β2-microglobulin is a systemic pro-aging factor that impairs cognitive function and neurogenesis

Abstract

Aging drives cognitive and regenerative impairments in the adult brain, increasing susceptibility to neurodegenerative disorders in healthy individuals. Experiments using heterochronic parabiosis, in which the circulatory systems of young and old animals are joined, indicate that circulating pro-aging factors in old blood drive aging phenotypes in the brain. Here we identify β2-microglobulin (B2M), a component of major histocompatibility complex class 1 (MHC I) molecules, as a circulating factor that negatively regulates cognitive and regenerative function in the adult hippocampus in an age-dependent manner. B2M is elevated in the blood of aging humans and mice, and it is increased within the hippocampus of aged mice and young heterochronic parabionts. Exogenous B2M injected systemically, or locally in the hippocampus, impairs hippocampal-dependent cognitive function and neurogenesis in young mice. The negative effects of B2M and heterochronic parabiosis are, in part, mitigated in the hippocampus of young transporter associated with antigen processing 1 (Tap1)-deficient mice with reduced cell surface expression of MHC I. The absence of endogenous B2M expression abrogates age-related cognitive decline and enhances neurogenesis in aged mice. Our data indicate that systemic B2M accumulation in aging blood promotes age-related cognitive dysfunction and impairs neurogenesis, in part via MHC I, suggesting that B2M may be targeted therapeutically in old age.

Figure 1. Systemic B2M increases with age and impairs hippocampal-dependent cognitive function and neurogenesis

a,b, Schematic of unpaired young versus aged mice (a), and young isochronic versus heterochronic parabionts (b). Changes in plasma concentration of B2m with age at 3, 6, 12, 18 and 24 months (a) and between young isochronic and young heterochronic parabionts five weeks after parabiosis (b). Data from 5 mice per group. c,d, Changes in plasma (c; r=0.51; p<0.0001; 95% confidence interval=0.19–0.028) and CSF (d) B2M concentrations with age in healthy human subjects. Data from 318 individuals (c), 8 young (20–45) individuals (d), and 22 old (65–90) individuals (d). e, Young adult (3 months) mice were injected intraorbitally with B2M or PBS (vehicle) control five times over 12 days. e, Schematic of chronological order used for B2M treatment and cognitive testing. e, Hippocampal learning and memory assessed by RAWM (number of entry arm errors prior to finding platform) and contextual fear conditioning (percent freezing time 24h after training). Data from 10 mice per group. f, Representative field of Dcx-positive cells for each treatment group (scale bar: 100μm). g, Quantification of neurogenesis in the dentate gyrus (DG) after treatment. Data from 7 B2M treated and 8 vehicle treated mice. All data represented as dot plots with Mean or bar graphs with Mean±SEM; *P < 0.05; **P < 0.01; ***P < 0.001 t-test (b,d,e,g), ANOVA, Tukey’s post-hoc test (a), Mann-Whitney U Test (c) and repeated measures ANOVA, Bonferroni post-hoc test (e).

Figure 2. B2M expression increases in the aging hippocampus and impairs hippocampal-dependent cognitive function and neurogenesis

a,b, Representative Western blot and quantification of hippocampal lysates probed with anti-B2m and anti-Actin antibodies from young (3 months) and aged (18 months) unpaired animals (a), or young isochronic and young heterochronic parabionts five weeks after parabiosis (b). c–e, Young (3 months) adult wild type (WT; c,d) or Tap1-deficient (e) mice were given bilateral stereotaxic injections of B2M or vehicle six days (c,e) or 30 days (d) prior to behavioral testing. c–e, Schematics illustrating chronological order used for local B2M administration and cognitive testing. c–e, Learning and memory as assessed by RAWM and contextual fear conditioning following stereotaxic injections. Data from 10 animals per genotype and treatment group. f–i, Young adult (3 months) WT and Tap1^−/− mice were given unilateral stereotaxic injections of B2M or vehicle control. f, Schematic illustrating injection paradigm. g, Representative field of Dcx-positive cells in adjacent sides of the DG within the same section are shown for WT and Tap1^−/− treatment groups. h,i, Quantification of neurogenesis in the DG of WT (h) and Tap1^−/− (i) mice after stereotaxic B2M administration. Data from five mice per genotype and treatment group. All data represented as Mean±SEM; *P<0.05; **P<0.01; n.s. not significant; ANOVA, t-test (a–e,h,i); repeated measures ANOVA, Bonferroni post-hoc test (c–e).

Figure 3. Reducing MHC I surface expression mitigates the negative effects of heterochronic parabiosis on neurogenesis

a, Schematic of young (3 months) wild type (WT) and Tap1^−/− isochronic parabionts and young WT and Tap1^−/− heterochronic parabionts. b,c Representative images of the dentate gyrus (b) and quantification (c) of Doublecortin immunostaining of young isochronic and heterochronic parabionts five weeks after parabiosis (arrowheads point to individual cells, scale bar: 100μm). d, Prior to euthanasia animals were injected with Bromodeoxyuridine (BrdU) for three days, and proliferating cells having incorporated BrdU were quantified in DG after parabiosis. Data from 8 young isochronic WT, 6 young isochronic Tap1^−/−, 8 young heterochronic WT, and 8 young heterochronic Tap1^−/− parabionts. All data represented as Mean±SEM; *P< 0.05; ANOVA, Tukey’s post-hoc test.

Figure 4. Absence of B2m enhances hippocampal-dependent cognitive function and neurogenesis in aged animals

a–d, Learning and memory in young (3 months) and aged (17 months) wild type (WT) and B2m knockout (B2m^−/−) mice by RAWM (a,c) and contextual fear conditioning (b,d). Data from 10 young WT, 10 young B2m^−/−, 8 aged WT, and 12 aged B2m^−/− mice. e–j, Neurogenesis as assessed by immunostaining for Dcx-positive cells in the DG of young and aged WT and B2m^−/− mice. Representative field and quantification of Dcx-positive cells are shown for young (e,f) and aged (e,g) WT and B2m^−/− animals (arrowheads point to individual immature neurons, scale bar: 100μm). Data from 8 young and 10 aged mice per genotype. h–j, WT and B2m^−/− mice were administered BrdU by intraperitoneal injections for six days and euthanized 28 days later. h, Representative confocal images from the DG of brain sections immunostained for BrdU (red) in combination with NeuN (green; scale bar: 25μm). i,j, Quantification of the relative number of BrdU and NeuN-double positive cells out of the total BrdU-positive cells in the young (i) and aged (j) DG of WT and B2m^−/− animals. Data from 8 mice per group (3 sections per mouse). All data represented as Mean±SEM; *P< 0.05; **P<0.01; n.s. not significant; t-test (b,d,f,i,j); repeated measures ANOVA, Bonferroni post-hoc test (a,c).