Segregation of mice with chronic pain into susceptible and unsusceptible subpopulations to memory impairment.

(A) Timeline of CCI surgery, pain threshold tests, Y-maze test and MWM training. (B) Representative travelling traces and statistical results of Y maze test showing distance and percent time in the novel arm (Red) in CCI-Acute mice (6d post CCI, n = 8-21). (C) Representative travelling traces and statistical results of MWM training showing escape latency and percent time in the quadrant in CCI-Acute mice (11d post CCI, n = 8-21). (D) Representative travelling traces and statistical results of Y maze test showing distance and percent time in the novel arm (Red) in CCI-Chronic mice (22d post CCI, n = 8-21). (E) Horizontal scatterplot depicting the distribution of ratio of time in novel arm for Sham, susceptible (S), and unsusceptible (U) mice in Y maze test. Bar graph represents the ratio of S and U mice in CCI-Chronic mice (22d post CCI, n = 71-152). (F) Representative travelling traces and statistical results of MWM training showing escape latency and percent time in the quadrant in CCI-Chronic mice (27d post CCI, n = 8-21). (G) Horizontal scatterplot depicting the distribution of ratio of time in the quadrant for Sham, S, and U mice in MWM training. Bar graph represents the ratio of S and U mice in CCI-Chronic mice (27d post CCI, n = 71-152). (H) Time in novel arm versus percent time in quadrant, for 152 CCI-Chronic mice. Each dot corresponds to one mouse. Colors of dots correspond to the groups of U in Y maze test but S in MWM (Red), U in both Y maze and MWM (Grey), S in Y maze test but U in MWM (Blue), and S in both Y maze and MWM(Yellow), respectively. Bar graph represents the ratio of each group in CCI-Chronic mice (n=152). (I) PWLs before and after administration of meloxicam (10mg/kg, i.p., n = 10-20). (J) Statistical results of MWM training showing escape latency and percent time in the quadrant in 27d CCI mice before meloxicam administration (10mg/kg, i.p., n = 10-20). (K) Performance of CCI mice in Y maze test before and after meloxicam administration (10mg/kg, i.p.). Bar graph (right) represents the ratio of U and S on 22d, 36d and 43d after CCI (n = 10-20). Data were analyzed by unpaired t test or one-way analysis of variance (one-way ANOVA), followed by post hoc Tukey’s multiple comparisons between multiple groups when appropriate. All data are presented as the mean ± s.e.m. ns, not significant; *p < 0.05; **p < 0.01; ***p < 0.001. CCI, chronic constrictive injury; d, day; MWM, Morris water maze; PWL, paw withdrawal latency; PWT, paw withdrawal threshold; U, unsusceptible; S, susceptible.

S1PR1 expression is decreased in the hippocampal DG of susceptible mice.

(A) Volcano plot showing RNA-Seq data for DG from Sham versus susceptible mice. DEGs are designated in red (upregulation [up]) and blue (downregulation [down]) and defined as having an FDR of less than 0.05. (B) Bar plot showing significant enrichment of DEGs in various pathways related with metabolism for Sham versus susceptible mice. (C) Relative expression levels are shown for genes related with lipid metabolism upon susceptible as compared with Sham. (D) Trend pattern used for analysis of Sham versus U versus S. (E) Bubble diagram represents the top 10 enrichment of KEGG pathways. (F-I) Example Western bands (F) and densitometric comparison (H) of the average expression of S1PR1 in DG lysates from Sham and CCI-Acute mice (7d post CCI). Lane1-6 represent Sham, Lane 7-12 represent CCI-Acute (n = 6); Example Western bands (G) and densitometric comparison (I) of the average expression of S1PR1 in DG lysates from Sham, U and S mice. Lane1-4 represent Sham, Lane 5-8 represent U, and Lane 9-12 represent S (n = 4). Data were analyzed by unpaired t test or one-way analysis of variance (one-way ANOVA), followed by post hoc Tukey’s multiple comparisons between multiple groups when appropriate. All data are presented as the mean ± s.e.m. ns, not significant; ***p < 0.001. CCI, chronic constrictive injury; d, day; U, unsusceptible; S, susceptible.

S1PR1 knockdown in the DG induces memory impairment.

(A) Timeline of intra-DG virus injection, CCI surgery, pain threshold tests, Y maze test and MWM training. (B) A confocal image showing virus expression in the DG (Scale bar, 100 μm). (C) Example Western bands showing efficient S1PR1 knockdown in the DG lysates from Sham-Scramble, Shamshs1pr1, CCI-Scramble, and CCI-shs1pr1. Densitometric comparison of the average expression of S1PR1 (n = 6). (D) Pain threshold in Sham- and CCI-treated mice subjected to Scramble/shs1pr1 in the DG (n = 8-10). (E) Quantitative summary of Y-maze showing distances traveled and time spent in the novel arm in Sham- and CCI-treated mice subjected to Scramble/shs1pr1 in the DG (n = 10-16). (F) Quantitative summary of MWM training showing escape latency and time spent in the quadrant in Sham- and CCI-treated mice subjected to Scramble/shs1pr1 in the DG (n = 10-16). (G) Ratio of U and S in CCI-Scramble and CCI-shs1pr1 mice. Data were analyzed by one-way analysis of variance (one-way ANOVA), followed by post hoc Tukey’s multiple comparisons between multiple groups. All data are presented as the mean ± s.e.m. ns, not significant; *p < 0.05; **p < 0.01; ***p < 0.001. CCI, chronic constrictive injury; MWM, Morris water maze; PWL, paw withdrawal latency; PWT, paw withdrawal threshold; U, unsusceptible; S, susceptible.

Overexpression of S1PR1 in the DG rescues chronic pain-induced memory impairment.

(A) Timeline of intra-DG virus injection, CCI surgery, pain threshold tests, Y maze test and MWM training. (B) A confocal image showing virus expression in the DG (Scale bar, 100 μm). (C) Example Western bands showing efficient S1PR1 overexpression in the DG lysates from Sham-Scramble, Sham-mimic, CCI-Scramble, and CCI-mimic. Densitometric comparison of the average expression of S1PR1 (n = 6). (D) Pain threshold in Sham- and CCI-treated mice subjected to Scramble/mimic in the DG (n = 8-10). (E) Quantitative summary of Y-maze showing distances traveled and time spent in the novel arm in Sham- and CCI-treated mice subjected to Scramble/mimic in the DG (n = 10-16). (F) Quantitative summary of MWM training showing escape latency and time spent in the quadrant in Sham- and CCI-treated mice subjected to Scramble/mimic in the DG (n = 10-16). (G) Ratio of U and S in CCI-Scramble and CCI-shs1pr1 mice. Data were analyzed by one-way analysis of variance (one-way ANOVA), followed by post hoc Tukey’s multiple comparisons between multiple groups. All data are presented as the mean ± s.e.m. ns, not significant; *p < 0.05; **p < 0.01; ***p < 0.001. CCI, chronic constrictive injury; MWM, Morris water maze; PWL, paw withdrawal latency; PWT, paw withdrawal threshold; U, unsusceptible; S, susceptible.

Local infusion of SEW2871 in the DG prevented the presence of chronic pain-induced memory impairment.

(A) Timeline of cannula implant, CCI surgery, pain threshold tests, Y maze test and MWM training. (B) Confocal image showing cannula implanted in the DG (Scale bar, 100 μm). (C) Example Western bands showing expression of S1PR1 in DG lysates from Sham-Vehicle, Sham-SEW2871, CCI-Vehicle, and CCI-SEW2871. Densitometric comparison of the average expression of S1PR1 (n = 6). (D) Pain threshold in Sham- and CCI-treated mice subjected to local infusion of vehicle/SEW2871 in the DG (n = 8-10). (E) Quantitative summary of Y-maze showing distances traveled and time spent in the novel arm in Sham- and CCI-treated mice subjected to local infusion of vehicle/SEW2871 in the DG (n = 10-16). (F) Quantitative summary of MWM training showing escape latency and time spent in the quadrant in Sham- and CCI-treated mice subjected to local infusion of vehicle/SEW2871 in the DG (n = 10-16). (G) Ratio of U and S in CCI-vehicle and CCI-SEW2871 mice. Data were analyzed by one-way analysis of variance (one-way ANOVA), followed by post hoc Tukey’s multiple comparisons between multiple groups. All data are presented as the mean ± s.e.m. ns, not significant; *p < 0.05; **p < 0.01; ***p < 0.001. CCI, chronic constrictive injury; DG, dentate gyrus; U, unsusceptible; S, Susceptible.

Susceptible mice exhibit altered excitatory synaptic plasticity in the hippocampal dentate gyrus.

(A) Representative TEM images of synapses in the DG in Sham and CCI-Chronic mice (27d post CCI). Blue indicates presynaptic site and yellow indicates postsynaptic sites of excitatory synapses, respectively. Synaptic densities are bracketed by arrows (Scale bar, 500 nm). (B) Mean number of excitatory synapses per μm2 of DG in Sham, unsusceptible and susceptible mice (n = 18-24 from 4 mice/group). (C) Cumulative distribution plots for the lengths and widths of postsynaptic density in the DG in Sham, unsusceptible and susceptible mice (n = 102-156 from 4 mice/group). (D) Representative Golgi-staining images of dendritic spine morphology from the DG in Sham, unsusceptible and susceptible mice (Scale bar, top:50 μm; bottom:10μm). (E) The number of intersections of all dendritic branches in Sham, unsusceptible and susceptible mice (n = 15-18 from 4 mice/group). (F) Violin plots indicate the total dendritic length. (G) Violin plots indicate the number of mushroom/stubby type dendritic spines (left), and the number of thin/filopodia type dendritic spines (right) in Sham, unsusceptible and susceptible mice (n = 15-18 from 4 mice/group). Data were analyzed by one-way analysis of variance (one-way ANOVA), followed by post hoc Tukey’s multiple comparisons between multiple groups when appropriate. All data are presented as the mean ± s.e.m. ns, not significant; *p < 0.05; **p < 0.01; ***p < 0.001. CCI, chronic constrictive injury; TEM, transmission electron microscope.

Overexpression of S1PR1 or local infusion of SEW2871 in the DG maintained the synaptic structural plasticity.

(A) Representative TEM images of synapses in the DG in Sham- and CCI-treated mice subjected to Scramble/mimic in the DG. Blue indicates presynaptic site and yellow indicates postsynaptic sites of excitatory synapses, respectively. Synaptic densities are bracketed by arrows (Scale bar, 500 nm). (B) Mean number of excitatory synapses per μm2 of DG in Sham- and CCI-treated mice subjected to Scramble/mimic (n = 18-24 from 4 mice/group). (C) Cumulative distribution plots for the lengths and widths of postsynaptic density in the DG in Sham- and CCI-treated mice subjected to Scramble/mimic in the DG (n = 121-162 from 4 mice/group). (D) Representative Golgi-staining images of dendritic spine morphology from the DG in Sham- and CCI-treated mice subjected to Scramble/mimic in the DG (Scale bar, top: 50 μm; bottom:10 μm). (E) The number of intersections of all dendritic branches in Sham- and CCI-treated mice subjected to Scramble/mimic in the DG (n = 18-24 from 4 mice/group). (F) Violin plots indicate the total dendritic length (n = 18-24 from 4 mice/group). (G) The number of mushroom/stubby type dendritic spines (left), and the number of thin/filopodia type dendritic spines (right) in the DG of in Sham- and CCI-treated mice subjected to Scramble/mimic in the DG (n = 18-24 from 4 mice/group). (H) Representative TEM images of synapses in the DG in Sham- and CCI-treated mice subjected to local infusion of vehicle/SEW2871 in the DG. Blue indicates presynaptic site and yellow indicates postsynaptic sites of excitatory synapses, respectively. Synaptic densities are bracketed by arrows (Scale bar, 500 nm). (I) Mean number of excitatory synapses per μm2 of DG in Sham- and CCI-treated mice subjected to local infusion of vehicle/SEW2871 (n = 18-24 from 4 mice/group). (J) Cumulative distribution plots for the lengths and widths of postsynaptic density in the DG in Sham- and CCI-treated mice subjected to local infusion of vehicle/SEW2871 in the DG (n = 121-162 from 4 mice/group). (K) Representative Golgi-staining images of dendritic spine morphology from the DG in Sham- and CCI-treated mice subjected to local infusion of vehicle/SEW2871 in the DG (Scale bar, top: 50 μm; bottom:10 μm). (L) The number of intersections of all dendritic branches in Sham- and CCI-treated mice subjected to local infusion of vehicle/SEW2871 in the DG (n = 18-24 from 4 mice/group). (M) Violin plots indicate the total dendritic length (n = 18-24 from 4 mice/group). (N) The number of mushroom/stubby type dendritic spines (left), and the number of thin/filopodia type dendritic spines (right) in the DG of in Sham- and CCI-treated mice subjected to local infusion of vehicle/SEW2871 in the DG (n = 18-24 from 4 mice/group). Data were analyzed by one-way analysis of variance (one-way ANOVA), followed by post hoc Tukey’s multiple comparisons between multiple groups. All data are presented as the mean ± s.e.m. ns, not significant; *p < 0.05; **p < 0.01; ***p < 0.001. CCI, chronic constrictive injury; DG, dentate gyrus; TEM, transmission electron microscope.

Susceptible mice, S1PR1-knockdown mice and S1PR1 overexpression mice exhibit altered CDC42/RAC1 activity, ARP2/3-dependent actin signaling and engage ITGA2 in the DGCs.

(A) Bubble diagram showing significant enrichment of DEGs in top 20 KEGG pathways for Sham versus susceptible mice. (B) Bubble diagram represents the top 20 enrichment of KEGG pathways using analysis of Sham versus U versus S. (C) Example Western bands showing expression of RAC1, CDC42, ARP2, ARP3, and ITGA2 in DG lysates from Sham and CCI-Acute mice (7d post CCI). (D) Densitometric comparison of the average expression of RAC1, CDC42, ARP2, ARP3 and ITGA2 (n = 6). (E) Example Western bands showing expression of RAC1, CDC42, ARP2, ARP3, and ITGA2 in DG lysates from Sham, U and S mice. (F) Densitometric comparison of the average expression of RAC1, CDC42, ARP2, ARP3 and ITGA2 (n = 6). (G) Example Western bands showing expression of RAC1, CDC42, ARP2, ARP3, and ITGA2 in DG lysates from in Sham- and CCI-treated mice subjected to Scramble/shs1pr1 in the DG. (H) Densitometric comparison of the average expression of RAC1, CDC42, ARP2, ARP3 and ITGA2 (n = 6). (I) Example Western bands showing expression of RAC1, CDC42, ARP2, ARP3, and ITGA2 in DG lysates from in Sham- and CCI-treated mice subjected to Scramble/s1pr1-mimic in the DG. (J) Densitometric comparison of the average expression of RAC1, CDC42, ARP2, ARP3 and ITGA2 (n = 6). Data were analyzed by unpaired t test or one-way analysis of variance (one-way ANOVA), followed by post hoc Tukey’s multiple comparisons between multiple groups when appropriate. All data are presented as the mean ± s.e.m. ns, not significant; *p < 0.05; **p < 0.01; ***p < 0.001. CCI, chronic constrictive injury; DG, dentate gyrus; U, unsusceptible; S, Susceptible.

S1PR1 regulates actin polymerization by interaction with ITGA2. (A-B)

Quantification of F/G-actin ratio in dentate gyrus of Sham, U and S mice by Western blot. (C) Phalloidin staining of F-actin showing the simple cytoskeleton of s1pr1-/- knockdown HT-22 cells and primary hippocampal neurons in comparison to highly organized actin fibers present within scramble HT-22 cells and primary hippocampal neurons (Scale bar, 50 μm). (D) Interaction between s1pr1 and itga2 in a yeast two-hybrid system. pNubG-Fe65 and pTSU2-APP were used as a pair of positive control. pPR3-N and pTSU2-APP were used as a negative control. DDO, SD/-Trp/-Leu; QDO, SD/-Trp/-Leu/-His/-Ade. (E) In vivo co-immunoprecipitation assay shows that S1PR1 interacts with ITGA2 in the dentate gyrus of mice. Total protein extracts were immunoprecipitated by the anti-S1PR1-specific antibody and analyzed by immunoblot probed with the anti-S1PR1 and anti-ITGA2 antibodies. Immunoglobulin G was used as the negative control. (F) Timeline of intra-DG virus injection, CCI surgery, pain threshold tests, Y maze test and MWM training. (G) A confocal image showing virus expression in the DG (Scale bar, 100 μm). (H) Example Western bands showing efficient ITGA2 knockdown in the DG lysates from Sham-Scramble, Sham-shitga2, CCI-Scramble, and CCI-shitga2. Densitometric comparison of the average expression of S1PR1 (n = 3). (I) Pain threshold in Sham- and CCI-treated mice subjected to Scramble/shitga2 in the DG (n = 10-20). (J) Quantitative summary of Y-maze showing distances traveled and time spent in the novel arm in Sham- and CCI-treated mice subjected to Scramble/shitga2 in the DG (n = 10-19). (K) Quantitative summary of MWM training showing escape latency and time spent in the quadrant in Sham- and CCI-treated mice subjected to Scramble/shitga2 in the DG (n = 9-20). (L) Timeline of intra-DG virus injection, Y maze test and MWM training. (M) A confocal image showing virus expression in the DG (Scale bar, 100 μm). (N) Quantitative summary of Y-maze showing distances traveled and time spent in the novel arm in WT mice subjected to Scramble/shs1pr1/shitga2/shs1pr1+shitga2 in the DG (n = 10). (O) Quantitative summary of MWM training showing escape latency and time spent in the quadrant in WT mice subjected to Scramble/shs1pr1/shitga2/shs1pr1+shitga2 in the DG (n = 10). Data were analyzed by one-way analysis of variance (one-way ANOVA), followed by post hoc Tukey’s multiple comparisons between multiple groups. All data are presented as the mean ± s.e.m. ns, not significant; *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001. CCI, chronic constrictive injury; d, day; MWM, Morris water maze; PWL, paw withdrawal latency; PWT, paw withdrawal threshold; DG, dentate gyrus; U, unsusceptible; S, susceptible.

Schematic representation of S1P/S1PR1 mediated susceptibility to chronic pain related memory impairment.

(A) Mice subjected to chronic pain can be separated into memory impairment susceptible and unsusceptible subpopulations 21 days post CCI. (B) Structural synaptic plasticity is mainly regulated by the actin cytoskeleton organization. In the DGCs of Sham/unsusceptible mice, S1P/S1PR1 signaling mediates actin dynamics via Itga2-dependent activation of the Rac1/Cdc42 signaling cascade and Arp2/3 dependent actin polymerization, whereas in susceptible mice dysregulation of S1P/S1PR1 signaling in the DGCs leads to defective actin cytoskeleton organization which alters the synaptic plasticity. CCI, chronic constrictive injury; DGCs, dentate granule cells.