Aerobic exercise reverses aging-induced depth-dependent decline in cerebral microcirculation

Abstract

Aging is a major risk factor for cognitive impairment. Aerobic exercise benefits brain function and may promote cognitive health in older adults. However, underlying biological mechanisms across cerebral gray and white matter are poorly understood. Selective vulnerability of the white matter to small vessel disease and a link between white matter health and cognitive function suggests a potential role for responses in deep cerebral microcirculation. Here, we tested whether aerobic exercise modulates cerebral microcirculatory changes induced by aging. To this end, we carried out a comprehensive quantitative examination of changes in cerebral microvascular physiology in cortical gray and subcortical white matter in mice (3-6 vs. 19-21 months old), and asked whether and how exercise may rescue age-induced deficits. In the sedentary group, aging caused a more severe decline in cerebral microvascular perfusion and oxygenation in deep (infragranular) cortical layers and subcortical white matter compared with superficial (supragranular) cortical layers. Five months of voluntary aerobic exercise partly renormalized microvascular perfusion and oxygenation in aged mice in a depth-dependent manner, and brought these spatial distributions closer to those of young adult sedentary mice. These microcirculatory effects were accompanied by an improvement in cognitive function. Our work demonstrates the selective vulnerability of the deep cortex and subcortical white matter to aging-induced decline in microcirculation, as well as the responsiveness of these regions to aerobic exercise.

[Figure 1]

Animal preparation and experiment design. (a) Timeline of the study. Optical measurements and behavioral testing were performed after 5 months of voluntary exercise when the animals were 19–21 months of age. (b) Average daily running distance in km per day for each mouse in the exercise group, calculated as the sum of daily running distance divided by a total running period of 5 months. Data are shown as mean ± SD. (c) Daily running distance for two representative mice in the exercise group across time.

[Figure 2]

Experimental setup and imaging protocols. (a) Schematic of our home-built multimodal optical system featuring primary components of the system. A 50 kHz spectral-domain OCT system was designed to partially share the imaging optics with the two-photon microscope. A Hg:Xe arc lamp in combination with a CCD camera was used for OISI. EOM: electro-optic modulator, SH: shutter, GM: galvanometer mirror pair. (b) A CCD image of brain surface vasculature in the mouse barrel cortex showing the ROIs where various optical measurements were performed (red ROI: capillary RBC flux, pO₂, and microvasculature imaging, blue ROI: Doppler OCT imaging, and green ROI: OCT intensity imaging). (c) A representative OCT intensity B-scan image extracted from a volumetric OCT image. White arrows indicate the boundary between the gray matter (GM) and the corpus callosum (CC), which appears as a bright band in the image. (d) Survey scan images of cerebral microvasculature of the region outlined with the red square in (b) obtained by two-photon microscope at two imaging depths (z=0.4 and 0.9 mm). Two representative fluorescent intensity time courses acquired within the capillaries at the locations indicated by the red dots in the survey angiograms are presented on the right. (e) A 3D angiogram of the mouse cortex acquired by the two-photon microscope at the location outlined by the red square in (b). One representative 2D plane from the angiogram acquired at a depth of 200 μm showing pO₂ measurements from different capillary segments. pO₂ values were color-coded (in mmHg) and spatially co-registered with the angiogram. (f) A 3D Doppler OCT image showing an axial velocity map of the diving vessels at the location outlined by the blue square in (b). (g) An OIS image of the cranial window obtained by calculating the relative intensity difference between the post-stimulus response image and pre-stimulus baseline. The region of activation is manually selected from the OIS image as indicated by a black square. The lower panel shows a time course of the relative intensity change due to sensory-evoked hemodynamic response induced by a 2-s-long whisker stimulation, averaged over the selected region of interest. Scale bars: 400 μm for (b) and (c), 100 μm for (d), (e), and (f), and 500 μm for (g).

[Figure 3]

Aging and exercise induced alterations in cerebral microcirculation. (a) Mean capillary RBC flux in young-adult sedentary mice. The data are from 226 and 218 capillaries in three young sedentary mice in the gray matter (GM) and white matter (WM), respectively. Comparison was made between young and aged sedentary group (shown in (b)) in each gray (layers II/III, IV in aged sedentary group) and white matter region. (b) Capillary RBC flux across cortical layers II/III and IV, and subcortical white matter in aged sedentary and exercise group. (c) Histograms of capillary RBC flux in the gray and white matter in each animal group. (d) The coefficient of variance (CV) of capillary RBC flux across cortical layers II/III and IV, and subcortical white matter in sedentary and exercise group. (e) Venular flow versus vessel diameter. Different symbols represent different animals. The red dashed and blue solid line is the best fit result of each linear regression for sedentary and exercise group, respectively. (f) Mean venular flow in ascending venules in (d) in sedentary and exercise group. The data in (b-d) are from 921, 486, and 112 capillaries in 7 mice in the sedentary group and 1046, 465, and 238 capillaries in 8 mice in the exercise group, in cortical layers II/III, IV, and the white matter, respectively. The data in (e) and (f) are from 14 and 7 ascending venules in 9 and 6 mice in the sedentary and exercise group, respectively. Statistical analysis was carried out using Two-way ANOVA with post hoc Tukey’s in (a,b) and (d) and Student’s t-test in (f). *p<0.05; **p<0.01. Additional details on boxplots and animals excluded from the analyses are provided in the Supplementary document.

[Figure 4]

Exercise-induced changes in microvascular pO₂ across cortical layers in old mice. (a) Capillary mean-pO₂ across cortical layers in sedentary controls and exercising mice. (b) Histograms of capillary pO₂ in layers I, II/III, and IV. (c) The mean Hct levels from sedentary (n=8) and exercise group (n=5). (d) and (e) Intravascular pO₂ and SO₂ in the diving arterioles across cortical layers I-IV in sedentary (blue boxplots) and exercise (red boxplots) group, respectively. (f) and (g) Intravascular pO₂ and SO₂ in the surfacing venules across cortical layers I-IV in sedentary (blue boxplots) and exercise (red boxplots) group, respectively. (h) Depth-dependent OEF in sedentary (blue boxplots) and exercise (red boxplots) group. The analysis in (a) and (b) was made with 1224, 2601, and 922 capillaries across n=9 mice in sedentary group and 1334, 2840, and 1078 capillaries across n=9 mice in exercise group in cortical layers, I, II/III, and IV, respectively. The analysis in (d-h) was made with 13 arterioles and 12 venules from n=9 mice in sedentary group and 14 arterioles and 12 venules from n=9 mice in exercise group. Statistical analysis was carried out using Two-way ANOVA with post hoc Tukey’s in (a) and (d-h) and Student’s t-test in (f). *p<0.05; **p<0.01. Additional details on boxplots and exclusions are provided in the Supplementary document.

[Figure 5]

Effects of aging and exercise on functional hemodynamic response. (a, b) Optical intrinsic signal time courses in the whisker barrel cortex of individual old mice in aged sedentary (blue; n=9) and exercise (red; n=8) mice (a), and younger (7 months old) sedentary mice (n=8) (b). Thick curves represent averages. (c) Average changes in the peak intensity in old sedentary, old exercise, and young group. One-way ANOVA with Tukey post hoc test. *** p<0.001. Please see Supplementary document for exclusions.

[Figure 6]

Cortical microvascular density in aged mice. (a) Representative MIP images of the three-dimensional angiograms of three mice in the sedentary group and three mice in the exercise group, over the cortical depth range from 50 μm to 400 μm, and 200 × 200 μm² FOV. Scale bars: 50 μm. (b) Vessel segment density and (c) vessel length density of cortical capillaries from sedentary (n=19,669 segments; n=9 mice) and exercise (n=18,044 segments; n=8 mice) group. Student’s t-test. **p<0.01; *** p<0.001. Please see Supplementary document for exclusions.

[Figure 7]

Effect of exercise in the old mice on cognitive performance. (a) DI scores in NORT, calculated with four different exploration time periods in the sedentary (n=9) and exercise (n=7) group. The calculated DI values at each time interval were subsequently averaged across animals. (b-d) Correlations between the daily average running distance and four different DI scores: 1, 2, 3, and 4 minutes, respectively. (f) Spontaneous alteration scores in the Y-maze test in the sedentary (n=9) and exercise (n=7) group. Statistical analysis was carried out using Two-way ANOVA with post hoc Tukey’s in (a) and Student’s t-test in (f). **p<0.01. Please see Supplementary document for exclusions.