Fig. 3

Insulin administration rescued central atrophy, neuronal and synaptic loss. a Brain weight was significantly reduced in DMO at P1 (**p < 0.01 vs. Control) and P7 (**p < 0.01 vs. Control). Insulin administration reversed this effect at 10 weeks of age (**p < 0.01 vs. rest of the groups, ††p < 0.01 vs. Control and Control + ICV-Ins) (P1 Control n = 6 and SZT n = 4; P7 Control n = 7 and STZ n = 9; 10 weeks Control n = 27, Control + ICV-Ins n = 17, Control + IN-Ins n = 8, STZ n = 18, STZ + ICV-Ins n = 10 and STZ + IN-Ins n = 6). b Cortical (**p = 0.03 vs. rest of the groups) and hippocampal (**p < 0.01 vs. rest of the groups) thickness are maintained after insulin treatment to DMO (n = 5 in all groups). c Illustrative example of cortical thickness after cresyl violet staining. Scale bar = 125 μm. d Cortical caspases 3/7 are activated in DMO at P1 (**p < 0.01 vs. Control) and P7 (*p = 0.016 vs. Control) while differences are no longer detectable at 10 weeks of age (p = 0.074). A similar profile is observed in the hippocampus at P1 (*p = 0.014 vs. Control), P7 (**p = 0.005 vs. Control) and at 10 weeks of age (p = 0.386) (P1a and P7 n = 9, 10 weeks n = 10 in all groups). e Insulin treatment maintained NeuN-positive cells in DMO (**p < 0.01 vs. rest of the groups, ††p < 0.01 vs. Control, Control + ICV-Ins, Control + IN-Ins and STZ+ IN-Ins) (Control n = 4, Control + ICV-Ins n = 5, Control + IN-Ins n = 5, STZ n = 5, STZ + ICV-Ins n = 5 and STZ + IN-Ins n = 5). f Illustrative example of NeuN immunostaining and DAPI counterstain. Scale bar = 125 μm. g Neuritic curvature was not affected in any of the groups under study (p = 0.649) (n = 4 in all groups). h Spine density was reduced in DMO and insulin treatment reverted this effect (**p < 0.01 vs. rest of the groups, ‡‡p < 0.01. vs. Control + IN-Ins and STZ + IN-Ins) (Control n = 4, Control + ICV-Ins n = 4, Control + IN-Ins n = 5, STZ n = 3, STZ + ICV-Ins n = 3 and STZ + IN-Ins n = 6). i Neuronal complexity was significantly reduced in DMO and insulin administration reversed this limitation (<10 μm: ††p < 0.01 vs. Control, Control + ICV-Ins, Control + IN-Ins, STZ + IN-Ins, ‡‡p < 0.01. vs. Control + IN-Ins and STZ + IN-Ins], 10–20 μm: **p < 0.01 vs. rest of the groups, ‡‡p < 0.01. vs. Control + IN-Ins and STZ + IN-Ins, 20–30 μm: **p < 0.01 vs. rest of the groups, ‡‡p < 0.01. vs. Control + IN-Ins and STZ + IN-Ins, 40–50 μm: ††p < 0.01 vs. Control, Control + ICV-Ins, Control + IN-Ins, STZ + IN-Ins, ‡‡p < 0.01. vs. Control + IN-Ins and STZ + IN-Ins, 50–60 μm: ††p < 0.01 vs. Control, Control + ICV-Ins, Control + IN-Ins, STZ + IN-Ins, ‡‡p < 0.01. vs. Control + IN-Ins and STZ + IN-Ins) Control n = 4, Control + ICV-Ins n = 4, Control + IN-Ins n = 5, STZ n = 3, STZ + ICV-Ins n = 3 and STZ + IN-Ins n = 6). j Illustrative example of neuronal complexity (Scale bar = 25 μm) and spine density (Scale bar = 10 μm). l Cortical synaptophysin levels were reduced in DMO at P1 (**p = 0.001 vs. control) and at P7 (**p < 0.001 vs. control). Insulin administration rescued synaptophysin levels at 10 weeks of age (‡‡p = 0.002 vs. Control, Control + ICV-Ins, Control + IN-Ins and STZ + IN-Ins) (P1 Control n = 6 and SZT n = 4; P7 Control n = 6 and STZ n = 6; 10 weeks Control n = 6, Control + ICV-Ins n = 6, Control + IN-Ins n = 5, STZ n = 6, STZ + ICV-Ins n = 6 and STZ + IN-Ins n = 4). A similar profile was observed in hippocamal synaptophysin levels at P1 (**p < 0.001 vs. control), P7 (**p < 0.001 vs. control) and 10 weeks of age (♯♯p = 0.06 vs. Control + IN-Ins and Control + IN-Ins) (P1, P7 and 10 weeks n = 6 in all groups). Student t-test for independent samples was used when 2 populations were under study (P1 and P7) and one-way ANOVA, followed by Tuckey b or Tamhane test, was used when 6 groups were under study (5 and 10 weeks). k Illustrative example of cortical synaptophysin levels in all groups under study