A Potential Mechanism Underlying the Therapeutic Effects of Progesterone and Allopregnanolone on Ketamine-Induced Cognitive Deficits

Abstract

Ketamine exposure can model cognitive deficits associated with schizophrenia. Progesterone (PROG) and its active metabolite allopregnanolone (ALLO) have neuroprotective effects and the pathway involving progesterone receptor membrane component 1 (PGRMC1), epidermal growth factor receptor (EGFR), glucagon-like peptide-1 receptor (GLP-1R), phosphatidylinositol 3 kinase (PI3K), and protein kinase B (Akt) appears to play a key role in their neuroprotection. The present study aimed to investigate the effects of PROG (8,16 mg kg^-1) and ALLO (8,16 mg kg^-1) on the reversal of cognitive deficits induced by ketamine (30 mg kg^-1) via the PGRMC1 pathway in rat brains, including hippocampus and prefrontal cortex (PFC). Cognitive performance was evaluated by Morris water maze (MWM) test. Western blot and real-time quantitative polymerase chain reaction were utilized to assess the expression changes of protein and mRNA. Additionally, concentrations of PROG and ALLO in plasma, hippocampus and PFC were measured by a liquid chromatography-tandem mass spectrometry method. We demonstrated that PROG or ALLO could reverse the impaired spatial learning and memory abilities induced by ketamine, accompanied with the upregulation of PGRMC1/EGFR/GLP-1R/PI3K/Akt pathway. Additionally, the coadministration of AG205 abolished their neuroprotective effects and induced cognitive deficits similar with ketamine. More importantly, PROG concentrations were markedly elevated in PROG-treated groups in hippocampus, PFC and plasma, so as for ALLO concentrations in ALLO-treated groups. Interestingly, ALLO (16 mg kg^-1) significantly increased the levels of PROG. These findings suggest that PROG can exert its neuroprotective effects via activating the PGRMC1/EGFR/GLP-1R/PI3K/Akt pathway in the brain, whereas ALLO also restores cognitive deficits partially via increasing the level of PROG in the brain to activate the PGRMC1 pathway.

Keywords: PGRMC1 signaling; allopregnanolone; cognitive deficits; ketamine; neuroprotection; progesterone.

FIGURE 1

Cognitive performance in the hidden platform trials of MWM test after ketamine exposure. (A) A diagram of the time course illustrating when the procedures took place. (B) Learning ability manifested as escape latency and path length of ketamine-exposed rats (n = 42) vs. NC (n = 7). (C) The swimming traces of the rats are illustrated (NC vs. Ket). ^{^} p < 0.05, ^{^^} p < 0.01, and ^{^^^} p < 0.001 for Ket vs. NC.

FIGURE 2

Cognitive performance in the hidden platform trials of MWM test after add-on PROG and ALLO treatments. (A) Escape latency (B) Path length (C) Escape latency on the third day and (D) path lengths on the third day were shortened by PROG (all p < 0.01) or ALLO treatment (all p < 0.0001) in comparison with ketamine treatment. Data are expressed as mean ± SD, n = 7 for each experimental group. ^* p < 0.05, ^** p < 0.01 and ^*** p < 0.001 for add-on of PROG and ALLO vs. Ket. ^# p < 0.05, ^## p < 0.01 and ^### p < 0.001 for PROG (8 mg kg⁻¹)+AG205 vs. PROG (8 mg kg⁻¹). ⁺ p < 0.05, ⁺⁺ p < 0.01 and ⁺⁺⁺ p < 0.001 for PROG (16 mg kg⁻¹) vs. PROG (8 mg kg⁻¹) and ALLO (16 mg kg⁻¹) vs. ALLO (8 mg kg⁻¹).

FIGURE 3

Basal protein expression profile of the PGRMC1/EGFR/GLP-1R/PI3K/Akt pathway in the hippocampus. (A) PGRMC1, (B) EGFR, (C) p-EGFR, (D) GLP-1R, (E) PI3K, (F) p-PI3K, (G) Akt, (H) p-Akt. Data are expressed as mean ± SD, n = 7 for each experimental group. ^{^} p < 0.05, ^{^^} p < 0.01, and ^{^^^} p < 0.001 for Ket vs. NC. ^* p < 0.05, ^** p < 0.01 and ^*** p < 0.001 for add-on of PROG and ALLO vs. Ket. ^# p < 0.05, ^## p < 0.01 and ^### p < 0.001 for PROG (8 mg kg⁻¹)+AG205 vs. PROG (8 mg kg⁻¹). ⁺ p < 0.05, ⁺⁺ p < 0.01 and ⁺⁺⁺ p < 0.001 for PROG (16 mg kg⁻¹) vs. PROG (8 mg kg⁻¹) and ALLO (16 mg kg⁻¹) vs. ALLO (8 mg kg⁻¹). The relative expression data are presented as the ratio to the β-actin protein level. Note: In our study, the prototype protein and its phosphorylated form were separated in one gel simultaneously. Therefore, the loading control images of β-actin are re-used for illustrative purposes in B and C, E and F, G and H.

FIGURE 4

Basal CREB/BDNF/caspase-3/9 protein expression in the hippocampus. (A) CREB, (B) truncated BDNF, (C) mature BDNF, (D) cleaved caspase-3 (17 kDa)/caspase-3, (E) cleaved caspase-3 (19 kDa)/caspase-3, (F) cleaved caspase-9/caspase-9. A representative blot is shown in (G). Data are expressed as mean ± SD, n = 7 for each experimental group. ^{^} p < 0.05, ^{^^} p < 0.01, and ^{^^^} p < 0.001 for Ket vs. NC. ^* p < 0.05, ^** p < 0.01 and ^*** p < 0.001 for add-on of PROG and ALLO vs. Ket. ^# p < 0.05, ^## p < 0.01 and ^### p < 0.001 for PROG (8 mg kg⁻¹)+AG205 vs. PROG (8 mg kg⁻¹). ⁺ p < 0.05, ⁺⁺ p < 0.01 and ⁺⁺⁺ p < 0.001 for PROG (16 mg kg⁻¹) vs. PROG (8 mg kg⁻¹) and ALLO (16 mg kg⁻¹) vs. ALLO (8 mg kg⁻¹). The relative expression data are presented as the ratio to the β-actin protein level.

FIGURE 5

Hippocampal Nrf2 mRNA expression and the enzyme activity of CAT, SOD, GSH-Px and the concentrations of PROG and ALLO in hippocampus, PFC and plasma. (A) Nrf2, (B) CAT, (C) GSH-Px, (D) SOD, (E) PROG concentrations, (F) ALLO concentrations. ^* p < 0.05, ^** p < 0.01 and ^*** p < 0.001 for add-on of PROG and ALLO vs. Ket. ⁺ p < 0.05, ⁺⁺ p < 0.01 and ⁺⁺⁺ p < 0.001 for PROG (16 mg kg⁻¹) vs PROG (8 mg kg⁻¹) and ALLO (16 mg kg⁻¹) vs. ALLO (8 mg kg⁻¹).

FIGURE 6

Illustrative model of the mechanism underlying the neuroprotective effects of PROG and ALLO against ketamine-induced cognitive deficits. Under certain conditions, the PI3K/Akt pathway can be activated to exert its neuroprotective function by phosphorylating a battery of protein substrates, including Nuclear factor erythroid-2-related factor 2 (Nrf2), caspase-3/9, cAMP response element-binding protein (CREB) and brain-derived neurotrophic factor (BDNF). Furthermore, the PI3K/Akt pathway is a putative downstream signaling pathway regulated by EGFR and GLP-1R to elicit multiple biological responses, especially cognitive function. Intriguingly, PGRMC1 co-precipitates and co-localizes with EGFR in cytoplasmic vesicles in cells and also serves as a novel component of the liganded GLP-1R complex. In the present study, the therapeutic effect of PROG or ALLO at least in part rely on the activation of PGRMC1/EGFR/GLP-1R/PI3K/Akt pathway in the brain. Moreover, the PRGMC1-specific inhibitor AG205 significantly antagonized the therapeutic effects of PROG in the MWM task and downregulated the PGRMC1/EGFR/GLP-1R pathway, which supports a role for PGRMC1 in the functional regulation of EGFR and GLP-1R, as well as their relation to neuroprotective effects.