How muscle talks to brain: apelin protein mediates exercise-induced antidepressant effects

Figures

• 

  Fig. 1. Muscle-specific apelin knockout diminishes the beneficial effects of voluntary running on antidepressant-like behaviors and adult neurogenesis. a Experimental timeline of 4‑week chronic unpredictable stress (CUS) and voluntary running (Run). CUS increased depression-like behavior which could be restored by running, as shown by b increasing sucrose preference in sucrose preference test (SPT): Interaction F(1, 47) = 3.866, P > 0.05; CUS F(1, 47) = 35.94, P < 0.001; Run F(1, 47) = 30.35, P < 0.001. c CUS reduced self‑grooming time in sucrose splash test (SST), but running showed opposite effects. Interaction F(1, 47) = 0.717, P > 0.05; CUS F(1, 47) = 14.51, P < 0.001; Run F(1, 47) = 34.84, P < 0.001. d CUS promoted immobility time in forced swimming test (FST) but running showed opposite effect. Interaction F(1, 47) = 3.123, P > 0.05; CUS F(1, 47) = 107.0, P < 0.001; Run F(1, 47) = 58.87, P < 0.001. e Neither CUS nor running affected time spent in the center in the open field test (OPT). Interaction F(1, 47) = 0.836, P = 0.365; CUS F(1, 47) = 1.636, P > 0.05; Run F(1, 47) = 0.171, P > 0.05. f Running and CUS decreased distance travelled in the OFT: Interaction F(1, 47) = 6.384, P <…

• 

  Fig. 2. Muscle-secreted apelin is involved in the antidepressant effect of voluntary running. a Treatment timeline for overexpression of apelin by intramuscular injections of adeno-associated viruses (AAV) in wildtype mice. AAV‑GFP (AAV-MCK-Control-P2A-GFP): control; AAV‑APLN (AAV-MCK-APLN-3×FLAG-P2A-GFP): muscle specific apelin overexpression. b Confirmation of intramuscular injection of AAV showing reporter gene expression of GFP in TA by the IVIS Spectrum system (PerkinElmer). Signal intensity is presented as a heat map and quantified as total radiant efficiency (p/s/cm²/sr)/(μW/cm²) using Living Image software. c-d Western blotting analysis confirming overexpression of apelin in muscle (P < 0.05). e Overexpression of apelin in muscles increased serum apelin levels (P < 0.01). f-h Western blotting analysis. Overexpression of muscle apelin increased protein expression of apelin (P < 0.01) and APJ in the hippocampi (P < 0.01). i Muscle apelin overexpression increased self‑grooming time in SST (P < 0.01), and j decreased immobility time in FST (P < 0.05), k but did not alter sucrose preference in SPT (P > 0.05). l-n Immunohistochemical staining showed that muscle apelin overexpression…

• 

  Fig. 3. Muscle apelin overexpression enhances excitatory neuronal transmission and NMDA receptor expression in the hippocampus. a Representative traces of sEPSCs in CA1 pyramidal neurons from AAV‑GFP (Control) and AAV‑APLN mice. b Cumulative probability plots for inter‑event intervals (s); d Amplitudes (pA) of sEPSCs. Muscle apelin overexpression increased: c sEPSC frequency (P < 0.05). e sEPSC amplitude (P < 0.01). f Maximal AMPA and NMDA currents in CA1 of AAV‑GFP and AAV‑APLN mice. Overexpression of muscle apelin increased: g NMDA/AMPA ratio (P < 0.05) and i NMDA current amplitude (P < 0.05) without altering h AMPA current amplitude (P > 0.05). j Muscle apelin overexpression decreased pair-pulse ratio at k 50 ms inter-pulse interval (P < 0.01), with no significant effects at l 100 ms or m 200 ms interval (P > 0.05). n Western blotting analysis using hippocampal synaptoneurosome proteins. Apelin overexpression increased expression levels of o phospho‑GluN2B (S1480, P < 0.01) and q phospho‑GluN2A (Y1246, P < 0.05) but did not affect p total GluN2B or r total GluN2A (P > 0.05). There was no significant effect on s phospho‑GluA1 (S831) and t total GluA1 (P > 0.05). Data are shown as mean…

• 

  Fig. 4. Knockdown of APJ in hippocampal glutamatergic neurons diminished anti-depressant effects of running. a APJ expressions primarily co-labelled with Camk2a+ neurons (glutamatergic neurons), rather than GAD67+ neurons (GABAergic neurons) in hippocampal dentate gyrus of wild‑type C57BL/6J mice (scale bars, 20 μm at 400×). b Treatment timeline of Camk2a‑Cre mice with APJ knockdown using intra‑ventral hippocampal injections with AAV. AAV‑RFP: AAV‑CMV‑DIO‑RFP-shControl; AAV‑shAPLNR: AAV‑CMV‑DIO‑RFP-shAPLNR; Sed: Sedentary; Run: Runner. c Confirmation of AAV injection site with RFP expression (scale bars, 500 μm at 40×), and d knockdown of APJ mRNA expression in ventral hippocampus (P < 0.05). APLNR mRNA expression normalized to GAPDH as the housekeeping gene. (e-g) Ventral hippocampal APJ knockdown in glutamatergic neurons reduced the antidepressant effect of running, as shown by absence of running-induced increase in e sucrose preference in SPT: Interaction F(1,43) = 7.453, P < 0.01; shAPLNR F(1,43) = 2.014, P > 0.05; Run F(1,43) = 7.256, P < 0.01. f Running increased grooming time in SST, but APJ knockdown did not show significant effect. Interaction F(1,43) = 2.327, P > 0.05; shAPLNR…

• 

  Fig. 5. Knockdown of APJ in hippocampal glutamatergic neurons diminishes the beneficial effects of running on enhancing synaptic NMDA receptor expression. a Western blotting analysis using hippocampal synaptoneurosome. AAV‑RFP: Camk2a‑Cre + AAV‑DIO‑RFP; AAV‑shAPLNR: Camk2a‑Cre + AAV‑DIO‑shAPLNR; Sed: Sedentary; Run: Runner. b Knockdown of APJ attenuated running‑induced increases in phosphorylation of GluN2B S1480: Interaction F(1,26) = 4.535, P < 0.05; shAPLNR F(1,26) = 5.129, P < 0.05; Run F(1,26) = 9.211, P < 0.01. Running increased phosphorylation of d GluN2A, but APJ knockdown had no significant effects: Interaction F(1,26) = 2.694, P > 0.05; shAPLNR F(1,26) = 1.555, P > 0.05; Run F(1,26) = 20.38, P < 0.001. f GluA1 S831: Interaction F(1,26) = 3.157, P > 0.05; shAPLNR F(1,26) = 0.399, P > 0.05; Run F(1,26) = 5.100, P < 0.05. Neither APJ knockdown nor running altered total protein levels of: c GluN2B: Interaction F(1,26) = 0.264, P > 0.05; shAPLNR F(1,26) = 6.513e‑007, P > 0.05; Run F(1,26) = 2.037, P > 0.05. e GluN2A: Interaction F(1,26) = 0.177, P > 0.05; shAPLNR F(1,26) = 2.189, P > 0.05; Run F(1,26) = 2.435, P > 0.05. g GluA1: Interaction F(1,26) = 5.522, P > 0.05; shAPLNR F(1,26)…

• 

  Fig. 6. Apelin/APJ interacts with Casein Kinase 2 to phosphorylate GluN2B-pS1480 of NMDA receptor. a Mechanistic hypothesis: Apelin binding to APJ interacts with casein kinase 2 (CK2), leading to phosphorylates GluN2B at S1480, to potentiates NMDAR function, consequently upregulate calpain‑2 to enhance hippocampal plasticity. Figure created with BioRender.com. b Western blotting analysis showing hippocampal synaptoneurosome CK2 levels in AAV‑GFP (Control) mice and AAV-APLN mice. c Muscle apelin overexpression increases CK2α in hippocampal synaptoneurosome (P < 0.05). d Muscle apelin overexpression promotes CK2 activity in hippocampal synaptoneurosome (P < 0.05). e Western blotting analysis shows CK subunits in ventral hippocampal synaptoneurosome. AAV‑RFP: Camk2a‑Cre + AAV‑DIO‑RFP; AAV‑shAPLNR: Camk2a‑Cre + AAV‑DIO‑shAPLNR; Sed: sedentary; Run: runner. Knockdown of APJ attenuates running‑induced increases in CK2α, CK2β expression: f CK2α: Interaction F(1, 20) = 4.769, P < 0.05; shAPLNR F(1, 20) = 11.61, P < 0.01; Run F(1, 20) = 5.119, P < 0.05. g CK2β: Interaction F(1, 20) = 2.288, P > 0.05; shAPLNR F(1, 20) = 4.311, P > 0.05; Run F(1, 20) = 4.037, P > 0.05. h Running promoted…