Article Figures & Data
Figures
- Figure 1.
Running increases old adult-born neuron survival and modifies their afferent input in middle-aged mice. A, Timeline of the experiment. Adult young control (CON; n = 4) and runner (RUN; n = 4) mice (1.5 months old) were injected with retrovirus expressing nuclear GFP, avian TVA receptor, and rabies glycoprotein (RV-SYN-GTRgp; first injection) into the dentate gyrus (DG) to label proliferating neural progenitor cells. Mice were housed in either control or voluntary running conditions for six to nine months. Thereafter, mice were injected with EnvA-pseudotyped rabies virus expressing MCh (EnvA-ΔG-MCh, second injection) into the same sites to trace the inputs to the old adult-born neurons. Double infected (GFP+ + MCh+ = yellow nuclei) cells are adult-born neurons from which the tracing originated, termed “starter cells,” while cells synaptically connected to old adult-born neurons expressing MCh (red) only are termed “traced cells.” B, C, Photomicrographs of the dorsal DG showing starter (⬆) and traced («) cells in hippocampal sections derived from (B) CON and (C) RUN mice. Nuclei were stained with DAPI (blue). Scale bar: 50 μm. D, Long-term running increases the number of starter cells (t(6) = 8.514; p = 0.0001). E, Distribution analysis showed that running increased starter cell number in the dorsal (dDG) but not intermediate (iDG) or ventral (vDG) dentate gyrus (F(2,12) = 25.04, p < 0.0001). F, The total traced cell number is increased in RUN as compared with CON mice (t(6) = 3.676, p = 0.0104). G, The connectivity index measured as the traced/starter cell ratio is significantly reduced by long-term running (t(6) = 2.790, p = 0.0316). Statistical scores are listed in Extended Data Table 1-1. See Extended Data Figure 1-1 for distance run, and GFP+ adult-born neuron numbers. Data are means ± SEM *p < 0.05, **p < 0.0001. H, hilus; GCL, granule cell layer; ML, molecular layer.
- Figure 2.
Running increases inhibitory input onto old adult-born neurons. A, Schematic representation of the hippocampal areas and layers. ML, molecular layer; GCL, granule cell layer; L-M, stratum lacunosum moleculare; Rad, stratum radiatum; Luc, stratum lucidum; Oriens, stratum oriens. Pyr, stratum pyramidale. B, Long-term running increases the total number of traced interneurons (INT; t(6) = 3.258, p = 0.0173), but not the number of mature granule cells (mGC; t(6) = 0.8441, p = 0.4310), mossy cells (MC; t(6) = 1.476, p = 0.1904), pyramidal cells (PYR; t(6) = 1.348, p = 0.2262), or astrocytes (AS; t(6) = 0.1345, p = 0.8974). See Extended Data Figure 2-1 for the innervation ratio and subregional distribution of mGC, MC, and AS. C, Percentage of traced INT per hippocampal area. D, Photomicrograph of the dentate gyrus (DG) showing traced INT and MC. INT (⬆) colabeling MCh+ + GABA (red + green; insets). Traced mossy cell expressing MCh only (*). E, Traced INT (⬆) located in the molecular layer (ML) of the dentate gyrus. Inset, Overview of the dentate gyrus. F, Photomicrographs of CA3 area showing mossy fibers (MF) and traced INT located in the stratum oriens (*), lucidum (#), and radiatum (⬆) expressing MCh. Inset, Overview of CA3 area. G, Photomicrograph of area CA1 showing traced INT located in the stratum lacunosum-moleculare (⬆). Inset, Overview of CA1 area. H, Subfield analysis of traced INT number revealed an increase in RUN versus CON in the DG, areas CA3 and CA1, but not in CA2 (F(3,18) = 7.005, p < 0.0026). I, The ratio of connectivity between INT and starter cells (INT/SC) was not modified by running (F(1,6) = 2.74, p > 0.14). DG INT/SC ratio was higher than that of the other subfields (F(3,18) = 6.5, p < 0.036). J, The majority of traced INT are located in the dorsal hippocampus in both CON and RUN mice. K, Percentage of traced INT in the DG, area CA3 and CA1 per layer. L, Analysis within each subfield shows that running increased traced INT number in the DG (F(1,6) = 8.13, p < 0.029), area CA3 (F(1,6) = 11.79, p < 0.013), and area CA1 (F(1,6) = 11.33, p < 0.015). D, Scale bar: 20 μm; inset, 10 μm. Photomicrographs (E–G): Scale bar: 50 μm; inset, 100 μm. Nuclei were stained with DAPI (blue). Data are means ± SEM *p < 0.05.
- Figure 3.
Running increases the connectivity of dorsal CA3 PYR onto old adult-born neurons. A, The percentage of traced pyramidal cells (PYR) was highest in the CA3 area, followed by CA1 and a low percentage in CA2, in both control (CON) and runner mice (RUN). B, E, Photomicrographs of dorsal hippocampal horizontal sections derived from CON (B) and RUN (E) mice show MCh+ PYR (red) located in the CA3 area (arrowhead). C, F, Higher magnification of MCh+ PYR in CA3 area from panels (B) and (E), respectively. D, Schematic representation of a dorsal horizontal brain section (adapted from Paxinos and Franklin, 2007) showing the location of CA3–CA1 areas. G, Long-term running does not modify the number of traced PYR cells in areas CA3–CA1 (F(1,6) = 1.84, p > 0.22). H, Percentage of traced PYR in the dorsal, intermediate (inter), and ventral hippocampus in CA3–CA1 areas. I, Dorsal-ventral distribution of analysis of area CA3 revealed that long-term running increases the number of traced PYR in the dorsal area CA3 (F(2,12) = 5.37, p < 0.021). There were no changes in the dorso-ventral distribution in area CA2 (F(2,12) = 0.97, p > 0.40) or area CA1 (F(2,12) = 2.20, p > 0.15). Data are means ± SEM *p < 0.05. Scale bar: 100 μm. Nuclei were stained with DAPI (blue).
- Figure 4.
Running recruits noncanonical subicular inputs to old adult-born neuron network. A–C, Schematic representations of horizontal slices of the hippocampal formation showing the location of traced cells in the subiculum (SUB; A, B; left) and parasubiculum (PARA, C; left; red dots). Photomicrographs of the SUB (A, B, right) and PARA (C, right) areas show traced cells expressing MCh. Insets: An overview of the hippocampal formation showing the location of the traced cells of the subicular complex (⬆). Scale bar: 50 μm; inset, 200 μm (A, B) and 100 μm (C). Nuclei were stained with DAPI (blue). D, Long-term running increases the total number of traced cells in the SUB-C (t(6) = 3.979, p = 0.0073). E, Percentage of traced cells in subicular complex areas [SUB, PARA, and presubiculum (PRE)] in control and long-term running mice. F, Long-term running increased SUB traced cell number (F(2,12) = 9.42, p < 0.0035). G, The highest percentage of traced cells are located in the dorsal SUB. H, Distribution analysis revealed increased traced cell number in the dorsal SUB with running (F(2,12) = 9.28, p < 0.0037). I, The ratio of connectivity between SUB and starter cells (SUB/SC) was not modified by running (t(6) = 1.388, p = 0.2145). Data are means ± SEM *p < 0.05.
- Figure 5.
Running preserves PRh input and modifies the distribution of entorhinal cortex afferents onto old adult-born neurons. A, Schematic representation of the cortical input areas in a horizontal slice. Percentages of cortical innervation onto old adult-born neurons in mice housed under control (CON; n = 4) or long-term running (RUN; n = 4) conditions [entorhinal cortex (EC): caudomedial EC (CEnt), lateral EC (LEC) and medial EC (MEC); perirhinal cortex (PRh) and sensory cortices (SCtx)]. B, PRh input to adult-born neurons is only present in RUN mice (Kolmogorov–Smirnov test p = 0.0286). C, Running increases the total number of traced cells in the EC (t(6) = 3.234, p = 0.0178). D, Distribution analysis revealed that long-term running increased CEnt traced cell number (F(2,12) = 16.031, p < 0.0004). E, Analysis of the TC/SC ratios showed that the CEnt/SC ratio in the CON group is lower than in LEC/SC, whereas in the RUN group the CEnt/SC ratio is higher than the LEC/SC ratio. Running reduced the LEC/SC ratio as compared with the CON group (F(2,12) = 19.54, p < 0.002). F, Modified mouse brain atlas images (Paxinos and Franklin, 2007) showing the dorso-ventral horizontal section depth (distance from bregma) corresponding to photomicrographs below. G,H, Photomicrographs of horizontal sections showing traced cells (MCh+) derived from CON (G) and RUN (H) mice. Scale bar: 250 μm. Nuclei were stained with DAPI (blue). Extended Data Figure 5-1 depicts the locations and number of sparse inputs from sensory cortices. Data are means ± SEM *p < 0.05.
- Figure 6.
Old adult-born neurons receive inputs from subcortical areas. A, Diagram of a horizontal mouse brain section (adapted from Paxinos and Franklin, 2007) showing the location of the diagonal band of Broca (DBB) and medial septum (MS) of the basal forebrain. Right, Photomicrographs of MCh+ cells in the MS. Scale bar: 50 μm. B, The highest percentage of traced cells was located in the DBB. C, Running did not modify the number of traced cells in either region (F(1,6) = 0.61, p < 0.46). D, Analysis of the ratio of TC/SC showed that there was no effect of running (F(1,6) = 1.13, p > 0.32). E, G, I, Schematic representation of horizontal mouse brain sections showing the location of MCh+ traced cells in the (E) medial mammillary nucleus, lateral part (MML), (G) Raphe nucleus (RN) and (I) thalamic nucleus (ThN; adapted from Paxinos and Franklin, 2007). Right, Photomicrographs of MCh+ cells (arrowhead) in the MML (E), RN (G), and ThN (I). The inset shows a higher magnification of MCh+ cells in their respective panels. Scale bar: 200 μm; inset 50 μm. Nuclei were stained with DAPI (blue). F, H, J, Running did not modify the number of traced (F) medial mammillary nucleus cells (MMN; t(6) = 1.309, p = 0.2386), (H) RN cells (t(6) = 1.698, p = 0.1405), or (J) ThN cells (t(6) = 2.178, p = 0.0723). Data are means ± SEM *p < 0.05. MMM, medial mammillary nucleus, medial part; DG, dentate gyrus; GN, dorsal geniculate nucleus; ThN ld, laterodorsal thalamic nucleus; ThN lp, lateral posterior thalamic nucleus.
- Figure 7.
Running throughout middle-age maintains old adult-born neuron wiring. A, Percentages of traced cells in intrahippocampal, cortical, and subcortical areas projecting onto old adult-born neurons in control (CON) and running (RUN) middle-aged mice. B, Long-term running-induced fold changes in old adult-born neurons, starter cells (SC), and in intrahippocampal, cortical and subcortical inputs. Only the subiculum (SUB; ∼6-fold), CEnt (∼4-fold), and PRh exceeded (∼170-fold) the fold change of the starter cells (∼3-fold). mGC, Mature granule cells; MC, Mossy cells; AS, astrocytes; INT, interneurons; PYR, pyramidal cells; PRE, presubiculum; PARA, parasubiculum; DBB, diagonal band of Broca; MS, medial septum; DG, dentate gyrus.
Extended Data
Extended Data Figure 1-1
Running increases the survival of old adult-born neurons born during early adulthood. A, Long-term running distance, there is a significant reduction in monthly distance run over time in male C57Bl/6 mice (n = 9; F(1,7) = 12.86, p < 0.0089). B, C, Photomicrographs of the dorsal dentate gyrus showing GFP+ cells from control (CON; B) and long-term running mice (RUN; C). Scale bar: 50 μm. Nuclei were stained with DAPI (blue). D, Long-term running increases the number of total GFP+ cells in the dentate gyrus compared to control middle-aged mice (t(6) = 5.196, p = 0.002; CON, n = 4; RUN, n = 4). E, Dorsal to ventral distribution analysis shows that running increases the number of GFP+ cells in the dorsal (dDG) but not intermediate (iDG) or ventral (vDG) dentate gyrus (F(2,12) = 8.64, p < 0.0047). F, The proportion of double-labeled cells (starter cells) with respect to the total number of GFP+ cells is similar between groups (t(6) = 0.7366, p = 0.4891). Data are mean ± SEM *p < 0.05 Download Figure 1-1, TIF file.
Extended Data Table 1-1
Statistical information Download Table 1-1, xlsx file.
Extended Data Figure 2-1
Old adult-born neurons receive intrahippocampal inputs. A, Photomicrograph of starter old adult-born neurons (SC; red + nuclear yellow) and traced mature GCs (mGCs; red only; arrowhead). B, The ratio of mGC/SC is similar in middle-aged long-term running (RUN) and control (CON) mice (t(6) = 1.984, p = 0.0945). C, D, Traced mGCs were located mainly in the dorsal dentate gyrus (dDG) rather than the intermediate (iDG) or ventral (vDG) dentate gyrus and this distribution did not differ between the groups (F(2,12) = 0.272, p > 0.77). E, Photomicrograph of traced mossy cells (MC) expressing MCh (red; arrowheads). F, The ratio of MC/SC is not modified by long-term running in middle-aged mice (t(6) = 1.408, p = 0.2089). G, H, Traced MC are homogeneously distributed through the dorso-ventral dentate gyrus in middle-aged CON and RUN mice (F(2,12) = 0.062, p > 0.94). I, Photomicrograph showing traced astrocytes (red, arrowheads) surrounding the dendritic tree of an old adult-born SC neuron (red + yellow nuclei). J, The ratio of AS/SC was not modified by long-term running (t(6) = 1.872, p = 0.1104). K, L, Traced AS are homogeneously distributed through the dorso-ventral dentate gyrus in both groups (F(2,12) = 0.744, p > 0.49). Data are mean ± SEM. Scale bar: 10 μm. Nuclei were stained with DAPI (blue). Download Figure 2-1, TIF file.
Extended Data Figure 5-1
Old adult-born neurons receive sparse inputs from sensory cortices. A, Diagram of a horizontal section (adapted from Paxinos and Franklin, 2007) showing the location of the primary visual cortex (V1), secondary visual cortex (V2), primary auditory cortex (AuD), and primary somatosensory cortex (S1). B Photomicrographs of a dorsal brain slice showing MCh+ cells located in the V1 (), V2 (▲), and AuD (*). C, Long-term running does not modify the inputs from sensory cortices onto the old adult-born neurons (t(6) = 0.5774, p = 0.5847). D,E, High magnification photomicrographs derived from B showing traced MCh+ cells located in (D) V1 and V2 and (E) AuD. Data are mean ± SEM. Scale bars: 100 μm. Nuclei were stained with DAPI (blue). Download Figure 5-1, TIF file.
In this issue
