Circulating insulin-like growth factor I mediates effects of exercise on the brain

Abstract

Physical exercise increases brain activity through mechanisms not yet known. We now report that in rats, running induces uptake of blood insulin-like growth factor I (IGF-I) by specific groups of neurons throughout the brain. Neurons accumulating IGF-I show increased spontaneous firing and a protracted increase in sensitivity to afferent stimulation. Furthermore, systemic injection of IGF-I mimicked the effects of exercise in the brain. Thus, brain uptake of IGF-I after either intracarotid injection or after exercise elicited the same pattern of neuronal accumulation of IGF-I, an identical widespread increase in neuronal c-Fos, and a similar stimulation of hippocampal brain-derived neurotrophic factor. When uptake of IGF-I by brain cells was blocked, the exercise-induced increase on c-Fos expression was also blocked. We conclude that serum IGF-I mediates activational effects of exercise in the brain. Thus, stimulation of the uptake of blood-borne IGF-I by nerve cells may lead to novel neuroprotective strategies.

Figures

Fig. 1.

Physical exercise and intracarotid injection of IGF-I produce similar effects in the brain. A, The same brain areas show labeling of neurons with IGF-I after treadmill running (a–c) and intracarotid injection of IGF-I (d–f). Three representative areas are shown. Nonexercised, saline-injected rats show almost undetectable brain IGF-I staining (B). Str, Striatum;Cx, cerebral cortex; RN, red nucleus. Biotinylated anti-rabbit IgG followed by Cy3-streptavidin was used after incubation with a polyclonal anti-IGF-I antibody.B, Digoxigenin (DIG) and IGF-I colocalize within the same neurons after intracarotid injection of DIG–IGF-I. A representative field in the brainstem is shown. a, Low magnification (10×) of IGF-I staining in the inferior olive nucleus (IO) of a saline-injected rat. Note the absence of signal. Inset, Higher magnification (40×) of the IO field. b, The same field showing IGF-I staining in an IGF-I-injected rat. Inset, High magnification showing IGF-I-positive cells. c, High magnification (40×) of IO neurons stained with a monoclonal anti-DIG antibody (green). d, The same field stained with a polyclonal anti-IGF-I antibody (red).e, Colocalization of DIG and IGF-I within the same IO neurons. Scale bars: a, b, 500 μm;c–e, 50 μm. Primary antibody incubation was followed by an anti-rabbit Cy2 and anti-mouse Cy5, respectively. C, Exercise or intracarotid injection of IGF-I elicits a similar pattern of increased c-Fos staining throughout the brain. Only the piriform cortex (Pir) is shown as a representative area. a, Control animals show no c-Fos staining. b, c-Fos staining after 1 hr of intracarotid injection of IGF-I. c, c-Fos staining after 1 hr of running. Scale bar (a–c): 500 μm.d, Higher magnification of the field in cshowing nuclear localization of the c-Fos signal. Scale bar, 50 μm.Arrows indicate immunoreactive cells. A monoclonal anti-c-Fos antibody followed by a biotinylated anti-mouse IgG and Cy3-streptavidin was used. D, Blockade of the exercise-induced capture of IGF-I by brain cells results in absence of c-Fos labeling after exercise. a, IGF-I labeling in the hippocampus of a rat that ran for 1 hr. c, Chronic intracerebroventricular delivery of a combination of an anti-IGF-I antibody and an IGF-I receptor antagonist results in absence of IGF-I staining after 1 hr of running exercise. Scale bar (a,c): 50 μm. b, c-Fos staining is induced in the hippocampus by 1 hr of running. c, No c-Fos labeling is seen in exercised animals in which brain uptake of IGF-I is blocked. Scale bar (b, d): 500 μm. The hippocampus is shown as a representative area, but absence of labeling for IGF-I and c-Fos was found in all brain areas. E, Expression of BDNF in the hippocampus is increased by running and by intracarotid injection of IGF-I. Control: background BDNF RNA staining in brain slices incubated with excess unlabeled probe. Saline: animals injected with saline show weak BDNF expression in the hippocampus. Exercise: running induces increased expression of BDNF in the hippocampus. IGF-I: injection of IGF-I produces a similar increase in hippocampal expression of BDNF. Cx, Cortex;Hy, hippocampus.

Fig. 2.

Blood-borne IGF-I is taken up by brain cells. A, Calbindin and IGF-I colocalize within neurons after intracarotid injection of DIG–IGF-I. A representative field in the cerebellar cortex is shown. a, Purkinje neurons stained with a polyclonal anti-IGF-I antibody (red).b, The same field stained with a monoclonal anti-calbindin antibody (green).c, Colocalization of calbindin and IGF-I within cerebellar Purkinje cells. Primary antibody incubation was followed by an anti-rabbit Cy2 and anti-mouse Cy5, respectively. B, Uptake of DIG–IGF-I from serum is abolished by coinjection of excess unlabeled IGF-I but not by excess insulin. a, Inferior olive (IO) neurons of a rat after intracarotid injection with DIG–IGF-I (10 μg). b, IO neurons of an animal injected with DIG–IGF-I (10 μg) plus unlabeled IGF-I (100 μg) show almost no immunoreactivity for digoxigenin. c, Coinjection of insulin (2 mg) with DIG–IGF-I (10 μg) results in partial displacement of digoxigenin label. Intracarotid injections were performed 1 hr earlier. C, Accumulation of DIG–IGF-I by the brain is time-dependent. A representative field of the cerebellar cortex is shown. a, No digoxigenin immunoreactivity is present in a saline-injected rat. Purkinje cells (PC) in the cerebellar cortex stained for digoxigenin 5 min (b), 1 hr (c), and 3 hr (d) after intracarotid injection of DIG–IGF-I; 6 hr (e) later, staining was absent.GL, Granule cell layer; ML, molecular layer. Scale bars, 50 μm. Polyclonal anti-DIG antibody followed by biotinylated anti-rabbit IgG and Cy3-streptavidin was used.D, Brain levels of IGF-I are increased after exercise (p < 0.05) or after intracarotid injection of IGF-I (p = 0.05), as compared with control levels.

Fig. 3.

Serum IGF-I enters into the brain through the blood–CSF pathway. A, Epithelial cells of the choroid plexus (CP) accumulate DIG–IGF-I and show IGF-I receptor immunoreactivity. a, Digoxigenin labeling of CP cells in the lateral cerebral ventricle (LV) of a rat injected with DIG–IGF-I 5 min before. b, Ependymal cells (EC) lining the wall of the ventricle show digoxigenin staining 1 hr after injection of DIG–IGF-I.c, IGF-I receptor immunoreactivity in CP cells.B, Levels of immunoreactive IGF-I in the CSF were significantly increased 15 min after intracarotid injection of the peptide. *p = 0.03 versus saline-injected control rats by Whitney test (n = 3). C, Injection of excess unlabeled IGF-I into the CSF displaces the uptake of serum DIG–IGF-I. Cerebellar Purkinje cells (a) or IO neurons (c) stained with DIG–IGF-I 1 hr after intracarotid injection. DIG–IGF-I staining of Purkinje cells (b) or IO neurons (d) is absent after intracerebroventricular injection of 100 μg unlabeled IGF-I. D, The pattern of neuronal uptake of IGF-I is similar after either intracerebroventricular (a, b) or intracarotid injection of DIG–IGF-I (c,d). Arrows indicate examples of positive cells. Scale bar, 50 μm. Antibodies that were used are as in previous figures.

Fig. 4.

Intracarotid injection of IGF-I elicits long-lasting electrophysiological changes in cerebellar Purkinje cells.A1, Increase over time of mean spontaneous activity of Purkinje cells after intracarotid injection of IGF-I (n = 11). A2, Control saline injection did not modify the firing rate over time (n = 3). B, Evoked field potentials in the cerebellum after parallel fiber stimulation. Two negative waves with short latencies were elicited. Arrowheads indicate the trace showing increased evoked potential 15 min after injection of IGF-I. C, Mean area of the first (1) and second (2) negative waves increase after IGF-I injection (n = 11). Baseline (control) levels were obtained before intracarotid injection of IGF-I. *p < 0.05.

Fig. 5.

Intracarotid injection of IGF-I elicits electrophysiological changes in brainstem DCN neurons.A, DCN cells show intense IGF-I labeling after intracarotid injection of IGF-I. Similar IGF-I staining of these neurons is seen after exercise. Arrow indicates an IGF-I-accumulating neuron. Scale bar, 50 μm. B1, Mean spontaneous activity of DCN neurons increases 15–30 min after IGF-I injection and tends toward baseline levels 1 hr later.B2, Control saline injection did not modify the firing rate over time (n = 4). C, PSTHs (20 stimuli) of a representative cuneate nucleus cell before (1) and after (2) 15 min of intracarotid injection of IGF-I. The neuron responded to somatosensory stimulation delivered on the second digit of the hindpaw. Note the increase in the number of stimulus-evoked spikes after IGF-I.D, The number of spikes evoked by each somatosensory stimulus increased over time after injection of IGF-I (n = 12), indicating increased effectiveness of sensory stimulation. Control refers to values before injection of IGF-I. *p < 0.05; **p < 0.01.