Physical activity attenuates intermittent hypoxia-induced spatial learning deficits and oxidative stress

Abstract

Rationale: Exposure to intermittent hypoxia (IH), such as occurs in sleep-disordered breathing, is associated with substantial cognitive impairments, oxidative stress and inflammation, and increased neuronal cell losses in brain regions underlying learning and memory in rats. Physical activity (PA) is now recognized as neuroprotective in models of neuronal injury and degeneration.

Objectives: To examine whether PA will ameliorate IH-induced deficits.

Methods: Young adult Sprague-Dawley rats were randomly assigned to one of four treatment groups including normal activity (NA) or PA for 3 months and then subjected to either normoxia (RA) or exposure to IH during the light phase during the last 14 days.

Measurements and main results: Significant impairments in IH-exposed rats emerged on both latency and pathlength to locate the hidden platform in a water maze and decreased spatial bias during the probe trials. These impairments were not observed in PA-IH rats. In addition, the PA-IH group, relative to NA-IH, conferred greater resistance to both lipid peroxidation and 8-hydroxy-2'-deoxyguanosine (DNA damage) in both the cortex and hippocampus. In support of a neuroprotective effect from PA, PA-IH versus NA-IH rats showed greater AKT activation and neuronal insulin growth factor-1 in these regions.

Conclusions: Behavioral modifications such as increased physical activity are associated with decreased susceptibility to IH-induced spatial task deficits and lead to reduced oxidative stress, possibly through improved preservation of insulin growth factor-1-Akt neuronal signaling. Considering the many advantages of PA, interventional strategies targeting behavioral modifications leading to increased PA should be pursued in patients with sleep-disordered breathing.

Figure 1.

Mean changes in body weights of physically active (PA) and non–physically active (NA) rats (n = 24 per group, *P < 0.01).

Figure 2.

(A) Mean latencies (s) and pathlengths (cm) to locate the target platform during place training in physically active (PA) and non–physically active (NA) rats either exposed to intermittent hypoxia (IH) or maintained in room air (RA) (n = 24 per group, *P < 0.05). (B) Mean latencies (s) to locate the target platform during cued training in PA and NA rats either exposed to IH or maintained in RA (n = 24 per group). (C) Mean probe trial pathlength in the target quadrant after completion of water maze testing in PA and NA rats either exposed to IH or maintained in RA (n = 24 per group, *P < 0.01).

Figure 3.

Relative fold increase in malondialdehyde (MDA) production, an indicator of lipid peroxidation, in cerebral cortex of physically active (PA) and non–physically active (NA) rats exposed to IH (n = 6 per group, *P < 0.05, PA vs. NA). MDA ratios of IH:RA in PA and NA groups are shown.

Figure 4.

(A) Hippocampal and cortical sections illustrating 8-OHDG labeling, a marker of DNA oxidative damage, in physically active (PA) and non–physically active (NA) rats exposed to intermittent hypoxia (IH). (B) Mean counts of 8-OHDG positively labeled cells in hippocampus of PA and NA rats exposed to IH or room air (RA) (n = 6 per experimental group; 10 sections were counted by blinded investigator per animal). Arrows indicate examples of positively labeled cells. Cx = cortex; HPC = hippocampus.

Figure 5.

(Top panel) Western blot of phospho-AKT and total AKT in hippocampal lysates from rats exposed to physical activity (PA) or no physical activity (NA) and to intermittent hypoxia (IH) or room air (RA). (Left bottom panel) Mean ratio of phospho-AKT to total AKT in lysates of the hippocampal CA1 region of PA and NA rats exposed to RA or IH, (n = 6 per group). (Right bottom panel) Akt activity in lysates of the hippocampal CA1 region of PA and NA rats exposed to RA or IH (n = 6 per group).

Figure 6.

(A) Representative double-labeled immunohistochemical staining of hippocampal sections for insulin growth factor (IGF)-1 (red) and Neu-N (green) in non–physically active (NA; upper panel) and physically active (PA; lower panel) rats exposed to intermittent hypoxia (IH). (B) Upper panels: Higher magnification of a representative double-labeled immunohistochemical staining of hippocampal sections for IGF-1 (red) and NeuN (green) in PA and NA rats exposed to IH. Bottom panel: Representative double-labeled immunohistochemical staining of hippocampal sections for IGF-1 (red) and glial fibrillary acidic protein (GFAP) (green), showing that IGF-1 is preferentially expressed in neurons and not in glia. Similar findings were consistently found in four rats in each experimental group.