Fig 2: AKT isoform abundance in human skeletal muscle. (A) Amino acid sequence alignment among human AKT isoforms. Sequences represent residues 359–418 from AKT1 (NCBI Reference Sequence: NP_001014432.1), residues 360–419 from AKT2 (NCBI Reference Sequence: NP_001617.1), and residues 356–415 from AKT3 (NCBI Reference Sequence: NP_005456.1). Regions in red font indicate residues of peptide sequences used in LC-PRM/MS. (B) Representative chromatograms for AKT isoforms using purified peptides and following immunoprecipitation of protein lysates using rabbit (left) and mouse (right) pan AKT antibodies. Asterisks denote the stable labeled synthetic internal standard peptide. (C) LC-PRM/MS was performed on eluates derived from immunoprecipitations of human vastus lateralis lysates using either rabbit or mouse pan AKT antibodies, as indicated (means ± SEMs; n = 4 subjects; *, P < 0.05 versus AKT1 and AKT3; ***, P < 0.001 versus AKT1 and AKT3). “amol;” attomoles; “µL;” microliter. (D) LC-PRM/MS was performed on eluates derived from immunoprecipitations of human astrocyte lysates using rabbit pan AKT antibody.

Fig 3: Blocking AKT activity leads to enhanced mGluR-LTD.(a,b) Blocking AKT activity using AZD5363 or MK2206 resulted in enhanced mGluR-LTD induced with 3,5 RS-DHPG (100 µM) in hippocampal area CA1 (AZD F(1,23)=5.010, p=0.035; MK F(1,40)=7.216, p=0.010), n = 12–20 slices/group, 5–6 mice/group. (c) Blocking PI3K using LY294002 resulted in enhanced mGluR-LTD (F(1,27)=10.809, p=0.003), n = 13–16 slices/group, 7–8 mice/group. (d–f) Akt1, Akt2 or Akt3 KO mice show normal mGluR-LTD compared to WT controls (p>0.05), n = 13–20 slices/group, 5–7 mice/group. *p<0.05.

Fig 4: AKT is not involved in LFS-LTD.(a) pGSK-3ß levels were significantly reduced in the hippocampus of Akt3 KO mice but normal in Akt1 KO mice (Akt3 KO t(20)=-4.725, p<0.001; Akt1 KO t(14)=0.104, p=0.918), 4–6 mice/group. (b–d) Hippocampal slices from Akt1 KO, Akt2 KO or Akt3 KO mice showed normal LFS-LTD induced by 900 stimuli of 1 Hz compared to WT slices (p>0.05 for all genotypes), n = 12–20 slices/group, 5–7 mice/group. (e) Blocking AKT activity in hippocampal slices with AZD5363 (30 µM) prior to 900 stimuli of 1 Hz did not affect LFS-LTD compared to vehicle-treated slices (p>0.05), n = 12–13 slices/group, 6 mice/group. (f) AKT activity blocked by MK2206 (30 µM) also resulted in normal LFS-LTD in hippocampal slices compared to vehicle-treated slices (p>0.05), n = 10–12 slices/group, 5 mice/group.

Fig 5: AKT1 and AKT3 are involved in mGluR-LTD.(a) Double Akt1 and Akt3 mutant (cA1F/A3K) mice with Cre-mediated removal of both Akt1 alleles in the Akt3 KO background were generated with WT littermate controls by breeding Camk2a-Cre::Akt1loxP/+/Akt3+/- female mice with Akt1loxP/+/Akt3+/- males. (b,c) Western blot analysis showing significantly reduced AKT1 levels and no AKT3 expression in the hippocampus of cA1F/A3K mice compared with WT mice (AKT1 levels: t(6)=3.802, p=0.008). (d) Immunostaining confirming reduced neuronal AKT1 expression by Camk2a-driven Cre removal of Akt1 in the hippocampus, especially in the pyramidal cell body layer of CA1. (e) mGluR-LTD is enhanced in cA1F/A3K hippocampal slices (F(1,30)=7.923, p=0.009), n = 16 slices/group, 5 mice/group. (f,g) Western blot analysis showing significantly increased phosphorylation of ERK Thr202/Tyr204 and S6 S235/236 but not TSC2 S939 in cA1F/A3K hippocampal slices at 60 min post-DHPG, while WT slices had no differences (pERK: WT Ctrl vs. DHPG t(22)=1.814, p=0.083, cA1F/A3K Ctrl vs. DHPG t(14)=2.918, p=0.011; pS6: WT Ctrl vs. DHPG t(22)=1.252, p=0.224, cA1F/A3K Ctrl vs. DHPG t(14)=2.647, p=0.019; pTSC2: WT Ctrl vs. DHPG t(20)=1.252, p=0.225, cA1F/A3K Ctrl vs. DHPG t(14)=0.374, p=0.714). n= 8–12 slices/group, 4–6 mice/group, *p<0.05.