Insulin-Like Growth Factor (IGF)-I Modulates Endothelial Blood-Brain Barrier Function in Ischemic Middle-Aged Female Rats

Abstract

In comparison with young females, middle-aged female rats sustain greater cerebral infarction and worse functional recovery after stroke. These poorer stroke outcomes in middle-aged females are associated with an age-related reduction in IGF-I levels. Poststroke IGF-I treatment decreases infarct volume in older females and lowers the expression of cytokines in the ischemic hemisphere. IGF-I also reduces transfer of Evans blue dye to the brain, suggesting that this peptide may also promote blood-brain barrier function. To test the hypothesis that IGF-I may act at the blood-brain barrier in ischemic stroke, 2 approaches were used. In the first approach, middle-aged female rats were subjected to middle cerebral artery occlusion and treated with IGF-I after reperfusion. Mononuclear cells from the ischemic hemisphere were stained for CD4 or triple-labeled for CD4/CD25/FoxP3 and subjected to flow analyses. Both cohorts of cells were significantly reduced in IGF-I-treated animals compared with those in vehicle controls. Reduced trafficking of immune cells to the ischemic site suggests that blood-brain barrier integrity is better maintained in IGF-I-treated animals. The second approach directly tested the effect of IGF-I on barrier function of aging endothelial cells. Accordingly, brain microvascular endothelial cells from middle-aged female rats were cultured ex vivo and subjected to ischemic conditions (oxygen-glucose deprivation). IGF-I treatment significantly reduced the transfer of fluorescently labeled BSA across the endothelial monolayer as well as cellular internalization of fluorescein isothiocyanate-BSA compared with those in vehicle-treated cultures, Collectively, these data support the hypothesis that IGF-I improves blood-brain barrier function in middle-aged females.

Figure 1.

IGF-I regulates immune cell infiltration into ischemic cortex. A, Transient ischemia (90 minutes)/reperfusion (48 hours) resulted in an increased proportion of CD4⁺ cells in the ischemic cortex of vehicle-treated middle-aged females (top left plot) compared with those of IGF-I–treated females (top right plot). B, IGF-I significantly reduced the ratio of CD4⁺ cells in the ischemic vs nonischemic hemisphere compared with those in vehicle-treated animals. C and D, the ratio of Treg cells (cells stained positive for CD4/CD25/FoxP3) was also significantly decreased in the IGF-I–treated group compared with that in controls, suggesting IGF-I–mediated reduction in blood-brain barrier permeability in the ischemic cortex. n = 5 to 6; *, P < .05.

Figure 2.

Characterization of BMECs from adult rat brain cortex. A, BMECs were immuno-labeled for the endothelial marker factor VIII (vWF). Virtually all cells were immunopositive for factor VII (red; ai), whereas no label was seen in cultures in which the primary antibody was omitted (aii). BMECs were counterstained with the nuclear dye, Hoechst dye, seen in blue. B, Cultures were exposed to fluorescence-labeled Ac-LDL in medium, which was internalized by all cells, and is visible here as red punctate labeling. C, BMECs were not immunopositive for the astrocytic marker GFAP (ci) or the microglial marker Iba-1 (cii), although <0.002% cells were positively labeled for α-smooth muscle actin, a pericyte marker (ciii). Cells were counterstained with a nuclear dye shown in blue. Bar, 20 μm (A and B); 50 μm (C).

Figure 3.

Effect of OGD, OGD and reoxygenation, and IGF-I on BMECs from middle-aged females. A, Compared with normoxic controls (ai), OGD resulted in morphological changes in endothelial cells including cell elongation, shrunken cell profiles, and reduced intercellular contact (aii). IGF-I–treated cultures showed less cellular elongation and fewer intercellular spaces indicating, greater cell-cell communication after OGD (aiii). Reoxygenation did not improve cell morphology and cell contacts in either untreated (av) or IGF-I–treated cultures (avi). B, LDH was measured in BMEC cultures exposed to IGF-I or vehicle after OGD and OGD and reoxygenation (Re-OX). OGD significantly increased medium LDH levels. However, IGF-I did not attenuate OGD-induced increase in LDH levels. LDH levels after reoxygenation were no different in vehicle and IGF-I–treated cultures. The data shown represent 3 separate runs, with each run consisting of cells derived from 4 animals. *, P < .05 (compared to normoxia). Bar, 50 μm.

Figure 4.

OGD and reoxygenation induced changes in barrier properties of endothelial cells. A, OGD induced changes in FITC-BSA transfer: FITC-BSA transport was measured across the monolayer of cells after OGD and reoxygenation. FITC-BSA transfer was significantly lower in cultures treated with IGF-I compared with vehicle treatment. B, FITC-BSA in the endothelial monolayer was normalized to nuclear dye (Hoechst) to determine the extent of BSA that was internalized by endothelial cells. There was a 50% decrease in the FITC-BSA/Hoechst ratio in IGF-I–treated cultures, suggesting reduced cellular uptake compared with that for vehicle treatment. The data represent the average from 3 independent runs, each run consisting of cells derived from 4 middle-aged females. *, P < .05.