PMC

Fig. 3L-EV treatment in CRST-treated mice reversed stress-induced decreased expression of BDNF in the hippocampus and rescued stress-induced depressive-like behaviors. (A) Experimental design (Exp.2). Mice exposed to CRST (2h×14 d RST) were treated with saline, imipramine (20 mg/kg), and L-EV (0.1 µg/kg for days 1~5; 0.18 µg/kg for days 6~7; 0.27 µg/kg for day 8 and thereafter). Behavior tests were performed on post-stress days 14~16 (p14~p16; behavior tests, #1) and post-stress days 28~30 (p28~p30; behavior tests, #2). Mice were sacrificed on post-stress day 35 (p35). Control mice injected with saline (CON+veh), mice treated with L-EV (CON+L-EV), mice treated with repeated restraint and injected with saline (CRST+veh), and mice treated with repeated restraint and injected with L-EV (CRST+L-EV) and mice treated with repeated restraint and injected with imipramine (CRST+IMI) were prepared. (B) Body weight changes of CON+veh, CON+L-EV, CRST+veh, CRST+L-EV, and CRST+IMI (n=8~12 animals per group). (C) Expression levels of Ngf, tBdnf, Bdnf1, Bdnf4, Nt3, Nt4/5, and Trkb in the hippocampus of CON+veh, CON+L-EV, CRST+veh, CRST+L-EV, and CRST+IMI (n=6~8 animals and 4~6 PCR repeats per group). For the comparison of CON+veh, CON+L-EV, CRST+veh, and CRST+L-EV groups, two-way ANOVA followed by Bonferroni post hoc test was used (Ngf, F(1,14)=1.549 and p=0.2337 for CRST main effect; F(1,14)=0.3197 and p=0.5808 for L-EV main effect; F(1,14)=0.1826 and p=0.6757 for CRST×L-EV; tBdnf, F(1,14)=4.871 and p=0.0445 for CRST main effect; F(1,14)=4.274 and p=0.0577 for L-EV main effect; F(1,14)=4.752 and p=0.0468 for CRST×L-EV; Bdnf1, F(1,12)=0.5907 and p=0.4570 for CRST main effect; F(1,12)=6.682 and p=0.0239 for L-EV main effect; F(1,12)=8.992 and p=0.0111 for CRST×L-EV; Bdnf4, F(1,13)=25.35 and p=0.0002 for CRST main effect; F(1,13)=7.794 and p=0.0153 for L-EV main effect; F(1,13)=17.60 and p=0.0010 for CRST×L-EV; Nt3, F(1,14)=43.45 and p<0.0001 for CRST main effect; F(1,14)=3.274 and p=0.0919 for L-EV main effect; F(1,14)=22.21 and p=0.0003 for CRST×L-EV; Nt4/5, F(1,11)=0.0645 and p=0.8042 for CRST main effect; F(1,11)=6.108 and p=0.0310 for L-EV main effect; F(1,11)=15.76 and p=0.0022 for CRST×L-EV; Trkb, F(1,14)=0.3743 and p=0.5505 for CRST main effect; F(1,14)=7.106 and p=0.0185 for L-EV main effect; F(1,14)=1.669 and p=0.2173 for CRST×L-EV). For the comparison of CON+veh, CRST+veh, CRST+L-EV, and CRST+IMI groups, one-way ANOVA followed by Newman-Keuls post hoc test was used (Ngf, F(3, 14)=1.495 and p=0.2591; tBdnf, F(3, 14)=2.938 and p=0.0699; Bdnf1, F(3, 11)=14.81 and p=0.0004; Bdnf4, F(3, 13)=5.176 and p=0.0143; Nt3, F(3, 12)=9.203 and p=0.0020; Nt4/5, F(3, 11)=2.170 and p=0.1493; Trkb, F(3, 14)=3.280 and p=0.0527). (D) Western blot analysis of proBDNF and BDNF expression in the hippocampus of CON+veh, CON+L-EV, CRST+veh, and CRST+L-EV (n=8 animals and 3~5 WB repeats per group). Two-way ANOVA followed by Bonferroni post hoc test (proBDNF, F(1,16)=22.62 and p=0.0002 for CRST main effect, F(1,16)=4.830 and p=0.0430 for L-EV main effect, and F(1,16)=3.681 and p=0.0731 for CRST×L-EV; BDNF, F(1,8)=1.329 and p=0.2823 for CRST main effect, F(1,8)=10.25 and p=0.0126 for L-EV main effect, and F(1,8)=0.1521 and p=0.7067 for CRST×L-EV). (E~H) Behavior tests (#1) on post-stress days 14~16. Representative tracking of mice placed in the sociability test (E). The percent time (%) spent in the target or non-target fields (F), and immobility time in the TST (G) and FST (H) of CON+veh, CON+L-EV, CRST+veh, CRST+L-EV, and CRST+IMI (n=6~12 animals per group). Two-way ANOVA followed by Bonferroni post hoc test was used for the comparison of CON+veh, CON+L-EV, CRST+veh, CRST+L-EV (target field in the sociability test, F(1,27)=2.243 and p=0.1458 for CRST main effect, F(1,27)=0.6705 and p=0.4200 for L-EV main effect, and F(1,27)=9.550 and p=0.0046 for CRST×L-EV; non-target field in the sociability test, F(1,27)=2.420 and p=0.1315 for CRST main effect, F(1,27)=0.5370 and p=0.4700 for L-EV main effect, and F(1,27)=9.869 and p=0.0041 for CRST×L-EV; TST, F(1,34)=7.324 and p=0.0.106 for CRST main effect, F(1,34)=0.6786 and p=0.4158 for L-EV main effect, and F(1,34)=5.308 and p=0.0275 for CRST×L-EV; FST, F(1,35)=29.00 and p<0.0001 for CRST main effect, F(1,35)=5.846 and p=0.0210 for L-EV main effect, and F(1,35)=18.19 and p=0.0001 for CRST×L-EV). One-way ANOVA followed by Neuman-Keuls post hoc test was for the comparison of CON+veh, CRST+veh, CRST+L-EV and CRST+IMI (target field in the sociability test, F(3,27)=6.470 and p=0.0019; non-target field in the sociability test, F(3,27)=6.435 and p=0.0020; TST, F(3,33)=6.772 and p=0.001; FST, F(3,34)=21.57 and p<0.0001). (I~L) Behavior tests (#2) on post-stress days 28~30. Representative tracking of mice placed in the sociability test (I). The percent time (%) spent in the target or non-target fields (J), and immobility time in the TST (K) and FST (L) of CON+veh, CON+L-EV, CRST+veh, CRST+L-EV, and CRST+IMI (n=6~12 animals per group). Two-way ANOVA followed by Bonferroni post hoc test was used for the comparison of CON+veh, CON+L-EV, CRST+veh, CRST+L-EV (target field in the sociability test, F(1,28)=1.580 and p=0.2192 for CRST main effect, F(1,28)=9.744 and p=0.0041 for L-EV main effect, and F(1,28)=1.332 and p=0.2582 for CRST×L-EV; non-target field in the sociability test, F(1,28)=1.557 and p=0.2224 for CRST main effect, F(1,28)=9.706 and p=0.0042 for L-EV main effect, and F(1,28)=1.304 and p=0.2632 for CRST×L-EV; TST, F(1,36)=2.902 and p=0.0971 for CRST main effect, F(1,36)=0.6042 and p=0.4421 for L-EV main effect, and F(1,36)=6.737 and p=0.0136 for CRST×L-EV; FST, F(1,35)=6.714 and p=0.0139 for CRST main effect, F(1,35)=5.713 and p=0.0224 for L-EV main effect, and F(1,35)=5.783 and p=0.0216 for CRST×L-EV). One-way ANOVA followed by Neuman-Keuls post hoc test was for the comparison of CON+veh, CRST+veh, CRST+L-EV and CRST+IMI (target field in the sociability test, F(3,27)=4.567 and p=0.0103; non-target field in sociability test, F(3,27)=4.535 and p=0.0106; TST, F(3,31)=6.029 and p=0.0023; FST, F(3,34)=5.762 and p=0.0027). Data are presented as mean±SEM. *p<0.05; **p<0.01.

Fig. 1L. plantarum-derived EVs (L-EVs) increased BDNF expression in HT22 cells. (A) Expression levels of tBdnf, Bdnf1, Bdnf4, and Ngf in cultured HT22 cells in the presence of GC (400 ng/ml), L-EV (20 µg/ml) or GC+L-EV for 24 h (n=8~12 per group). Two-way ANOVA followed by Bonferroni post hoc test (tBdnf, F(1,37)=0.0052 and p=0.9429 for GC main effect, F(1,37)=11.19 and p=0.0019 for L-EV main effect, and F(1,37)=14.75 and p=0.0005 for GC×L-EV; Bdnf1, F(1,32)=1.390 and p=0.2471 for GC main effect, F(1,32)=13.60 and p=0.0008 for L-EV main effect, and F(1,32)=7.166 and p=0.0116 for GC×L-EV; Bdnf4, F(1,35)=0.2697 and p=0.6068 for GC main effect, F(1,35)=22.60 and p<0.0001 for L-EV main effect, and F(1,35)=12.49 and p=0.0012 for GC×L-EV; Ngf, F(1,34)=27.95 and p<0.0001 for GC main effect, F(1,34)=24.33 and p<0.0001 for L-EV main effect, and F(1,34)=2.639 and p=0.1135 for GC×L-EV). (B) Western blot data showing the expression level of proBDNF in HT22 cells treated with GC (400 ng/ml), L-EV (20 µg/ml) or GC+L-EV for 24 h (n=6 per group). Two-way ANOVA followed by Bonferroni post hoc test (F(1,20)=14.19 and p=0.0012 for GC main effect, F(1,20)=3.819 and p=0.0648 for L-EV main effect, and F(1,20)=3.819 and p=0.0648 for GC×L-EV). (C) Expression levels of Sirt1, Hdac2, and Creb1 in HT22 cells treated with GC (400 ng/ml), L-EV (20 µg/ml) or GC+L-EV for 24 h (n=6~10 per group). Two-way ANOVA followed by Bonferroni post hoc test (Sirt1, F(1,27)=0.0168 and p=0.8979 for GC main effect, F(1,27)=65.48 and p<0.0001 for L-EV main effect, and F(1,27)=22.96 and p<0.0001 for GC×L-EV; Hdac2, F(1,22)=5.034 and p=0.0353 for GC main effect, F(1,22)=0.2970 and p=0.5913 for L-EV main effect, and F(1,22)=3.793 and p=0.0643 for GC×L-EV; Creb1, F(1,26)=0.0513 and p=0.8227 for GC main effect, F(1,26)=4.846 and p=0.0368 for L-EV main effect, and F(1,26)=0.0056 and p=0.9412 for GC×L-EV). (D, E) Expression levels of Sirt1(D), tBdnf, Bdnf1, Bdnf4, and Creb1 (E) in HT22 cells treated with siRNA-CON (50 pmol), siRNA-Sirt1(50 pmol), L-EV (20 µg/ml) or siRNA-Sirt1+L-EV for 24 h (n= 5~10 per group). Two-way ANOVA followed by Bonferroni post hoc test (Sirt1, F(1,23)=22.48 and p<0.0001 for siSirt1 main effect, F(1,23)=0.5684 and p=0.4586 for L-EV main effect, and F(1,23)=0.3164 and p=0.5792 for siSirt1×L-EV; Hdac2, F(1,28)=1.194 and p=0.2839 for siSirt1 main effect, F(1,28)=0.4969 and p=0.4867 for L-EV main effect, and F(1,28)=2.108 and p=0.1576 for siSirt1×L-EV; Creb1 , F(1,19)=47.84 and p<0.0001 for siSirt1 main effect, F(1,19)=0.0141 and p=0.9069 for L-EV main effect, and F(1,19)=0.1306 and p=0.7218 for siSirt1×L-EV; tBdnf, F(1,24)=3.681 and p=0.0670 for siSirt1 main effect, F(1,24)=1.632 and p=0.2137 for L-EV main effect, and F(1,24)=1.446 and p=0.2409 for siSirt1×L-EV; Bdnf1, F(1,21)=5.607 and p=0.0276 for siSirt1 main effect, F(1,21)=0.7048 and p=0.4106 for L-EV main effect, and F(1,21)=0.7991 and p=0.3815 for siSirt1×L-EV; Bdnf4, F(1,23)=25.64 and p<0.0001 for siSirt1 main effect, F(1,23)=0.3326and p=0.5697 for L-EV main effect, and F(1,23)=0.1315 and p=0.7202 for siSirt1×L-EV). GC, glucocorticoid; veh, vehicle. Data are presented as mean±SEM. *p<0.05; **p<0.01.

Extended Data Figure 5:TeNT-mediated silencing of DMHLepr, DMHPdyn, and LH→DMH neurons significantly attenuates learning a cue-initiated food acquisition task by increasing mistakes made during the task but TeNT-mediated silencing of DMHLepr neurons does not affect learning a cue-initiated water acquisition task.

Fig. 2L-EV treatment during the stress-treatment period blocked the stress-induced decrease in the expression of neurotrophic factors in the hippocampus and inhibited stress-induced depressive-like behaviors. (A) Experimental design (Exp. 1). Mice were treated with restraint for 2-h daily for 14 days (2h×14 d RST). Saline or L-EV (0.1 µg/kg) were i.p. injected 30 min before restraint treatment each day. Behavioral tests were performed on post-stress days 1-3 (p1~p3) and mice were sacrificed on post-stress day 7 (p7). Control mice injected with saline (CON+veh), mice treated with L-EV (CON+L-EV), mice treated with repeated restraint and injected with saline (CRST+veh), and mice treated with repeated restraint and injected with L-EV (CRST+L-EV) were prepared. (B) Body weight (g) changes of CON+veh, CON+L-EV, CRST+veh, and CRST+L-EV (n=8~12 animals per group). (C) Expression levels of Ngf, tBdnf, Bdnf1, Bdnf4, Nt3, Nt4/5, and Trkb in the hippocampus of CON+veh, CON+L-EV, CRST+veh, and CRST+L-EV (n=6 animals and 4~8 PCR repeats per group). Two-way ANOVA followed by Bonferroni post hoc test (Ngf, F(1,28)=2.900 and p=0.0997 for CRST main effect, F(1,28)=0.6187 and p=0.4381 for L-EV main effect, and F(1,28)=1.024 and p=0.3202 for CRST×L-EV; tBdnf, F(1,25)=0.2452 and p=0.6248 for CRST main effect, F(1,25)=15.43 and p=0.0006 for L-EV main effect, and F(1,25)=1.932 and p=0.1768 for CRST×L-EV; Bdnf1, F(1,12)=9.542 and p=0.0094 for CRST main effect, F(1,12)=15.01 and p=0.0022 for L-EV main effect, and F(1,12)=2.324 and p=0.1533 for CRST×L-EV; Bdnf4, F(1,12)=13.95 and p=0.0028 for CRST main effect, F(1,12)=9.418 and p=0.0097 for L-EV main effect, and F(1,12)=8.003 and p=0.0152 for CRST×L-EV; Nt3, F(1,28)0.9424 and p=0.3400 for CRST main effect, F(1,28)=0.1996 and p=0.6585 for L-EV main effect, and F(1,28)=1.974 and p=0.1710 for CRST×L-EV; Nt4/5, F(1,24)=14.45 and p=0.0009 for CRST main effect, F(1,24)=1.654 and p=0.2106 for L-EV main effect, and F(1,24)=8.419 and p=0.0078 for CRST×L-EV; Trkb, F(1,14)=0.0650 and p=0.8025 for CRST main effect, F(1,14)=1.768 and p=0.2049 for L-EV main effect, and F(1,14)=2.969 and p=0.1069 for CRST×L-EV). (D) Western blot data showing the expression level of proBDNF in the hippocampus of CON+veh, CON+L-EV, CRST+veh, and CRST+L-EV (n=6 animals and 4 WB repeats per group). Two-way ANOVA followed by Bonferroni post hoc test (F(1,12)=12.29 and p=0.0043 for CRST main effect, F(1,12)=1.697 and p=0.2172 for L-EV main effect, and F(1,12)=8.952 and p=0.0112 for CRST×L-EV). (E) Expression levels of Sirt1, Hdac2, and Creb1 in the hippocampus of CON+veh, CON+L-EV, CRST+veh, and CRST+L-EV (n=6 animals and 6–10 PCR repeats per group). Two-way ANOVA followed by Bonferroni post hoc test (Sirt1, F(1,23)=6.483 and p=0.0181 for CRST main effect, F(1,23)=12.17 and p=0.0020 for L-EV main effect, and F(1,23)=2.365 and p=0.1377 for CRST×L-EV; Hdac2, F(1,25)=10.31 and p=0.0036 for CRST main effect, F(1,25)=5.110 and p=0.0327 for L-EV main effect, and F(1,25)=1.340 and p=0.2280 for CRST×L-EV; Creb1, F(1,22)=0.2456 and p=0.6251 for CRST main effect, F(1,22)=2.829 and p=0.1067 for L-EV main effect, and F(1,22)=1.358 and p=0.2564 for CRST×L-EV). (F, G) Immobility time in the tail suspension test (TST; F) and forced swim test (FST; G) of CON+veh, CON+L-EV, CRST+veh, and CRST+L-EV (n=7~12 animals per group). Two-way ANOVA followed by Bonferroni post hoc test (TST, F(1,33)=11.90 and p=0.0016 for CRST main effect, F(1,33)=6.235 and p=0.0177 for L-EV main effect, and F(1,33)=1.665 and p=0.2059 for CRST×L-EV; FST, F(1,36)=23.84 and p<0.0001 for CRST main effect, F(1,36)=3.265 and p=0.0792 for L-EV main effect, and F(1,36)=12.17 and p=0.0013 for CRST×L-EV). Data are presented as mean±SEM. *p<0.05; **p<0.01.