A functional role of meningeal lymphatics in sex difference of stress susceptibility in mice

Abstract

Major depressive disorder is one of the most common mental health conditions. Meningeal lymphatics are essential for drainage of molecules in the cerebrospinal fluid to the peripheral immune system. Their potential role in depression-like behaviour has not been investigated. Here, we show in mice, sub-chronic variable stress as a model of depression-like behaviour impairs meningeal lymphatics in females but not in males. Manipulations of meningeal lymphatics regulate the sex difference in the susceptibility to stress-induced depression- and anxiety-like behaviors in mice, as well as alterations of the medial prefrontal cortex and the ventral tegmental area, brain regions critical for emotional regulation. Together, our findings suggest meningeal lymphatic impairment contributes to susceptibility to stress in mice, and that restoration of the meningeal lymphatics might have potential for modulation of depression-like behaviour.

Fig. 1. Sub-chronic variable stress (SCVS) impairs meningeal lymphatics in female but not male mice.

a The experimental timeline of the SCVS paradigm, behavioral tests, injection of the 70kD dextran tracer into the cisterna magna (i.c.m.), and tissue collection (sac). b Schematic diagram of the i.c.m tracer injection. c The schematic diagram of dura mater, with dotted red line bordering the superior sagittal sinus (SSS) as well as the confluence of sinus and transverse sinus (COS + TS) areas chosen for image analyses. d, e Representative images depicting the LYVE1 staining (gray) and the dextran tracer (red) in the SSS and COS + TS areas of dura mater of female (d) and male (e) mice, comparing non-stressed naive group (N) and the SCVS group. Scale bars: 500 µm. f, k Representative images depicting LYVE1-labeled meningeal lymphatic vessels (mLV) of female (f) and male (k) mice at higher magnification for analysis of the diameters. Scale bars: 200 µm. g–j, l–o Quantification of the fluorescence intensity of the LYVE1 staining (g, l), the area covered by mLV (h, m), and the diameter of LYVE1-labeled mLV (i, n) in the SSS and COS + TS areas of dura mater in female (g–i: n = 9–12 per group; results from three independent experiments) and male (l–n: n = 8–10 per group; results from two independent experiments) mice. j, o Quantification of the fluorescence intensity of the dextran tracer in the SSS and COS + TS areas of dura mater of female (j: n = 8–9 per group) and male (o: n = 7-8– per group) mice. p, r Representative images depicting the dextran tracer (red) and DAPI (blue) in the deep cervical lymph nodes (dCLN) of female (p) and male (r) mice. Scale bars: 500 µm. q, s Quantification of the fluorescence intensity of dextran tracer in the dCLN of female (q: n = 8–9 per group) and male (s: n = 7–8 per group) mice. All data are presented as mean ± s.e.m. and analyzed by unpaired Student’s t tests (g–j, m–o, q, s) or Mann–Whitney test (l). Source data are provided as a Source data file.

Fig. 2. Intracisternal delivery of AAV-VEGFC improves meningeal lymphatics and alleviates stress-induced depression-like behaviors in female mice.

a The experimental timeline of the intracisternal AAV infusion, sub-chronic variable stress (SCVS), behavioral tests, tracer injection into the cisterna magna (i.c.m.), and tissue collection (sac). b Representative images depicting eGFP (green)-labeled AAV1-infected cells surrounding LYVE1 (gray)⁺ meningeal lymphatics (mLV). Nuclei: DAPI (blue). left: whole mount dura mater, scale bars: 2000 µm. right: enlarged views of the boxed areas from the image on the left, scale bars: 200 µm. c Representative images depicting the LYVE1 staining (gray) and the dextran tracer (red) in the SSS and COS + TS areas of dura mater, comparing female mice injected with AAV-VEGFC versus those injected with AAV-eGFP. Both groups received SCVS. Scale bars: 500 µm. d Representative images depicting LYVE1⁺ mLV at higher magnification. Scale bars: 200 µm. e–h Quantification of the fluorescence intensity of the LYVE1 staining (e), the area covered by mLV (f), the diameter of LYVE1⁺ mLV (g), and the fluorescence intensity of the dextran tracer in the SSS and COS + TS areas of dura mater (h). i Representative images depicting the dextran tracer (red) and DAPI (blue) in the deep cervical lymph nodes (dCLN). Scale bars: 500 µm. j Quantification of the fluorescence intensity of dextran tracer in the dCLN. k Quantification of changes in the body weight. l Quantification of grooming duration in the splash test (ST). m Quantification of the immobile time in the forced swim test (FST). n Representative traces of animal’s paths in the open field (OF). o–q Quantification of the traveled distance in the center zone as a percentage of the total traveled distance (o), the time spent in the center zone as a percentage of the total time (p), and total traveled distance (q) in the OF. r Quantification of the latency to eat in the novelty-suppressed feeding test (NSF) (e–h, j–m, o–r: n = 8 per group; results from two independent experiments). All data are presented as mean ± s.e.m. and analyzed by unpaired Student’s t tests (e, g, j–m, o, p) or Mann–Whitney tests (f, h, q, r). Source data are provided as a Source data file.

Fig. 3. Intracisternal delivery of AAV-VEGFC in female mice rescues sub-chronic variable stress (SCVS)-induced impaired expression of astrocytic markers in the mPFC and cFOS expression in the VTA dopaminergic neurons.

a Upper panel, the coronal atlas, with the brain regions of interest highlighted in orange. Lower panel, Representative images of intracisternally-injected OVA-Alexa Fluor 647 tracer (red) distribution in the brain sections at 15, 30, and 60 min post-injection, with DAPI staining (blue). Scale bars: 1000 μm. Yellow arrowheads indicated accumulation of tracer in the mPFC at 15 and 30 min post-injection, and in the VTA at 30 and 60 min post-injection. b Quantification of the distribution of tracer in multiple brain regions at different time points post-intracisternal injection (n = 4 per group). c The experimental timeline of the intracisternal AAV infusion, sub-chronic variable stress (SCVS) paradigm, and tissue collection (sac). d, g Representative images depicting the S100β (d: scale bars: 50 µm) and GFAP (g: scale bars: 200 µm) staining in the mPFC. e, h Quantification of the percentage of covered area by S100β (e) and GFAP (h) staining in mPFC. f Quantification of the density of S100β-labeled astrocytes in the mPFC. i Representative images of the TH (red) and c-FOS (gray) staining in the VTA. White arrowheads denote cells dual-labeled by c-FOS and TH. Scale bars: 100 µm. j Quantification of the percentage of c-FOS⁺ neurons in TH-labeled dopaminergic neurons in the VTA (e–j: n = 5–6 per group; results from two independent experiments). All data are presented as mean ± s.e.m. and analyzed by one-way ANOVA followed by Tukey’s post hoc tests. Source data are provided as a Source data file.

Fig. 4. Intracisternal delivery of AAV-VEGFR3_d1-4 impairs meningeal lymphatics and promotes stress-induced depression-like behaviors in male mice.

a The experimental timeline of the AAV infusion, sub-chronic variable stress (SCVS), behavioral tests, trace injection into the cisterna magna (i.c.m.), and tissue collection (sac). b Representative images of eGFP (green)-labeled AAV9-infected cells surrounding LYVE1 (gray)-positive meningeal lymphatics (mLV). Nuclei: DAPI (blue). Left: whole mount dura mater, scale bars: 2000 µm. Right: enlarged views of the boxed areas from the image on the left, scale bars: 200 µm. c Representative images of the LYVE1 staining (gray) and the dextran tracer (red) in the SSS and COS + TS areas of dura mater, comparing male mice injected with AAV-VEGFR3_d1-4 (VEGFR3_d1-4 + SCVS) with those injected with AAV-VEGFR3_d4-7 (VEGFR3_d4-7 + SCVS). Both groups experienced SCVS. Scale bars: 500 µm. d Representative images of LYVE1⁺ mLV at higher magnification. Scale bars: 200 µm. e–h Quantification of the fluorescence intensity of the LYVE1 staining (e), the area covered by mLV (f), the diameter of LYVE1⁺ mLV (g), and the fluorescence intensity of the dextran tracer in the SSS and COS + TS areas of dura mater (h). i Representative images of dextran tracer (red) and DAPI (blue) in the dCLNs. Scale bars: 500 µm. j Quantification of the fluorescence intensity of dextran tracer in the dCLNs. k–n Quantification of grooming duration in the splash test (ST, k), the immobile time in the forced swim test (FST, l), the latency to eat in the novelty-suppressed feeding test (NSF, m) and changes in the body weight (n). o Representative traces of animal’s paths in the open field (OF). The center zone: the gray box in the center. p–r Quantification of the traveled distance in the center zone as a percentage of the total traveled distance (p), the time spent in the center zone as a percentage of the total time (q), and total traveled distance (r) in the OF (e–g, k–n, p–r: n = 8–9 per group; h, j: n = 7–8 per group; results from two independent experiments). All data are presented as mean ± s.e.m. and analyzed by unpaired Student’s t tests (e–h, j–l, n, p–r) or Mann–Whitney test (m). Source data are provided as a Source data file.

Fig. 5. Intracisternal delivery of AAV-VEGFR3_d1-4 in male mice promotes sub-chronic variable stress (SCVS)-induced impaired expression of astrocytic markers in the mPFC and cFOS expression in the VTA dopaminergic neurons.

a Upper panel, the coronal atlas, with the brain regions of interest highlighted in orange. Lower panel, Representative images of intracisternally-injected OVA-Alexa Fluor 647 tracer (red) distribution in the brain sections at 15, 30, and 60 min post-injection, with DAPI staining (blue). Scale bars: 1000 μm. Yellow arrowheads indicated accumulation of tracer in the mPFC at 15 and 30 min post-injection, and in the VTA at 30 and 60 min post-injection. b Quantification of the distribution of tracer in multiple brain regions at different time points post-intracisternal injection (n = 4 per group). c The experimental timeline of the intracisternal AAV infusion, SCVS paradigm and tissue collection (sac). d, g Representative images depicting the S100β (d: scale bars: 50 µm) and GFAP (g: scale bars: 200 µm) staining in the mPFC. e, h Quantification of the percentage of covered area by S100β (e) and GFAP (h) staining in mPFC. f Quantification of the density of S100β-labeled astrocytes in the mPFC. i Representative images of the TH (red) and c-FOS (gray) staining in the VTA. White arrowheads denote cells dual-labeled by c-FOS and TH. Scale bars: 100 µm. j Quantification of the percentage of c-FOS⁺ neurons in TH-labeled dopaminergic neurons in the VTA (e, f, h, j: n = 7–8 per group; results from two independent experiments). All data are presented as mean ± s.e.m. and analyzed by one-way ANOVA followed by Tukey’s post hoc tests. Source data are provided as a Source data file.

Fig. 6. Surgical ligation of the lymphatic vessels afferent to the deep cervical lymph nodes (dCLN) in male mice promotes sub-chronic variable stress (SCVS)-induced depression-like behaviors as well as reduction of astrocytic protein expression in the mPFC and c-FOS expression in the VTA dopaminergic neurons.

a The experimental timeline of the surgical ligation of the lymphatic vessels afferent to the dCLN, sub-chronic variable stress (SCVS) paradigm, behavioral tests, injection of the dextran tracer into the cisterna magna (i.c.m.), and tissue collection (sac). b Representative images depicting the dextran tracer (red) and DAPI (blue) in the dCLN. Scale bars: 500 µm. c Quantification of the fluorescence intensity of dextran tracer in the dCLN. d Quantification of grooming duration in the splash test (ST). e Quantification of the immobile time in the forced swim test (FST). f Quantification of the latency to eat in the novelty-suppressed feeding test (NSF). g Quantification of changes in the body weight by SCVS. h Representative traces of animal’s paths in the open field (OF). i–k Quantification of the traveled distance in the center zone as a percentage of the total traveled distance (i), the time spent in the center zone as a percentage of the total time (j), and total traveled distance (k) in the OF. l, o Representative images of the S100β (l: scale bars: 50 µm) and GFAP (o: scale bars: 200 µm) staining in the mPFC. m, p Quantification of the percentage of covered area by S100β (m) and GFAP (p) staining in mPFC. n Quantification of the density of S100β-labeled astrocytes in mPFC. q Representative images of the TH (red) and c-FOS (gray) staining in the VTA. White arrowheads denote cells dual-labeled by c-FOS and TH. Scale bars: 100 µm. r Quantification of the percentage of c-FOS⁺ neurons in TH-labeled dopaminergic neurons in the VTA (c–g, i–k, p, r: n = 8 per group; m, n: n = 7–8 per group; results from two independent experiments). All data are presented as mean ± s.e.m. and analyzed by unpaired Student’s t tests. Source data are provided as a Source data file.

Fig. 7. Sub-chronic variable stress (SCVS) induces transcriptional changes of meningeal lymphatic endothelial cells in female mice.

a Schematic of harvesting the dura mater, isolating meningeal lymphatic endothelial cells (LECs) by fluorescence-activated cell sorting (FACS) and RNA sequencing. b Representative dot and contour plots showing the gating strategy used to isolate meningeal LECs (CD45⁻PDPN⁺CD31⁺). c–f Quantification of the percentages of LECs (c), blood endothelial cells (BECs, gated as CD45⁻PDPN⁻CD31⁺) (d), CD45⁺ leukocytes (e) and other stromal cells (CD45⁻CD31⁻) (f) in viable cells, comparing female SCVS mice with non-stressed naive mice (N) (5 biological replicates per group. Each replicate was pooled from 3 to 4 animals). Data are presented as mean ± s.e.m. and analyzed by unpaired Student’s t tests. g Heatmap showing relative expression levels of differentially expressed genes (DEGs, adjusted p < 0.1 by Deseq2) in the meningeal LECs, comparing naive (N) with SCVS groups. Color scale bar values represent standardized rlog-transformed values across samples. h Volcano plot showing gene expression changes by comparing naive (N) with SCVS groups. Blue and red dots represent significantly downregulated and upregulated genes, respectively. i Top enriched (normalized enrichment score, NES > 0; false-discovery rate, FDR < 0.25) and de-enriched (NES < 0, FDR < 0.25) gene sets in the SCVS samples revealed by gene set enrichment analysis (GSEA). GO-BP: gene ontology – biological process, GO-CC: gene ontology – cellular component, GO-MF: gene ontology – molecular function, KEGG: the Kyoto Encyclopedia of Genes and Genomes signaling pathway. j Heatmap of DEGs enriched in the serotonin uptake pathway. k Heatmap of DEGs co-regulated by stress and aging. l Heatmap of DEGs shared by LECs of female mice after SCVS and brain regions of depressed humans. AI anterior insula, NAc nucleus accumbens, vSUB ventral subiculum, dlPFC dorsolateral PFC, vmPFC ventromedial PFC, OFC orbitofrontal cortex. Source data are provided as a Source data file.

Fig. 8. Intracisternal delivery of CCL6-neutralizing antibody (CCL6 mAb) alleviates sub-chronic variable stress (SCVS)-induced depression-like behaviors and impairment of meningeal lymphatics (mLV) in female mice.

a The experimental timeline of the intracisternal CCL6 mAb or control immunoglobulin (IgG), SCVS paradigm, behavioral tests, and tissue collection (sac). b Representative heatmaps of animals’ trace in the social interaction (SI) test. c, d Quantification of the total exploration time and differential index (DI) in the SI, comparing female mice intracisternally injected with CCL6 mAb or control IgG. Both groups of mice experienced SCVS. e Quantification of the immobile time in the forced swim test (FST) (c–e: n = 7–9 per group; results from two independent experiments). f Representative images of the LYVE1 staining (gray) in the SSS and COS + TS areas of dura mater. Scale bars: 500 µm. g Representative images of LYVE1-labeled mLV at higher magnification. Scale bars: 200 µm. h–j Quantification of the fluorescence intensity of the LYVE1 staining (h), the area covered by mLV (i), and the diameter of LYVE1-labeled mLV (j) in the SSS and COS + TS areas of dura mater (h–j: n = 4–5 per group). All data are presented as mean ± s.e.m. and analyzed by unpaired Student’s t tests. Source data are provided as a Source data file.