New neurons migrate along blood vessels with abundant flow.

(A) Experimental scheme.

(B, C) Three-dimensional reconstructed images of a new neuron (green) and vasculature (red) in the rostral migratory stream (RMS) (B) and glomerular layer (GL) (C).

(D) Distance between new neurons and nearest vessels in the olfactory bulb and RMS (one-way repeated measures ANOVA followed by Bonferroni’s test; 3 and 4 mice for the analysis in the OB and RMS, respectively).

(E) Transmission electron microscopy image of a new neuron (green) and a blood vessel (red) in the GL. Astrocytes (clear arrowheads).

(F) Time-lapse images of a migrating neuron (asterisks) in the Dcx-EGFP mouse GL. Red blood cell (RBC) flow is recorded as a two-photon line-scan image as shown in the right panel. Stationary cells are indicated by sequential numbers.

(G) Average distance between migrating cells and nearest blood vessels (41 cells from 38 mice).

(H) Density of perivascular migrating cells (Wilcoxon signed-rank test; 10 mice).

(I) Proportions of new neurons migrating in different directions along vessels (39 interactions from 26 mice).

(J) Average migration speed (Welch’s t-test; low, 19 cells, high, 25 cells from 39 mice).

(K) Percentage of migratory period (Mann-Whitney U-test; low, 19 cells, high, 24 cells from 38 mice).

(L) Maximum migration speed (unpaired t-test; low, 22 cells, high, 24 cells from 39 mice).

(M) Fluorescent images in the NG2-DsRed mouse GL. Arterioles, capillaries, and venules were characterized by band-like smooth muscle cells (solid arrowhead), pericytes (arrows), and fenestrated smooth muscle cells (clear arrowhead), respectively. CD31 (blue), DsRed (red).

(N) Fluorescent image of a Dcx+/BrdU+ new neuron (solid arrowhead) attached to a capillary. Dcx (green), BrdU (blue), CD31 (blue, tube-like structures), DsRed (red).

(O) Density of BrdU+/Dcx+ cells in the perivascular region of arteriole-side and venule-side capillaries (paired t-test; 3 mice).

(P) Two-photon images of GABAergic neurons (white) and a blood vessel (red) in the VGAT-Venus mouse GL. Circles show positions of added (yellow) and lost (pink) Venus+ cells. Added and lost neurons are indicated by yellow and pink arrows, respectively. RBC flow on Day 21 is shown in the right panel.

(Q) Density of newly added neurons in the perivascular region (paired t-test; 7 mice).

Data are presented as the means ± standard error of the mean (SEM). Scale bars: B, 30 μm; C, 40 μm; E, 1 μm; F, 10 μm; M, 20 μm; N, 20 μm; P, 10 μm. See also Figure. S1 and S2.

Blood flow inhibition attenuates neuronal migration.

(A, D) Experimental schemes.

(B) Fluorescent images of Venus+ new neurons (green) in the rostral migratory stream and olfactory bulb (OB).

(C) Proportion of Venus+ cells in the OB in the Sham and bilateral carotid artery stenosis groups (unpaired t-test; Sham, 6 mice, BCAS, 5 mice).

(E) Two-photon images of neuronal migration (arrows) before and after photothrombotic clot formation in a Dcx-EGFP mouse. A new neuron (green), a blood vessel (red).

(F) Line-scan images from a blood vessel shown in (E).

(G, H) Comparison of migration speed before and after laser irradiation in the control (G) (paired t-test; 7 cells from 7 mice) and photothrombosis groups (H) (paired t-test; 4 cells from 4 mice).

Data are presented as the means ±SEM. Scale bars: B, 100 μm; E, 10 μm.

Ghrelin is delivered from the bloodstream to the RMS and OB.

(A) Representative images of the OB and the cortex from fluorescent ghrelin-infused mice. CD31 (red), Dcx (magenta), fluorescent ghrelin (green).

(B) Fluorescent images of neuronal migration along blood vessels in the EPL and the RMS. CD31 (red), Dcx (magenta), fluorescent ghrelin (green).

(C) Fluorescent images of blood vessels in the GL (C). CD31 (white), endomucin (red), fluorescent ghrelin (green).

(D) Normalized fluorescence signal intensity in vascular endothelial cells (paired t-test; 3 mice). Data are presented as the means ±SEM. Scale bars: A, 50 μm; B, 20 μm (EPL), 10 μm (RMS); C, 20 μm. See also Figure. S3.

Ghrelin promotes neuronal migration by activation of actin cup formation.

(A) Fluorescent images of Matrigel culture. Dcx (white).

(B) Percentage of Dcx+ cells > 200 μm distant from the edge of pellets (unpaired t-test; 3 independent cultures prepared on different days).

(C) Time-lapse images of cultured new neurons expressing DsRed (red). The number above each panel indicates minutes after initiation of migration.

(D–J) Migration speed (D), percentage of migratory phase (E), migration cycle (F), length/speed of leading process extension (G, H), and stride/speed of somal translocation (I, J) in neuronal migration (one-way ANOVA followed by Turkey-Kramer test; D-F; control / Ghrelin (-), 15 cells, control / Ghrelin (+), 13 cells, KD / Ghrelin (-), 13 cells, KD / Ghrelin (+), 18 cells, I, J; control / Ghrelin (-), 17 events, control / Ghrelin (+), 18 events, KD / Ghrelin (-), 14 events, KD / Ghrelin (+), 23 events).

(K) Time-lapse images of actin cup formation (arrowheads) in the cell soma of new neurons. EGFP-UtrCH (green). Condensed dots of F-actin were scattered throughout the elongated cell soma in a control cell with ghrelin application.

(L, M) Average duration of actin cups (N) and migration distance during actin cup formation (O) in new neurons (Kruskal-Wallis test followed by the Steel-Dwass test; control / Ghrelin (-), 79 cells, control / Ghrelin (+), 31 cells, KD / Ghrelin (-), 39 cells, KD / Ghrelin (+), 44 cells). Data are presented as the means ±SEM. Scale bars: A, 100 μm; C, 5 μm; K, 5 μm.

Ghrelin signaling promotes neuronal migration in the adult brain.

(A, D) Experimental schemes.

(B) Fluorescent images of new neurons in the olfactory bulb in (A).

(C) Proportion of labeled cells in the GL at 5 dpi. in (A) (paired t-test; 3 mice).

(E, G) Fluorescent images of new neurons in the glomerular layer (GL) (E) and rostral migratory stream (RMS) (G) for the experiments shown in (D). Control cells (white arrowheads), ghsr1a-KD cells (clear arrowheads).

(F, H) Proportion of labeled cells in the GL (F) and the RMS (H) at 10 dpi. in (D) (paired t-test; 4 mice).

(I, J) Proportion of labeled cells in the GL at 8 dpi in the ad libitum (I) and calorie restriction (J) groups (Control, unpaired t-test; AL, 4 mice, DR, 3 mice) (KD, unpaired t-test; AL, 4 mice, CR, 3 mice).

Control (green), ghsr1a-KD (red). GL (glomerular layer), EPL (external plexiform layer), MCL (mitral cell layer), IPL (internal plexiform layer), GCL (granule cell layer), RMS (rostral migratory stream). AL (ad libitum), CR (calorie restriction). Data are presented as the means ±SEM. Scale bars: B, 100 μm; E, 40 μm; G, 40 μm.

New neurons migrate along endomucin-negative vessels.

(A, B) Representative images of vasculature in the glomerular layer (GL) of the olfactory bulb (OB). Red blood cell (RBC) flow was recorded in a live animal (A), followed by immunostaining of endomucin/CD31 in a fixed brain section (B). Identical vessels are indicated by different numbers (endomucin-positive; 2, 6, 7, endomucin-negative; 1, 3, 4, 5).

(C) Average RBC flow in endomucin-positive and endomucin-negative vessels (Mann-Whitney U-test; endomucin-negative, 24 vessels, endomucin-positive, 46 vessels).

(D) Fluorescent image of new neurons distributed in the vasculature in the GL. BrdU+/Dcx+ cells are shown in the perivascular region of endomucin-negative vessels (white arrowheads), endomucin-positive vessels (clear arrowhead), and distant from vessels (arrow). CD31 (magenta), endomucin (green, tube-like structures), Dcx (green), BrdU (red).

(E–G) Density of BrdU+/Dcx+ cells in the vicinity of endomucin-positive and endomucin- negative vessels in the GL (E), granule cell layer (F), and rostral migratory stream (OB core) (G) (paired t-test; 4 mice).

(H) Density of BrdU+ mature neurons at 28 dpi. in the vicinity of endomucin-positive and endomucin-negative vessels in the GL (paired t-test; 4 mice).

Data are presented as the means ±SEM. Scale bars: A, B, 20 μm; D, 20 μm.

New neurons exhibit a preference for arteriole-side vessels.

(A) Fluorescent image of immunostained tissue sections from the glomerular layer of the olfactory bulb. Dcx (green), BrdU (deep blue), CD31 (red), SLC16A1 (green, tube-like structures).

(B) Density of perivascular new neurons in the vicinity of SLC16A1-positive and SLC16A1- negative vessels (paired t-test; 4 mice).

(C) Schematic illustration of the distribution of new neurons and vessel identification.

(D) Fluorescent images of the ventral striatum from a 4-month-old common marmoset. Immature neurons are indicated by solid arrowheads. Dcx (green), SLC16A1 (deep blue), CD31 (red).

(E, F) Density of BrdU+/Dcx+ cells in the vicinity of SLC16A1-positive and SLC16A1-negative vessels in the ventral striatum (E) (paired t-test; 5 animals) and in the neocortex (F).

Data are presented as the means ±SEM. Scale bars: A, 20 μm; D, 20 μm.

Blood-derived ghrelin enters the RMS and OB.

(A, B) Representative images of the OB from mice with saline injection (A) and with fluorescent ghrelin injection (B). High-magnification images are shown in (A’), (B’). It was found that the experimental process did not affect the brightness of sections. CD31 (red), Dcx (magenta), fluorescence (647 nm) (green).

Scale bars: A,B, 100 μm.

Calorie restriction promotes neuronal maturation in the OB.

(A) Fluorescent images of new neurons in the OB. BrdU (green), NeuN (red), Dcx (magenta). A NeuN-/Dcx+ cell (white arrow), NeuN+/Dcx+ cells (white arrowheads), NeuN+/Dcx-cells (yellow arrowheads).

(B) Proportion of NeuN+/Dcx-cells among total BrdU+ cells in the OB (KD, unpaired t-test; AL, 3 mice, CR, 4 mice).

IPL (internal plexiform layer), GCL (granule cell layer). Data are presented as the means ±SEM. Scale bars: A, 50 μm.