nonobstructive sleep apnea. Semin. Neurol. 24, 261–269. Alger, B.E., Teyler, T.J., 1976. Long-term and short-term plasticity in the CA1, CA3, and dentate regions of the rat hippocampal slice. Brain Res. 110, 463–480. Andersen, P., 1960. Interhippocampal impulses. II. Apical dendritic activation of CA1 neurons. Acta Physiol. Scand. 48, 178–208. Andersen, P., Bliss, T.V., Lomo, T., Olsen, L.I., Skrede, K.K., 1969. Lamellar organization of hippocampal excitatory pathways. Acta Physiol. Scand. 76, 4A–5A. Balestrino, M., Somjen, G.G., 1988. Concentration of carbon dioxide, interstitial pH and synaptic transmission in hippocampal formation of the rat. J. Physiol. 396, 247–266. Beebe, D.W., Gozal, D., 2002. Obstructive sleep apnea and the prefrontal cortex: towards a comprehensive model linking nocturnal upper airway obstruction to daytime cognitive and behavioral deficits. J. Sleep Res. 11, 1–16. Choi, D.W., 1988. Glutamate neurotoxicity and diseases of the nervous system. Neuron 1, 623–634. Coelho, J.E., de Mendonça, A., Ribeiro, J.A., 2000. Presynaptic inhibitory receptors mediate the depression of synaptic transmission upon hypoxia in rat hippocampal slices. Brain Res. 869, 158–165. Creager, R., Dunwiddie, T., Lynch, G., 1980. Paired-pulse and frequency facilitation in the CA1 region of the in vitro rat hippocampus. J. Physiol. 299, 409–424. de Pina-Benabou, M.H., Szostak, V., Kyrozis, A., Rempe, D., Uziel, D., Urban-Maldonado, M., Benabou, S., Spray, D.C., Federoff, H.J., Stanton, P.K., Rozental, R., 2005. Blockade of gap junctions in vivo provides neuroprotection after perinatal global ischemia. Stroke 36, 2232–2237. Engleman, H.M., Kingshott, R.N., Martin, S.E., Douglas, N.J., 2000. Cognitive function in the sleep apnea/hypopnea syndrome (SAHS). Sleep 23 (Suppl. 4), S102–S108. Fenik, V.B., Fung, S.J., Chase, M.H., 2010. An animal model to study hypoglossal motoneuron excitability during sleep and wakefulness. Soc. Neurosci. Abstr., 817.4. Fenik, V.B., Lim, V., Fung, S.J., Chase, M.H., 2011. Excitatory inputs to hypoglossal motoneurons are differentially depressed during REM sleep in rats. Sleep 34, A31 (Suppl.). Flick, M.R., Block, A.J., 1977. Continuous in-vivo monitoring of arterial oxygenation in chronic obstructive lung disease. Ann. Intern. Med. 86, 725–730. Forder, J.P., Tymianski, M., 2009. Postsynaptic mechanisms of excitotoxicity: involvement of postsynaptic density proteins, radicals, and oxidant molecules. Neuroscience 158, 293–300. Frankfurt, O.S., Robb, J.A., Sugarbaker, E.V., Villa, L., 1996. Monoclonal antibody to single-stranded DNA is a specific and sensitive cellular marker of apoptosis. Exp. Cell Res. 226, 387–397. Fung, S.J., Xi, M.C., Zhang, J.H., Sampogna, S., Yamuy, J., Morales, F.R., Chase, M.H., 2007. Apnea promotes glutamate-induced excitotoxicity in hippocampal neurons. Brain Res. 1179, 42–50. Fung, S.J., Xi, M.C., Zhang, J.H., Yamuy, J., Sampogna, S., Tsai, K.L., Lim, V., Morales, F.R., Chase, M.H., 2009. Eszopiclone prevents excitotoxicity and neurodegeneration in the hippocampus induced by experimental apnea. Sleep 32, 1593–1601. Fung, S.J., Xi, M.C., Lim, V., Bronson, D.R., Wilson, C.L., Chase, M.H., 2011. Apnea produces hyperactive synaptic activities in the hippocampus in in vivo rats. Soc. Neurosci. Abstr., program no. 502.8. Fung, S.J., Xi, M.C., Fenik, V.B., Lopatto, J., Chase, M.H., in press. Recurrent apnea disrupts long-term potentiation (LTP) in the hippocampus in in vivo rats. Soc. Neurosci. Abstr. Fung, S.J., Zhang, J.H., Xi, M.C., Sampogna, S., Chase, M.H., 2012. Recurrent Apnea Induces Apoptosis in Hypoglossal Motoneurons in In Vivo Rats. Sleep 35, A44 (Suppl.). Goldbart, A., Cheng, Z.J., Brittian, K.R., Gozal, D., 2003. Intermittent hypoxia induces time-dependent changes in the protein kinase B signaling pathway in the hippocampal CA1 region of the rat. Neurobiol. Dis. 14, 440–446.

113

Gozal, D., Daniel, J.M., Dohanich, G.P., 2001. Behavioral and anatomical correlates of chronic episodic hypoxia during sleep in the rat. J. Neurosci. 21, 2442–2450. Green, D.R., Kroemer, G., 2004. The pathophysiology of mitochondrial cell death. Science 305, 626–629. Gries, R.E., Brooks, L.J., 1996. Normal oxyhemoglobin saturation during sleep. How low does it go? Chest 110, 1489–1492. Gu, X.Q., Kanaan, A., Yao, H., Haddad, G.G., 2007. Chronic high-inspired CO2 decreases excitability of mouse hippocampal neurons. J. Neurophysiol. 97, 1833–1838. Harris, E.W., Cotman, C.W., 1985. Effects of synaptic antagonists on perforant path pairedpulse plasticity: differentiation of pre- and postsynaptic antagonism. Brain Res. 334, 348–353. Henderson, L.A., Woo, M.A., Macey, P.M., Macey, K.E., Frysinger, R.C., Alger, J.R., Yan-Go, F., Harper, R.M., 2003. Neural responses during Valsalva maneuvers in obstructive sleep apnea syndrome. J. Appl. Physiol. 94, 1063–1074. Hota, K.B., Hota, S.K., Chaurasia, O.P., Singh, S.B., 2012. Acetyl-L-carnitine-mediated neuroprotection during hypoxia is attributed to ERK1/2-Nrf2-regulated mitochondrial biosynthesis. Hippocampus 22, 723–736. Jelev, A., Sood, S., Liu, H., Nolan, P., Horner, R.L., 2001. Microdialysis perfusion of 5-HT into hypoglossal motor nucleus differentially modulates genioglossus activity across natural sleep–wake states in rats. J. Physiol. 532, 467–481. Johnston, D., Amaral, D.G., 1998. Hippocampus, In: Shepherd, G.M. (Ed.), The Synaptic Organization of the Brain, 4th edition. Oxford University Press, New York, pp. 417–458. Katz, B., Miledi, R., 1968. The role of calcium in neuromuscular facilitation. J. Physiol. 195, 481–492. Klausberger, T., Somogyi, P., 2008. Neuronal diversity and temporal dynamics: the unity of hippocampal circuit operations. Science 321, 53–57. Kumar, G.K., Kim, D.K., Lee, M.S., Ramachandran, R., Prabhakar, N.R., 2003. Activation of tyrosine hydroxylase by intermittent hypoxia: involvement of serine phosphorylation. J. Appl. Physiol. 95, 536–544. Lau, A., Tymianski, M., 2010. Glutamate receptors, neurotoxicity and neurodegeneration. Pflugers Arch. 460, 525–542. Macey, P.M., Henderson, L.A., Macey, K.E., Alger, J.R., Frysinger, R.C., Woo, M.A., Harper, R.K., Yan-Go, F.L., Harper, R.M., 2002. Brain morphology associated with obstructive sleep apnea. Am. J. Respir. Crit. Care Med. 166, 1382–1387. Manabe, T., Wyllie, D.J., Perkel, D.J., Nicoll, R.A., 1993. Modulation of synaptic transmission and long-term potentiation: effects on paired pulse facilitation and EPSC variance in the CA1 region of the hippocampus. J. Neurophysiol. 70, 1451–1459. Morrell, M.J., Jackson, M.L., Twigg, G.L., Ghiassi, R., McRobbie, D.W., Quest, R.A., Pardoe, H., Pell, G.S., Abbott, D.F., Rochford, P.D., Jackson, G.D., Pierce, R.J., O'Donoghue, F.J., Corfield, D.R., 2010. Changes in brain morphology in patients with obstructive sleep apnoea. Thorax 65, 908–914. Sateia, M.J., 2003. Neuropsychological impairment and quality of life in obstructive sleep apnea. Clin. Chest Med. 24, 249–259. Sebastião, A.M., de Mendonca, A., Moreira, T., Ribeiro, J.A., 2001. Activation of synaptic NMDA receptors by action potential-dependent release of transmitter during hypoxia impairs recovery of synaptic transmission on reoxygenation. J. Neurosci. 21, 8564–8571. Smith, A.J., Alder, L., Silk, J., Adkins, C., Fletcher, A.E., Scales, T., Kerby, J., Marshall, G., Wafford, K.A., McKernan, R.M., Atack, J.R., 2001. Effect of alpha subunit on allosteric modulation of ion channel function in stably expressed human recombinant gamma-aminobutyric acid(A) receptors determined using (36)Cl ion flux. Mol. Pharmacol. 59, 1108–1118. Tafil-Klawe, M., Klawe, J.J., Sikorski, W., Drzewiecka, B., 2004. Reflex respiratory responses to progressive hyperoxic hypercapnia and normocapnic hypoxia in normocapnic and hypercapnic obstructive sleep apnea patients. J. Physiol. Pharmacol. 55, 135–138. Vannucci, R.C., Brucklacher, R.M., Vannucci, S.J., 1997. Effect of carbon dioxide on cerebral metabolism during hypoxia–ischemia in the immature rat. Pediatr. Res. 42, 24–29. Veasey, S.C., Zhan, G., Fenik, P., Pratico, D., 2004. Long-term intermittent hypoxia: reduced excitatory hypoglossal nerve output. Am. J. Respir. Crit. Care Med. 170, 665–672. Wang, J.H., Kelly, P.T., 1996. Regulation of synaptic facilitation by postsynaptic Ca2+/CaM pathways in hippocampal CA1 neurons. J. Neurophysiol. 76, 276–286. Wang, J.H., Kelly, P.T., 1997. Attenuation of paired-pulse facilitation associated with synaptic potentiation mediated by postsynaptic mechanisms. J. Neurophysiol. 78, 2707–2716. Waters, K.A., Machaalani, R., 2004. NMDA receptors in the developing brain and effects of noxious insults. Neurosignals 13, 162–174. Weiss, J.W., Launois, S.H., Anand, A., 2000. Cardiorespiratory changes in sleep-disordered breathing, In: Kryger, M.H., Roth, T., Dement, W.C. (Eds.), Principles and Practice of Sleep Medicine, 3rd ed. W.B. Saunders Co., Philadelphia, pp. 859–868. Wynne, J.W., Block, A.J., Hemenway, J., Hunt, L.A., Shaw, D., Flick, M.R., 1978. Disordered breathing and oxygen desaturation during sleep in patients with chronic obstructive pulmonary disease. Chest 73, 301–303. Zhang, J.H., Fung, S.J., Xi, M.C., Sampogna, S., Chase, M.H., 2009. Recurrent apnea induces neuronal apoptosis in the guinea pig forebrain. Exp. Neurol. 216, 290–294. Zhang, J.H., Fung, S.J., Xi, M.C., Sampogna, S., Chase, M.H., 2010. Apnea produces neuronal degeneration in the pons and medulla of guinea pigs. Neurobiol. Dis. 40, 251–264.